This report contains the collective views of an international group of experts anddoes not necessarily represent the decisions or the stated policy of the World Health Organization

WHO Technical Report Series951

THE SCIENTIFIC BASIS OFTOBACCO PRODUCT

REGULATION

Second report of a WHO study group

WHO Library Cataloguing-in-Publication Data

(WHO technical report series ; no. 951)

1.Tobacco use disorder - prevention and control. 2.Tobacco industry - legislation. 3.Tobacco controlcampaigns. 4. Tobacco-derived products labelling. 5.Tobacco-derived products packing. I.World HealthOrganization. II.WHO Tobacco Free Initiative. III.Series.

ISBN 978 92 4 120951 9 (NLM classification: QV 137)

ISSN 0512-3054

© World Health Organization 2008

All rights reserved. Publications of the World Health Organization can be obtained from WHO Press,World Health Organization, 20 Avenue Appia, 1211 Geneva 27, Switzerland (tel.: +41 22 791 3264;fax: +41 22 791 4857; e-mail: bookorders@who.int). Requests for permission to reproduce or translateWHO publications – whether for sale or for noncommercial distribution – should be addressed to WHOPress, at the above address (fax: +41 22 791 4806; e-mail: permissions@who.int).

The designations employed and the presentation of the material in this publication do not imply theexpression of any opinion whatsoever on the part of the World Health Organization concerning the legalstatus of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiersor boundaries. Dotted lines on maps represent approximate border lines for which there may not yet befull agreement.

The mention of specific companies or of certain manufacturers’ products does not imply that they areendorsed or recommended by the World Health Organization in preference to others of a similar naturethat are not mentioned. Errors and omissions excepted, the names of proprietary products aredistinguished by initial capital letters.

All reasonable precautions have been taken by the World Health Organization to verify the informationcontained in this publication. However, the published material is being distributed without warranty of anykind, either expressed or implied. The responsibility for the interpretation and use of the material lies withthe reader. In no event shall the World Health Organization be liable for damages arising from its use.

This publication contains the collective views of an international group of experts and does not necessarilyrepresent the decisions or the policies of the World Health Organization.

The scientific basis of tobacco product regulation: second report of a WHO study group.

Printed in Switzerland

Contents

WHO Study Group on Tobacco Product Regulation

Preface

1. Advisory note on smokeless tobacco products: health effects,implications for harm reduction and research1.1 Purpose1.2 Background1.3 Types of smokeless tobacco product1.4 Health effects1.5 Regional and global patterns of use1.6 Reducing harm due to use of smokeless tobacco1.7 Conclusions1.8 Recommendations1.9 Gaps in knowledge and research needsReferences

2. Advisory note on ‘fire-safer’ cigarettes: approaches to reducedignition propensity2.1 Purpose2.2 Background and history2.3 Regulatory responses

2.3.1 Effectiveness of regulatory measures in populations2.3.2 Regulatory considerations

2.4 Research needs2.4.1 Techniques2.4.2 Testing methods2.4.3 Surveillance and monitoring2.4.4 Exposure to emissions and smoking behaviour

2.5 Findings and recommendationsReferences Annex 2.1 Model legislation

3. Mandated lowering of toxicants in cigarette smoke:tobacco-specific nitrosamines and selected other constituents3.1 Executive summary and recommendations3.2 Introduction3.3 Background3.4 First report on mandated lowering of tobacco-specific

nitrosamines3.5 Use of mandated lowering of toxicant levels in a product

regulatory strategy3.5.1 Selection of machine testing method3.5.2 Sources of data on toxicants3.5.3 Criteria for selecting toxicants for mandatory lowering

regulation3.6 Recommended toxicants for reporting and regulation

vii

ix

1123577

10121215

171717202324272727228293033

45454852

53

576166

6678

iii

8

3.6.1 Toxicity of the selected constituents3.6.2 Methods for measuring the toxicants3.6.3 Existing techniques to reduce specific toxicant

emissions in cigarette smoke3.6.4 Considerations in applying mandatory lowering to

multiple toxicants3.7 Recommended toxicants and recommended mandated limits3.8 Implementing mandatory reductions in toxicant yields

3.8.1 Considerations for modified cigarettes and potentialreduced exposure products

3.8.2 Communication of regulatory values and testing resultsto the public

3.9 Issues for regulators mandating lower levels of toxicants3.10 Recommendations for follow-up and future work

3.10.1 Toxicants in tobacco smoke3.10.2 Smokeless tobacco products

References Annex 3.1. Levels of toxicants per milligram of nicotine for

international Philip Morris brands, Canadian brandsand Australian brands

Annex 3.2. Basis for calculation of toxicant animal carcinogenicityindex and toxicant non-cancer response index

Annex 3.3. Calculation of toxicant animal carcinogenicity indexand toxicant non-cancer response index forInternational Philip Morris brands, Canadian brandsand Australian brands

Annex 3.4. Correlation of toxicant yields from international brands,Canadian brands and Canadian brands with less than100 ng/mg nicotine in the modified intense smokingregimen with machine measurement

8087

93

104111113

114

115117119119121125

137

199

213

229

iv

Annex 3.5. Variation in yield of selected toxicants by brand

4. Recommendation on cigarette machine smoking regimens

5. Overall recommendations5.1 Harm reduction and smokeless tobacco products: regulatory

recommendations and research needs5.2 ‘Fire-safer’ cigarettes: approaches to reducing ignition

propensity5.3 Mandated lowering of toxicants in cigarette smoke: tobacco-

specific nitrosamines and selected other constituents5.4 Cigarette machine smoking regimens

260

269

273

273

274

275277

Acknowledgements

The WHO Study Group on Tobacco Product Regulation gratefully acknowl-edges the valuable contributions made to its work by the WHO TobaccoFree Initiative and International Agency for Research on Cancer workinggroup, established to provide scientific guidance on a strategy for regu-lating toxicant levels in cigarette smoke. The members of the group wereDr David Ashley, Centers for Disease Control and Prevention, USA,Dr David Burns, Dr Mirjana Djordjevic, Tobacco Control ResearchBranch of the National Cancer Institute, Bethesda, Maryland, UnitedStates, Dr Carolyn Dresler, Head of the Tobacco and Cancer Group, In-ternational Agency for Research on Cancer, France, Dr Erik Dybing,Dr Nigel Gray, Dr Pierre Hainaut, Dr Stephen Hecht, University ofMinnesota, USA, Dr Richard O’Connor, Assistant Professor of Oncology,Department of Health Behavior, Roswell Park Cancer Institute, Dr AntoonOpperhuizen, Head of the Laboratory of Toxicology, Pathology, andGenetics, of the National Institute of Public Health and the Environmentand Dr Kurt Straif, Scientist. The report submitted to TobReg aftermeetings over 18 months to define maximal limits for toxicants served asthe basis for discussion on the issue during the study group’s fourth meet-ing, held in Stanford, California, United States of America, from25 to 27 July 2007.

TobReg also expresses gratitude to the following, who were eithercommissioned to write background papers or lent their expertise on otherareas of tobacco product regulation reviewed by the study group at itsfourth meeting: Dr Gregory N. Connolly, Professor of Public HealthPractice at the Harvard School of Public Health, Boston, Massachusetts,United States, and Mr Denis Choinière, Director of Regulations andCompliance, Tobacco Control Programme, Health Canada, Ottawa,Canada, for their work on reduced ignition propensity; Dr MarkParascandola, epidemiologist with the Tobacco Control Research Branchof the National Cancer Institute, Bethesda, Maryland, United States,Mr Mitch Zeller, Vice President for Policy and Strategic Communications,Pinney Associates, Bethesda, Maryland Dr Dorothy Hatsukami, ForsterFamily Professor in Cancer Prevention, Department of Psychiatry,

v

University of Minnesota, USA for their contributions to the chapter onharm reduction; as well as Christy Anderson and Tore Sanner for theircontribution to the report on the regulation of products based on reducingthe toxicant levels in cigarettes smoke.

vi

WHO Study Group on Tobacco Product RegulationStanford University, California, United States of America, 25–27July 2007

Members

Dr D.L. Ashley, Chief, Emergency Response and Air ToxicantsBranch, Centers for Disease Control and Prevention, Atlanta,Georgia, United States of America

Dr D. Burns, Professor of Family and Preventive Medicine, Schoolof Medicine, University of California at San Diego, San Diego,California, United States of America

Dr M. Djordjevic, Program Director, National Cancer Institute,Division of Cancer Control and Population Sciences, TobaccoControl Research Branch, Bethesda, Maryland, United Statesof America

Dr E. Dybing (Chair), Director, Division of EnvironmentalMedicine, Norwegian Institute of Public Health, Oslo, Norway

Dr N. Gray, Scientist, International Agency for Research onCancer, Lyon, France

Dr S.K. Hammond, Professor of Environmental Health Sciences,School of Public Health, University of California at Berkeley,Berkeley, California, United States of America

Dr J. Henningfield, Professor (Adjunct), Behavioral Biology,Johns Hopkins University School of Medicine; Vice President,Research and Health Policy, Pinney Associates, Bethesda,Maryland, United States of America

Dr M. Jarvis, Principal Scientist, Cancer Research UK, HealthBehaviour Unit, Royal Free and University College LondonMedical School, London, England

Dr K.S. Reddy, Professor of Cardiology, All India Institute of Med-ical Sciences, New Delhi, India

vii

Dr C. Robertson, Senior Associate Dean for Faculty and Aca-demic Affairs, School of Engineering, Stanford University,California, United States of America

Dr G. Zaatari, Professor, Department of Pathology and Labora-tory Medicine, American University of Beirut, Beirut, Lebanon

Secretariat

Ms E. Tecson, Administrative Assistant, Tobacco Free Initiative,WHO, Geneva, Switzerland

Ms G. Vestal, Legal Officer and Scientist, Framework Conventionon Tobacco Control, Tobacco Free Initiative, WHO, Geneva,Switzerland

viii

Dr D.W. Bettcher, Coordinator, Framework Convention onTobacco Control, and Director, a.i., Tobacco Free Initiative,WHO, Geneva, Switzerland

Preface

Tobacco product regulation — regulating the contents and emissions oftobacco products by testing, mandating disclosure of test results and reg-ulating the packaging and labelling of tobacco products — is a pillar ofany comprehensive tobacco control programme. The Contracting Partiesto the World Health Organization Framework Convention on TobaccoControl (WHO FCTC) are legally bound by the treaty’s provisions ontobacco product regulation, contained in its Articles 9, 10 and 11.

The information provided by the ad hoc WHO Scientific Advisory Commit-tee on Tobacco Product Regulation, established in 2000 to fill gaps inknowledge on tobacco product regulation, served as the basis for thenegotiations and the subsequent consensus on the language of the afore-mentioned articles of the treaty.

In November 2003, in recognition of the importance of regulating tobaccoproducts, the WHO Director-General formalized the Scientific AdvisoryCommittee into the WHO Study Group on Tobacco Product Regulation(TobReg). TobReg’s membership comprises national and internationalexperts on product regulation, treatment of tobacco dependence, labora-tory analysis of tobacco contents and emissions and design features. Itswork is based on current research, and it also conducts research andproposes testing to fill regulatory gaps in tobacco control. The Director-General reports to the Executive Board on the results and recommenda-tions of the Study Group in order to draw the attention of Member Statesto WHO’s efforts in tobacco product regulation.

This technical report was prepared by TobReg in accordance with the prior-ities of the WHO Tobacco Free Initiative and the provisions of the WHOFCTC concerning tobacco product regulation, in response to requests fromMember States in which the population is affected by the issues addressed.The fourth meeting of TobReg was held at Stanford University, California,United States of America, on 25–27 July 2007. The agenda was preparedto respond partly to Decision 15 of the first session of the Conference ofParties to the WHO FCTC, held in Geneva, Switzerland, on 6–17 February2006, when the Parties adopted templates for guidelines for implementing

ix

x

Articles 9 and 10 of the Framework Convention. According to the templateon regulations, the guidelines should be based on work performed byTobReg and the WHO Tobacco Free Initiative, which serves as TobReg’ssecretariat and coordinating body.

This report presents the conclusions and recommendations of the WHO StudyGroup on Tobacco Product Regulation (TobReg) from its fourth meetingwhich was held at Stanford University, California, United States ofAmerica (USA), on 25–27 July 2007. The agenda was prepared to respondpartly to Decision 15 of the first session of the Conference of Parties tothe WHO Framework Convention on Tobacco Control, held in Geneva,Switzerland, on 6–17 February 2006, when the Parties adopted a templatefor the development of guidelines for implementing Articles 9 and 10 ofthe Framework Convention. At this fourth meeting of WHO TobReg,the Study Group deliberated on a number of topics in the field of tobaccoproduct regulation and produced the following advisory notes andrecommendations:

• an advisory note on smokeless tobacco products: health effects, implica-tions for harm reduction and research needs;

• an advisory note on ‘fire safer’ cigarettes: approaches to reduced ignitionpropensity;

• a recommendation on mandated lowering of toxicants in cigarette smoke:tobacco-specific nitrosamines and selected other constituents; and

• a recommendation on cigarette machine smoking regimens.

The four sections of this report address these four issues, and the StudyGroup’s recommendations are set out at the end of each section. Its overallrecommendations are summarized in section 5.

The Study Group’s members serve without remuneration in their personalcapacities and not as representatives of governments or other bodies; theirviews do not necessarily reflect the decisions or the stated policy of WHO.

1. Advisory note on smokelesstobacco products: health effects,implications for harm reduction andresearch

1.1 Purpose

Since the recommendation on smokeless tobacco by the predecessor of theWHO Study Group on Tobacco Product Regulation (TobReg), the ScientificAdvisory Committee on Tobacco Product Regulation, in 2003, the healtheffects of smokeless tobacco products have been addressed in numerous pub-lished reports and in reports of scientific advisory groups. Smokeless tobaccohas also been increasingly discussed with respect to its potential to reduce theharm caused by cigarette smoking.

It is beyond the scope of this advisory note to summarize all these reports indetail. It was, however, evident to TobReg that the recommendation onsmokeless tobacco of the 2003 Scientific Advisory Committee on TobaccoProduct Regulation should be revised to reflect advances in knowledge. Fur-thermore, since the 2003 recommendation, the WHO Framework Conventionon Tobacco Control (WHO, 2005) has come into force and has been ratifiedby more than 150 Parties, representing more than 80% of the world’s popu-lation. The Framework Convention is intended to reduce the prevalence ofand harm caused by tobacco use by regulating product communications,marketing, smuggling and patterns of use. To the greatest extent possible,implementation of the Framework Convention is based on scientific dataconcerning regulated products.

The scientific advisory notes of WHO TobReg are intended to facilitate im-plementation of the Framework Convention by evaluating the scientific basisof communications and regulation on tobacco products. For example, the2005 TobReg advisory note on waterpipe smoking made clear that this in-creasingly prevalent form of tobacco use is harmful and addictive andwarrants inclusion in comprehensive tobacco control efforts (TobReg, 2005).The aim of that advisory note was to address the common misconception thatwaterpipe smoking is a safe form of tobacco use and one that should beexempt from tobacco product regulation. There is little question that, in gen-eral, smokeless tobacco products are less harmful than combusted tobaccoproducts such as cigarettes; however, whether smokeless tobacco products

1

contribute to continuation or reduction of the global tobacco epidemic de-pends in part on their nature, how their health effects are communicated, howthey are marketed and how they are used. These factors must be consideredas the Framework Convention is implemented. This advisory note providesscientifically based guidance for implementation of the Framework Conven-tion that will contribute to reducing the use of tobacco and the harm it causes.

1.2 Background

Scientific publications, including reports by several expert panels, publishedsince 2003 have provided a strong foundation for the present advisory note.This note essentially replaces the 2003 recommendation by Scientific Advi-sory Committee on Tobacco Product Regulation (2004). The main expertpanels that have addressed smokeless tobacco and health are a working groupconvened by the International Agency for Research on Cancer (IARC, 2007),the initial report of the Scientific Committee on Emerging and Newly Iden-tified Health Risks of the European Commission (2008), the United States(US) National Institutes of Health panel addressing tobacco use (NationalInstitutes of Health, 2006), and the Royal College of Physicians and Surgeons(2007). The ‘strategic dialogue on tobacco harm reduction’ supported by theAmerican Legacy Foundation and the Robert Wood Johnson Foundationpublished conclusions that agree in general with the previous reports (Zeller,Hatsukami, Strategic Dialogue on Harm Reduction Group, 2007). Consider-ation was also given to the findings of an expert panel convened withunrestricted support by Star Scientific, Inc. to examine smokeless tobaccoand health issues; the panel included several international leaders in drugaddiction science and tobacco and health science and policy (Savitz et al.,2006). In addition, this advisory note relied on the conclusions of a WHOconsultation on tobacco harm reduction convened subsequent to the 13thWorld Conference on Tobacco and Health in Washington DC, USA, in July2006.

The scope of the subjects discussed by these expert panels on smokeless to-bacco differed somewhat, but their findings generally supported the broadconclusion that use of smokeless tobacco could reduce the harm due to to-bacco in people who would otherwise have smoked cigarettes and otherproducts. A concern expressed in several reports was that increased smoke-less tobacco use might undermine smoking cessation or be a risk factor forsmoking initiation and use of both products. This concern (and othersaddressed in this advisory note) must be resolved, so that such unintendedconsequences of the promotion of smokeless tobacco use do not contributeto the overall harm due to tobacco. Similar priorities for scientific researchare therefore addressed in this note. The conclusions also reflect the

2

deliberations of the TobReg and additional invited experts who met in July2007 (http://www.who.int/tobacco/global_interaction/tobreg/en/).

1.3 Types of smokeless tobacco product

Products referred to as ‘smokeless tobacco’ differ widely in form, contentand manufacture, with substantial regional and national variation. For exam-ple, in India, products are made and marketed by large national and multi-national companies as well as by street vendors and small unregistered familybusinesses. In contrast, the Swedish smokeless tobacco market is dominatedby the multinational Swedish Match company and a small number of nationalcompanies.

The report of the Scientific Committee on Emerging and Newly IdentifiedHealth Risks (European Commission, 2008) provides a comprehensiveoverview of product types, which is summarized briefly in Table 1.1. Theoriginal table contains additional information on product contents and useand the brand names of commonly marketed products. The reasons and pat-terns of use vary, some products being taken mainly after meals, some beingused at least partly to clean the teeth, some being used almost continuouslyand at least one product being used to calm teething infants. A common factoris that all appear capable of inducing and sustaining nicotine addiction, pri-marily by delivering nicotine to the oral cavity and nasal passages, althoughsome absorption undoubtedly occurs through the gastrointestinal tract fromswallowed nicotine-infused saliva. The nicotine concentrations in the prod-ucts used usually vary by more than 100-fold. The plasma levels of nicotineand the speed of delivery depend on pH and buffering capacity: raising theoral pH into the alkaline range results in more rapid nicotine absorptionthrough the buccal mucosa (Food and Drug Administration, 1995, 1996; Fantet al., 1999; Stratton et al., 2001; IARC, 2007; European Commission, 2008).The buffering agents include slaked lime, bicarbonate and ash. As discussedelsewhere and in detail in reports published by IARC (2007) and the ScientificCommittee on Emerging and Newly Identified Health Risks (European Com-mission, 2008), the products also vary by several thousandfold in theirconcentrations of carcinogens such as tobacco-specific nitrosamines(TSNA), benzo[a]pyrene, arsenic, nickel, and formaldehyde, radioactivepolonium-210, and other toxicants.

The wide diversity of products makes it difficult to make generalizationsabout the potential health hazards of products designated ‘smokelesstobacco’.

3

Table 1.1Some types of smokeless tobacco product

Common name Where used How used

Chewing tobacco Europe, USA Chewed tobacco can also besmoked in pipes.

Chimo Venezuela Tobacco, other plants, ash,used orally in the past

Creamy snuff India Tobacco, clove oil, mentholand other ingredients in apaste for mouth and toothcleaning

Dry rape snuff Brazil Dry, peppery tobacco powdertaken by inhalation

Dry snuff Georgia, Germany, UnitedKingdom

Dry tobacco powder taken byinhalation

Gul or gadakhu Central and eastern India Tobacco powder, molassesand other ingredients usedorally and for cleaning teeth

Gutkha Europe and South-East Asia Betel nut, tobacco, slakedlime and flavours, chewed orheld in the mouth

Khaini India Tobacco, lime paste andsometimes areca nut

Kiwan India Tobacco, slaked lime andspices, chewed or held in themouth

Loose-leaf chew USA Coarse shredded tobaccoand flavours for chewing

lq’mik Alaska Tobacco and punk ash usedorally and for teething infants

Mawa or kiwam India Tobacco, slaked lime andareca nut for chewing

Mishri, masheri or misheri India Tobacco applied to teeth andheld in the mouth

Moist plug, plug or twistchew

USA Flavoured tobacco leavesheld in the mouth or chewed

Moist snuff Sweden, Norway, USA Moist, finely ground tobaccoheld in the mouth

Nass, naswar or niswar Afghanistan, India, IslamicRepublic of Iran, Pakistan

Tobacco, ash, cotton orsesame oil, water, slakedlime, menthol and otheringredients, held in the mouthand sometimes chewed

Pan masala South-East Asia Tobacco, areca nut, slakedlime, betel leaf and flavours,sometimes chewed aftermeals

4

Red tooth powder India Tobacco powderShammah Saudi Arabia Tobacco, ash and slaked lime

for oral useSnuff South Africa Dry tobacco powder for

sniffingToombak Sudan Tobacco and sodium

bicarbonate, held in themouth

Zarda India Processed tobacco, betel nut,spices and dyes for oral use

Adapted from Table 1 in the report of the Scientific Committee on Emerging and Newly IdentifiedHealth Risks (European Commission, 2008)

1.3 Health effects

The report of an advisory committee to the US Surgeon General and theresults of a consensus conference convened by the US National Institutes ofHealth concluded that smokeless tobacco causes cancer in the oral cavity, theoesophagus and several other sites, dental disease and addiction (technicallydesignated as ‘disorders of dependence and withdrawal’), and is probably arisk factor for cardiovascular disease (Department of Health and Human Ser-vices, 1986). Since that report, the scientific evidence for these conclusionshas been strengthened considerably.

The conclusions of these scientific groups cannot be summarized briefly be-cause different approaches were used, and they addressed different questions.TobReg found, however, that some of their conclusions on the nature andhealth effects of smokeless tobacco products were relevant to its evaluationof potential strategies for reducing harm:

• All smokeless tobacco products are hazardous; they contain known toxi-cants, in addition to addicting levels of nicotine.

• The concentrations of toxicants and nicotine vary widely among marketedproducts.

• The extent and nature of the risk due to use of different marketed smokelesstobacco products vary widely.

• Smokeless tobacco products cause cancer, the site of cancer and the levelof risk depending on product characteristics, patterns and extent of use.

• Smokeless tobacco products cause various oral diseases.

• Smokeless tobacco products are addictive and cause the diseases of de-pendence and withdrawal listed by WHO in the tenth revision of the

5

International Classification of Diseases (WHO, 1992) and the AmericanPsychiatric Association Diagnostic and Statistical Manual (American Psy-chiatric Association, 1994).

• Smokeless tobacco increases the risk for death after myocardial infarctionbut does not increase the risk for myocardial infarction. Experiments inanimals and studies in humans indicate that oral tobacco use has short-termeffects on blood pressure and heart rate, which might contribute to cardio-vascular disease.

• There is mixed evidence concerning the causation of problems in preg-nancy, although a study in India indicated a dose–response relation be-tween the amount used daily by pregnant women and adverse pregnancyoutcomes; complications of pregnancy might vary with product charac-teristics and patterns and level of use.

• Smokeless tobacco products do not cause the lung diseases causally asso-ciated with use of combusted tobacco products such as in cigarettes, pipesand cigars.

• Smokeless tobacco products are not effective for smoking cessation, by thestandards required for endorsement of efficacy for smoking cessation; theresults of surveys suggest that some cigarette smokers give up smoking bysubstituting smokeless tobacco.

• Many smokeless tobacco products are marketed for use in situations inwhich smoking is not allowed, indicating that they may delay smokingcessation.

• Some smokeless tobacco products are manufactured and marketed in sucha way as to appeal to young people and thereby stimulate initiation of use.

• Smokeless tobacco products might delay smoking cessation or stimulatewider use of tobacco.

• Smokeless tobacco might have a role in reducing the harm due to tobaccofor people who cannot completely abstain from tobacco and who switchfrom combusted tobacco to smokeless products; however, the individualbenefits may depend on product characteristics and patterns of use.

• The benefits and risks for populations depend on the characteristics andpatterns of use of the products and the extent to which promotion of suchproducts increases the numbers of smokers of combusted tobacco or delayscessation of use of smoked tobacco products.

6

1.4 Regional and global patterns of use

Whereas cigarette smoking is a common form of tobacco use in every StateParty to the Framework Convention on Tobacco Control, the prevalence ofuse of smokeless tobacco products, the types of products and the patterns ofuse vary widely. In the European Union for example, use of oral (moist snuff)smokeless tobacco is low, its sale being prohibited in most countries. Sweden,however, is exempted, and the prevalence of use among men exceeds that ofsmoking. More than 20% of young men in Norway, which is not a memberof the European Union, report daily use of smokeless tobacco. The prevalenceof use is much higher in the USA than in Canada, Mexico and South America.It is high in North Africa and low in most of the rest of the African continent.The overall prevalence of smokeless tobacco use is highest in South-EastAsia, where use is also common among women.

In Norway, Sweden and the USA, virtually all smokeless tobacco productsare manufactured and marketed commercially. In contrast, many of thoseused in India and other parts of South-East Asia are manufactured locally,often by small vendors, with little or no standardization or possibility forregulating the contents, method of manufacture or packaging.

South-East Asia has some of the highest rates of oral cancer due to use ofsmokeless tobacco (IARC, 2007), and adverse reproductive effects, such aslow birth weight, lowered gestational age and more stillbirths, have beendocumented (Gupta, Sreevidya, 2004; Gupta, Subramoney, 2006).

1.5 Reducing harm due to use of smokeless tobacco

It is theoretically possible to reduce the risk for harm to tobacco users whoare unable or unwilling to quit, by changing the product, its characteristics orthe mode of intake by which they satisfy their need for nicotine, because mostof the harm is due to contents other than nicotine (TobReg, 2007). The re-duction of harm due to use of tobacco can be viewed from a broad publichealth perspective, in which the risks for disease and premature mortality arereduced by reducing exposure to toxicants and pathogens. For instance,malaria can be controlled by reducing the numbers of malaria spirochaete-carrying mosquitoes, respiratory disease can be controlled by reducing airpollution, and HIV/AIDS can be controlled by preventing unprotected sexand needle-sharing by intravenous injection drug users. The aim of the WHOFramework Convention on Tobacco Control is to contribute to reducing theharm due to tobacco by preventing use, encouraging cessation, protectingother people from exposure to second-hand tobacco smoke and regulatingproduct contents.

7

The definitions and goals of harm reduction vary according to the area ofpubic health but provide a useful perspective for tobacco. For example, theInternational Harm Reduction Association was established to address theharm caused by addictive drugs other than tobacco and alcohol, although ithas included alcohol and tobacco in its mandate since 2004. It defines harmreduction as “policies and programmes which attempt primarily to reduce theadverse health, social and economic consequences of all psychoactive sub-stances to individuals drug users, their families and their communi-ties” (http://www.ihra.net/). The definition given in a report of the USInstitute of Medicine (Stratton et al., 2001) is more specific, referring tomodification of conventional tobacco products, new products and pharma-ceutical products that “lower total tobacco-related mortality and morbidityeven though use of that product may involve continued exposure to tobacco-related toxicants”.

From the perspective of global harm reduction in the context of the WHOFramework Convention on Tobacco Control, TobReg considers that preven-tion, cessation and reduction of exposure to secondhand tobacco smoke arethe best-validated means of preventing disease and improving public health.TobReg also agrees, however, with the earlier expert panels, that tobaccousers who are unable or unwilling to quit could reduce harm by changingproducts, their characteristics and the form of intake. Like the expert groups,TobReg emphasizes that any such action must not undermine prevention,cessation and reduction of exposure to second-hand smoke and, ideally,should support them.

For the purpose of this advisory note, TobReg concluded that the objectiveof reducing harm due to tobacco use is to decrease morbidity and mortalityamong persons who continue to use tobacco and nicotine and are unwillingor unable to quit, with due consideration of the effects at population level.With this objective in mind, conclusions were reached, recommendationsmade and research needs identified, as described below.

TobReg established principles for harm reduction from smokeless tobacco,in order to make its goals transparent. Other expert groups, such as that of theUS Institute of Medicine (Stratton et al., 2001), have taken a similar approach.The principles listed below are generally consistent with those of the Instituteof Medicine but adapted to the principles and objectives of the FrameworkConvention.

• The risks associated with use of nicotine-containing products vary, the useof medicinal nicotine products entailing the lowest risk and use ofcigarettes generally entailing the highest risk.

8

• The risks associated with smokeless tobacco use vary according to theproduct used, from risks that might be important for harm reduction strate-gies to risks that are too small to be meaningful for such strategies.

• The differences in risks associated with use of different smokeless tobaccoproducts mean that it would be scientifically inappropriate to considersmokeless tobacco as a single product for the purposes of estimating riskor setting policies.

• The risk to a user of a particular product varies with the pattern and intensityof use.

• Apple-type services for unicode imaging (ATSUI) are needed to presentproduct characteristics and patterns of use and their effects on levels ofrisk. This technique gives unicode-encoded text advanced typographicalfeatures and automatically handles many of the complexities inherent intext layout, including correct rendering of text in bidirectional and verticalscript systems.

• Shifts in product use could reduce individual harm but increase harm tothe population as a whole or to segments of the population; therefore care-ful scientific assessment is needed of both individual risks and changes inuse at the population level when considering harm reduction approaches.

• Consideration of the harm-reducing potential for individuals of changingproducts and their characteristics should include the physical and chemicalcharacteristics of the product and its emissions, how it is used, the exposureof users to toxicants, measures of product toxicity, addiction potential andthe risk for disease related to use.

• Consideration of the harm-reducing potential for the population of strate-gies for shifting products or mandating changes in product characteristicsshould include the possibility of initiation of product use by non-users, thepossibility that the product might be used as a transition to a more harmfulproduct or might lead to cessation of all use, the possibility that use of twotobacco products will result in greater exposure to toxicants or prevent ordelay cessation, with relapse to more harmful products by people whowould otherwise quit, and the exposure of non-smokers.

• Claims that products reduce harm can have adverse effects on individualsand the population and should be made only on the basis of compellingscientific evidence.

• Demonstrations of altered product contents, lowered emissions or reducedintake of toxicants are not sufficient to support claims or implications ofreduced toxicity or harm.

9

• No specific or implied claim of harm reduction should be permitted withregard to any tobacco product before it has been validated by an indepen-dent regulatory agency on the basis of sufficient scientific data.

• Postmarketing surveillance of use patterns, consumer perceptions andhealth effects should be conducted in order to assess population effects.

• As claims of exposure reduction can be interpreted as claims of harm re-duction, the former should be allowed only when it can be demonstratedthat reductions in the levels of exposure to specific toxicants result in re-ducted risk at the population level.

• Claims of reduced harm based on evidence for a particular population canhave unintended consequences for vulnerable groups that would not oth-erwise have used the product.

1.7 Conclusions

In accordance with the WHO Framework Convention on Tobacco Controland the principles and objectives stated in this advisory note, and with dueconsideration for the conclusions of other expert groups, the WHO HarmReduction Consultation and the participants in the fourth meeting of TobReg(July 2007), TobReg came to the following conclusions about smokeless to-bacco and harm reduction.

• The composition, the hazardous properties and the manner in whichsmokeless tobacco products are used are widely diverse.

• Current evidence indicates that all smokeless tobacco products are haz-ardous to health.

• All smokeless tobacco products are addictive.

• Scientifically documented risks for disease associated with differentsmokeless tobacco products varies according to product, pattern of use andgeographical region.

• Users of smokeless tobacco products generally have lower risks for to-bacco-related morbidity and mortality than users of combustible tobaccoproducts such as cigarettes.

• People who would otherwise continue to smoke and who switch fromcigarette smoking to a smokeless product with a risk that is scientificallydocumented as being at the low end of the range are likely to have a reducedrisk for subsequent disease.

10

• People who continue to use smokeless tobacco products the characteristicsof which are changed to those with a risk that is scientifically documentedas being at the low end of the range are likely to benefit.

• Harm reduction strategies that involve encouraging smokeless tobacco usein regions where smokeless tobacco is the predominant type of tobaccoused may cause harm, particularly when the product used predominantlyhas substantial hazardous properties.

• The design and characteristics of smokeless tobacco products affect theirappeal by changing the taste, flavour and ease of use, the amount of nicotinedelivered, its addictiveness, and therefore the potential for harm.

• Contents of smokeless tobacco products other than tobacco affect the ap-peal of the product by changing the taste, flavour and ease of use, theamount of nicotine delivered and its addictiveness, and therefore the po-tential for harm.

• Packaging, labelling and marketing of smokeless tobacco products affecttheir appeal, the prevalence of their use, initiation, addiction and thereforethe potential for harm.

• Conflicting conclusions have been reached about whether initiation of to-bacco use with smokeless tobacco products results in a higher prevalenceof subsequent use of combustible tobacco products.

• The evidence that smokeless tobacco products are effective for smokingcessation does not meet the standards required, although the results of sur-veys suggest that some cigarette smokers have given up smoking bychanging to smokeless tobacco.

• Evidence that adult use of smokeless tobacco increases cessation ofcigarette smoking at the population level remains inconclusive and variesby country.

• Some smokeless tobacco products are marketed for use by cigarette smok-ers in situations where smoking is not allowed; the effect of such use oncessation is not known.

• The effects of concurrent use of combustible and smokeless tobacco prod-ucts on exposure to toxicant are not known.

• In countries where smokeless tobacco use is common, the prevalence ishighest among adolescents and young men.

• The pattern of smokeless tobacco use by adult populations varies by coun-try, with substantial adult use in some countries and use largely confinedto young men in others; the roles of marketing and social differences re-main uncertain.

11

1.8 Recommendations

TobReg recognized that recommendations on the possible negative or posi-tive role of smokeless tobacco in reducing the harm due to tobacco might beuseful for setting policies and implementing the Framework Convention. Therecommendations below were proposed for consideration, although somemay be irrelevant or inappropriate for some Parties. TobReg did not take aposition about the introduction of smokeless tobacco products in regionswhere they are not marketed, as that would involve complex health, politicaland other considerations.

• All tobacco products, including smokeless tobacco, should be subjectedto comprehensive regulatory control by an independent, scientificgovernment agency. Such control requires disclosure of ingredients bymanufacturers.

• Any health claim made for a smokeless tobacco product must be supportedby sufficient scientific data, as determined by an independent, scientificgovernment regulatory agency.

• As claims for reduced exposure could be interpreted as claims for harmreduction, no claims for reduced exposure should be permitted in the ab-sence of evidence for reduced risk.

• The contents and emissions of smokeless tobacco products must be testedand measured continuously in order to detect national and regional varia-tions and changes over time.

• Research on the health hazards and risks to individuals and populations ofuse of smokeless tobacco products is essential for governments and forimplementation of the Framework Convention.

• Research on smokeless tobacco products, their effects and how their designand manufacture are modified in order to alter their effects is essential foradequate testing and measurement, to provide information for governmentsand for implementation of the Framework Convention.

• As the risks posed by tobacco products vary, tobacco control policy shouldbe based on these differences.

1.9 Gaps in knowledge and research needs

As also concluded by other expert groups that have considered the potentialrole of smokeless tobacco products and other types of tobacco products (con-ventional and modified) in reducing the harm due to tobacco (e.g. Stratton etal., 2001; Scientific Advisory Committee on Tobacco Product Regulation,

12

2003; TobReg, 2004, 2007), TobReg recognized certain gaps in knowledge,which indicate that a cautious approach is appropriate.

Research on and testing of smokeless tobacco products, their use, the conse-quences of their use and the implications of different policies are essential,so that policies can be set that will improve health and minimize unintendedconsequences. The extent of the public health epidemic due to tobacco mar-keting and use makes it essential to consider alternative approaches, toaddress gaps in knowledge and to avoid taking actions that could worsen theepidemic regionally or globally. General principles for research and testingare given in the TobReg recommendations Guiding principles for the devel-opment of tobacco product research and testing (TobReg, 2004). There areboth many uncertainties and numerous areas in which there is a strong sci-entific foundation. For example, the conclusion that people who use onlysmokeless tobacco products have lower overall risks for disease and prema-ture mortality than cigarette smokers can be reached with a high degree ofconfidence. There is, however, considerable debate about the population im-pact of promoting smokeless tobacco use for harm reduction in cigarettesmokers, and about the polices and marketing approaches that lead to im-proved health rather than undermine health promotion efforts.

Resolution of these issues is complicated by the wide array of smokelesstobacco products, marketing approaches and social, cultural and regionalfactors that influence patterns of use. For example, substantial differences arefound in the patterns of use and the health consequences in South-East Asiaand Sweden (Foulds et al. 2003; Cnattingius, 2005) and in the USA (Depart-ment of Health and Human Services, 1986; National Institutes of Health,2006). The Scientific Committee on Emerging and Newly Identified HealthRisks (European Commission, 2008) concluded that “It is not possible toextrapolate the patterns of tobacco use from one country where oral tobaccois available to other countries due to societal and cultural differences.”

Thus, categories of products that meet specified standards and are marketedunder certain conditions in particular countries are more likely to benefitpublic health than products that meet other standards and conditions. Thereis no strong scientific basis for identifying product types or product perfor-mance standards that could be promoted as substitutes for smoking, for theforms of promotion that could be allowed or how cultural differences shouldbe addressed. Nevertheless, this information is necessary in order to makesignificant progress beyond policies in which all tobacco products are con-sidered to have the same or similar risks. In order to guide short- and long-term policy, research is needed:

13

• to characterize more thoroughly the diversity of ingredients used in dif-ferent product categories and brands and to assess their toxicity andpotential for causing disease;

• to understand better how addictiveness and product appeal are encouragedby physical features, such as tobacco cut size, nicotine content, free-base fraction, flavourings, other design and sensory characteristics andpackaging;

• to understand better which products are made by vendors and in homes,especially in South-East Asia, and their toxic and addictive potential, toprovide a basis for education, thus reducing toxicity and addictiveness;

• on how the toxicity, harmfulness and addiction potential of a product canchange over time under different storage conditions after manufacture;

• to understand better the relations between the amount and length of expo-sure and risk for disease in order to define performance standards forproduct constituents;

• on the reductions in risk when product use is discontinued completely orpartially;

• on the reduction in risk due to replacement of use of cigarettes and othercombustible forms of tobacco by use of smokeless tobacco products;

• on the characteristics of products, packaging and labelling, marketing andother forms of communication that contribute to the risk for addiction bypeople who are experimenting with the product and to the promotion ofcontinued use rather than cessation by regular users;

• to characterize the risk that use of and addiction to smokeless tobacco willlead to tobacco smoking;

• to characterize the risk that initiation of an addiction to smokeless tobaccowill lead to use of and escalation of use of addictive drugs, including al-cohol, by people who are already using them;

• to characterize the risk that initiation of smokeless tobacco use and addic-tion will delay or prevent smoking cessation rather than substitute forsmoking;

• to determine if and under what conditions smokeless tobacco use might aidsmoking cessation;

• on the potential or actual effects of various policies on smokeless tobaccouptake, overall tobacco use and risk for disease;

14

• to understand which regional populations, by gender and ethnic group, aremost likely to benefit from or be adversely affected by promotion ofsmokeless tobacco for reducing the harm due to smoking;

• to address how social, ethnic and marketing forces influence initiation ofthe use of smokeless tobacco and the trajectory from use to addiction andcessation; and

• to understand better the effects of smokeless tobacco use on vulnerablepopulations, such as adolescents, women of child-bearing age and im-munologically compromised persons, who might initiate use of smokelesstobacco products as a result of claims for reduced harm.

References

American Psychiatric Association (1994) Diagnostic and statistical manual ofmental disorders. 4th ed. Washington DC.

Cnattingius S (2005) Hälsorisker med svenskt snus. [Health risks of Swedishsnus.] Stockholm, National Institute of Public Health, Karolinska Institute.

Department of Health and Human Services (1986) Health consequences ofusing smokeless tobacco. A report of the Surgeon General. Bethesda,Maryland (NIH Publication No. 86-2874).

European Commission (2008) Health effects of smokeless tobacco products.Brussels, Scientific Committee on Emerging and Newly Identified Health Risks.

Fant RV et al. (1999) Pharmacokinetics and pharmacodynamics of moist snuffin humans. Tobacco Control, 8:387–392.

Food and Drug Administration (1995) Regulations restricting the sale anddistribution of cigarettes and smokeless tobacco products to protect childrenand adolescents; proposed rule analysis regarding FDA’s jurisdiction overnicotine-containing cigarettes and smokeless tobacco products; notice. FederalRegister, 60:41314–41792.

Food and Drug Administration (1996) Regulations restricting the sale anddistribution of cigarettes and smokeless tobacco to protect children andadolescents; final rule. 21 CFR Part 801. Federal Register, 61:44396-45318.

Foulds J et al. (2003) Effect of smokeless tobacco (snus) on smoking and publichealth in Sweden. Tobacco Control, 12:349–359.

Gupta PC, Sreevidya S (2004) Smokeless tobacco use, birth weight, andgestational age: population based prospective cohort study of 1217 women inMumbai (Bombay), India. British Medical Journal, 328:1538–1540.

Gupta PC, Subramoney S (2006) Smokeless tobacco use during pregnancyand risk of stillbirth: a cohort study in Mumbai, India. Epidemiology, 17:47–51.

IARC (2007) Smokeless tobacco and some tobacco-specific N-nitrosamines.Lyon, International Agency for Research on Cancer (IARC Monographs on theEvaluation of Carcinogenic Risks to Humans, Vol. 89).

National Institutes of Health (2006) State-of-the-science conference statementon tobacco use: prevention, cessation, and control. Bethesda, Maryland.

15

Royal College of Physicians and Surgeons (2007) Harm reduction in nicotineaddiction: helping people who can’t quit. A report by the Tobacco AdvisoryGroup of the Royal College of Physicians. London.

Savitz et al. (2006) Public health implications of smokeless tobacco use as aharm reduction strategy. American Journal of Public Health, 96:1934–1939.Erratum in: American Journal of Public Health, 2007, 97:202.

Scientific Advisory Committee on Tobacco Product Regulation (2003)Recommendation on smokeless tobacco products. Geneva, World HealthOrganization.

Scientific Advisory Committee on Tobacco Product Regulation (2004)Recommendation on tobacco product ingredients and emissions. Geneva,World Health Organization.

Stratton et al. (2001) Clearing the smoke: assessing the science base fortobacco harm reduction. Washington DC, National Academy Press.

TobReg (2004) Recommendation: guiding principles for the development oftobacco product research and testing capacity and proposed protocols for theinitiation of tobacco product testing. Geneva, World Health Organization.

TobReg (2005) WHO TobReg advisory: waterpipe tobacco smoking: healtheffects, research needs and recommended actions by regulators. Geneva,World Health Organization.

TobReg (2007) The scientific basis of tobacco product regulation. Geneva,World Health Organization (WHO Technical Report Series, No. 945).

WHO (1992) The ICD-10 classification of mental and behavioural disorders:clinical descriptions and diagnostic guidelines. Geneva, World HealthOrganization.

WHO (2005) WHO Framework Convention on Tobacco Control. Geneva, WorldHealth Organization.

Zeller M, Hatsukami D, Strategic Dialogue on Harm Reduction Group (2007)The strategic dialogue on tobacco harm reduction: a vision and blueprint foraction. (in press)

16

2. Advisory note on ‘fire-safer’cigarettes: approaches to reducedignition propensity

2.1 Purpose

This advisory note, formulated by TobReg, addresses growing concern aboutthe loss of life, injury and property damage due to fires ignited by combustedtobacco products, and particularly by cigarettes. Its purposes are to provideguidance to WHO and its Member States about the risks related to fires causedby cigarettes and the measures that can be taken to mitigate those risks. Thenote also gives guidance to researchers and research agencies interested infacilitating better understanding of fire-related deaths and injuries associatedwith cigarettes.

2.2 Background and history

Cigarettes and other lighted tobacco products are a leading cause of fire-related deaths and injuries in countries throughout the world. In 2003, 25600 cigarette-induced fires occurred in North America, resulting in an esti-mated 760 deaths, 1520 injuries and US$ 481 million in direct propertydamage (Hall, 2006). A survey in 14 Member States of the European Unionand Norway in 2005–2006 showed that in the countries that respondedthere were about 11 000 cigarette-caused fires, 520 deaths, 1600 injuries and€ 13 million material damages each year (J. Vogelgesang, unpublished data,2006). Extrapolation to the 25 countries of the European Union and Norwayindicates that 12 900 fires, 650 deaths, 2400 injuries and € 48 million inmaterial damages could be prevented. In New South Wales, Australia, 32 of233 fire deaths were directly attributed to cigarettes, with an additional 63possibly due to cigarettes. Annually, cigarettes cause 4574 fires acrossAustralia and may be responsible for up to 78 894 more. Australia’s NationalCoroners Information system attributed 67 of 678 fire deaths in the period2000–2005 directly to cigarettes. Further, an estimated 7% of all bush-fires in Australia are attributable to discarded cigarettes. Cigarette-relatedfires cost Australia AUS$ 80.6 million in 1998, which were projected toAUS$ 124 million in 2006 terms on the basis of the Consumer Price Index.In Canada, 3000 fires are started by smoking articles annually, which are

17

responsible for 70 fatalities, 300 injuries and CDN$ 40 million in propertydamage (D. Choinière, unpublished data, 2006).

A significant proportion of the deaths, injuries and property destruction couldbe prevented by the introduction of fire-safety standards for cigarettes, whichwould mean that they were either self-extinguishing, i.e. would go out whennot being puffed, or had altered smouldering characteristics, making a fireless likely. Cigarettes designed to comply with these standards are commonlyreferred to as ‘fire-safe’ or ‘reduced ignition propensity’ cigarettes.

In the USA, Congress enacted the Cigarette Safety Act of 1984 that requiredthe creation of a technical study group within the Consumer ProtectionAgency to determine the technical, economic and commercial feasibility ofdesigning a cigarette with minimum ignition propensity and to report itsfindings to Congress. In its final report, released in 1987, the group concludedthat the goal was technically feasible and might be economically feasible.Congress subsequently passed the Fire Safe Cigarette Act of 1990, whichcharged the National Institute of Standards and Technology to design a stan-dard method for determining the ignition propensity of cigarettes. It did not,however, give any government agency the authority to regulate the reductionof the propensity of cigarettes to cause fires.

The first performance standard is known as the ‘mock-up furniture ignitiontest method’, in which fabric and foam are used to simulate a piece of furnitureand in which a burning cigarette is tested to determine whether it transfersenough heat to ignite these materials. The second performance standard isknown as the ‘cigarette extinction method’, in which a set number of layersof filter paper are used to absorb heat, and a cigarette is tested to determinewhether it generates enough heat to continue to burn on the paper. Thecigarette extinction method is readily reproducible and takes less time per testthan the mock-up furniture ignition test method. The cigarette extinctionmethod was therefore refined and published by the American Society forTesting and Materials (2004) as the standard test method for measuring theignition strength of cigarettes (ASTM E2187).

The tobacco industry claimed for years that cigarettes with reduced ignitionpropensity could not be made and even bribed fire service organizations tothwart the passage of laws. The tobacco industry itself, however, establishedthat such cigarettes could be made and that their performance could be eval-uated. More than 80 years of research by the industry and over 300 patentshave addressed the design of ‘fire-safe’ cigarettes. The scientific basis is welladvanced, and the tobacco industry and cigarette paper manufacturers con-tinue their research and development.

18

Philip Morris began exploring the design of ‘fire-safe’ cigarettes in 1974.Both RJ Reynolds and Brown & Williamson have had extensive testing pro-grammes since the late 1970s or early 1980s. Lorillard began testing itscigarettes for ignition propensity at least as early as 1980. RJ Reynolds iden-tified means of changing the burning rate of cigarette paper, which affectsignition propensity and developed prototypes throughout the 1980s thatsuccessfully reduced ignition propensity, using cigarette papers produced bythe Ecusta Paper Company. The factors identified by RJ Reynolds in 1979are nearly identical to those identified by the technical study group a decadelater in their final report, released in 1987, which concluded that a ‘fire-safe’cigarette was technically feasible and might be economically feasible (Gunjaet al., 2002).

An internal Philip Morris document stated that: “Historical treatments ofignition-propensity results show that time to ignition measurements are re-lated to the maximum temperatures which smouldering cigarettes willachieve on a standard fabric. Further analysis indicates that these maximumtemperatures scale with the mass burn rates of the isolated cigarettes. Thisreduces the design problem to that of achieving target MBR’s [mass burnrates].” (Philip Morris, 1987).

The cigarette construction parameters identified by the technical study groupand by industry as affecting the burning rate are wrapping paper properties,such as permeability, porosity, oxygen diffusion, chemical additives (e.g.citrate or chalk), cigarette circumference and tobacco density (Ohlemilleret al., 1993).

After the release of the technical study group report, RJ Reynolds refocusedtheir research on ‘fire-safe’ cigarettes so as to target consumer acceptability.Other companies made similar progress in their ‘fire-safe’ cigarette projects.Brown & Williamson, Philip Morris and RJ Reynolds all obtained low-ignition paper from the Ecusta and Schweitzer paper companies from theearly 1980s. In the 1980s, Brown & Williamson designed two cigarette pro-totypes with Schweitzer papers and tested a Kimberly-Clark banded cigarettepaper. The method of banding most commonly used to reduce ignitionpropensity is that in which ultra-thin concentric bands are applied to tradi-tional cigarette paper (Figure 2.1). These bands, sometimes compared with‘speed bumps’, cause the cigarette to go out if it is not smoked, by restrictingoxygen to the burning ember (Connolly et al., 2005). Banded cigarette paperis manufactured either in a water-based online process, referred to as ‘paperbanding’, or by additional water- or solvent-based printing, referred to as‘print banding’ (Thelen, 2006). In the early 1990s, Philip Morris designed acigarette with bands that would extinguish the cigarette if it was not puffed.Banded cigarettes were tested as early as 1985, and in 2000 Philip Morris

19

released a banded cigarette with a ‘fire-safe paper select’ wrapper in the Meritbrand (Gunja et al., 2002). Internal industry testing demonstrated that thewidth and location of the bands can be used to control ignition propensity,wider bands and lower inter-band width being associated with the greatestreduction. Internal studies by Philip Morris showed that the technique usedto place the paper bands is very precise.

2.3 Regulatory responses

The first law to regulate ignition propensity was passed by the State of NewYork, USA, which mandated that all cigarettes sold in the State had to havereduced ignition propensity. The New York Fire Safety Standards forCigarettes (Part 429, Title 18 of the Official Compilation of Codes, Rules,and Regulations of the State of New York) came into force on 28 June 2004.Since then, Canada and 19 states of the USA have mandated reduced ignitionperformance standards for cigarettes. All existing fire safety standards forcigarettes are modelled after the New York standards and require that ciga-rettes be manufactured or sold to meet an ignition propensity performancestandard that makes them significantly less likely to cause fires if leftunattended. Recently, the Australian Government prescribed regulationsmandating a safety standard for cigarettes, covering performance, testing,packaging and marking requirements for cigarettes manufactured or im-ported into Australia. The Trade Practices (Consumer Product Safety Stan-dard) (Reduced Fire Risk Cigarettes) Regulations 2008 came into force on23 September 2008. South Africa’s Tobacco Products Control Act was

Figure 2.1Composition of a reduced ignition cigarette

20

amended on 23 February 2008 to include authority to set regulationsmandating a standard on the ignition propensity of cigarettes. Other countries,including New Zealand and Member States of the European Union, are con-sidering similar policies, and the European Commission is examining thefeasibility of proposing a standard.

Both the Canadian and the New York State laws incorporate the ASTM stan-dard test method in which a lit cigarette is placed on multiple layers ofstandard filter paper in a draft-free environment (Figure 2.2). The paper can-not ignite to smouldering and draws its heat from the cigarette. The persis-tence of burning is an indication of the energy available to ignite softfurnishings. Thus, the indicator is whether the cigarette burns to its full length.The standard adopted in these two jurisdictions requires that no more than25% of 40 test cigarettes placed on a thickness of 10 layers of filter paperundergo full-length burning. The techniques used by manufacturers to meetthe standard are generally unrestricted. In 2006, Standards Australia pub-lished a draft standard test method for determining the extinction propensityof cigarettes, which is also based on the ASTM test method.

The New York Fire Safety Standards for Cigarettes include provisions re-garding the reporting and investigation of cigarette-caused fires, the testingand certification of cigarettes, package labelling requirements, tax stamps andenforcement. The fire services must report all suspected cigarette-related fireswithin 14 days of completing an investigation and must provide informationon the brand and style, marking as compliant, and the location and mannerof purchase of the cigarette. Cigarette manufacturers are responsible underthe New York Standards for testing each of their brands and for providingwritten certification to the Office of Fire Prevention and Control and to theAttorney General. The Office of Fire Prevention and Control is required totest cigarettes suspected of igniting fires and to retest any cigarettes to whicha manufacturer makes a change that is likely to alter its compliance. Taxstamps may not be affixed to cigarette packages in New York State unlessthe cigarettes have been certified as meeting the Standards (J. Mueller, un-published data, 2006).

The New York Office of Fire Prevention and Control is authorized to examinebooks, papers, invoices and other records and to impose civil penalties andsuspensions. Enforcement includes penalties for false certification and forsale of non-compliant cigarettes. Public health officers are authorized toimpose penalties on retail dealers. Officers of the Office of Fire Preventionand Control and of the Taxation and Finance Office are authorized to seizecigarettes not marked as compliant, and the seized cigarettes are to be de-stroyed (J. Mueller, unpublished data, 2006). Tobacco companies are re-quired to pay for testing and stamping in all states of the USA. The fees for

21

certification range from US$ 100 to US$ 1000 per cigarette brand or brandfamily and could increase.

Canada adopted the same standard as New York State when it introducedregulations requiring all cigarettes manufactured in or imported into Canadaas of 1 October 2005 to satisfy the reduced ignition propensity standard.

Figure 2.2Standard method for testing the ignition propensity performance standard ofcigarettes, in which a lit cigarette is placed on multiple layers of standard filter paper(A) in a draught-free environment (B)

A.

B.

22

Canada’s law applies at the manufacturing and importation levels, whereasthe laws of states in the USA apply to the sale of cigarettes by retailers.

Approximately 1200 cigarette brands have been certified as compliant in NewYork State (New York Office of Fire Prevention and Control, 2008). HealthCanada (2008a) has been sampling cigarettes from manufacturers and im-porters to determine whether they comply with the standard outlined in itsregulations. As a result of ‘fire-safe’ cigarette laws, cigarette manufacturersin Canada have modified nearly all their brands (D. Choinière, unpublisheddata, 2006). The results of laboratory analyses of samples collected by HealthCanada are posted on the Internet and updated periodically.

The ignition propensity of five brands of cigarettes sold in New York Stateafter implementation of the Fire Safety Standards was tested with the cigaretteextinction method; the full-length burning per brand was found to be 2.5–30.0% (Connolly et al., 2005). In contrast, the full-length burning of the samebrands of cigarettes sold in Massachusetts and California, in New York beforethe law was passed, and in Australia and Thailand was 100% (Figure 2.3).

2.3.1 Effectiveness of regulatory measures in populations

Current regulatory measures are not expected to eliminate cigarette fire-related deaths but are intended to reduce such incidents over time, in the

Figure 2.3Cigarette ignition propensity of major brands in California, Massachusetts and NewYork, USA, and in Australia and Thailand

0

25

50

75

100

Perc

enta

ge fu

ll-le

ngth

bur

ning

Marlboro Reds Marlboro Lights

New York

New York

New York

New York

Massa

chuse

tts

Thailan

d

Thailan

d

Australi

a

Massa

chuse

tts

Californ

ia

Californ

ia

From Tobacco Control Research Program, Harvard School of Public Health

23

expectation that future changes in cigarette design will continue to do so. Atthe same time, governments must improve data collection on the frequencyof cigarette-related fires and design regulations to address the problem. Pre-liminary data suggest that governments will benefit by implementing regu-lations to reduce ignition propensity.

Compliance with standards based on the ASTM method should reduce thenumbers of cigarettes that cause smouldering combustion and consequentlyignite fires. The effectiveness of the standard should be monitored continu-ously by recording the incidence of cigarette-caused fires and the associateddeaths, injuries and costs. Fire reporting is fraught with problems of quality,and the methods should perhaps be reconsidered. Reliable information onreductions in cigarette consumption, improved mattress standards and chang-ing fire prevention standards is also difficult to obtain.

Preliminary data on the population effect of cigarettes with reduced ignitionpropensity that are on the market suggests that the numbers of smoking-related fires and fire deaths declined in New York during the first 2 yearsafter implementation of the Fire Safety Standards (Figure 2.4).

In Canada, the regulatory impact assessment of reduced ignition propensitycigarettes predicted that a fire safety standard for cigarettes would reducecigarette-caused fires by 34–68% (Health Canada, 2008b). As more countriesregulate ignition propensity, it will become easier to evaluate the efficacy ofsuch cigarettes.

2.3.2 Regulatory considerations

Emissions and biological assays

One concern is that changes in the design of cigarettes might lead to changesin exposure (by topography, burn temperature and emissions) that might

Figure 2.4Fires (left-hand panel) and deaths (right-hand panel) in fires due to smoking materialsin New York State before and after application of fire-safety standards for cigarettes

Fires2750

2500

2250 2223 2279

26182456

20352000

2001 2002 2003 2004 2005

1750

1500

1250

1000

Deaths

49

3840

45 45

39 38

3133

55

50

45

40

20012000199919981997 2002 2003 2004 2005

35

30

25

20

From Mueller (2006)

24

make these products more harmful than they already are. The preliminarydata available do not indicate that this is a serious problem.

In the USA, under the Fire-safe Cigarette Act of 1990, the National Instituteof Standards and Technology compared the yields of tar, nicotine and carbonmonoxide of cigarettes with reduced ignition propensity from the TobaccoInstitute Testing Laboratory with the yields of the 14 best-selling commercialcigarette brands. No significant differences were found (Ohlemiller et al.,1993). Preliminary data from Canada indicated small changes in the deliveryof carbonyls, tar, carbon monoxide and nicotine (M.J. Kaiserman, unpub-lished data, 2006).

Internal industry testing of banded cigarettes has also shown them to be sub-stantially the same as regular cigarettes with regard to a number of toxico-logical end-points, including mutagenicity and the concentrations of toxicchemicals in emissions (Theophilus et al., 2007a,b). Philip Morris assessedsome toxicological aspects of banded cigarettes and found “no significantdifferences between the two cigarettes based on the chemical and biologicalassays used”. Similar findings have been published and presented at scientificmeetings by other companies (Patskan et al., 2000; Appleton, Krauuter,Lauterbach, 2003; Misra et al., 2005), including RJ Reynolds, which has longopposed regulations on reduced ignition propensity, claiming increased risk(Theophilus et al., 2007a,b). The tobacco industry claims that there is anunintended additional risk of ‘coal’ (a lightweight, short-lived bit) droppingoff from banded cigarettes and has stated that 11% of consumer complaintsabout banded cigarettes in the USA in 2000 were related to coal drop-off. Apublication from British American Tobacco in 1988 concluded that paperpermeability had no influence on coal retention in the range tested (Dittrich,1988).

When 42 smokers in Ontario, Canada, were asked to compare smoking theirown brands before and after implementation of the cigarette ignition propen-sity law in 2005, no significant differences in puffing behaviour or exhaledcarbon monoxide were found (Hammond et al., 2007).

Sense of security

Cigarette manufacturers have asserted that smoking ‘fire-safe’ cigarettescould give a false sense of security, which might increase fire-risk behaviour.According to a survey conducted before the coming into force of the cigaretteignition propensity regulation in Canada, 12% of current smokers had smokeda cigarette in bed in the past week, and 17% reported that they left lit ciga-rettes burning unattended on a daily basis (M.J. Kaiserman, unpublisheddata, 2006). In another survey in Ontario, Canada, nearly one in four smokersleft burning cigarettes unattended, and 15% had smoked in bed in the past

25

30 days, indicating a high frequency of fire-risk behaviour (O’Connor et al.,2007); early data from follow-up studies showed little change in such be-haviour after 1 year (O’Connor, 2008).

Economic effects

Research by the Harvard School of Public Health, USA, showed no declinein cigarette sales in New York State after implementation of the Fire SafetyStandards for Cigarettes, confirming the conclusions of the technical studygroup in 1987 (Connolly et al., 2005). The results of a nationwide survey inthe USA also showed that the New York Fire Safety Standards appeared tohave had no discernable effect on how smokers perceived the taste of theircigarettes, smoking behaviour or intention to quit, countering argumentsmade by cigarette manufacturers that the law would have a negative effecton consumer acceptability (O’Connor et al., 2006).

Health Canada (2008c) estimated that if the cost of complying with measuresfor cigarette ignition propensity was absorbed entirely by cigarette manufac-turers, the companies’ operating profits would be reduced by 2.9–5.9%; theycould raise their prices to offset the increased costs. While some price increaseis likely, the extent to which individual manufacturers would raise their pricesis uncertain and would depend on competitive forces in the tobacco productsmarket. Given the degree of competition in that market, it is unlikely thatprices would rise by the full amount of the estimated cost increase (i.e. US$0.13–0.26 per carton).

Implementation and compliance

Approximately 1200 cigarette brands have been certified as compliant in NewYork State (New York Office of Fire Prevention and Control, 2008). HealthCanada sampled products from manufacturers and importers to determinewhether cigarettes in Canada comply with the standard outlined in its regu-lations and found that ‘fire-safe’ cigarette laws have resulted in modificationof nearly all brands (D. Choinière, unpublished data, 2006). The results oflaboratory analyses of samples collected by Health Canada are posted on theInternet and updated periodically.

New York State has taken the lead in validating industry reports by indepen-dent testing of cigarettes every 3 years and comparing industry reports withtheirs. The cost of testing is US$ 400–700 per brand but should fall as morecountries become involved. Other states in the USA rely on New York Stateand have not conducted testing. Efforts are being made in the USA to coor-dinate state reporting and testing. The National Institute of Standards andTechnology gives technical support to laboratories, including a referencecigarette (http://firesafecigarettes.org/assets/files/niststandard.pdf) and small

26

grants. There are currently six laboratories for testing. Brands that have beentested in New York State are listed online at http://www.dos.state.ny.us/fire/cigarette.htm.

The International Organization for Standardization (ISO) may adopt a stan-dard that is identical to ASTM E2187, except in format. The alternative ofdrafting a guidance document that refers to the test method in the ASTMstandard could take 1–2 years.

2.4 Research needs

2.4.1 Techniques

Research is needed to ensure the effectiveness of any regulations on reducedignition propensity and to provide a basis for future policy. The main ap-proach used to modify burning rate and consequently to reduce ignitionpropensity is to decrease oxygen diffusion by lowering the permeability ofcigarette paper. The techniques used by manufacturers to achieve this shouldbe monitored, including the effects of measured differences among brands inband placement and other design features. Some researchers are using reverseengineering of products to examine their banding characteristics, such as thepresence, number, width and spacing; filter ventilation and pressure drop;paper porosity and citrate content; tobacco weight and density and cigarettecircumference.

Cigarette companies and paper manufacturers are conducting research in in-dustry-based research and development and programmes to follow up theachievements of the National Institute of Standards and Technology and theNew York State standards. Further research and reviews of the scientific lit-erature, industry documents and other sources are important for monitoringindustry findings on cigarette ignition propensity and performance.

Some of the patented designs for reducing ignition propensity are paper withvery low porosity and added perforations, addition of fire retardant to thecentre of the tobacco rod, cellulose bands on paper, application of chemicalsoutside the paper and addition of intumescent powder to the tobacco column.The last reduces the ignition propensity of tobacco by decreasing its densityduring heating (Stevenson, Graham, 1988).

2.4.2 Testing methods

Methods are needed for testing ignition propensity performance, such asthermal imaging. Their potential use in effective, efficient testing might beincluded in regulations.

27

2.4.3 Surveillance and monitoring

Fires and subsequent losses due to cigarettes should be surveyed and moni-tored in order to judge the success of policies and to determine whether thestandards should be adopted. The New York State measures appear to bereducing the number of deaths due to cigarette-related fires, but high-qualityfire incidence reporting and data are needed. The data must be accurate andtimely and based on large enough numbers so that statistical significance canbe assessed. The outcomes that should be monitored are the incidence of firesand the associated losses, injuries and deaths (D. Hemenway, unpublisheddata, 2006). The capacity of investigators at the scene of a fire must be im-proved to allow them to ascertain whether the fire was started by a cigaretteand what other factors contributed to the severity of the fire.

The impact of measures to reduce ignition propensity should be followed upover time, on the basis of criteria such as population health and the optimumpercentage reduction in fires.

The Fire Safety Standards for Cigarettes in New York State contain a provi-sion that allows the Office of Fire Prevention and Control to review infor-mation on the incidence of fires in the light of technological changes after aperiod of 3–4 years and to consider revising the Standards. Other jurisdictionsmight wish to adopt a similar approach.

2.4.4 Exposure to emissions and smoking behaviour

Further assessment of exposure to emissions and changes in smoking be-haviour should include consideration of product design, emissions of tar,nicotine and carbon monoxide, puffing behaviour, filter analyses and bio-markers of exposure.

Population surveys with baseline and follow-up measures, such as those beingconducted by the Roswell Park Cancer Institute and the Harvard School ofPublic Health in the USA, should include questions such as “Has a cigaretteever started a fire in your home?”, “How often does your cigarette go out onits own?” and “How often does the lit end or ash fall off your cigarette on itsown?”. Such analyses should also assess the prevalence of fire-risk events inthe 30 days before the survey. Information on fire-risk behaviour should in-clude instances of burnt clothes, burnt furniture, burning cigarettes leftunattended, dozing off and falling asleep while smoking and smoking in bed.

The tobacco industry has claimed that some methods of reducing the ignitionpropensity of cigarettes could increase the toxicity by increasing smoke de-livery. There is no evidence that cigarettes with reduced ignition propensityincrease the risk for disease from smoking. Cigarette smoke is a highly com-plex mixture, containing over 4000 chemicals, and the links between these

28

chemicals and the toxicity of the smoke are not well defined. It is probablethat the smoke from cigarettes with reduced ignition propensity is just as toxicas that of unmodified cigarettes.

2.5 Findings and recommendations

Fires and fire deaths are caused by cigarettes.

Fires due to cigarettes and the related deaths are a major global public healthproblem. Although the number of deaths is far lower than that caused bysmoking (900 deaths in the USA from fires and 460 000 from smoking), it isstill high, and policies are needed to reduce the number.

Cigarettes with reduced ignition propensity should be mandatory.

As cigarettes are the principal cause of residential fires and related deaths andtechniques exist to reduce ignition propensity and thus the likelihood of acigarette igniting a fire, Member States should require reduced ignitionpropensity cigarettes, in line with the standard of the National Institute ofStandards and Technology or any other that has been shown to be effective.Countries and jurisdictions within countries should retain the right to alterthe standard on the basis of population-based data on its effectiveness.

While Canada has adopted measures for reducing ignition propensity withinpublic health laws, Australia and most states of the USA implement suchmeasures within laws on fire safety. In the European Union, such measuresare being considered in the framework of consumer protection legislation.

The products covered by these measures should include not only cigarettesbut also cigars and any other combusted tobacco product if evidence indicatesthat their ignition strength should be regulated. Considerations could includestate or national monetary appropriations, identification of parties or agenciesresponsible for certification, the delay required for re-certification, identifi-cation of the agency responsible for auditing brands, the scope and frequencyof audits, evaluation of the population impact, fees and fines, advisory com-mittees and whether the regulations can be superseded by federal law.Countries should require tobacco manufacturers to test ignition strength, re-port the results to the responsible authority and pay a fee for implementationof the measures.

Independent laboratory testing capacity is currently minimal. It could be in-creased if countries adopted measures that require testing of ignition propen-sity by independent laboratories accredited according to ISO standard 17025,General requirements for the competence of calibration and testing labora-tories. Industry-generated results should be validated by independent tests.

29

Legislation and regulatory measures should give the responsible authority themeans to take appropriate legal action to ensure compliance with the standard.

No risk claims are permissible.

As reduced ignition propensity cigarettes must be made available to an entirepopulation, manufacturers canot be allowed to claim that they reduce the riskof fire. If they did, consumers might conclude that they reduced overall healthrisks. Public education campaigns are needed as part of any reduced ignitionpropensity programme, to inform consumers that all cigarettes are lethal andthat smokers should quit. Such programmes should also include educationcampaigns to teach the public how to prevent fires.

The effectiveness of reduced ignition propensity cigarettes must bemonitored.

Adequate, appropriate monitoring, reporting and archiving are needed torecord the effectiveness of techniques for reducing ignition propensity fordecreasing deaths, injuries and property damage due to cigarette-inducedfires. Such assessments will increase public assurance and lead to more ef-fective means of diminishing the needless losses that occur as a result ofcigarette-ignited fires.

International collaboration is necessary.

International collaboration between interested institutions and authorities isneeded to coordinate education, advocacy, testing, research and evaluationof reduced ignition propensity cigarettes and for implementation of suchmeasures in all WHO regions.

References

American Society for Testing and Materials International (2004) ASTME2187-04. Standard test method for measuring the ignition strength ofcigarettes. West Conshohocken, Pennsylvania, ASTM International.

Appleton S, Krauuter GR, Lauterbach JH (2003) Toxicological evaluation of anew cigarette paper designed for lowered ignition propensity. In: 57th TobaccoScience Research Conference, Program Booklet and Abstracts, No. 16, p. 27.Raleigh, North Carolina, North Carolina State University, Tobacco LiteratureService.

Connolly GN et al. (2005) Effect of the New York State cigarette fire safetystandard on ignition propensity, smoke constituents, and the consumer market.Tobacco Control, 14:321–327.

Dittrich DJ (1988) Influence of cigarette design on coal retention. Southampton,BAT (UK and Export) Ltd. Research and Development Centre (Report No. RD.2125).

30

Gunja M et al. (2002) The case for fire cigarettes made through industrydocuments. Tobacco Control, 11:346–353.

Hall J (2006) The smoking-material fire problem. Quincy, Massachusetts,National Fire Protection Association, Fire Analysis and Research Division.

Hammond D et al. (2007) The impact of Canada’s lowered ignition propensityregulations on smoking behaviour and consumer perceptions. In: Society forResearch on Nicotine and Tobacco 13th Annual Meeting Proceedings.Available at: http://www.srnt.org/meeting/2007/pdf/onsite/2007srntabstracts-final.pdf. Accessed 23 October 2008.

Health Canada (2008a) Laboratory analysis of cigarette for ignition propensity.http://www.hc-sc.gc.ca/hl-vs/tobac-tabac/legislation/reg/ignition-alllumage/index_e.html. Accessed 23 October 2008.

Health Canada (2008b) Industrial Economics, Inc. (2004) Economic evaluationof Health Canada’s regulatory proposal for reducing fire risks from cigarettes.Prepared for Health Canada. Available at: http://www.hc-sc.gc.ca/hl-vs/pubs/tobac-tabac/evaluation-risks-risques/benefits-avantages5-eng.php#3.Accessed 23 October 2008.

Health Canada (2008c) Industrial Economics, Inc. (2004) Economic evaluationof Health Canada’s regulatory proposal for reducing fire risks Fromcigarettes. Prepared for Health Canada. Available at: http://www.hc-sc.gc.ca/hl-vs/pubs/tobac-tabac/evaluation-risks-risques/ecoimpacts-impactseco7-eng.php#2. Accessed 23 October 2008.

Mueller J (2006) Cigarette fire safety standards. In: Second internationalconference on fire safer cigarettes. Boston, Massachsetts, Harvard School ofPublic Health.

Misra M et al. (2005) Toxicological evaluation of a cigarette paper with reducedignition propensity: in vitro and in vivo tests. Toxicologist, 84 (Suppl. 1):1186.

New York Office of Fire Prevention and Control (2008) List of cigarettes certifiedby manufacturers. Available at: http://www.dos.state.ny.us/fire/pdfs/cigaretteweblist.pdf. Accessed 23 October 2008.

O’Connor RJ (2008) New research on fire safe cigarettes. In: Third internationalconference on fire ‘safer’ cigarettes. Norwood, Massachusetts, National FireProtection Association and Harvard School of Public Health.

O’Connor RJ et al. (2006) Smokers’ reactions to reduced ignition propensitycigarettes. Tobacco Control, 15:45–49.

O’Connor RJ et al. (2007) Prevalence of behaviors related to cigarette-causedfires: a survey of Ontario smokers. Injury Prevention, 13:237–242.

Ohlemiller TJ et al. (1993) Test methods for quantifying the propensity ofcigarettes to ignite soft furnishings. Gaithersburg, Pennsylvania, TechnologyAdministration, National Institute of Standards and Technology, Department ofCommerce (NIST Special Publication 851).

Patskan G et al. (2000) Toxicological characterization of a novel cigarettepaper. Toxicologist, 54:398.

Philip Morris (1987) Project Tomorrow—status and plans, 25 November 1987.Richmond, Virginia (Bates No. 1002816087-6088).

Stevenson WW, Graham J (1988) Reduced ignition propensity smokingarticle. United States Patent No. 4 776 355, 11 October 1988.

31

Thelen VK (2006) Fire-safe cigarettes: an overview. Top paper: performanceby understanding. Traun, Delfort Group. Available at: http://www.delfortgroup.com/toppaper/archive/toppaper_0206.pdf.

Theophilus EH et al. (2007a) Toxicological evaluation of cigarettes with twobanded cigarette paper technologies. Experimental Toxicology andPathology, 59:17–27.

Theophilus EH et al. (2007b) Comparative 13-week cigarette smoke inhalationstudy in Sprague-Dawley rats: evaluation of cigarettes with two bandedcigarette paper technologies. Food Chemistry and Toxicology, 45:1076–1090.

32

THE CIGARETTE FIRE SAFETY STANDARDAND FIREFIGHTER PROTECTION ACT

1. Title. This Act shall be known and may be cited as the ‘Fire SafetyStandard and Firefighter Protection Act’.

2. Findings.

The Legislature finds and declares that:

a. Cigarettes are the leading cause of fire deaths in this State and thenation;

b. Each year in the United States, 700–900 persons are killed due tocigarette fires and 3000 are injured in fires ignited by cigarettes, whilein this State [ ] residential fires and [ ] fatalities were attributable tocigarettes in years [_ _ _ _-2005];

c. A high proportion of the victims of cigarette fires are non-smokers,including senior citizens and young children;

d. Cigarette-caused fires result in billions of dollars of property lossesand damages in the United States and millions of dollars in this State;

e. Cigarette fires unnecessarily jeopardize firefighters and result inavoidable emergency response costs for municipalities;

f. In 2004, New York State implemented a cigarette fire safety regula-tion requiring cigarettes sold in that State to meet a fire safetyperformance standard; in 2005, Vermont and California enactedcigarette fire safety laws directly incorporating New York’s regula-tion into statute; and, in 2006, Illinois, New Hampshire and Mas-sachusetts joined these states in enacting such laws.

g. In 2005, Canada implemented the New York State fire safety standardcontained in the other state laws, becoming the first nation to have acigarette fire safety standard;

h. New York State’s cigarette fire safety standard is based upon decadesof research by the National Institute of Standards and Technology,Congressional research groups, and private industry;

i. This cigarette fire safety standard minimizes costs to the State andminimally burdens cigarette manufacturers, distributors and retailsellers, and, therefore, should become law in this State; and

33

Annex 2.1 Model legislation

j. It is therefore fitting and proper for this State to adopt the cigarettefire safety standard that is in effect in New York State to reduce thelikelihood that cigarettes will cause fires and result in deaths, injuriesand property damages.

3. Definitions. For the purposes of this Act:

(a) ‘Agent’ shall mean any person authorized by the [State entity thatadministers cigarette tax stamps] to purchase and affix stamps onpackages of cigarettes.

(b) ‘Cigarette’ shall mean:

(1) any roll for smoking, whether made wholly or in part of tobaccoor any other substance, irrespective of size or shape, and whetheror not such tobacco or substance is flavored, adulterated or mixedwith any other ingredient, the wrapper or cover of which is madeof paper or any other substance or material, other than leaf to-bacco; or

(2) any roll for smoking wrapped in any substance containing tobaccowhich, because of its appearance, the type of tobacco used in thefiller, or its packaging and labeling, is likely to be offered to, orpurchased by, consumers as a cigarette as described in subpara-graph 1 above.

(c) ‘Director’ shall mean the Director of the [State entity responsible foradministering the provisions of this Act].

(d) ‘Manufacturer’ shall mean:

(1) any entity which manufactures or otherwise produces cigarettesor causes cigarettes to be manufactured or produced anywherethat such manufacturer intends to be sold in this State, includingcigarettes intended to be sold in the United States through an im-porter; or

(2) the first purchaser anywhere that intends to resell in the UnitedStates cigarettes manufactured anywhere that the original manu-facturer or maker does not intend to be sold in the United States;or

(3) any entity that becomes a successor of an entity described inparagraph (1) or (2) of this subsection.

(e) ‘Quality control and quality assurance program’ shall mean the lab-oratory procedures implemented to ensure that operator bias, system-atic and nonsystematic methodological errors, and equipment-related

34

problems do not affect the results of the testing. Such a program en-sures that the testing repeatability remains within the required repe-atability values stated in paragraph (6) of subsection (a) of Section 4of this Act for all test trials used to certify cigarettes in accordancewith this Act.

(f) ‘Repeatability’ shall mean the range of values within which therepeat results of cigarette test trials from a single laboratory will fall95 percent of the time.

(g) ‘Retail dealer’ shall mean any person, other than a manufacturer orwholesale dealer, engaged in selling cigarettes or tobacco products.

(h) ‘Sale’ shall mean any transfer of title or possession or both, exchangeor barter, conditional or otherwise, in any manner or by any meanswhatever or any agreement therefor. In addition to cash and creditsales, the giving of cigarettes as samples, prizes or gifts, and the ex-changing of cigarettes for any consideration other than money, areconsidered sales.

(i) ‘Sell’ shall mean to sell, or to offer or agree to do the same.

(j) ‘Wholesale dealer’ shall mean any person other than a manufacturerwho sells cigarettes or tobacco products to retail dealers or other per-sons for purposes of resale, and any person who owns, operates ormaintains one or more cigarette or tobacco product vending machinesin, at or upon premises owned or occupied by any other person.

4. Test Method and Performance Standard.

(a) Except as provided in subsection (g) of this section, no cigarettes maybe sold or offered for sale in this State or offered for sale or sold topersons located in this State unless the cigarettes have been tested inaccordance with the test method and meet the performance standardspecified in this section, a written certification has been filed by themanufacturer with the [State entity responsible for administering theprovisions of this Act] in accordance with section 5 of this Act, andthe cigarettes have been marked in accordance with section 6 ofthis Act.

(1) Testing of cigarettes shall be conducted in accordance with theAmerican Society of Testing and Materials (‘ASTM’) standardE2187-04, “Standard Test Method for Measuring the IgnitionStrength of Cigarettes.”

(2) Testing shall be conducted on 10 layers of filter paper.

35

(3) No more than 25 percent of the cigarettes tested in a test trial inaccordance with this section shall exhibit full-length burns. Fortyreplicate tests shall comprise a complete test trial for eachcigarette tested.

(4) The performance standard required by this section shall only beapplied to a complete test trial.

(5) Written certifications shall be based upon testing conducted by alaboratory that has been accredited pursuant to standard ISO/IEC17025 of the International Organization for Standardization(‘ISO’), or other comparable accreditation standard required bythe [State entity responsible for administering the provisions ofthis Act].

(6) Laboratories conducting testing in accordance with this sectionshall implement a quality control and quality assurance programthat includes a procedure that will determine the repeatability ofthe testing results. The repeatability value shall be no greaterthan 0.19.

(7) This section does not require additional testing if cigarettes aretested consistent with this Act for any other purpose.

(8) Testing performed or sponsored by the [State entity responsiblefor administering the provisions of this Act] to determine acigarette’s compliance with the performance standard requiredshall be conducted in accordance with this section.

(b) Each cigarette listed in a certification submitted pursuant to section 5of this Act that uses lowered permeability bands in the cigarette paperto achieve compliance with the performance standard set forth in thissection shall have at least two nominally identical bands on the papersurrounding the tobacco column. At least one complete band shall belocated at least 15 millimeters from the lighting end of the cigarette.For cigarettes on which the bands are positioned by design, there shallbe at least two bands fully located at least 15 millimeters from thelighting end and 10 millimeters from the filter end of the tobaccocolumn, or 10 millimeters from the labeled end of the tobacco columnfor non-filtered cigarettes.

(c) A manufacturer of a cigarette that the [State entity responsible foradministering the provisions of this Act] determines cannot be testedin accordance with the test method prescribed in paragraph (1) ofsubsection (a) of this section shall propose a test method and perfor-mance standard for the cigarette to the [State entity responsible for

36

administering the provisions of this Act]. Upon approval of the pro-posed test method and a determination by the [State entity responsiblefor administering the provisions of this Act] that the performancestandard proposed by the manufacturer is equivalent to the perfor-mance standard prescribed in subsection (a) (3) of this section, themanufacturer may employ such test method and performance standardto certify such cigarette pursuant to section 5 of this Act. If the [Stateentity responsible for administering the provisions of this Act] deter-mines that another state has enacted reduced cigarette ignition propen-sity standards that include a test method and performance standardthat are the same as those contained in this Act, and the [State entityresponsible for administering the provisions of this Act] finds that theofficials responsible for implementing those requirements have ap-proved the proposed alternative test method and performance standardfor a particular cigarette proposed by a manufacturer as meeting thefire safety standards of that state’s law or regulation under a legalprovision comparable to this section, then the [State entity responsiblefor administering the provisions of this Act] shall authorize that man-ufacturer to employ the alternative test method and performancestandard to certify that cigarette for sale in this State, unless the [Stateentity responsible for administering the provisions of this Act] demon-strates a reasonable basis why the alternative test should not beaccepted under this Act. All other applicable requirements of this sec-tion shall apply to the manufacturer.

(d) Each manufacturer shall maintain copies of the reports of all testsconducted on all cigarettes offered for sale for a period of three years,and shall make copies of these reports available to the [State entityresponsible for administering the provisions of this Act] and the At-torney General upon written request. Any manufacturer who fails tomake copies of these reports available within sixty days of receivinga written request shall be subject to a civil penalty not to exceed$10,000 for each day after the sixtieth day that the manufacturer doesnot make such copies available.

(e) The [State entity responsible for administering the provisions of thisAct] may adopt a subsequent ASTM Standard Test Method for mea-suring the Ignition Strength of Cigarettes upon a finding that suchsubsequent method does not result in a change in the percentage offull-length burns exhibited by any tested cigarette when compared tothe percentage of full-length burns the same cigarette would exhibitwhen tested in accordance with ASTM Standard E2187-04 and theperformance standard in subsection (a)(3) of this section.

37

(f) The [State entity responsible for administering the provisions of thisAct] shall review the effectiveness of this section and report everythree years to the Legislature [the State entity’s] findings and, if ap-propriate, recommendations for legislation to improve the effective-ness of this Act. The report and legislative recommendations shall besubmitted no later than June thirtieth following the conclusion of eachthree-year period.

(g) The requirements of subsection (a) of this section shall not prohibit:

(1) wholesale or retail dealers from selling their existing inventory ofcigarettes on or after the effective date of this Act if the wholesaleor retailer dealer can establish that State tax stamps were affixedto the cigarettes prior to the effective date and the wholesale orretailer dealer can establish that the inventory was purchased priorto the effective date in comparable quantity to the inventory pur-chased during the same period of the prior year; or

(2) the sale of cigarettes solely for the purpose of consumer testing.For purposes of this subsection, the term ‘consumer testing’ shallmean an assessment of cigarettes that is conducted by a manu-facturer (or under the control and direction of a manufacturer),for the purpose of evaluating consumer acceptance of suchcigarettes, utilizing only the quantity of cigarettes that is reason-ably necessary for such assessment, and in a controlled settingwhere the cigarettes are either consumed on-site or returned to thetesting administrators at the conclusion of the testing.

(h) This Act shall be implemented in accordance with the implementationand substance of the New York Fire Safety Standards for Cigarettes.

5. Certification and Product Change

(a) Each manufacturer shall submit [to the State entity responsible foradministering the provisions of this Act] a written certification attest-ing that:

(1) each cigarette listed in the certification has been tested in accor-dance with section 4 of this Act; and

(2) each cigarette listed in the certification meets the performancestandard set forth in section 4.

(b) Each cigarette listed in the certification shall be described with thefollowing information:

(1) brand, or trade name on the package;

38

(2) style, such as light or ultra light;

(3) length in millimeters;

(4) circumference in millimeters;

(5) flavor, such as menthol or chocolate, if applicable;

(6) filter or non-filter;

(7) package description, such as soft pack or box;

(8) marking pursuant to section 6 of this Act;

(9) the name, address and telephone number of the laboratory, if dif-ferent than the manufacturer that conducted the test; and

(10) the date that the testing occurred.

(c) The certifications shall be made available to the Attorney General forpurposes consistent with this Act and the [State entity responsible foradministering the State cigarette tax act] for the purposes of ensuringcompliance with this section.

(d) Each cigarette certified under this section shall be re-certified everythree years.

(e) For each cigarette listed in a certification, a manufacturer shall pay tothe [State entity responsible for administering the provisions of thisAct] a $250 fee. The [State entity responsible for administering theprovisions of this Act] is authorized to annually adjust this fee to en-sure it defrays the actual costs of the processing, testing, enforcementand oversight activities required by this Act.

(f) There is established in the [State treasury] a separate, nonlapsing fundto be known as the ‘Fire Safety Standard and Firefighter ProtectionAct Enforcement Fund’. The fund shall consist of all certificationfees submitted by manufacturers, and shall, in addition to any othermonies made available for such purpose, be available to the [Stateentity responsible for administering the provisions of this Act] solelyto support processing, testing, enforcement and oversight activitiesunder this Act.

(g) If a manufacturer has certified a cigarette pursuant to this section,and thereafter makes any change to such cigarette that is likely to alterits compliance with the reduced cigarette ignition propensity stan-dards required by this Act, that cigarette shall not be sold or offeredfor sale in this State until the manufacturer retests the cigarette inaccordance with the testing standards set forth in section 4 of this Act

39

and maintains records of that retesting as required by section 4 of thisAct. Any altered cigarette which does not meet the performance stan-dard set forth in Section 4 of this Act may not be sold in this State.

6. Marking of Cigarette Packaging

(a) Cigarettes that are certified by a manufacturer in accordance withsection 5 of this Act shall be marked to indicate compliance with therequirements of section 4 of this Act. The marking shall be in eightpoint type or larger and consist of:

(1) Modification of the product UPC Code to include a visible markprinted at or around the area of the UPC Code. The mark mayconsist of alphanumeric or symbolic characters permanentlystamped, engraved, embossed or printed in conjunction with theUPC; or

(2) Any visible combination of alphanumeric or symbolic characterspermanently stamped, engraved or embossed upon the cigarettepackage or cellophane wrap; or

(3) Printed, stamped, engraved or embossed text that indicates thatthe cigarettes meet the standards of this Act.

(b) A manufacturer shall use only one marking, and shall apply thismarking uniformly for all packages, including but not limited to packs,cartons, and cases, and brands marketed by that manufacturer.

(c) The [State entity responsible for administering the provisions of thisAct] shall be notified as to the marking that is selected.

(d) to the certification of any cigarette, a manufacturer shall present itsproposed marking to the [State entity responsible for administeringthe provisions of this Act] for approval. Upon receipt of the request,the [State entity responsible for administering the provisions of thisAct] shall approve or disapprove the marking offered, except that the[State entity responsible for administering the provisions of this Act]shall approve:

(1) any marking in use and approved for sale in New York pursuantto the New York Fire Safety Standards for Cigarettes, or

(2) the letters ‘FSC’, which signifies Fire Standards Compliant ap-pearing in 8 point type or larger and be permanently printed,stamped, engraved or embossed on the package at or near theUPC code.

40

Proposed markings shall be deemed approved if the [State entity re-sponsible for administering the provisions of this Act] fails to actwithin 10 business days of receiving a request for approval.

(e) No manufacturer shall modify its approved marking unless the mod-ification has been approved by the [State entity responsible for ad-ministering the provisions of this Act] in accordance with this section.

(f) Manufacturers certifying cigarettes in accordance with section 5 ofthis Act shall provide a copy of the certifications to all wholesaledealers and agents to which they sell cigarettes, and shall also providesufficient copies of an illustration of the package marking utilized bythe manufacturer pursuant to this section for each retail dealer towhich the wholesale dealers or agents sell cigarettes. Wholesale deal-ers and agents shall provide a copy of these package markings re-ceived from manufacturers to all retail dealers to which they sellcigarettes. Wholesale dealers, agents and retail dealers shall permitthe [State entity responsible for administering the provisions of thisAct], the [State entity responsible for administering the provisions ofthe State cigarette tax act], the Attorney General, and their employeesto inspect markings of cigarette packaging marked in accordance withthis section.

7. Penalties.

(a) A manufacturer, wholesale dealer, agent or any other person or entitywho knowingly sells or offers to sell cigarettes, other than throughretail sale, in violation of section 4 of this Act, shall be subject to acivil penalty not to exceed one hundred ($100) dollars for each packof such cigarettes sold or offered for sale provided that in no case shallthe penalty against any such person or entity exceed one hundredthousand ($100,000) dollars during any thirty-day period.

(b) A retail dealer who knowingly sells or offers to sell cigarettes in vi-olation of section 4 of this Act shall be subject to a civil penalty notto exceed one hundred ($100) dollars for each pack of such cigarettessold or offered for sale, provided that in no case shall the penaltyagainst any retail dealer exceed twenty-five thousand ($25,000) dol-lars for sales or offers to sell during any thirty-day period.

(c) In addition to any penalty prescribed by law, any corporation, part-nership, sole proprietor, limited partnership or association engaged inthe manufacture of cigarettes that knowingly makes a false certifica-tion pursuant to section 5 of this Act shall be subject to a civil penaltyof at least seventy-five thousand ($75,000) dollars and not to exceed

41

two-hundred fifty thousand ($250,000) dollars for each such falsecertification.

(d) Any person violating any other provision in this Act shall be subjectto a civil penalty for a first offense not to exceed one thousand ($1,000)dollars, and for a subsequent offense subject to a civil penalty not toexceed five thousand ($5,000) dollars for each such violation.

(e) Any cigarettes that have been sold or offered for sale that do not com-ply with the performance standard required by section 4 of this Actshall be subject to forfeiture [under the pertinent provision of Statelaw having to do with forfeiture of contraband]. Cigarettes forfeitedpursuant to this section shall be destroyed; provided, however, thatprior to the destruction of any cigarette forfeited pursuant to theseprovisions, the true holder of the trademark rights in the cigarettebrand shall be permitted to inspect the cigarette.

(f) In addition to any other remedy provided by law, the [State entityresponsible for administering the provisions of this Act] or AttorneyGeneral may file an action in [name of court] for a violation of thisAct, including petitioning for injunctive relief or to recover any costsor damages suffered by the State because of a violation of this Act,including enforcement costs relating to the specific violation and at-torney’s fees. Each violation of this Act or of rules or regulationsadopted under this Act constitutes a separate civil violation for whichthe [State entity responsible for administering the provisions of thisAct] or Attorney General may obtain relief.

(g) Whenever any law enforcement personnel or duly authorized repre-sentative of the [State entity responsible for administering the provi-sions of this Act] shall discover any cigarettes that have not beenmarked in the manner required by section 6 of this Act, such personnelis hereby authorized and empowered to seize and take possession ofsuch cigarettes. Such cigarettes shall be turned over to the [departmentof taxation and finance], and shall be forfeited to the State. Cigarettesseized pursuant to this section shall be destroyed; provided, however,that prior to the destruction of any cigarette seized pursuant to theseprovisions, the true holder of the trademark rights in the cigarettebrand shall be permitted to inspect the cigarette.

8. Implementation.

(a) The [State entity responsible for administering the provisions of thisAct] may promulgate rules and regulations, pursuant to the [State ad-ministrative procedures act], necessary to effectuate the purposes ofthis Act.

42

(b) The [State entity responsible for administration of the State cigarettetax act] in the regular course of conducting inspections of wholesaledealers, agents and retail dealers, as authorized under the [Statecigarette tax act] may inspect such cigarettes to determine ifthe cigarettes are marked as required by section 6 of this Act. If thecigarettes are not marked as required, the [State entity responsible foradministration of the State cigarette tax act] shall notify the [Stateentity responsible for administering the provisions of this Act].

9. Inspection. To enforce the provisions of this Act, the Attorney General,the [State department of taxation and finance] and the [State entity re-sponsible for administering the provisions of this Act], their duly autho-rized representatives and other law enforcement personnel are herebyauthorized to examine the books, papers, invoices and other records ofany person in possession, control or occupancy of any premises wherecigarettes are placed, stored, sold or offered for sale, as well as the stockof cigarettes on the premises. Every person in the possession, control oroccupancy of any premises where cigarettes are placed, sold or offeredfor sale, is hereby directed and required to give the Attorney General,the [State department of taxation and finance] and the [State entity re-sponsible for administering the provisions of this Act], their duly autho-rized representatives and other law enforcement personnel the means,facilities and opportunity for the examinations authorized by this section.

10. Cigarette Fire Safety Standard and Firefighter Protection ActFund. There is hereby established in the State Treasury a special fundto be known as the ‘Cigarette Fire Safety Standard and Firefighter Pro-tection Act Fund’. The fund shall consist of all monies recovered aspenalties under section 7 of this Act. The monies shall be deposited tothe credit of the fund and shall, in addition to any other monies madeavailable for such purpose, be made available to the State entity respon-sible for administering the provisions of this Act to support fire safetyand prevention programs.

11. Sale Outside of [State name]. Nothing in this Act shall be construed toprohibit any person or entity from manufacturing or selling cigarettesthat do not meet the requirements of section 4 of this Act if the cigarettesare or will be stamped for sale in another state or are packaged for saleoutside the United States and that person or entity has taken reasonablesteps to ensure that such cigarettes will not be sold or offered for sale topersons located in this State.

12. Preemption. This Act shall be repealed if a federal reduced cigaretteignition propensity standard that preempts this Act is adopted and be-comes effective.

43

13. Effective Date. This Act shall take effect on the first day of the thirteenthmonth after enactment.

Hyperlink to New York State Regulations:http://www.dos.state.ny.us/fire/amendedcigaretterule.htm

Hyperlink to Canada Cigarette Ignition Propensity Regulations:http://canadagazette.gc.ca/partII/2005/20050629/html/sor178-e.html

44

3. Mandated lowering of toxicants incigarette smoke: tobacco-specificnitrosamines and selected otherconstituents

3.1 Executive summary and recommendations

The WHO Framework Convention on Tobacco Control recognizes the needfor tobacco product regulation in Articles 9 and 10. Existing product regula-tory strategies based on the machine-measured tar, nicotine and carbonmonoxide (CO) yields per cigarette with the current ISO regimen are causingharm. By allowing communication of the yields as measures of exposure orrisk, they mislead smokers into believing that low-yield cigarettes carry lessrisk and are a reasonable alternative to cessation. This harm precludes con-tinued acceptance of strategies of product regulation based on per-cigarettemachine-measured tar and nicotine and necessitated the development of anew approach. A WHO Study Group on Tobacco Product Regulation/Inter-national Agency for Research on Cancer (TobReg/IARC) working group wasformed to consider an alternative approach.

This report recommends a strategy for regulation based on product perfor-mance measures with the goal of reducing toxicant levels in mainstreamcigarette smoke. It recommends establishing levels for selected mainstreamsmoke toxicants per milligram of nicotine and prohibiting the sale or importof cigarette brands that have yields above these levels. The purpose of nor-malizing toxicant levels per milligram of nicotine is to shift the interpretationof the measurement away from the quantity of smoke generated per cigaretteand away from the misleading use of ISO tar, nicotine and CO values asmeasures of human exposure and risk. It moves towards product characteri-zation of the toxicity of smoke generated under standardized conditions.

Prohibiting consumer communications based on any machine measurementswill also be necessary. This will minimize the likelihood that consumers willbe misled into thinking that the machine- measured yields represent exposureof the smoker, as has been occurring with existing tar, nicotine and CO values.

TobReg recognizes that the ideal product regulatory strategy would includemeasures of human exposure and injury; however, examination of the exist-ing evidence on use of biomarkers of exposure or harm for product regulationsuggests that validated biomarkers of harm do not currently exist. Validated

45

biomarkers of exposure exist, but differences in these biomarkers with theproduct used cannot be separated from differences due to the characteristicsof the smokers who use the product, making use of biomarkers of exposurefor product regulation problematic. Given these constraints, TobReg con-cluded that currently available regulatory strategies are limited to productperformance standards based on product ingredients and emissions in ma-chine-generated smoke rather than strategies based on measures of exposureor harm.

The list of toxicants was selected on the basis of an assessment that includedconsideration of: data on animal and human toxicity, toxicity indices, varia-tion in toxicants across brands, the potential for the toxicant to be loweredand the inclusion of constituents from both the particulate and the gas phasesof smoke and from different chemical classes in cigarette smoke. Consider-ation was also given to selecting compounds implicated in cardiovascular andpulmonary toxicity as well as carcinogenicity. The most important criterionfor selecting compounds for regulation was evidence of toxicity.

Available data on variations in toxicant levels in cigarette brands providedan initial set of observations to identify the levels of reduction that have al-ready been achieved for some products on the market. A means for using suchdata to identify appropriate levels for the market being regulated is presented.The initial levels are intended to be a first step in an overall strategy to reducelevels of toxicants in tobacco smoke further as our understanding of what ispossible expands and new technology develops. The levels recommended inthis report represent TobReg’s judgement from the available data on the mostpractical trade-off, considering the need to regulate a range of toxicants, tomandate substantial lowering of those toxicants and yet not to recommendelimination of most of the brands in the data sets used to set the levels. Reg-ulators are encouraged to obtain data from their own markets and of coursemay select different levels that are more appropriate for their circumstances.They are encouraged to set levels on the basis of the yields of the brands soldin their markets, using the principles set out in this report.

The recommended regulatory strategy should be implemented in phases, be-ginning with a period of required annual reporting of toxicant levels bycigarette manufacturers to the regulatory authority. This should be followedby the promulgation of levels for toxicants above which brands cannot beoffered for sale. Finally, the established levels would be enforced.

It is expected that regulators will take additional action to reduce the man-dated levels of toxicants further as this regulatory strategy is fully imple-mented. These actions can take the form of setting targets or milestones basedon what may be achievable with new technology or product designs.

46

The toxicants recommended for mandatory lowering and the initial regulatorylevels recommended from existing data are listed in Table 3.1. A modifiedmachine-testing regimen with more intense puffing parameters, in which allthe holes in cigarette filters are blocked, was used to generate these values,and TobReg recommends use of that regimen by regulators in implementingthe proposed regulatory strategy.

Table 3.1Toxicants recommended for mandated lowering

Toxicant Level in μg/mg nicotine Criterion for selecting value

Internationalbrandsa

Canadianbrandsb

NNK 0.072 0.047 Median value of data setNNN 0.114 0.027 Median value of data setAcetaldehyde 860 670 125% of median value of data setAcrolein 83 97 125% of median value of data setBenzene 48 50 125% of median value of data setBenzo[a]pyrene 0.011 0.011 125% of median value of data set1,3-Butadiene 67 53 125% of median value of data setCarbon monoxide 18 400 15 400 125% of median value of data setFormaldehyde 47 97 125% of median value of data set

NNK, 4-(N-nitrosomethylamino)-1-(3-pyridyl)-1-butanone; NNN, N´-nitrosonornicotinea Based on data from Counts et al. (2005)b Based on data reported to Health Canada minus brands with > 0.1 NNN per miligram of nicotine,

which eliminates most United States and Gauloise brands (http://www.hc-sc.gc.ca/hl-vs/tobac-tabac/legislation/reg/indust/constitu_e.html)

The levels given in Table 3.1 in the column ‘International brands’ are basedon an international sample (Counts at al., 2005) of brands of US-stylecigarettes with a blend of tobaccos. The figures in the column labelled ‘Cana-dian brands’ are based on a set of brands reported to Health Canada that reflectcigarettes containing predominantly flue-cured (bright) tobacco. These twostyles of cigarettes are similar to those marketed in many other countries.Regulators should use data reported for their own markets (if available) to setlevels, or they may select values from the data set that conforms most closelyto the cigarettes available on the markets being regulated. The differences inthe reported levels in the two styles of cigarettes are particularly large for thetobacco-specific nitrosamines (TSNA) N´-nitrosonornicotine (NNN) and 4-(N-nitrosomethylamino)-1-(3-pyridyl)-1-butanone (NNK).

As presented in the first report of TobReg on this regulatory approach (WHO,2007), the median values for NNN and NNK are used as the recommended

47

levels because there is strong evidence that the amounts of these toxicants intobacco smoke can be lowered dramatically with existing technology. Aninitial level of 125% of the median value is recommended for the other tox-icants, reflecting the somewhat greater uncertainty about the extent to whichthese toxicants can be reduced with existing approaches. Substantially lowerlevels should be the ultimate goal of this regulatory strategy.

The proposed regulatory strategy focuses on the product and the smoke gen-erated under standardized conditions. Mandated lowering of levels of toxi-cants per milligram of nicotine in cigarette smoke will make regulation ofcigarettes consistent with other regulatory approaches to mandate reductionof known toxicants in products used by humans. Reducing the level of toxi-cants in products intended for human use is a widely accepted regulatorypractice. The anticipated outcome is a marketplace that excludes those brandswith the highest levels of toxicants. This strategy is neither based on nor relieson measures of actual or estimated human exposure or risk, and so cannot beused to quantify reductions in human exposure, risk or disease. It does relyon measures of clearly established toxicants as they appear in tobacco smokegenerated under standardized conditions.

Regulatory authorities have an obligation to ensure that the public is not mis-led by the results of the recommended machine testing and mandated regu-latory values. This is because the public was misled by communication of thevalues of machine testing for tar, nicotine and CO yields with the current ISOregimen. Science has not established that reduction of any individual toxicantin machine- measured cigarette smoke, including those proposed in this re-port, will reduce actual human exposure or disease risk. Mandating lowerlevels and removing some brands with higher levels from the market do notconstitute a statement that the remaining brands are safe or less hazardousthan the brands removed, nor does it represent government approval of thesafety of the products that remain on the market.

TobReg recommends that any regulatory approach based on machine-generated yields specifically prohibit use of the results of testing, or relativeranking of brands by testing levels, as indicators of risk or exposure in mar-keting, promotion or other communications with the public, including prod-uct labelling. Statements that a brand has met government regulatorystandards should also be prohibited.

3.2 Introduction

Preventing initiation of tobacco product use, promoting cessation of tobaccouse and protecting the public from exposure to second-hand smoke are rec-ognized by the WHO Framework Convention on Tobacco Control as the mosteffective approaches to reducing tobacco-related morbidity and mortality.

48

The Convention also recognizes, however, the need for tobacco product reg-ulation, as stated in Articles 9 and 10 of the Treaty. TobReg has prepared aseries of reports to provide the scientific foundation for tobacco product reg-ulation (WHO, 2007). This report is the second of two defining how manda-tory lowering of toxicant levels in machine-measured cigarette smoke mightbe used as a product regulatory approach. The report is the result of a col-laborative effort by TobReg and IARC, which convened a working group toreport to TobReg.

Regulation of tobacco products requires some metric, or set of metrics, bywhich tobacco products can be assessed. The most common measurementsthat have been used for cigarettes have been machine-measured tar, nicotineand CO yields per cigarette based on the testing regimen of ISO and the USFederal Trade Commission (FTC). There is scientific consensus that per-cigarette yields do not provide valid estimates of human exposure or ofrelative human exposure during smoking of different brands of cigarettes(Stratton et al., 2001; National Cancer Institute, 2001; Scientific AdvisoryGroup on Tobacco Product Regulation, 2002; Department of Health and Hu-man Services, 2004a). Communication of these measures to smokers createsharm by leading them to believe that their exposures and risk will be differentif they switch to cigarette brands with different machine-measured yields.Any new product regulatory strategy should take into account the harm cur-rently being done by reporting tar, nicotine and CO values per cigarette andshould ideally help remedy that harm. This ongoing harm precludes accept-ance of current regulatory strategies based on per-cigarette machine-measuredtar, nicotine and CO levels and necessitates new regulatory approaches.

Machine smoking regimens other than the ISO/FTC regimen have also beenexamined, particularly ones with more intense puffing parameters, in whichsome or all of the ventilation holes in cigarette filters are blocked. Examplesinclude those developed by the State of Massachusetts in the USA and theCanadian Government. These regimens generally produce higher yields percigarette and reduce the differences between brands. Nevertheless, the regi-mens continue to rank brands by tar and nicotine yield per cigarette. Therankings by yield per cigarette with these more intense regimens do not pro-vide valid estimates of human exposure or of the relative exposure of smokerswhen they smoke different brands of cigarettes. Thus, even yields with thesemore intense smoking regimens have the potential to mislead smokers whenexpressed per cigarette.

A single machine-testing regimen produces a single set of toxicant yields. Incontrast to machines, individual smokers vary the pattern with which theypuff different cigarettes of the same brand, and cigarette design changes canlead smokers to systematically change how they puff cigarettes of different

49

designs. These variations in smokers’ puffing behaviour limit the use ofmachine-generated smoke yields as estimates of exposure. Individual smok-ers seek to achieve nicotine intakes that are sufficient to satisfy their addiction,but the level of nicotine intake varies substantially across the general popu-lation of smokers (Benowitz et al., 1983; Jarvis et al., 2001). There is alsosubstantial variation in the pattern of puffing (puff size, puff duration, inter-puff interval and depth of inhalation) used to achieve the desired level ofnicotine, and smokers may vary their pattern of puffing from one cigarette toanother and from one puff to another (Djordjevic, Stellman, Zang, 2000;Melikian et al., 2007a,b). These sources of variation mean that tar, nicotineand CO yields from a single standardized machine measurement regimen areunreliable estimates of the exposure of smokers.

Cigarette design changes, particularly the extent of filter ventilation, are as-sociated with a substantial difference in the average pattern of puffing ofpopulations of smokers who use them, and smokers who switch brands tendto alter their pattern of puffing in order to preserve their level of nicotineintake (National Cancer Institute, 2001). Tar, nicotine, CO and other toxicantyields vary considerably with differences in machine smoking parameters,and variations in other toxicant levels exist both when the yields are expressedper cigarette and when standardized per milligram of tar or nicotine. Sys-tematic differences in how products with different designs are used bysmokers and differences in yields when the cigarettes are puffed differentlymake comparisons between cigarette brands of per-cigarette machine-smoked yields virtually meaningless as estimates of human exposure. As aresult, machine testing with the regimens currently in widespread use cannotprovide estimates of human exposure and should not be used to support claimsof reduced exposure or risk.

The limitations of machine measurements led to efforts to quantify the actualexposure of smokers by measuring biomarkers in blood, urine and saliva. Asthese biomarkers are measured in individual smokers, however, they are in-fluenced by personal characteristics and the characteristics of the person’ssmoking behaviour, as well as by the characteristics of the product smoked(National Cancer Institute, 2001; Scientific Advisory Group on TobaccoProduct Regulation, 2002). Distinguishing the differences in biomarker levelsdue to variations between products from the differences due to smokers’ be-haviour (e.g. who uses the product and how they use it) is a formidablescientific challenge. Research is needed to resolve these issues so that expo-sure biomarkers can become effective tools for product regulation. Themultiplicity of brands on the market, self-selection of smokers who use dif-ferent products and differences in how smokers of different products use themmake the use of biomarkers of exposure in a regulatory strategy to monitor

50

cigarette product differences problematic, given the current level of scientificknowledge.

Characterization of differences in harm caused by different cigarettes wouldbe a powerful metric for product regulation. Markers of biologically effectivedose (levels of toxicants in critical target organs or tissues) are likely to bedeveloped and validated in the future, and these are expected to offer moreprecise measures of smoke uptake and better prediction of smoke toxicity(Stratton et al., 2001). Measures of injury or validated biomarkers of diseaserisk are also likely to be validated in the near future. They will allow morerapid assessment of differences in disease risks than is currently possible withepidemiological approaches to measuring disease outcomes. These advancesmay allow assessment of differences in risk between tobacco products. At themoment, however, none of these measures has been validated as a reliable,independent predictor of differences in tobacco-related disease risk amongsmokers using different products (Hatsukami et al., 2004).

The limitations of measures of human exposure and human injury suggestthat for the near future product regulatory approaches may be limited to mea-sures of differences between products in design characteristics and emissionsrather than measures derived from their human use. Chemical measurementsof the smoke produced by machines and their use as inputs for product hazardassessment, with all of their limitations, may represent the limit of currentscientific assessment of differences between brands that can be used for reg-ulatory assessment of product toxicity.

Measurements based on machine-generated smoke can be made simply andconsistently, and they provide information about the effect of design charac-teristics on the toxic compounds in cigarette smoke when cigarettes aresmoked under standard conditions. What machine-generated smoke mea-surements do not provide is an understanding of how smokers respond todesign changes, measures of actual or relative consumer exposure, or an as-sessment of the consequences of smokers’ responses for exposure or risk.Nevertheless, measurements of machine-generated smoke can be a usefulinterim regulatory tool before metrics of exposure and harm become availablethat can be used for product regulation. Because of the misleading history ofper-cigarette yields, TobReg recommends that toxicant yields be normalizedper milligram of nicotine, to shift the interpretation of measurements awayfrom the quantity of smoke generated per cigarette and the misleading com-munication that the machine-measured yields represent exposure of thesmoker, as has been occurring with existing ISO tar, nicotine and CO values.Expressing toxicant values per milligram of nicotine allows examination ofcigarettes as they perform under standardized conditions to generate smokeand characterizes that smoke by its toxicity normalized to the quantity of

51

nicotine generated. This change in characterization should shift the interpre-tation towards product toxicity characterization and away from exposure, buta ban on communicating these values as measures of exposure to risk willalso be needed as part of this regulatory approach.

3.3 Background

TobReg, at its meeting in Montebello, Canada, on 26–28 October 2004(WHO, 2004), concluded that machine-measured tar, nicotine and CO yieldsbased on the ISO/FTC method were misleading smokers and most regulators,and thereby causing harm; they also recommended that these measures shouldno longer be communicated to smokers as measures of exposure or risk. Atthe same time, TobReg recognized that abandoning measurements of tar,nicotine and CO with the ISO method as regulatory tools, before developmentof validated biomarkers of risk, would leave a regulatory and informationvoid that would not be in the best interests of WHO Member States.ISO Technical Committee 126 (TC126) recognized this misuse of machine-measured yields, and by a formal vote passed a resolution adopting as arationale for all machine-smoking testing standards the following statements:

No machine smoking regime can represent all human smoking behaviour.

Methods are recommended that test the product under conditions of dif-ferent intensities of machine smoking in order to collect mainstreamsmoke.

Machine smoking testing is useful to characterize cigarette emissions fordesign and regulatory purposes, but communication of machine measure-ments to smokers can result in misunderstanding about differences inexposure and risk across brands.

Smoke emission data from machine measurements may be used as inputsfor product hazard assessment, but they are not intended to be nor are theyvalid measures of human exposure or risks. Communicating differencesbetween products in machine measurements as differences in exposure orrisk is a misuse of testing using ISO standards (ISO, 2006).

As an interim step in the regulation of tobacco products, before the develop-ment of approaches to assess differences in actual exposure, harm or risk fromdifferent cigarette brands, TobReg recommended assessment of differentproducts by an examination of the toxicity of the products when tested understandardized conditions. This approach allows application of product regu-latory strategies similar to those used for other consumer products, where thefocus of the regulatory standard is to reduce the levels of known toxicantspresent in the product (in this case in the smoke) rather than requiring expo-sure measures derived from human populations. This approach is limited in

52

that it offers little human exposure information, but it does allow applicationof well-established regulatory strategies.

Despite these limitations, a change in regulatory strategy is needed, ascigarettes are enormously harmful, and the status quo of using tar and nicotineyields on a per-cigarette basis is causing harm by deceiving consumers intobelieving that cigarettes with lower machine-measured yields are less risky.Establishment of the proposed interim regime is, therefore, urgent, as mea-surements based on actual exposure and harm to the smoker require consid-erable additional research.

The TobReg recommendation is to quantify machine-measured levels of spe-cific toxicants per milligram of nicotine in the smoke generated by the testingmethod described below. Standardizing the levels of toxicants per milligramof nicotine shifts the focus of the measurement away from a metric of humanexposure and towards characterization of the product’s performance understandardized conditions. This shift may reduce the misleading effect of dif-ferences between levels of toxicants when they are expressed per cigarette.As nicotine is the principal addictive substance in smoke sought by smokers,it was decided to standardize per milligram of nicotine rather than per mil-ligram of tar. These standardized measures of toxicant yield will facilitateexamination of the relation between cigarette design and the composition ofcigarette smoke. They also provide regulators with a mechanism for reducingthe level of identified toxicants in tobacco smoke and allow regulatory actionwhile the science of assessing harm from tobacco smoke toxicants developsfurther.

In order to prepare scientific guidance on how best to implement the strategydefined by TobReg, the WHO Tobacco Free Initiative and IARC establisheda working group to define mandated reductions for tobacco smoke toxicants.Limits for TSNA and the methods by which they could be applied in a reg-ulatory strategy were the focus of the first report (WHO, 2007), prepared ata meeting in Lyon, France, 9–10 April 2006. The current report addressessmoke toxicants, including nitrosamines, and was prepared at meetings of theworking group in Geneva, Switzerland, 12–13 October 2006 and in SanDiego, California, USA, on 22–23 March 2007. A summary of the first reporton TSNA is included in this report for completeness.

3.4 First report on mandated lowering of tobacco-specificnitrosamines

The first report (WHO, 2007) recommended that mandated lowering of ma-chine yields per milligram of nicotine for NNN and NNK be adopted as aWHO recommendation. Levels were proposed for adoption as a WHO stan-dard and for use by individual countries in mandating lower TSNA levels.

53

TSNA are potent carcinogens, and there is clear evidence that alteration ofthe curing methods for tobacco and other changes in manufacturing ap-proaches can substantially lower the levels in smoke (Peele, Riddick, Ed-wards, 2001; IARC, 2004). There is marked variation across brands withincountries in the levels of these nitrosamines, as demonstrated by data fromAustralia and Canada (Australian Department of Health and Aging, 2001;Health Canada, 2004). A sample of international brands from one manufac-turer also showed substantial variation in TSNA levels (Counts et al., 2005).The potency of TSNA as carcinogens and the evidence that levels could besubstantially reduced with existing technology, thereby limiting the numberof cigarette brands that would be unable to achieve the regulated level bymodifying the tobacco used, identified these compounds as ideal toxicantsfor early application of mandated lowering as a regulatory strategy.

The carcinogenicity of NNK and NNN has been firmly established by ex-tensive studies in laboratory animals. On the basis of animal carcinogenicitydata for these toxicants, human exposure data and mechanistic studies, NNNand NNK together are classified as a human carcinogen by IARC (Group 1)(Cogliano et al., 2004; IARC, 2007).

The levels of NNK and NNN in unburnt tobacco contribute significantly toand are correlated with the levels in smoke. NNK and NNN form during thecuring and processing of tobacco. Modifications of these processes, particu-larly removal of propane heating as part of the curing process, are nowavailable that can substantially reduce the levels in tobacco and therefore insmoke (Peele, Riddick, Edwards, 2001; IARC, 2004). Independent of theeffects of the heat source used in curing, it is also known that NNK and NNNlevels vary significantly with tobacco type, being higher in air-cured, pro-cessed burley tobacco than in flue-cured bright tobacco. Other studies showthat tobacco nitrate levels contribute to the smoke levels of NNK and NNN.Collectively, the available evidence strongly indicates that the technology isavailable to reduce NNK and NNN levels in cigarette smoke significantly.

The variation in NNN and NNK per milligram of nicotine is defined in thefirst report on the basis of data for an international sample of Philip Morrisbrands published by Counts and colleagues (2004). While that study is basedon what appear to be the best international data currently available, it obvi-ously reflects a selected sample of brands from a single manufacturer con-taining tobacco blends that are much higher in nitrosamines than brands thatare currently on the market in some countries, notably Australia, Canada andthe United Kingdom. Therefore, the guidance on levels per milligram ofnicotine presented here is intended to inform countries in which the data arenot available or are inadequate. Countries where the principal brands on themarket contain flue-cured bright tobacco with low levels of nitrosamines may

54

be well advised to establish their own regulatory levels on the basis of mea-surements made on the cigarettes actually sold in their market.

Examination of the levels of NNN and NNK (Counts et al., 2005) revealedthat the median concentration of NNN in the Philip Morris brands tested wasapproximately 114 ng/mg of nicotine in smoke emissions, with a range of16–189 ng/mg. The median level of NNK in the brands tested was 72 ng/mgof nicotine, with a range of 23–111 ng/mg. These data suggest that a level ofNNN in smoke emissions of 114 ng (0.114 g) per milligram of nicotine orlower has already been achieved for half the cigarettes Philip Morris hasmarketed internationally. This level therefore reflects a readily achievablelevel to establish for this compound as an initial step in mandating lowerlevels and above which cigarettes should be excluded from the market.

A mandated limit for NNK of 72 ng (0.072 g) per milligram of nicotine insmoke emissions is similarly proposed.

These values were set on the basis of the median value reported for theseinternational brands, which are US blended cigarettes with higher nitrosaminelevels than blends used for cigarettes in Australia, Canada and the UnitedKingdom. Comparison of toxicant yields in those countries demonstrated thatthe tobacco used in the cigarettes smoked yielded levels of NNN and NNKthat are substantially lower than the median values measured by Counts andcolleagues in their sample of brands. Data on smoke toxicant yields forCanadian cigarettes (Health Canada, 2004), once US and Gauloise brandswere excluded, revealed a median NNN concentration of 26.9 ng/mg of nico-tine and a median NNK concentration of 46.6 ng/mg. The Canadian medianNNN level is less than one fourth that of the international brands used toestablish the mandated reduction recommendation, demonstrating both thatcigarettes with substantially lower levels of NNN can be manufactured andsuccessfully marketed and that individual countries may be well advised toset mandated reduction levels based on the products sold in their own markets.

Data from Australia are similar. The median smoke toxicant yield fromAustralian cigarettes (Australian Department of Health and Aging, 2001)measured with the Canadian intense method, was 19.5 ng/mg of nicotine forNNN and 25.6 ng/mg for NNK in the brands tested. This demonstrates thatthere is substantial scope for further reductions in the levels of nitrosaminesand that it is possible to manufacture cigarettes with low levels of ni-trosamines which have broad market appeal.

NNN and NNK levels are often closely correlated in individual brands, rais-ing the question of whether lower levels need to be mandated for bothnitrosamines. This correlation is not seen when US blended-style cigarettesthat contain burley tobacco in which the level of NNN is higher than that of

55

NNK are compared with Canadian brands containing predominantly flue-cured tobacco in which the level of NNN is lower than that of NNK.

Figure 3.1 presents data on the levels of NNN and NNK per milligram ofnicotine in individual Canadian brands plotted by increasing level of NNK.Many of the brands with high levels of NNN and NNK are US or Frenchbrand names, and the TSNA differences are presumably due to the differencein the tobacco blend used and the nitrosamine levels of the tobaccos in thoseblends. It is also evident, however, that the relation between NNN and NNKlevels is different in the US brands and most notably in the Gauloise brands.This suggests that setting regulatory levels for NNN and NNK may identifydifferent brands as being above the regulatory level, particularly when theblends of tobaccos used vary.

Figure 3.1NNN and NNK concentrations by brand of Canadian cigarettes

002040608001021041061081

MARK

TEN

PLA

IN K

ING

SIZE

MATI

NEE

SILV

ER R

EGUL

AR S

IZE

PLAY

ER'S

SPE

CIAL

BLE

ND R

EGUL

AR S

IZE

BELM

ONT

MILD

S RE

GULA

R SI

ZEPL

AYER

'S S

PECI

AL B

LEND

KIN

G SI

ZEPL

AYER

'S P

LAIN

REG

ULAR

SIZ

ESP

ORTS

MAN

PLAI

N RE

GULA

R SI

ZEVA

NTAG

E MA

X 15

KIN

G SI

ZEBE

LMON

T MI

LDS

KING

SIZ

EDU

MAU

RIER

EXT

RA LI

GHT

REGU

LAR

SIZE

CRAV

EN A

EXT

RA LI

GHT

KING

SIZ

ECR

AVEN

MEN

THOL

KIN

G SI

ZEDU

MAU

RIER

LIGH

T KI

NG S

IZE

NUMB

ER 7

KING

SIZ

EDU

MAU

RIER

KIN

G SI

ZEDU

MAU

RIER

LIGH

T RE

GULA

R SI

ZEMA

TINE

E EX

TRA

MILD

KIN

G SI

ZEDU

MAU

RIER

EXT

RA LI

GHT

KING

SIZ

ECR

AVEN

A K

ING

SIZE

DU M

AURI

ER R

EGUL

AR S

IZE

BENS

ON &

HED

GES

DELU

XE M

ENTH

OL U

LTRA

LIGH

TS 10

0'sRO

THMA

NS S

PECI

AL M

ILD K

ING

SIZE

PLAY

ER'S

EXT

RA LI

GHT

REGU

LAR

SIZE

VANT

AGE

RICH

12 K

ING

SIZE

PLAY

ER'S

REG

ULAR

SIZ

EPL

AYER

'S E

XTRA

LIGH

T KI

NG S

IZE

PLAY

ER'S

LIGH

T RE

GULA

R SI

ZEEX

PORT

'A' E

XTRA

LIGH

T KI

NG S

IZE

PLAY

ER'S

LIGH

T KI

NG S

IZE

EXPO

RT 'A

' MILD

KIN

G SI

ZEEX

PORT

'A' M

ILD R

EGUL

AR S

IZE

EXPO

RT 'A

' ULT

RA LI

GHT

KING

SIZ

EEX

PORT

'A' M

EDIU

M KI

NG S

IZE

VISC

OUNT

EXT

RA M

ILD K

ING

SIZE

EXPO

RT P

LAIN

REG

ULAR

SIZ

EEX

PORT

'A' F

ULL F

LAVO

UR K

ING

SIZE

CRAV

EN M

ILDS

ULTR

A MI

LD K

ING

SIZE

GAUL

OISE

S BL

ONDE

S KI

NG S

IZE

EXPO

RT 'A

' LIG

HT /

SPEC

IAL L

IGHT

KIN

G SI

ZEVA

NTAG

E SL

IMS

EXTR

A LIG

HT 4

100's

GAUL

OISE

S BL

ONDE

S LIG

HTS

KING

SIZ

EEX

PORT

'A' M

EDIU

M RE

GULA

R SI

ZEEX

PORT

'A' E

XTRA

LIGH

T RE

GULA

R SI

ZEEX

PORT

'A' U

LTRA

LIGH

T RE

GULA

R SI

ZEVA

NTAG

E UL

TRA

LIGHT

1 KI

NG S

IZE

EXPO

RT 'A

' LIG

HT / S

PECI

AL LI

GHT

REGU

LAR

SIZE

VANT

AGE

MEDI

UM 7

KING

SIZ

EGA

ULOI

SES

BLON

DES

ULTR

A LIG

HT K

ING

SIZE

EXPO

RT 'A

' FUL

L FLA

VOUR

REG

ULAR

SIZ

EDA

KAR

LIGHT

REG

ULAR

SIZ

EFI

NE LI

GHT

REGU

LAR

SIZE

GIPS

Y LIG

HT R

EGUL

AR S

IZE

RODE

O RE

GULA

R SI

ZESE

LEST

A LIG

HT R

EGUL

AR S

IZE

SMOK

ING

LIGHT

REG

ULAR

SIZ

ECA

NYON

REG

ULAR

SIZ

EDA

KAR

REGU

LAR

SIZE

DISC

RETI

ON R

EGUL

AR S

IZE

FINE

REG

ULAR

SIZ

EGI

PSY

REGU

LAR

SIZE

SELE

STA

REGU

LAR

SIZE

SMOK

ING

REGU

LAR

SIZE

TREM

BLAY

REG

ULAR

SIZ

ECA

NYON

KIN

G SI

ZEDA

KAR

KING

SIZ

EDI

SCRE

TION

KIN

G SI

ZEFI

NE K

ING

SIZE

GIPS

Y KI

NG S

IZE

SELE

STA

KING

SIZ

ESM

OKIN

G KI

NG S

IZE

TREM

BLAY

KIN

G SI

ZEDA

KAR

LIGHT

KIN

G SI

ZEFI

NE LI

GHT

KING

SIZ

EGI

PSY

LIGHT

KIN

G SI

ZERO

DEO

KING

SIZ

ESE

LEST

A LIG

HT K

ING

SIZE

SMOK

ING

LIGHT

KIN

G SI

ZETR

EMBL

AY LI

GHT

KING

SIZ

EMO

RE R

EGUL

AR 12

0'sMO

RE M

ENTH

OL 12

0'sW

INST

ON LI

GHT

100's

SALE

M LIG

HT K

ING

SIZE

WIN

STON

100's

CAME

L LIG

HT K

ING

SIZE

WIN

STON

LIGH

T KI

NG S

IZE

CAME

L KIN

G SI

ZESA

LEM

KING

SIZ

EW

INST

ON K

ING

SIZE

CAME

L PLA

IN R

EGUL

AR S

IZE

Yiel

d (n

g/m

g ni

cotin

e)

dnarB

NNN

KNN

NNN

KNN

NNK, 4-(N-nitrosomethylamino)-1-(3-pyridyl)-1-butanone; NNN, N´-nitrosonornicotine

In addition, as new curing and other techniques emerge with a potential toeliminate preformed TSNA (those nitrosamines contained in unburnt tobaccoleaf or blend), the issue of pyrosynthesis of TSNA will become more impor-tant. The nitrosation of nicotine, a tertiary amine, during smoking occurs ata much slower rate than that of nornicotine, a secondary amine. This willfavour formation of NNN, N´- nitrosoanatabine and N´-nitrosoanabasine overNNK. Therefore, mandated limits will have to be set for both NNN and NNK

56

(or for NNN plus NNK) to monitor unexpected shifts in the TSNA compo-sition of the smoke of cigarettes engineered to meet the mandated loweringgoals. As the data presented in this report demonstrate, NNN is the predom-inant TNSA in US blend cigarettes, while NNK predominates in Canadianflue-cured tobacco cigarettes, illustrating the need for caution in extrapolatingdata from the international sample of Philip Morris brands to individualcountry markets or to other manufacturers.

3.5 Use of mandated lowering of toxicant levels in a productregulatory strategy

The goal of the proposed regulatory strategy is to reduce the levels of toxicconstituents measured under standardized conditions in the smoke ofcigarettes allowed on the market. A secondary goal is to prevent the intro-duction onto a market of cigarettes with higher levels of smoke toxicants thanare present in brands already on the market.

The regulatory strategy recommended by TobReg is based on well-acceptedprecautionary approaches used in public health. These approaches move to-wards a general reduction of known harmful toxicants in any product byestablishing product performance standards and as part of good manufactur-ing processes. They do not require that, for the substance under consideration,there be proof of a specific link between a lower level (amount) of any indi-vidual toxicant and a lower level of human disease (response). They merelyrequire that the substance be known to be harmful and that processes existfor its diminution or removal. Evidence of actual reduction in harm is notrequired for this approach; correspondingly, compliance with these regula-tions does not support a claim that a given brand is safe or less hazardous thanother brands.

Expressing the level of toxicants in smoke per unit amount of nicotine allowsquantification of the levels of toxicants that accompany a specified amountof nicotine in the smoke of different brands, at least under the conditions ofsmoke generation with the machine testing regimen. This normalization ofother toxicants to the nicotine yield allows a focus on the mixture of toxicconstituents generated when a cigarette is smoked and moves away from per-cigarette yields, which are misleading when used as estimates of the yieldsthat will occur when smokers actually smoke the cigarettes. The focus be-comes the toxicity of the smoke generated under a standardized machinetesting regimen rather than the quantity of the smoke generated. This stan-dardized measure of toxicity can be used to regulate cigarettes as a product.

The proposed regulatory approach mandates lowering the levels of specifictoxicants per milligram of nicotine by exclusion from the market of brandswith levels in the smoke that exceed specified regulatory levels. The existing

57

variation in toxicant levels across the brands currently on the market is usedto show that reducing levels of toxicants is technically achievable, as a largenumber of existing brands have already achieved the lower levels. Exclusionfrom the market of those brands with high levels of a smoke toxicant permilligram of nicotine should lower the mean level of that machine- measuredtoxicant per milligram of nicotine among the brands remaining on the market.A progressive reduction in the amount of toxicants in smoke over time canalso be achieved by progressively lowering the regulatory levels or by settingtargets or goals for further reduction as the technology to reduce toxicantsadvances.

Use of the variation in toxicant levels in existing brands to set the mandatedlimits ensures that there are manufacturing approaches that allow productionof cigarettes which are both acceptable to the smoker and can achieve theregulatory limit. In addition, this approach may encourage manufacturers ofcigarettes to voluntarily decrease toxicant emissions to the lowest levelsachievable even for products below the established regulatory limits.

The initial recommendations for regulatory levels are derived from data ontoxicant yields for a sample of international brands (Counts et al., 2005) andthose reported to Health Canada (2004). Reliance on these two lists is merelyan interim step. The proposed approach includes a recommendation for amulti-year period of required reporting of the machine-measured yields foreach brand sold in a given market. Market-specific data can eventually beused to establish the actual variation occurring so as to more precisely defineappropriate regulatory levels.

The value of examining the variation in toxicant yields for the market ofinterest rather than using an international sample is exemplified by comparingthe benzo[a]pyrene yields in the international sample with those reported toHealth Canada for Canadian brands. When the international sample is rankedfrom the brand with the lowest level of benzo[a]pyrene per milligram ofnicotine to the brand with the highest value, there is a continuous, steadyincrease across brands, from 5.7 to 13.8 ng/mg of nicotine. In contrast, theranking of Canadian brands from the lowest to the highest values (Figure 3.2)reveals a level which rises to approximately 9.6 ng/mg of nicotine, then asudden jump to 16.8 ng/mg. The explanation for this sudden discontinuity inlevels is not immediately evident, but the value of examining the Canadianexperience rather than assuming that the sample of international brands wouldprovide an adequate description of the Canadian market for establishingmandated reductions is clear.

58

Figure 3.2Benzo[a]pyrene concentrations by brand of Canadian cigarettes

0

2

4

6

8

01

21

41

61

81

02

EX

POR

T 'A

' UL

TR

A L

IGH

T R

EG

UL

AR

SIZ

EV

AN

TA

GE

SL

IMS

EX

TR

A L

IGH

T 4

100

'sE

XPO

RT

'A' M

ED

IUM

KIN

G S

IZE

EX

POR

T 'A

' FU

LL

FL

AV

OU

R K

ING

SIZ

EC

RA

VE

N A

EX

TR

A L

IGH

T K

ING

SIZ

EE

XPO

RT

'A' M

ED

IUM

RE

GU

LA

R S

IZE

EX

POR

T 'A

' EX

TR

A L

IGH

T R

EG

UL

AR

SIZ

EM

OR

E R

EG

UL

AR

120

'sE

XPO

RT

'A' L

IGH

T /

SPE

CIA

L L

IGH

T K

ING

SIZ

EE

XPO

RT

'A' L

IGH

T /

SPE

CIA

L L

IGH

T R

EG

UL

AR

SIZ

EM

OR

E M

EN

TH

OL

120

'sV

AN

TA

GE

UL

TR

A L

IGH

T 1

KIN

G S

IZE

EX

POR

T 'A

' EX

TR

A L

IGH

T K

ING

SIZ

EE

XPO

RT

'A' M

ILD

KIN

G S

IZE

EX

POR

T 'A

' MIL

D R

EG

UL

AR

SIZ

EE

XPO

RT

'A' U

LT

RA

LIG

HT

KIN

G S

IZE

NU

MB

ER

7 K

ING

SIZ

EC

RA

VE

N M

ILD

S U

LT

RA

MIL

D K

ING

SIZ

EC

AM

EL

LIG

HT

KIN

G S

IZE

WIN

STO

N L

IGH

T K

ING

SIZ

EE

XPO

RT

PL

AIN

RE

GU

LA

R S

IZE

RO

TH

MA

NS

SPE

CIA

L M

ILD

KIN

G S

IZE

WIN

STO

N L

IGH

T 1

00's

WIN

STO

N 1

00's

GA

UL

OIS

ES

BL

ON

DE

S K

ING

SIZ

EPL

AY

ER

'S E

XT

RA

LIG

HT

RE

GU

LA

R S

IZE

CA

ME

L K

ING

SIZ

ESA

LE

M K

ING

SIZ

EW

INST

ON

KIN

G S

IZE

EX

POR

T 'A

' FU

LL

FL

AV

OU

R R

EG

UL

AR

SIZ

ED

U M

AU

RIE

R E

XT

RA

LIG

HT

RE

GU

LA

R S

IZE

CR

AV

EN

A K

ING

SIZ

ESA

LE

M L

IGH

T K

ING

SIZ

EM

AT

INE

E E

XT

RA

MIL

D K

ING

SIZ

EC

AM

EL

PL

AIN

RE

GU

LA

R S

IZE

DU

MA

UR

IER

LIG

HT

KIN

G S

IZE

VIS

CO

UN

T E

XT

RA

MIL

D K

ING

SIZ

ESP

OR

TSM

AN

PL

AIN

RE

GU

LA

R S

IZE

VA

NT

AG

E M

ED

IUM

7 K

ING

SIZ

EB

EN

SON

& H

ED

GE

S D

EL

UX

E M

EN

TH

OL

UL

TR

A L

IGH

TS

100'

sPL

AY

ER

'S L

IGH

T R

EG

UL

AR

SIZ

EV

AN

TA

GE

MA

X 1

5 K

ING

SIZ

EC

RA

VE

N M

EN

TH

OL

KIN

G S

IZE

MA

TIN

EE

SIL

VE

R R

EG

UL

AR

SIZ

EB

EL

MO

NT

MIL

DS

RE

GU

LA

R S

IZE

BE

LM

ON

T M

ILD

S K

ING

SIZ

EPL

AY

ER

'S L

IGH

T K

ING

SIZ

EG

AU

LO

ISE

S B

LO

ND

ES

LIG

HT

S K

ING

SIZ

EV

AN

TA

GE

RIC

H 1

2 K

ING

SIZ

EG

AU

LO

ISE

S B

LO

ND

ES

UL

TR

A L

IGH

T K

ING

SIZ

EM

AR

K T

EN

PL

AIN

KIN

G S

IZE

DU

MA

UR

IER

LIG

HT

RE

GU

LA

R S

IZE

DU

MA

UR

IER

KIN

G S

IZE

PLA

YE

R'S

RE

GU

LA

R S

IZE

DU

MA

UR

IER

EX

TR

A L

IGH

T K

ING

SIZ

EPL

AY

ER

'S S

PEC

IAL

BL

EN

D K

ING

SIZ

EPL

AY

ER

'S S

PEC

IAL

BL

EN

D R

EG

UL

AR

SIZ

EPL

AY

ER

'S P

LA

IN R

EG

UL

AR

SIZ

EPL

AY

ER

'S E

XT

RA

LIG

HT

KIN

G S

IZE

DU

MA

UR

IER

RE

GU

LA

R S

IZE

CA

NY

ON

RE

GU

LA

R S

IZE

DA

KA

R R

EG

UL

AR

SIZ

ED

ISC

RE

TIO

N R

EG

UL

AR

SIZ

EFI

NE

RE

GU

LA

R S

IZE

GIP

SY R

EG

UL

AR

SIZ

ESE

LE

STA

RE

GU

LA

R S

IZE

SMO

KIN

G R

EG

UL

AR

SIZ

ET

RE

MB

LA

Y R

EG

UL

AR

SIZ

EC

AN

YO

N K

ING

SIZ

ED

AK

AR

KIN

G S

IZE

DIS

CR

ET

ION

KIN

G S

IZE

FIN

E K

ING

SIZ

EG

IPSY

KIN

G S

IZE

SEL

EST

A K

ING

SIZ

ESM

OK

ING

KIN

G S

IZE

TR

EM

BL

AY

KIN

G S

IZE

DA

KA

R L

IGH

T K

ING

SIZ

EFI

NE

LIG

HT

KIN

G S

IZE

GIP

SY L

IGH

T K

ING

SIZ

ER

OD

EO

KIN

G S

IZE

SEL

EST

A L

IGH

T K

ING

SIZ

ESM

OK

ING

LIG

HT

KIN

G S

IZE

TR

EM

BL

AY

LIG

HT

KIN

G S

IZE

DA

KA

R L

IGH

T R

EG

UL

AR

SIZ

EFI

NE

LIG

HT

RE

GU

LA

R S

IZE

GIP

SY L

IGH

T R

EG

UL

AR

SIZ

ER

OD

EO

RE

GU

LA

R S

IZE

SEL

EST

A L

IGH

T R

EG

UL

AR

SIZ

ESM

OK

ING

LIG

HT

RE

GU

LA

R S

IZEYi

eld

(ng/

mg

nico

tine

Brand

It is not known whether reducing the levels of the high-priority toxicantsidentified in this report will actually reduce harm or even reduce actual ex-posure to these harmful compounds. Therefore, it is an essential part of thisproposal that regulators assume the responsibility to ensure that consumersnot be informed directly or indirectly, or be led to believe, that cigarettes thatmeet the toxic limits established pursuant to this proposal are less hazardous,have been approved by the government or meet government-establishedhealth or safety standards. In particular, ranking of cigarette brands by anyof the machine-generated smoke metrics proposed in this report has greatpotential to be interpreted by smokers as reliable differences in probable ex-posure or harm that will result from smoking different brands. Communicat-ing such rankings to consumers, or allowing them to be communicateddirectly or indirectly, is likely to influence smokers’ behaviour in ways thatwill cause harm, similar to the harm currently caused by communicating ISOtar, nicotine and CO ratings.

Any health or exposure claims based on machine testing must be prohibiteduntil scientifically validated measures of exposure and harm are developedthat will allow regulators to determine that differences between brands doreduce actual exposure and risk. Hence, the current strategy to prohibit claims

59

will limit the possibility that the new standards will become marketing toolsand be used to misinform consumers.

The measurements of toxicant levels by brand and the costs associated withtesting and reporting are expected to be the responsibility of the cigarettemanufacturers. The results should be reported to the regulatory authority, anda sample of those results should be verified by an independent laboratory.Alternatively, some regulators may conduct the testing themselves with fund-ing derived from taxation or licensing of tobacco products.

It is the addictive nature of tobacco use that results in the long-term, repetitiveexposure to tobacco toxicants which result in disease manifestation. Thecharacteristics of tobacco products that contribute to and facilitate addictionare a potential focus for tobacco product regulation that could have substantialeffects on the burden of disease. Methods to assess addiction and identifica-tion of the factors that influence addiction are the focus of other work beingconducted by TobReg. They will not be discussed in this report, but theirexclusion should not be taken to imply their lack of importance.

The selection of toxicants in this report is based on a number of factors, in-cluding the fact that they are constituents of smoke currently established astoxic, they are known to be potently toxic and they have been measured inmultiple brands of cigarettes. As our knowledge of smoke chemistry expandsand the demonstration of the toxicity of smoke toxicants becomes more com-plete, the list of priority toxicants may change.

The principal goal of the proposed mandated lowering is to reduce the toxi-cant yields of brands currently on the market. It is possible, however, that, inthe absence of effective product regulation, the toxicant yields for the mix ofbrands on a market may actually increase as new brands are introduced or thecharacteristics of existing brands are changed. This possibility is of particularconcern for those markets in which the current levels of toxicants are lowerthan those in brands sold in other markets. The data presented in Figure 3.1show that the brands sold in Canada with high NNK levels are largely brandsalso sold in France and the USA rather than those brands that are predomi-nantly sold in the Canadian market. A similar situation exists in Australia(Figure 3.3). The mean and range of NNN and NNK yields per milligram ofnicotine in the brands sold in Australia (including Philip Morris brands) andreported to the Australian Government in 2001 (Australian Department ofHealth and Aging, 2001) are contrasted with the much higher levels of NNNand NNK reported in a Philip Morris Marlboro brand identified as beingAustralian in the paper of Counts and colleagues (2005). As Marlboro be-comes a leading brand in the Australian market, as it has in many othermarkets, the difference in NNN and NNK levels would be expected to in-crease the average yield of these toxicants in the cigarettes on the Australian

60

market. Setting levels for toxicants would provide a regulatory strategy thatcould prevent the introduction of newer brands with greater toxicant yieldsthan the existing brands.

Figure 3.3Mean and range of NNN and NNK yields per milligram of nicotine in brands reportedto the Australian government in 1999, and levels of NNN and NNK reported for a PhilipMorris Marlboro brand identified as being Australian

3.5.1 Selection of machine testing method

Measurement of smoke toxicants requires machine-generated smoke. Dif-ferent cigarette smoking machine testing regimens result in different levelsof toxicants per milligram of nicotine, and the relative ranking of brands bytoxicant also varies with the testing regimen used. Where possible, it is usefulto make measurements with more than one regimen in order to examine howthe results and rankings by brand vary with the measurement approach.

Three standardized approaches to machine testing in which data on machinetesting of multiple brands are available were examined: the ISO/FTC regi-men, the Massachusetts Department of Health regimen (slightly larger puffvolumes and 50% of vents blocked) and the modified intense machine smok-ing regimen used by Health Canada. Each of these methods has advantagesand disadvantages, but the modified intense smoking regimen was selected

61

PH F SK oroblraMnailartsuA/ 52

PH F SK oroblraMnailartsuA/ 52

detropeRnaem

detropeRnaemxaM

xaM

niMniM

0

03

06

09

021

051

KNNNNN

Yiel

d (n

g/pe

r mg

nico

tine)

NNK, 4-(N-nitrosomethylamino)-1-(3-pyridyl)-1-butanone; NNN, N´-nitrosonornicotine; KS, King Size;F, filter; HP, hard pack

as the method with the best fit for measuring toxicants for use in the proposedregulatory strategy. This regimen is modified from the existing ISO regimenas follows: the puff volume is increased from 35 ml to 55 ml, the puff intervalis decreased from 60 s to 30 s, and all ventilation holes are blocked by placingover them a strip of Mylar adhesive tape (Scotch Brand product no. 600Transparent Tape), which must be cut so that it covers the circumference andis tightly secured from the end of the filter to the tipping overwrap seam, orby another method of equivalent efficiency (Health Canada, 2000). This test-ing regimen is called ‘modified (intense) ISO’ by the Canadian Governmentbut is referred to in this report as the ‘modified intense smoking’ regimen toavoid confusion between it and the existing ISO/FTC regimen.

The selection of the modified intense smoking regimen was based on severalcriteria. First, the larger quantity of smoke generated by this regimen reducesthe coefficient of variation of replicate measurements for TSNA, which arethe initial set of toxicants recommended for regulatory consideration.Figure 3.4 presents the mean coefficient of variation for four TSNA (N´-nitrosoanabasine, N´-nitrosoanatabine, NNN and NNK) in each of the inter-national brands of cigarettes measured by Counts et al. (2005), with the resultsfor all three machine testing regimens presented in the same figure and plottedagainst the tar yield for each brand measured with the machine testing method.

Figure 3.4Mean of coefficients of variation of replicate measurements of four nitrosaminesplotted by the tar level in individual brands measured by each of three machinetesting protocols

0

50.0

1.0

51.0

2.0

52.0

040302010

Coe

ffici

ent o

f var

iatio

n of

repl

icat

e m

easu

rem

ents

)gm(levelraT

OSI

deifidoMesnetni

ssaM

62

It is evident from the graph that the variation in replicate measurements ofthe TSNA increases when the testing regimen generates less than approxi-mately 10 mg of tar. The measurements of both the ISO and the Massachusettsregimen include substantial numbers of international brands that have lessthan 10 mg of tar and correspondingly show larger variation in replicatemeasurements. Only the modified intense smoking regimen yielded a stablevariation in replicate measurements across brands with different tar levels.

A second reason for selecting the modified intense smoking regimen was thatwith certain design features the more intense machine parameters may yieldlevels of individual smoke toxicants substantially above those that wouldresult when ISO smoking conditions are used, even when standardized byreporting them per milligram of nicotine. Higher yields are generated undermachine conditions that correspond to a more intense human smoking profileand therefore may better reflect how the product performs under these intenseconditions.

Third, in selecting a machine testing regimen, TobReg considered it impor-tant to select a regimen that could accurately characterize cigarette designchanges, other than filter ventilation, as correction of toxicant yields by ex-pressing them per milligram of nicotine alone was not sufficient to charac-terize the yields produced under conditions of more intense puffing.

The use of charcoal in cigarette filters is an example of a design change withan impact that is not well characterized by normalization to nicotine. Use ofthe ISO smoking regimen (35-ml puff, 60-s puff interval, 2-s puff, no ventblocking) to test the delivery of volatile (e.g. benzene, 1,3-butadiene, acry-lonitrile) components in smoke from cigarettes with charcoal filters showedthat the levels of these compounds are significantly lower than those of othersmoke toxicants, including nicotine.

As long as enough charcoal is included in the filter, these reductions arepresent even when an intense puffing regimen like the modified intensesmoking regimen (55-ml puff, 30-s puff interval, 2-s puff, 100% vent block-ing) is used. The newly introduced Marlboro UltraSmooth that is beingmarketed in Salt Lake City, Utah, USA, has an example of a filter in whichsufficient charcoal is present to maintain reductions in yields of volatile com-pounds even under the modified intense smoking regimen. The MarlboroUltraSmooth cigarette marketed in Atlanta, Georgia, USA, and the modifiedcharcoal filtered Marlboro Ultralight marketed in North Dakota, USA, haveless charcoal, however, and there is breakthrough of the volatiles with moreintense puffing regimens, resulting in higher yields. For these two products,smoking with the ISO regimen indicates that the levels of smoke volatiles aresignificantly lower than in other cigarettes with similar delivery that are notcharcoal-filtered, and the difference persists even when presented per

63

milligram of nicotine. When these products are smoked with the modifiedintense smoking regimen, however, the proportionate increase in the levelsof volatile components is much greater than the increase in nicotine. Nor-malization per milligram of nicotine with data from the ISO smoking regimenwould not correct for the increased smoke toxicant yields of these interme-diate level charcoal filter brands that results from more intense smokingpatterns.

Regulatory authorities need correct information about the products for salein their jurisdictions, and the regulatory smoking regimen selected should beable to characterize design changes that result in significantly higher smoketoxicant delivery relative to nicotine delivery with more intense smoking.While no one machine smoking regimen perfectly characterizes cigarettetoxicant yields, TobReg concluded that the modified intense smoking regi-men offers significant advantages over the other two regimens.

When cigarettes are tested with different smoking regimens and the smokedeliveries are normalized for nicotine, significant differences remain in therelative nicotine-normalized deliveries under different smoking conditions.Table 3.2 shows data taken from Counts and colleagues (2005) on the deliveryof three smoke toxicants of concern (acrolein, benzo[a]pyrene and NNN)measured with three different smoking regimens (ISO/FTC, MassachusettsDepartment of Health, modified intense), normalized for nicotine deliverydetermined under the same smoking conditions for five types of cigarettes(two full flavour, two light and one ultralight).

The two full-flavour cigarettes (E5 and E7) showed little change in the rela-tive delivery of all three toxicants in the three puffing regimens. Nicotine-normalized acrolein in the two ‘light’ cigarettes (E17 and E24) and the ‘ultra-light’ cigarette (E33) increased significantly as the puff parameters becamemore intense and ventilation holes were blocked. Nicotine-normalizedbenzo[a]pyrene in the two ‘light’ cigarettes and the ‘ultra-light’ cigaretteshowed virtually no change as the puff parameters became more intense andventilation holes were blocked. Nicotine-normalized NNN in the two ‘light’cigarettes showed an inconsistent pattern, but a small decrease was seen asthe puff parameters became more intense and ventilation holes were blocked.For the ‘ultra-light’ cigarette, the nicotine- normalized NNN decreased sub-stantially as the puff parameters became more intense and ventilation holeswere blocked.

These results show that the puffing regimen affects not only differences inthe dilution of smoke but also in the cigarette burning properties that alter therelative delivery of toxicants. This affects the relative levels of toxicantsmeasured in smoke with different smoking regimens. These results are inagreement with TobReg Recommendation 1 (WHO, 2004), which suggested

64

that, when the resources are available, a second set of puffing parametersshould be used to test cigarettes under more intense smoking conditions inaddition to the ISO/FTC regimen. Where possible, measurements should betaken over a range of puffing parameters to understand the delivery of toxi-cants in smoke emissions.

Table 3.2Delivery of specified toxicants under different machine smoking conditions

Lot Manufacturer and type Acrolein (μg/mg nicotine)

ISO/FTC MassachusettsState

Modifiedintense

E5 Marlboro KS F SP/US 54.31 54.10 58.13E7 L&M KS F HP/EU 71.11 66.04 74.73E17 L&M KS F Lt/EU 61.38 69.17 84.93E24 Marlboro KS F HP Lt/UK 56.79 63.93 78.86E33 Marlboro KS F HP Ult/EU 40.69 53.02 69.32

Lot Manufacturer and type Benzo[a]pyrene (ng/mg nicotine)

ISO/FTC MassachusettsState

Modifiedintense

E5 Marlboro KS F SP/US 11.7 9.0 9.9E7 L&M KS F HP/EU 12.9 11.6 10.0E17 L&M KS F Lt/EU 12.8 11.6 9.6E24 Marlboro KS F HP Lt/UK 11.1 9.5 10.0E33 Marlboro KS F HP Ult/EU 8.1 7.4 7.7

Lot Manufacturer and type NNN (ng/mg nicotine)

ISO/FTC MassachusettsState

Modifiedintense

E5 Marlboro KS F SP/US 154 145 151E7 L&M KS F HP/EU 96 84 93E17 L&M KS F Lt/EU 101 84 95E24 Marlboro KS F HP Lt/UK 85 77 64E33 Marlboro KS F HP Ult/EU 116 91 74

ISO, International Organization for Standardization; FTC, Federal Trade Commission; KS, king size;F, filter; SP, soft pack; US, United States; L&M, Liggett & Meyers; HP, hard pack; EU, EuropeanUnion; Lt, light; UK, United Kingdom; Ult, ultra-light; NNN, N´-nitrosonornicotine

From Counts et al. (2005)

65

3.5.3 Sources of data on toxicants

Comprehensive lists of toxicants measured in a consistent manner with themodified intense smoking regimen are available from three sources: a pub-lication by Counts and colleagues (2005) comparing a set of internationalbrands manufactured by Philip Morris, a set of Canadian brands (HealthCanada, 2004) reported by law to Health Canada for the year 2004, and a setof Australian brands (Australian Department of Health and Aging, 2001). Thedata from these three sources are listed in Annex 3.1. The data comparisonsare for measurements made with the modified intense smoking regimen withall ventilation holes blocked.The Canadian data were modified slightly to remove those brands that areeither US-style blended cigarettes or French-manufactured Gauloise brands.This modification was made to eliminate those high-TSNA brands (NNNlevels > 100 ng/mg of nicotine) which might distort the averages for brandsof more traditional Canadian style (Figure 3.1). In addition, the values fortoluene and styrene for the brands Vantage Rich 12 king size and VantageMax 15 king size were deleted from the analyses because they appeared tohave been transposed and, as a result, substantially distorted the maximumand minimum values of all brands reported for the same toxicants when theywere included.

3.5.4 Criteria for selecting toxicants for mandatory lowering regulation

The toxicants considered were those on the list reported to Health Canada(2004) and measured by Counts and colleagues (2005). The list was selectedby Health Canada to represent those constituents of smoke that contribute themost to its toxicity.

Priorities on the list were based on the known animal and human toxicity ofthe compounds, toxicity indices based on the concentration of the constituentmultiplied by its toxic potency, the variation in the toxicant across brands,the potential for the toxicant to be lowered in cigarette smoke with existingmethods and the need to have constituents that represent the different phasesof smoke (gas and particulate), different chemical families and toxicities thatreflect heart and lung disease as well as cancer. The most important criterionfor selecting compounds for regulation was the evidence on toxicity.

In addition, the yield per milligram of nicotine of a toxicant has to vary sub-stantially across the brands on the market, or little will be gained by removingthose with higher yields. The variation in toxicant yields across brands shouldbe substantially greater than the variation in repeated measurement in a singlebrand. Otherwise, more measurements would be required for each toxicantin each brand to have a precise estimate of the mean value, and the cost oftesting would increase proportionally.

66

The availability of technology or other approaches to reduce the level of aspecific toxicant per milligram of nicotine in smoke is also a consideration.When readily available alterations in tobacco processing or cigarette designand manufacture are known to reduce the level of certain toxicants in smoke,it is feasible to set limits on these toxicants, which are therefore of higherpriority. Toxicants are generated in smoke in at least three ways, which mayinfluence the ability to reduce their levels. The levels of toxicants such asTSNA that are generated mainly by the way in which the tobacco plant istreated (cured) can be lowered by changing the production process. For tox-icants derived mainly during the combustion of organic material, such aspolycyclic aromatic hydrocarbons (PAHs such as benzo[a]pyrene) and CO,changes in design characteristics, tobacco blends and manufacturing pro-cesses (e.g. catalysts) might be needed to substantially lower their levels.Other toxicants are generated by combustion from substances added to aproduct for various reasons (e.g. processing aids, sugars and flavours); theseinclude aldehydes and other volatile organic compounds, such as acetalde-hyde, formaldehyde, acrolein and acrylonitrile, which contribute substan-tially to the toxicity of smoke. The levels of these toxicants could be reducedby lowering or removing sugars as additives from the products.

A final consideration is the existence of a market for brands low in a particulartoxicant (such as Canadian brands low in TSNA). For these markets, man-dated limits might be set on toxicants to prevent the introduction of brandsfrom other markets that have substantially higher levels of that toxicant thanthe native brands.

A list of toxicants with all these characteristics would be very long, and thelarger the number of toxicants regulated, the more distortion there will be inthe existing market and the more complex will be regulatory oversight.Therefore, TobReg examined the list of reported toxicants in order to identifya smaller number of compounds that would balance the concerns identifiedwith the practical reality of a regulatory structure.

Examination of evidence for toxicity

A principal consideration in selecting compounds for regulation is theirknown toxicity. A review of the toxicity of all the compounds in smoke isbeyond the scope of this report, but the evidence for the toxicity of thosecompounds recommended for regulation or for reporting was reviewed.

Examination of toxicant levels and quantitative data on toxicity

Characterizing the hazards of chemical products and their emissions gener-ally involves estimating the levels of individual components, describing theinherent properties of those components to cause adverse effects (toxicity)

67

and relating this to their inherent strength to cause adverse effects (potency).A comprehensive hazard characterization of smoked tobacco products andtheir emissions would be very complex, in view of the very many componentsof smoke (more than 4000), limited toxicological information on many ofthese compounds and the possibility for interaction between the various tox-icants. Fowles and Dybing (2003) presented a simplified system for charac-terizing the hazards of the components of cigarette smoke by calculatingcancer risk and non-cancer risk indices. This system involves multiplying theyields of individual smoke toxicants derived with the standard ISO machinesmoking method by cancer and non-cancer potency factors.

The method devised by Fowles and Dybing (2003) is expanded in this report,in that yields of toxicologically important constituents of smoke were takenfrom published data based on the modified intense smoking regime and nor-malized per milligram of nicotine. The amount of smoke collected under themodified intense regimen is greater than under ISO machine conditions, andthe smoke is generated under more intense conditions of tobacco combustion.Normalizing levels of toxicants per milligram of nicotine has been identifiedas a means for reducing the misleading differences between levels of toxicantsexpressed per cigarette (WHO, 2004).

Cancer potency factors: The T25 cancer potency estimation method ofDybing et al. (1997) was used. The T25 value is the long-term daily dose thatwill produce tumours in 25% of animals above the background rate at a spe-cific tissue site. The T25 is determined by linear extrapolation from the lowestdose that results in a statistically significant increase in tumours. The pertinentdata from the cancer bioassays underlying calculation of the T25 values arepresented in Annex 3.2. For the present calculations, the T25 values wereconverted into cancer potency factors per milligram (1/T25).

Non-cancer potency factors: Long-term reference exposure levels publishedby the Office of Environmental Health Hazard Assessment, California En-vironmental Protection Agency, in February 2005 were used (http://www.oehha.ca.gov/air/chronic_rels/AllChrels.html). This level is an air-borne level of a chemical at or below which no adverse health effects areanticipated for individuals with long-term exposure to that level. Referenceexposure levels were derived from both human and animal toxicological dataand presented for the target system for each substance. For the present cal-culations, a reference exposure level is viewed as the inverse of the respectivetoxicant’s non-cancer potency factor, and the term ‘tolerable level’ is used inthis report instead of ‘reference exposure level’.

Toxicant animal carcinogenicity and toxicant non-cancer response indices:A toxicant animal carcinogenicity index is the numerical value of the amountof an individual component in cigarette smoke, normalized per milligram of

68

nicotine and multiplied by its cancer potency factor (1/T25). A toxicant non-cancer response index is the numerical value of the amount of an individualcomponent in cigarette smoke, normalized per milligram of nicotine andmultiplied by its non-cancer potency factor (or divided by its tolerable level).

The toxicant animal carcinogenicity and the toxicant non-cancer responseindices are calculated from the mean level of the individual toxicant per mil-ligram of nicotine in all brands in the data set and are presented in Annex 3.3Tables A3.1–A3.3. The indices are estimated from data sets of machine-generated smoke yields obtained by the modified intense smoking regimenfor international Philip Morris brands (Counts et al., 2005), Canadian brands(Health Canada, 2004) and Australian brands (Australian Department ofHealth and Aging, 2001). The indices were calculated from the reported levelsin the three data sets and are presented as means, 90th percentiles and maximaper milligram of nicotine per cigarette. The data sets include yield levels forup to 43 individual toxicants; some elements, such as arsenic, chromium andselenium, were not reported or quantified in some of the data sets. Informationon methyl ethyl ketone was not reported in the Canadian data.

Absolute T25 values and T25 per milligram of carcinogen per milligram ofnicotine are also presented in Annex 3.3 Tables A3.1–A3.3 for the 14 car-cinogens measured in tobacco smoke when such values could be calculated(see Annex 3.2 for the basis used to calculate T25 values). Toxicant animalcarcinogenicity indices are calculated by multiplying the yield per milligramof nicotine by the unit milligram T25 value. Toxicant non-cancer responseindices are calculated by multiplying the yield per milligram of nicotine bythe inverse tolerable level (or divided by the tolerable level).

Tables 3.3 and 3.4 list the mean toxicant animal carcinogenicity and toxicantnon-cancer response indices from the three data sets. A similar picture is seenin the ranking of toxicants with respect to both cancer and non-cancer re-sponses. With respect to cancer hazards, 1,3-butadiene, acetaldehyde, iso-prene, NNK, benzene and cadmium ranked highest. As the experimental dataon 4-aminobiphenyl did not allow proper estimation of a T25 value, a toxicantanimal carcinogenicity index value for this human carcinogen could not becalculated. With respect to toxicant non-cancer responses, the index wasmuch higher for acrolein (targets: respiratory system, eyes) than for the othertoxicants; acetaldehyde (target: respiratory system), formaldehyde (targets:respiratory system, eyes) and hydrogen cyanide (targets: cardiovascular sys-tem, nervous system, endocrine system) followed in ranking order.

69

Table 3.3Ranking of toxicants in smoke with respect to toxicant animal carcinogenicity indices

Toxicant Mean

Counts et al. (2005) Canadian data Australian data All

1,3-Butadiene 11.4 8.9 9.5 9.9Acetaldehyde 7.0 5.7 5.5 6.1Isoprene 4.6 2.9 3.6 3.7NNK 4.7 3.8 1.8 3.4Benzene 2.7 2.8 2.4 2.6Cadmium 1.6 2.4 1.2 1.7Acrylonitrile 1.7 1.4 1.2 1.4Hydroquinone 1.1 1.3 1.2 1.2Catechol 0.49 0.75 0.50 0.58NNN 0.55 0.22 0.10 0.29Benzo[a]pyrene 0.0082 0.0096 0.0081 0.00862-Aminonaphthalene 0.00081 0.00077 0.00047 0.000681-Aminonaphthalene 0.00049 0.00032 0.00028 0.000363Lead 0.00 0.00 0.00 0.00

NNK, 4-(N-nitrosomethylamino)-1-(3-pyridyl)-1-butanone

Table 3.4Ranking of toxicants in smoke with respect to toxicant non-cancer response indices

Toxicant Mean

Counts et al. (2005) Canadian data Australian data Mean

Acrolein 1127 1188 983 1099Acetaldehyde 77.2 62.9 61.1 67.1Formaldehyde 13.7 25.8 20.0 19.8Hydrogen cyanide 22.7 15.9 13.0 17.2Nitrogen oxides 5.0 2.2 2.1 3.1Cadmium 2.4 3.6 1.8 2.61,3-Butadiene 2.7 2.1 2.3 2.4Acrylonitrile 2.5 2.0 1.8 2.1Carbon monoxide 1.5 1.2 1.1 1.3Benzene 0.66 0.68 0.57 0.64Toluene 0.24 0.24 0.18 0.22Arsenic 0.16 – – 0.16Methyl ethyl ketone 0.09 – 0.07 0.08Ammonia 0.11 0.06 0.05 0.07Phenol 0.06 0.09 0.06 0.07Mercury 0.04 0.03 0.00 0.02Styrene 0.02 0.01 – 0.02m- and p-Cresol 0.01 0.02 0.01 0.01o-Cresol 0.01 0.01 0.00 0.01

70

This ranking of measured carcinogens and non-carcinogens in cigarettesmoke is based on integration of machine smoking yield data normalized permilligram of nicotine with information on the toxic potency of the measuredtoxicants. One approach to quantifying the potential independent contributionof individual toxicants to the toxicity of the products examined in the threedata sets is to compare the hazard indices. Generally, the same ranking orderwas achieved, whether the hazard indices were calculated per mean, 90thpercentile or maximum levels of toxicants. Another finding was that therelative ranking order, with respect to both carcinogenicity and non-carcinogenicity, was similar across the three data sets, although they coveredbrands with different tobacco blends. This suggests that the similar basicdesigns of the products and the general characteristics of the combustionprocess contribute relatively uniformly to the hazard of the products, and thatthe priority of compounds for regulatory consideration is similar for thebrands in all three data sets.

There are obvious limitations to the method used here for ranking cigarettesmoke toxicants. As each measured toxicant is treated individually, the pos-sibility of chemical interactions that either enhance or reduce the hazardousproperties of the smoke is not taken into account. Further, these calculationswere possible only for those toxicants for which T25 and tolerable level val-ues had been estimated, and not for the remaining some 4000 constituents ofcigarette smoke (IARC, 2004). In addition, potency factors were not availablefor many of the measured compounds, tolerable levels were lacking for 21 ofthe 43 compounds, and proper carcinogenicity bioassays have been carriedout for only some of the toxicants. As many of the potency factors werederived from animal experiments, the limitations in extrapolating from ani-mal models to the human situation also apply. These limitations preclude useof these indices as quantitative estimates of the likely harm or risk of exposureto these toxicants or of the risk or harm of different cigarette brands. Theydo, however, provide a useful set of metrics that can be considered in selectingwhich toxicants to regulate.

Variation in toxicants levels by brand

The variation in the level of toxicant per milligram of nicotine across differentbrands of cigarettes is a second criterion that can be used to identify toxicantsfor which mandated reductions are likely to have the greatest impact on low-ering the mean levels per milligram of nicotine. The primary criteria remainthose related to toxicity; however, little reduction in the average level of atoxicant will be achieved if the values for different brands are clustered tightlyaround the mean value. Correspondingly, mandated lowering of toxicantswith the broadest variation around the mid-point value will result in the largestlowering of the levels in brands containing levels above the mean value.

71

Measurement of variation among brands

A substantial proportion of the variation across cigarette brands in level oftoxicants produced by machine smoking is due to dilution of the smoke pro-duced by ventilation holes in the filter. Normalizing the yields per milligramof nicotine and blocking all the filter vents in the testing regimen are intendedto reduce the effect of ventilation on the values recommended as regulatorylevels.

The variation in levels of different toxicants can be expressed as the coeffi-cient of variation, which is the standard deviation of the measurement acrossbrands divided by the mean value for all the brands. The coefficient of vari-ation is then stated as the percentage variation for each toxicant around themean value, allowing direct comparison of the extent to which different tox-icants vary across brands. The coefficients of variation for individual toxi-cants in brands on the market are presented in Annex 3.4 for each of the datasets examined.

An issue of practical concern is the variation in repeated measurements of agiven toxicant in a given brand. The reproducibility of testing can vary sub-stantially for different toxicants and with different testing methods. Repeatedmeasurements are used to estimate the mean (true) value for a brand, and thevariation of the repeated measurements defines the confidence intervalaround that mean (true) value. When the variation of repeated measurementsis high, more measurements are needed to estimate the mean within a spec-ified confidence interval than when there is little variation. As a practicalmatter, a regulatory authority will be obliged to validate some of the valuesfor toxicants reported by tobacco companies, by comparing the reported andvalidated mean values for the toxicant in order to determine whether the valuereported by the tobacco company is outside the standard error of the estimateof the mean of the measurements made by the independent laboratory. If thevariation of the replicate measurements is high, either larger numbers ofreplicate measurements will be needed or a wide confidence interval aroundthat mean will result, making regulatory oversight of tobacco industry toxi-cant testing and reporting difficult and expensive.

Variation among brands and the variation of replicate measurements can becombined into a single measure for the purpose of examining different toxi-cants. The coefficient of variation across brands for the brand-specific meanvalues of a toxicant per milligram of nicotine can be presented as a ratio bydividing it by the mean of the coefficients of variation reported for the repli-cate measurements of that toxicant per milligram of nicotine by brand (thecoefficient of variation among brands over the replicate coefficient of varia-tion). This measure provides an estimate of the magnitude of the variationacross brands relative to the reproducibility of the measurement, and

72

potentially identifies those toxicants for which the brand variation is suffi-ciently large to support mandated reduction regulation and for which theburden of regulatory validation of tobacco industry smoke toxicant reportingwill be modest.

Table 3.5 presents the ratios of the coefficients of variation for Philip Morrisinternational brands, Canadian brands and Australian brands. The shadedcells of the table represent toxicants for which animal carcinogenicity andnon-cancer response indices have been calculated. Examination of the valuesfor the ratios and the ranking of the toxicants in each of the three data setsshows that there are substantial differences in the variation of toxicants incigarette brands in the three markets. For example, benzo[a]pyrene, whichranks high in the Canadian brands, is much lower in the Philip Morris inter-national brands. In contrast, some toxicants such as phenol and CO rank highin each of the three data sets.

The values of the ratios for individual toxicants found in the Philip Morrisinternational brands and the Canadian brands suggest that there is sufficientvariation across brands in most of the toxicant levels that mandated reductionswould have a substantial effect on the levels in the brands remaining on themarket.

A second approach to examining variation across brands is presented inFigures 3.5–3.7, which present the maximum and minimum values for eachtoxicant in the brand data set, expressed as a ratio of the median value for thattoxicant. The value for the 90th percentile value is also presented, as is theline for 125% of the median value. This format allows direct examination ofthe range of toxicant values across brands without any adjustment for varia-tion in replicate measurement for that toxicant. The 90% value is presentedto identify those toxicants for which a single brand or a few brands are outlierson the high side of the median value, and correspondingly for which only afew brands have a high level per milligram of nicotine. The figures indicatethat the Philip Morris international brands and the Canadian brands showmore variation than the Australian brands, which may be due partly to thesmall number of brands (15) reported to the Australian Government. Fur-thermore, in each data set, the variation for some toxicants is proportionallylarger than for other toxicants, suggesting that they may be toxicants for whichsetting mandated limits would have a relatively greater effect. For some tox-icants, the maximum value is substantially higher than the 90th percentilevalue, indicating that relatively few brands in the data set have dispropor-tionately high levels of these toxicants.

73

Tab

le 3

.5R

atio

of

coef

fici

ent

of

vari

atio

n in

to

xica

nt

leve

ls p

er m

illig

ram

of

nic

oti

ne

amo

ng

bra

nd

s to

th

e m

ean

of

rep

licat

e co

effi

cien

ts o

fva

riat

ion

Ph

ilip

Mo

rris

bra

nd

sC

anad

ian

bra

nd

s (e

xclu

din

gF

ren

ch a

nd

US

bra

nd

s)A

ust

ralia

n b

ran

ds

To

xica

nt

Rat

ioT

oxi

can

tR

atio

To

xica

nt

Rat

io

NN

N4.

89B

enzo

[a]p

yren

e4.

90R

esor

cino

l2.

75C

arbo

n m

onox

ide

4.83

Phe

nol

4.90

Phe

nol

2.71

N´-

Nitr

osoa

nata

bine

4.72

Isop

rene

4.87

o-C

reso

l2.

57C

adm

ium

4.19

m-

and

p-C

reso

l4.

274-

Am

inob

iphe

nyl

2.48

Phe

nol

3.93

Car

bon

mon

oxid

e4.

22C

adm

ium

2.41

Nitr

ic o

xide

3.84

NN

K4.

12N

NN

2.39

Nitr

ogen

oxi

des

3.74

2-A

min

onap

htha

lene

3.81

Car

bon

mon

oxid

e2.

27m

- an

d p-

Cre

sol

3.65

o-C

reso

l3.

783-

Am

inob

iphe

nyl

2.27

Tot

al h

ydro

gen

cyan

ide

3.55

4-A

min

obip

heny

l3.

75N

´-N

itros

oana

tabi

ne2.

07H

ydro

quin

one

3.5

T

olue

ne3.

58M

ercu

ry1.

96A

mm

onia

3.41

Hyd

roge

n cy

anid

e3.

39m

- an

d p-

Cre

sol

1.95

Lead

3.05

But

yral

dehy

de3.

33B

enzo

[a]p

yren

e1.

94Im

ping

er h

ydro

gen

cyan

ide

3.03

Cad

miu

m3.

312-

Am

inon

apht

hale

ne1.

90Q

uino

line

3.01

Ace

tone

3.28

Qui

nolin

e1.

88S

tyre

ne2.

98B

enze

ne3.

27S

tyre

ne1.

78F

orm

alde

hyde

2.97

For

mal

dehy

de3.

24H

ydro

gen

cyan

ide

1.75

Pad

hyd

roge

n cy

anid

e2.

93Le

ad3.

221,

3-B

utad

iene

1.67

o-C

reso

l2.

88A

ceta

ldeh

yde

3.21

Ben

zene

1.61

NN

K2.

86A

crol

ein

3.10

But

yral

dehy

de1.

60N

´-N

itros

oana

basi

ne2.

86N

´-N

itros

oana

tabi

ne3.

09N

NK

1.60

4-A

min

obip

heny

l2.

62Q

uino

line

2.89

Isop

rene

1.55

74

Pro

pion

alde

hyde

2.53

1,3-

But

adie

ne2.

88N

´-N

itros

oana

basi

ne1.

55A

ceta

ldeh

yde

2.52

Nitr

ogen

oxi

des

2.78

1-A

min

onap

htha

lene

1.53

Acr

olei

n2.

51P

ropi

onal

deyd

e2.

78A

mm

onia

1.53

Ace

tone

2.5

C

roto

nald

ehyd

e2.

74A

ceta

ldeh

yde

1.49

But

yral

dehy

de2.

49N

NN

2.72

Nitr

ic o

xide

1.45

Isop

rene

2.47

Hyd

roqu

inon

e2.

71M

ethy

l eth

yl k

eton

e1.

44C

atec

hol

2.44

Nitr

ic o

xide

2.56

Nitr

ogen

oxi

des

1.43

Pyr

idin

e2.

36P

yrid

ine

2.44

Pro

pion

alde

hyde

1.43

3-A

min

obip

heny

l2.

08A

cryl

onitr

ile2.

40A

crol

ein

1.42

Acr

ylon

itrite

2.05

Cat

echo

l2.

38P

yrid

ine

1.42

Cro

tona

ldeh

yde

2

A

mm

onia

2.36

Tol

uene

1.41

1,3-

But

adie

ne1.

92N

´-N

itros

oana

basi

ne2.

14A

cryl

onitr

ile1.

28R

esor

cino

l1.

9

Sty

rene

1.93

Ace

tone

1.28

Ben

zo[a

]pyr

ene

1.89

1-A

min

onap

htha

lene

1.75

Cat

echo

l1.

11M

ethy

l eth

yl k

eton

e1.

883-

Am

inob

iphe

nyl

1.74

Hyd

roqu

inon

e0.

992-

Am

inon

apht

hale

ne1.

73R

esor

cino

l1.

72F

orm

alde

hyde

0.98

Tol

uene

1.72

Mer

cury

1.59

Cro

tona

ldeh

yde

0.92

Mer

cury

1.62

Lead

0.43

Ben

zene

1.55

1-A

min

onap

htha

lene

1.53

Ars

enic

0.88

NN

N, N

´-ni

tros

onor

nico

tine;

NN

K, 4

-(N

-nitr

osom

ethy

lam

ino)

-1-(

3-py

ridyl

)-1-

buta

none

75

Figure 3.5Maximum, minimum and 90% values as ratios of the median of each constituent permilligram of nicotine for international Philip Morris brands, modified intense regimen

Rat

io

tneutitsnoC

0

1

2

3

Car

bon

mon

oxid

eA

ceta

ldeh

yde

Ace

tone

Acr

olei

nA

cryl

onitr

ite1-

Am

inon

apht

hale

ne2-

Am

inon

apht

hale

ne3-

Am

inob

iphe

nyl

4-A

min

obip

heny

lA

mm

onia

Ars

enic

Ben

zene

Ben

zo[a

]pyr

ene

1,3

But

adie

neB

utyr

alde

hyde

Cad

miu

mC

atec

hol

Cro

tona

ldeh

yde

Form

alde

hyde

Hy dr

oqui

none

Impn

gr h

y drog

en c

yani

deIs

opre

neLe

adm

-+p-

Cre

sol

Mer

cury

Met

hyl e

thyl

ket

one

NA

BN

AT

Nitr

ic o

xide

Nitr

ogen

oxi

des

NN

KN

NN

o-C

reso

lP

ad h

ydro

gen

cyan

ide

Phe

nol

Pro

pion

alde

hyde

Pyr

idin

eQ

uino

line

Res

orci

nol

Styr

ene

Tolu

ene

Ttl h

ydro

gen

cyan

ide

naidem/%09

naidem/xaM

naidem/niM

naidemfo%521

Figure 3.6Maximum, minimum and 90% values as ratios of the median of each constituent permilligram of nicotine for Canadian brands, minus French and US brands, modifiedintense regimen

0

1

2

3

4

Car

bon

mon

oxid

e

Am

mon

ia

1-A

min

onap

thal

ene

2-A

min

onap

thal

ene

3-A

mio

biph

enyl

4-A

mio

biph

enyl

Ben

zo{a

)pyr

ene

Form

alde

hyde

Ace

tald

ehyd

e

Ace

tone

Acr

olei

n

Pro

pion

alde

hyde

Cro

tona

ldeh

yde

But

yral

dehy

de

Hyd

roge

n cy

anid

e

Mer

cury

Lead

Cad

miu

m

NO

NO

X

NN

N

NN

K

NA

T

NA

B

Pyr

idin

e

Qui

nolin

e

Hyd

roqu

inol

ine

Res

orci

nol

Cat

echo

l

Phe

nol

M+

P C

reso

l

O-C

reso

l

1,3

But

adie

ne

Isop

rene

Acr

ylon

itrile

Ben

zene

Tolu

ene

Sty

rene

Rat

io

tneutitsnoC

naidem/%09

naidem/xaM

naidem/niM

naidemfo%521

76

Figure 3.7Maximum, minimum and 90% values as ratios of the median of each constituent permilligram of nicotine for Australian brands, modified intense regimen

Rat

io

tneutitsnoC

0

1

21,

3 B

utad

iene

1-A

min

onap

htha

lene

2-A

min

onap

htha

lene

3-A

min

obip

heny

l4-

Am

inob

iphe

nyl

Ace

tald

ehyd

eA

ceto

neA

crol

ein

Acr

ylon

itrile

Am

mon

iaB

enze

neB

enzo

[a]p

yren

eB

utyr

alde

hyde

Cad

miu

mC

arbo

n m

onox

ide

Cat

echo

lC

roto

nald

ehyd

eFo

rmal

dehy

deH

ydro

gen

cyan

ide

Hyd

roqu

inon

eIs

opre

neLe

adm

-+p-

Cre

sol

Mer

cury

Met

hyl e

thyl

ket

one

N n

itros

oana

basi

neN

nitr

osoa

nata

bine

Nic

otin

eN

itric

oxi

deN

NK

NN

NN

Ox

o-C

reso

lP

heno

lP

ropi

onal

dehy

deP

yrid

ine

Qui

nolin

eR

esor

cino

lS

tyre

neTo

luen

e

naidem/%09

naidem/xaM

naidem/niM

naidemfo%521

It is also worth noting that, even when the French and US brands are excluded,the Canadian data set showed substantial variation in the levels of TSNA permilligram of nicotine, suggesting that setting national levels for these com-pounds might be reasonable for countries that do not have a high fraction ofblended cigarettes on their market.

Variation across data sets

A further useful formulation of the variation in toxicant levels is a comparisonof the mean levels of toxicants in different data sets. This defines differencesin yield from the modified intense smoking regimen that result from differ-ences in the design of cigarettes offered for sale in different markets. Thesedifferences may allow identification of toxicants that can be lowered bychanging cigarette design or blends. If the mean level of toxicant per mil-ligram of nicotine in brands sold in one market, or by one manufacturer,differs substantially from that in other markets or from other manufacturers,it is clearly possible to manufacture cigarettes that yield lower levels of thattoxicant. This kind of evidence can be used to define targets for toxicant levelsper milligram of nicotine that might be set by regulators as mandatory goals,as it defines the levels that can be achieved by different cigarette designs morebroadly than with a single market or a single manufacturer.

77

Figure 3.8 presents the mean levels of each toxicant per milligram of nicotineacross brands for the Canadian data, the Canadian data minus French and USbrands and the Australian data. The means are presented as the ratio of themean value for the toxicant per milligram of nicotine in the specified data setto the mean levels reported by Counts et al. (2005) for the international PhilipMorris brands. The variation in toxicant levels between the data sets is aslarge as, and sometimes exceeds, the variation in levels across brands withina data set. The differences between the Philip Morris international brands andthe Canadian and Australian brands are not limited to the commonly recog-nized differences in TSNA levels, and the magnitude of the variationssuggests that examination of differences in toxicant yields from brands soldin different markets could provide better understanding of the levels of indi-vidual toxicants that could be achieved by changing cigarette design andmanufacturing practices.

Figure 3.8Ratios of means for constituents of brands in Australian and two Canadian samplesto the mean found in the sample of Counts et al. (2005)

0

5.0

1

5.1

2

5.2

Car

bon

mon

oxid

e A

mm

onia

1-A

min

onap

htha

lene

2-A

min

onap

htha

lene

3-A

min

obip

heny

l 4-

Am

inob

iphe

nyl

Ben

zo[a

]pyr

ene

Form

alde

hyde

A

ceta

ldeh

yde

Ace

tone

Acr

olei

nP

ropi

onal

dehy

de

Cro

tona

ldeh

yde

But

yral

dehy

de

Hyd

roge

n cy

anid

e M

ercu

ryLe

adC

adm

ium

Nitr

ic o

xide

N

Ox

NN

NN

NK

N n

itros

oana

tabi

ne

N n

itros

oana

basi

ne

Pyr

idin

e Q

uino

line

Hyd

roqu

inon

e R

esor

cino

lC

atec

hol

Phe

nol

m-+

p-C

reso

lo-

Cre

sol

1,3

But

adie

ne

Isop

rene

Acr

ylon

itrile

B

enze

neTo

luen

eS

tyre

ne

Met

hyl e

thyl

ket

one

Rat

io

tneutitsnoC

naemnailartsuAot)5002(.latestnuoCninaemfooitaR

lla(naemnaidanaCot)5002(.latestnuoCninaemfooitaR)sdnarb

dnarbnaemnaidanaCot)5002(.latestnuoCninaemfooitaRsdnarbNNNhgihsunim

NNN, N´-nitrosonornicotine

3.6 Recommended toxicants for reporting and regulation

A list of toxicants with high priority for consideration was identified on thebasis of the criteria described above (Table 3.6). Benzo[a]pyrene was

78

included despite its low toxicant animal carcinogenicity index because it is aproxy for a family of PAHs found in smoke and because there is a wealth ofevidence establishing the carcinogenicity of many of these PAHs. The toxi-cant 4- aminobiphenyl was added because it is a human carcinogen, althoughexperimental data did not allow a proper T25 calculation. Isoprene was omit-ted because of its structural similarity to 1,3-butadiene and its lower toxicantanimal carcinogenicity index. NNK and NNN were included as they had al-ready been identified in the first report (WHO, 2007) on mandating reductionsin toxicant yields. Crotonaldehyde was included because of its reactive

, -unsaturated aldehyde structure, although a tolerable level value was lack-ing. CO was also included even though it has a relatively low toxicant non-cancer response index, as it is thought to be mechanistically related tocardiovascular disease.

Table 3.6Initial list of priority toxicants

Toxicant Comments

Acetaldehyde TACI 6.1, TNCRI 67.1Acrolein TNCRI 1099Acrylonitrile TACI 1.4, TNCRI 2.14-Aminobiphenhyl Human carcinogen, but no experimental data for T25

calculation2-Aminonaphthalene TACI 0.68Benzene TACI 2.6Benzo[a]pyrene TACI 0.011,3-Butadiene TACI 9.9, TNCRI 2.6, Group 1, Human carcinogenCadmium TACI 1.7, TNCRI 2.6Carbon monoxide Relatively low TNCRI, but mechanistically related to

endothelial dysfunction in cardiovascular diseaseCatechol TACI 0.58Crotonaldehyde Aldehyde with reactive alkene structure, no threshold

limit value, inadequate evidence for carcinogenicityFormaldehyde TNCRI 19.8Hydrogen cyanide TNCRI 17.2Hydroquinone TACI 1.2Nitrogen oxides TNCRI 3.1NNN Identified by WHO (2007)NNK Identified by WHO (2007)

TACI, toxicant animal carcinogenicity index; TNCRI, toxicant non-cancer response index; NNN,N´-nitrosonornicotine; NNK, 4-(N-nitrosomethylamino)-1-(3-pyridyl)-1-butanone

79

The working group also considered other measured toxicants for which noT25 or tolerable level values were available. There are several metals incigarette smoke (arsenic, cadmium, chromium, nickel, lead, mercury, sele-nium) that occur either at low levels or could not be quantified in the threedata sets. Although these metals are quite toxic, only cadmium was found topresent at levels that contribute appreciably to the toxicant animal carcino-genicity and toxicant non-cancer response indices, and so was included.

The preliminary list of 16 toxicants in Table 3.6 satisfied the criteria for suf-ficient evidence of carcinogenicity or known respiratory or cardiac toxicity,variation in levels in different brands in different countries, being readilymeasurable and with possibilities for lowering yields in a product.

This initial list of 16 priority toxicants plus NNN and NNK was then discussedby the group at length to reduce it to a number that would be considered tobe manageable by regulators. As an overall expert assessment, it was con-cluded that seven (in addition to NNN and NNK) compounds are the mosthazardous toxicants in cigarette smoke for which there is potential for re-duction in cigarette smoke emissions, which represent different chemicalfamilies of toxicants and different phases of smoke and which show lung andcardiovascular toxicity as well as carcinogenicity. The seven toxicants are:

acetaldehyde

acrolein

benzene

benzo[a]pyrene

1,3-butadiene

carbon monoxide and

formaldehyde.

The remaining toxicants in the initial list of 16 should be measured andreported.

3.6.1 Toxicity of the selected constituents

Compounds proposed for regulation

Acetaldehyde: The concentrations of acetaldehyde in mainstream cigarettesmoke are similar to those of nicotine, more than 1000 times that of totalcarcinogenic TSNA and about 100 000 times the concentration of benzo[a]pyrene, although acetaldehyde is a weaker carcinogen. Acetaldehyde is found

80

widely in the diet and the environment and is the major metabolite of ethanol.It is also implicated as an enhancer of behavioural, endocrine and neuronalresponses to nicotine in rats (Cao et al., 2007). Feron et al. (1991) demon-strated the carcinogenicity of acetaldehyde administered by inhalation to ratsand hamsters (IARC, 1999a). Acetaldehyde is widely considered to be a ma-jor cause of head-and-neck cancers related to alcohol consumption (IARC,1999a; Brennan et al., 2004; Baan et al., 2007). The genetic effects of ac-etaldehyde have been summarized (IARC, 1999a). Positive responses havebeen obtained in a variety of asays for genotoxicity in eukaryotes. DNAadducts of acetaldehyde have been detected in leukocytes of smokers, whichdecrease upon smoking cessation (Chen et al., 2006). There is inade-quate evidence for the carcinogenicity of acetaldehyde in humans (IARC,1999a). Acetaldehyde has been evaluated as ‘possibly carcinogenic tohumans’ (Group 2B) by an IARC working group (IARC, 1999a) and as ‘rea-sonably anticipated to be a human carcinogen’ by the US Department ofHealth and Human Services (2004b).

Acrolein: Acrolein is a commonly occurring combustion product to whichthere is occupational and environmental exposure via inhalation (IARC,1995a). It is an intense irritant, and displays a range of toxic effects includingciliotoxicity (IARC, 1995a; Kensler, Battista, 1963). It causes eye and res-piratory irritation in humans, andepeated inhalation results in changes in theupper and lower respiratory tract. In dogs, acute congestion, changes in bron-chiolar epithelial cells and emphysema were found after inhalation of acrolein(IARC, 1995a). Acrolein potentiates the contractile response of passivelysensitized human bronchi to specific antigen stimulation (Roux et al., 1999).Acrolein has been identified as a hazardous air pollutant that is of particularconcern for people with asthma (Leikauf, 2002). With the exception of onestudy in which bladder tumours were produced in an initiation–promotionprotocol in rats treated with acrolein (by intraperitoneal injection) followedby dietary uracil, no carcinogenic effects of acrolein were reported (Cohenet al., 1992). Acrolein as a metabolite is widely considered to be responsiblefor the bladder toxicity of the chemotherapeutic agent cyclophosphamide.Acrolein–DNA adducts have been identified in a variety of human tissues,and their levels are higher in oral tissue from smokers than from non-smokers(Chung, Chen, Nath, 1996; Nath et al., 1998). The pattern of mutations in theP53 tumour suppressor gene in lung tumours from smokers is similar to thepattern of DNA damage caused by acrolein (Feng et al., 2006). An IARCworking group concluded that acrolein was not classifiable with respect to itscarcinogenicity to humans (Group 3) (IARC, 1995a).

Benzene: Humans are exposed to benzene mainly through inhalation of am-bient air. Sources of benzene include automobile engine exhaust, industrialemissions and evaporation from fuel during gasoline self-service. There are

81

also small amounts in food. Cigarette smoke is a major source of benzene,both for smokers and for non-smokers exposed to second-hand smoke (De-partment of Health and Human Services 2004b). Benzene induces chromo-somal aberrations, micronuclei and sister chromatid exchanges in bone-marrow cells of mice (IARC, 1987). Benzene causes multiple types oftumours in both rats and mice exposed by various routes, including oral ad-ministration, inhalation, injection and dermal application (IARC, 1987; De-partment of Health and Human Services, 2004b). Cohort epidemiologicalstudies studies of occupational exposure demonstrate that benzene causesacute and chronic myelogenic leukaemias (IARC, 1987; Department ofHealth and Human Services, 2004b). Benzene is classified as carcinogenicto humans by both IARC (Group 1) (IARC, 1987) and the US Department ofHealth and Human Services (2004b) and is probably a cause of leukaemia insmokers (IARC, 2004).

Benzo[a]pyrene: Benzo[a]pyrene is a prototypic representative of the PAHclass of carcinogens, which are products of the incomplete combustion oforganic matter and occur in cigarette smoke and the general environment asa mixture. Benzo[a]pyrene is always a component of this mixture and is con-sidered a surrogate for carcinogenic PAHs. Considerable evidence supportsan important role for PAHs as causes of lung cancer and other cancers insmokers (Hecht, 1999; Pfeifer et al., 2002; Hecht, 2003). CarcinogenicPAHs in cigarette smoke include benzo[a]pyrene, benz[a]anthracene,methylchrysenes, benzofluoranthenes, indeno[1,2,3-cd]pyrene and dibenz[a,h]anthracene, with total concentrations at least 5–10 times that of benzo[a]pyrene (IARC, 2004). Certain PAHs, including some of those in cigarettesmoke, are potent locally acting carcinogens that induce lung and trachealtumours in rodents and skin tumours in mice. Fractions of cigarette smokecondensate enriched in PAHs are carcinogenic on mouse skin (Hecht, 1999).The uptake of PAHs by smokers has been clearly demonstrated, and there issolid evidence for the presence of DNA adducts derived from benzo[a]pyrenein lung tissue from some smokers (Hecht, 2002; Boysen, Hecht, 2003; Belandet al., 2005). The pattern of mutations in the P53 tumour suppressor gene inlung tumours from smokers is similar to the pattern of DNA damage causedin vitro by diol epoxide metabolites of PAHs and by benzo[a]pyrene in cellculture (Denissenko et al., 1996; Smith et al., 2000; Pfeifer et al., 2002;Tretyakova et al. 2002). The pattern of KRAS oncogene mutations observedin lung tumours from smokers is also similar to that found in lung tumoursfrom animals treated with PAHs such as benzo[a]pyrene (Westra et al., 1993;Mills et al., 1995; Nesnow et al., 1998; Ahrendt et al., 2001). Collectively,these observations strongly support the role of PAHs as causes of lung cancerand perhaps other cancers in smokers. An IARC working group classifiedbenzo[a]pyrene as carcinogenic to humans (Group 1) (Straif et al., 2005).

82

1,3-Butadiene: Other than through cigarette smoking, exposure to1,3-butadiene occurs mainly in occupational settings and predominantly byinhalation (IARC, 1999c; Department of Health and Human Services 2004f).It is used as an intermediate in the manufacture of polymers and copolymersand in synthetic rubber production. 1,3-Butadiene is a multi-organ carcinogenin mice and rats. The sites of tumour induction in mice include thehaematopoietic system, heart, lung, forestomach, Harderian gland, preputialgland, liver, mammary gland, ovary and kidney, while in rats tumours havebeen observed in pancreas, testis, thyroid gland, mammary gland, uterus andZymbal gland (IARC, 1999a; Department of Health and Human Services2004a). Hprt mutations have been observed in lymphocytes isolated frommice exposed to 1,3-butadiene or its epoxide metabolites, and in workersexposed occupationally (Department of Health and Human Services, 2004b).Epoxide metabolites of 1,3-butadiene form DNA adducts, including cross-links (Park, Tretyakova, 2004). These metabolites are formed in rodents andhumans (Department of Health and Human Services, 2004b). The US De-partment of Health and Human Services (2004b) ranks 1,3-butadiene as a‘known human carcinogen’, and an IARC working group declared it a Group1 human carcinogen (IARC, in press).

Carbon monoxide: Half of the carbon monoxide produced is due to incom-plete combustion of organic material, automobile exhaust being a majorsource. CO is formed endogenously from catabolic degradation of haem, thecarboxyhaemoglobin level being about 0.7%. CO competes with oxygen forbinding to haemoglobin with an affinity that is 250 times greater than that ofoxygen. CO impairs the release of oxygen from haemoglobin, resulting intissue hypoxia. CO has neurotoxic effects that may be related to the releaseof nitric oxide. Exposure to CO at a level of 2000 mg/l of air for 3–4 h resultsin 70% carboxyhaemoglobin and death. Acute CO-related symptoms are un-likely to occur in smokers, who usually have carboxyhaemoglobin levels ofabout 5–6% (Scherer, 2006). CO in smokers is believed to reduce oxygendelivery, cause endothelial dysfunction and promote the progression ofatherosclerosis and other cardiovascular diseases (Department of Health andHuman Services, 2004a; Ludvig, Miner, Eisenberg, 2005; Scherer, 2006).

Formaldehyde: Exposure to formaldehyde occurs through endogenousmetabolism and in a variety of occupational and environmental settings inaddition to cigarette smoke. Formaldehyde is genotoxic in multiple in vitrosystems and in exposed humans and laboratory animals. Inhalation exposureto formaldehyde induces squamous cell carcinomas of the nasal cavities inrats, while studies in mice and hamsters by inhalation and by other routes ofadministration produced mixed results. Epidemiological studies of the rela-tion of occupational exposure to formaldehyde and cancer at various sitesprovide, according to an IARC working group (IARC, 2006), sufficient

83

evidence that formaldehyde causes nasopharyngeal cancer in humans, strongbut not sufficient evidence for a causal association between leukaemia andoccupational exposure to formaldehyde, and limited epidemiological evi-dence that formaldehyde causes sinonasal cancer in humans. Taking all theevidence together, the IARC working group concluded that formaldehyde iscarcinogenic to humans (Group 1) (IARC, 2006). The US Department ofHealth and Human Services (2004b) rates formaldehyde as ‘reasonably an-ticipated to be a human carcinogen’.

Compounds proposed for reporting

Acrylonitrile: Acrylonitrile is an important industrial chemical used in themanufacture of synthetic fibres, resins, plastics, elastomers and rubber (De-partment of Health and Human Services, 2004b). Inhalation and dermalexposure in occupational settings constitute the primary exposure circum-stances, in addition to cigarette smoke. Acrylonitrile induces tumours atmultiple sites, including the forestomach, central nervous system, mammarygland and Zymbal gland, in rats when administered orally or by inhalation.Increases in cancer frequency have not been found consistently in epidemi-ological studies of workers exposed to acrylonitrile (IARC, 1999a; Depart-ment of Health and Human Services, 2004b). Acrylonitrile is mutagenic insome genotoxicity assays (IARC, 1999a). It readily forms adducts with pro-teins, presumably through its epoxide metabolite, and acrylonitrile–haemoglobin adducts are found at higher levels in smokers than non-smokers(IARC, 1999a; Fennell et al., 2000). An IARC working group concluded thatacrylonitrile is possibly carcinogenic to humans (Group 2B) (IARC, 1999a),while the US Department of Health and Human Services (2004b) concludedthat it is ‘reasonably anticipated to be a human carcinogen’.

4-Aminobiphenyl: Commercial use and production of 4-aminobiphenyl havebeen widely discontinued, and cigarette smoking is presently probably themajor source of human exposure. 4- Aminobiphenyl induced bladder papil-lomas and carcinomas in rabbits and dogs after oral administration and causedneoplasms at various sites in mice, including angiosarcoma, hepatocellulartumours and bladder carcinomas. Subcutaneous administration to rats pro-duced tumours of the mammary gland and intestine. 4-Aminobiphenyl wasgenotoxic in short-term mutagenicity assays and formed DNA adducts in thebladder epithelium of dogs (IARC, 1987). 4-Aminobiphenyl– haemoglobinadducts have been quantified in smokers (Skipper, Tannenbaum, 1990) atlevels higher than in non-smokers, related to dose and decreasing upon smok-ing cessation (Maclure et al., 1990; Skipper, Tannenbaum, 1990; Castelao etal., 2001). Epidemiological studies conclusively demonstrate that 4-amino-biphenyl causes bladder cancer in humans (IARC, 1987; Department ofHealth and Human Services, 2004b). Both IARC and the US Department of

84

Health and Human Services classify 4-aminobiphenyl as carcinogenic to hu-mans (IARC Group 1) and as a probable cause of bladder cancer in smokers(IARC, 1987, 2004; Department of Health and Human Services, 2004b).

Cadmium: Exposure to cadmium can occur by inhalation of cadmium-containing particles in the ambient air or by consumption of food or drinking-water. Occupational and consumer exposure also occurs. The carcinogenicityof cadmium salts has been demonstrated in multiple studies in rats and otherspecies (IARC, 1999b; Department of Health and Human Services, 2004b).Inhalation of a variety of cadmium compounds caused lung tumours in rats,with a positive dose–response relation. Lung tumours were also induced byintratracheal administration of cadmium compounds to rats. When givenorally to rats, cadmium chloride caused dose-related increases in leukaemiaand testicular tumours. Cadmium induced DNA and chromosomal damagein several mammalian systems in vitro. Several epidemiological studies ofoccupationally exposed workers indicate that cadmium exposure is a causeof lung cancer. Cadmium is classified as carcinogenic to humans by bothIARC (Group 1) (IARC, 1999b) and the US Department of Health and HumanServices (2004b).

Catechol: Catechol occurs in various natural dietary items, such as fruits andvegetables. It may be released into the environment during chemical manu-facture, but it is not a common environmental pollutant. Catechol has not beenshown to induce tumours in mice but caused adenocarcinomas of the glan-dular stomach after oral administration to several strains of rat. It hasconsiderable co- carcinogenic activity on mouse skin (with benzo[a]pyrene)and in rat tongue, oesophagus and stomach. Catechol caused gene mutationsin mammalian cells in vitro. There are no available epidemiological data. AnIARC working group concluded that catechol is possibly carcinogenic to hu-mans (Group 2B) (IARC, 1999a).

Crotonaldehyde: Crotonaldehyde is a toxicant in gasoline and diesel engineexhaust and occurs in various foods (IARC, 1995b). It is also formedendogenously by lipid peroxidation (Chung, Chen, Nath, 1996).Crotonaldehyde induces neoplastic nodules of the liver after oral adminis-tration to rats, and it is mutagenic in various genotoxicity assays (IARC,1995b). Crotonaldehyde–DNA adducts have been detected in human tissues,including the oral mucosa and lungs of smokers (Chung et al., 1999; Zhanget al., 2006). These adducts can also be formed from acetaldehyde. An IARCworking group concluded that crotonaldehyde was not classifiable as to itscarcinogenicity to humans (Group 3) (IARC, 1995b).

Hydrogen cyanide: Cyanides occur as glycosides in various food productssuch as nuts and beans. Hydrogen cyanide and other cyanides are convertedmetabolically to thiocyanate, which occurs in cruciferous vegetables.

85

Hydrogen cyanide is very toxic, causing immediate death in humans after onoral dose of 0.5–3.5 mg/kg body weight (bw) or inhalation of 270 mg/l of airfor a few minutes. It acts by inhibiting cytochrome oxidase in the respiratorychain. Cigarette smoke can reduce detoxification of hydrogen cyanide, whichnormally occurs via formation of thiocyanate, leading to chronic hydrogencyanide exposure of smokers and consequent amblyopy, retrobulbar neuritisand sterility (Scherer, 2006). Hydrogen cyanide could be involved in im-paired wound healing in smokers (Silverstein, 1992).

Hydroquinone: Hydroquinone occurs as a conjugate in the leaves, bark andfruit of various plants and in some occupational settings, such as in photog-raphy. It is also used in the treatment of skin hyperpigmentation. Hydro-quinone is a metabolite of benzene and may be involved in leukaemiainduction by benzene. Hydroquinone was modestly carcinogenic in mice,inducing hepatocellular adenomas in females in one study and in males inanother. It caused renal tubule adenomas in male rats; it was devoid oftumour-promoting activity in most assays. There are no convincing epidemi-ological data implicating hydroquinone as a carcinogen. It is mutagenic in avariety of in vitro systems. Hydroquinone causes nephropathy in rats as wellas aplastic anaemia, liver cord-cell atrophy and ulceration of the gastric mu-cosa. An IARC working group concluded that hydroquinone is not classifi-able as to its carcinogenicity to humans (Group 3) (IARC, 1999a).Hydroquionone is classified in the European Union as a Category 3carcinogen (‘limited evidence of a carcinogenic effect’) (http://ecb.jrc.it/classification-labelling/).

2-Naphthylamine: Commercial use and production of 2-naphthylamine isbanned in several countries, and there is probably little exposure other thanfrom cigarette smoke (Department of Health and Human Services, 2004b).2-Naphthylamine is carcinogenic in many animal species. After oral admin-istration, it caused bladder neoplasms in hamsters, dogs and nonhumanprimates and liver tumours in mice. Bladder carcinomas have been observedin rats, and it gave positive results in the strain A/J mouse lung adenomabioassay. 2-Naphthylamine increased the incidence of sister chromatid ex-changes in mice and rabbits and was positive in a variety of other short-termassays for genotoxicity. It formed DNA adducts in bladder and liver cells ofdogs in vivo. Epidemiological studies conclusively demonstrate that2-naphthylamine causes bladder cancer in humans (IARC, 1987). Both IARC(1987) and the US Department of Health and Human Services (2004b) rank2- naphthylamine as carcinogenic to humans (IARC Group 1), and it is prob-ably a cause of bladder cancer in smokers (IARC, 2004).

Nitrogen oxides: Tobacco smoke contains nitric oxide, nitrogen dioxide andnitrous oxide, although freshly generated smoke contains almost exclusively

86

nitric oxide, while nitrogen dioxide is rapidly formed upon ‘ageing’ of thesmoke. The nitrate content of the tobacco determines the nitrogen oxide yieldsin smoke (IARC, 1986). Nitric oxide induces vasodilation and has been usedas a targeted pulmonary vasodilator in newborns (Weinberger et al., 2001).In aqueous solution, nitric oxide can decay to nitrite and nitrogen trioxide,which can nitrosate glutathione and amines. Nitrosation of amines can leadto the formation of carcinogenic nitrosamines. Such in vivo nitrosation hasbeen demonstrated in some studies of smokers (Hecht, 2002). Nitric oxidealso causes DNA strand breaks and base alterations, possibly leading directlyto mutagenesis and carcinogenesis. It reacts with superoxide to formperoxynitrite, an oxidant that can cause lipid peroxidation, may be involvedin DNA damage and could interfere with surfactant functioning (Weinbergeret al., 2001). Nitric oxide may contribute to nicotine addiction by increasingnicotine absorption, reducing symptoms of stress and increasing post-synap-tic dopamine levels (Vleeming, Rambali, Opperhuizen, 2002). Nitric oxideis readily oxidized to nitrogen dioxide, a pulmonary irritant. The WHO 1-hmean guideline value for nitrogen dioxide is 200 g/m3, and the annualmean guideline value is 40 g/m3 (http://www.who.euro.int/document/E87950.pdf).

3.6.2 Methods for measuring the toxicants

Sampling and reporting

The results of testing for toxicants should be reported yearly for each productfor regulation and reporting. Temporal and geographical differences, whichare important sources of product variability, should be included in any ex-amination of dissimilarities within and across products. Samples should becollected by the standard method described in ISO 8243. Cigarettes shouldbe obtained as part of a series of samplings at different times. The period forreporting (1 year) should be divided into at least five sub-periods. For eachbrand, 200 cigarettes (10 packs of 20 cigarettes per pack or equivalent) shouldbe taken in each sub-period. Each sample should be drawn from differentsampling points. Cigarettes for each analytical method should be assignedequally from each sub-period. Additional samples should be used if repeatedtesting is required due to any analytical problems. The results of the analysesof these samples should be combined for reporting purposes.

Samples can be taken from manufacturers or importers at their premises orfrom retail locations. Taking samples at business location allows easier accessto samples but presents other concerns. If the date of the visit is known, theproducts may be altered or chosen to ensure that they meet regulatory limits.If sampling at a business location is used, samples must be collected withoutprior notification, by independent investigators. Under no circumstances

87

should regulators allow samples to be delivered by the regulated business foranalysis. Obtaining samples at retail locations requires additional resources,but this approach ensures that the products will be those that are directlyavailable to smokers. Purchasing from multiple retail locations can be par-ticularly challenging if the country is large, has limited transport infrastruc-ture or comprises multiple islands or geographical regions that could havedifferent products that are difficult to access. In this case, regulatory author-ities may determine that broader sampling is required to obtain a representa-tive sample. Decisions about where to obtain samples are best taken bygovernment officials in each country.

In order to reduce variation in a product due to storage conditions, samplesshould be conditioned for at least 24 h before machine smoking according tothe ISO 3402 standard.

Analytical methods

Cigarettes should be analysed by the machine smoking method commonlyknown as the Canadian intense method, with a 55-ml puff volume, 2.0-s puffduration, 30-s puff interval and 100% vent blocking. The butt length is23 mm for non-filter cigarettes or the length of filter paper plus 3 mm forfilter brands. Smoke is collected on Cambridge filter pads in most cases.Alternative situations are described below.

Individual measurements are made for each pad, and the number of padsanalysed, the mean, the standard deviation and the minimum and maximumare reported for each brand. The dates and locations where each sample wastaken are reported with the final results.

Nicotine and carbon monoxide: Nicotine and CO are measured by ISO meth-ods 10315 and 8454, respectively, which have been well validated and are inuse in many laboratories worldwide. Cigarettes are smoked with the Canadianintense smoking regimen in standard commercial smoking machines, whichare generally equipped with an automated CO analyser as an option. In linearsmoking machines, three cigarettes are smoked on each of 20 Cambridgefilter pads. In rotary smoking machines, 10 cigarettes are smoked on each offive pads. Care must be taken that breakthrough of the pads does not occur.Nicotine is measured in the cigarette smoke particulate matter that is collectedon Cambridge filter pads, and CO is measured in the gas phase, collected ina vapour phase collection bag. The gas phase is automatically passed througha non-dispersive infra-red analyser, and the percentage CO is determined bycomparison with known standard CO gas concentrations. The Cambridgefilter pad is extracted with isopropanol containing anethole as the internalstandard. The extract is analysed for nicotine by packed column gas chro-matography with flame ionization detection. The gas chromatography

88

response is compared with that of known standard solutions to determinethe amount of nicotine in each pad. Details of this official Health Canadamethod can be found at http://www.qp.gov.bc.ca/stat_reg/regs/health/oic_94.pdf (p. 165; accessed 21 December 2006). The official ISO methodcan be obtained at http://www.iso.ch/iso/en/cataloguelistpage.cataloguelist?commid=3350&scopelist=all.Tobacco-specific nitrosamines: There is currently no ISO-validated methodfor measuring TSNA in mainstream tobacco smoke, but an ISO method isbeing developed. A method that has been in use in many tobacco testinglaboratories and is widely used to measure these compounds is gas chro-matography with thermal energy analysis. This technique is selective fornitroso-containing compounds and has been successfully used for many yearsfor measuring TSNA in tobacco smoke. In linear smoking machines, fivecigarettes are smoked on each of 20 Cambridge filter pads. In rotary smokingmachines, 20 cigarettes are smoked on each of five pads. TSNA are measuredin the mainstream cigarette smoke particulate that is captured on Cambridgefilter pads and are concentrated by extraction with dichloromethane, followedby column chromatography onto basic alumina. The fraction containing theTSNA is eluted and analysed by gas chromatography–thermal energy anal-ysis. A description of this official Health Canada method can be found athttp://www.qp.gov.bc.ca/stat_reg/ regs/health/oic_94.pdf (p. 119; accessed21 December 2006). Alternative methods, including liquid chroma-tography with tandem mass spectrometry (http://www.aristalabs.com/pdf/mainstreamanalysis.pdf (accessed 26 April 2006) or that described by Wu,Ashley and Watson (2003), have also been used, but any alternative techniqueused must be shown to be as accurate and reproducible as gaschromatography–thermal energy analysis before use. N´-Nitrosoanatabineand N´-nitrosoanabasine can also be determined with this method at minimaladditional cost.Carbonyls (formaldehyde, acetaldehyde, acrolein, crotonaldehyde): There iscurrently no ISO- validated method for measuring carbonyls in mainstreamtobacco smoke, but a method has been used successfully in many laboratoriesto measure these compounds. Two cigarettes are smoked in each of 25 smok-ing runs. The unfiltered mainstream cigarette smoke is bubbled throughimpingers containing 2,4-dinitrophenylhydrazine in acidified acetonitrile. Analiquot of the extract is filtered, diluted and subjected to reversed-phase gra-dient liquid chromatography with ultraviolet detection. Levels are quantifiedby external standard procedures. Care must be taken that breakthrough of theimpinger solution does not occur; if this occurs, the number of cigarettestested must be decreased. Retention times and analysis of pure compoundsare used to verify the identification of peaks of interest. Details of this officialHealth Canada method can be found at http://www.qp.gov.bc.ca/stat_reg/regs/health/oic_94.pdf (p. 41; accessed 21 December 2006). Acetone,

89

propionaldehyde and butyraldehyde can also be determined with this methodat minimal additional cost.

Volatiles (1,3-butadiene, acrylonitrile, benzene): There is currently no ISO-validated method for measuring volatiles in mainstream tobacco smoke, buta method has been used successfully in many laboratories to measure thesecompounds. The unfiltered mainstream cigarette smoke is bubbled throughcryogenic traps containing purge-and-trap grade methanol, withbenzene-D6 as an internal standard. Five cigarettes are smoked through eachof 10 traps. An aliquot of the solution is injected into a gas chromatograph–mass spectrometer for analysis. Quantification is carried out by comparingrelative peak areas to those of standard solutions of known concentrations ofthe analytes of interest. Retention times and analysis of pure compounds areused to verify the identification of peaks of interest. Details of this officialHealth Canada method can be found at http://www.qp.gov.bc.ca/stat_reg/regs/health/oic_94.pdf (p. 173; accessed 21 December 2006). Isoprene,toluene and styrene can also be determined with this method at minimal ad-ditional cost.

Benzo[a]pyrene: There is currently no ISO-validated method for measuringbenzo[a]pyrene in mainstream tobacco smoke, but an ISO method is beingdeveloped. A method has nevertheless been used successfully in many lab-oratories to measure this compound. In linear smoking machines, twocigarettes are smoked on each of 25 Cambridge filter pads. In rotary smokingmachines, 10 cigarettes are smoked on each of five pads. The mainstreamcigarette smoke particulate is captured on Cambridge filter pads, but caremust be taken that breakthrough of the pad does not occur. Particulate matteris extracted from the pad with cyclohexane. A portion of this extract is passedthrough silica and NH2 (amino) cartridges. Benzo[a]pyrene is eluted from thecartridge with hexane, blown down to dryness and reconstituted inacetonitrile. The sample is subjected to reverse-phase liquid chroma-tography and quantified by fluorescence detection. A description of thisofficial Health Canada method can be found at http://www.qp.gov.bc.ca/stat_reg/regs/health/oic_94.pdf (p. 27; accessed 21 December 2006). Alter-native methods, including gas chromatography–mass spectrometryhttp://www.aristalabs.com/pdf/mainstreamanalysis.pdf (accessed 26 April2006) or that described by Ding and colleagues (2005), have also been used.An ISO method being developed by ISO involving gas chromatography–mass spectrometry can be purchased (http://www.iso.ch/iso/en/cataloguelistpage.cataloguelist?commid=3350&scopelist=all). These gaschromatography–mass spectrometry methods can also be used to measure aseries of PAHs other than benzo[a]pyrene.

90

Aromatic amines (2-aminonaphthalene, 4-aminobiphenyl): There is currentlyno ISO-validated method for measuring aromatic amines in mainstream to-bacco smoke, but a method has been used successfully in many laboratories.In linear smoking machines, five cigarettes are smoked on each of 20Cambridge filter pads. In rotary smoking machines, 20 cigarettes are smokedon each of five pads. Care must be taken that breakthrough of the Cambridgefilter pad does not occur. The pad is extracted with 100 ml of 5% hydrochloricacid solution. The internal standard (D9-aminobiphenyl) is spiked into thesolution. The extract is washed with basic dichloromethane and extractedwith hexane. Sodium sulfate is used to dry the extracts, and the aromaticamines are derivitized with pentafluoropropionic acid anhydride andtrimethylamine. The resulting material is passed through a Florisil column.Aromatic amines are quantified by gas chromatography–mass spectrometry.A description of this official Health Canada method can be found at http://www.qp.gov.bc.ca/stat_reg/regs/health/oic_94.pdf (p. 16; accessed 21December 2006). 1-Aminonaphthalene and 3-aminobiphenyl can also be de-termined with this method at minimal additional cost.

Hydrogen cyanide: There is currently no ISO-validated method for measuringhydrogen cyanide in mainstream tobacco smoke, but a method has been usedsuccessfully in many laboratories. While two cigarettes are smoked, the par-ticulate is collected on a Cambridge filter pad, and the gas phase is trappedin an impinger containing 0.1N sodium hydroxide directly behind the pad.Twenty-five smoke sampling runs are carried out. The Cambridge filter padis extracted with 0.1N sodium hydroxide. Both the pad extract and theimpinger solution are analysed by continuous flow colorimetric analysis. Theanalysis step involves conversion of cyanide to cyanogen chloride and reac-tion with a pyrazolone reagent. Quantification is accomplished by compari-son with the response of known standards. Details of this official HealthCanada method can be found at http://www.qp.gov.bc.ca/stat_reg/regs/health/oic_94.pdf (p. 63; accessed 21 December 2006).

Cadmium: There is currently no ISO-validated method for measuring cad-mium in mainstream tobacco smoke, but methods have been used success-fully in many laboratories. The most challenging aspect of the analysis ofheavy metals is collecting the sample without contamination. The Cambridgefilter pads used in the usual mainstream smoke collection methods can besignificantly contaminated, invalidating the measurement. Scientists haveturned to alternative means of collecting smoke. Once the smoke has beencollected, a number of equally valid analytical instruments can be used forquantification. Ten cigarettes are smoked for each determination, and theparticulate is collected by electrostatic precipitation to avoid contamination.Ten separate smoking runs are carried out. The total particulate matter in theprecipitator is removed by washing with methanol. The methanol is removed

91

by gentle heating and blow-down. The remaining material is microwave-digested with hydrochloric acid, nitric acid and hydrogen peroxide. Theresulting solution is analysed by atomic absorption spectroscopy, inductivelycoupled argon plasma–atomic emission spectroscopy or inductively coupledargon plasma–mass spectrometry. Quantification is accomplished by com-parison with the response of known standards. Details of this official HealthCanada method can be found at http://www.qp.gov.bc.ca/stat_reg/regs/health/oic_94.pdf (p. 85; accessed 21 December 2006). Other heavy metals,such as nickel, lead, chromium, arsenic and selenium, can also be determinedwith this method. A similar method can be found at http://www.aristalabs.com/pdf/mainstreamanalysis.pdf (accessed 26 April 2006).An alternative means of collecting smoke particulate has been described byPappas et al. (2006). While requiring specially prepared filters, it does allowthe use of normal particulate collection procedures instead of an electrostaticanalyser.

Nitrogen oxides: There is no ISO-validated method for measuring oxides ofnitrogen in mainstream tobacco smoke, but a method has been used success-fully in many laboratories. One cigarette is smoked in the Canadian intensesmoking regimen, with 25 smoke sampling runs. The unfiltered smoke isallowed to mix in a smoke-mixing chamber and is then pulled, at regularintervals, into a chemiluminescence nitrogen oxide analyser for measurementof total nitrogen oxides. By reacting the smoke with ozone before determi-nation, the amount of nitric oxide can be determined. The difference in levelsof nitrogen oxides and nitric oxide is calculated as nitrogen dioxide. Quan-tification is accomplished by comparison with the response of known stan-dards. Details of this official Health Canada method can be found at http://www.qp.gov.bc.ca/stat_reg/regs/health/oic_94.pdf (p. 103; accessed 21December 2006).

Phenols (catechol, hydroquinone): There is currently no ISO-validatedmethod for measuring phenols in mainstream tobacco smoke, but a methodhas been used successfully in many laboratories. In linear smoking machines,two cigarettes are smoked on each of 25 Cambridge filter pads. In rotarysmoking machines, 10 cigarettes are smoked on each of five pads. The main-stream cigarette smoke particulate is captured on Cambridge filter pads; caremust be taken that breakthrough of the pad does not occur. The pad is ex-tracted with 40 ml of 1% acetic acid. An aliquot of the extract is filtered andsubjected to reverse-phase gradient liquid chromatography with selectivefluorescence detection. Phenols are quantified by gas chromatography–massspectrometry. A description of this official Health Canada method can befound at http://www.qp.gov.bc.ca/stat_reg/regs/health/oic_94.pdf (p. 153;accessed 21 December 2006). Resorcinol, phenol, m-cresol, o-cresol andp-cresol can also be determined with this method at minimal additional cost.

92

3.6.3 Existing techniques to reduce specific toxicant emissions in cigarettesmoke

An additional consideration in setting levels of toxicants per milligram ofnicotine is the ability of the tobacco industry to modify their products tocomply with lower toxicant levels. Examining the published literature andpublicly available tobacco industry documents and patents can help deter-mine the extent to which existing technology can be used to reduce theemissions of the specified toxicants in cigarette smoke. An important caveatis that there may be additional evidence of industry capacity to reduce toxicantlevels that is not yet publicly available.

Most of the data cited in publicly available reports and documents were ob-tained by machine smoking with the ISO regimen rather than with themodified intense smoking regimen recommended by TobReg. Nevertheless,the ISO data may provide some insight into the general design changes thatwould be required. Where possible, attempts have been made to standardizereductions per milligram of tar or nicotine or per unit volume. Most of thedata are presented as emissions per stick, but it is important to note that com-parisons of per-stick emissions can be highly misleading unless identicalproducts are being compared.

NNN and NNK: The evidence for differences in TSNA levels in differentcountries was reviewed above. Different approaches have been reported inthe literature for reducing the TSNA levels in mainstream smoke. These fallinto three broad categories: changing agricultural practices, curing and to-bacco blending.

The existing evidence indicates that reducing the use of nitrate fertilizerswould reduce TSNA in all types of tobacco cured by any method.

Changing the heating source for flue-curing from propane appears to havereduced TSNA in tobacco cured in this way. Air curing in well-ventilatedstructures can also lower TSNA in these types of tobacco. In general, coolertemperature and lower relative humidity at the time of curing favourthe production of air-cured tobaccos with low leaf TSNA concentrations(1.5–3.5 g/g), provided that the variety has a low nornicotine content,nitrogen fertilization is moderate, curing is performed in a well-ventilatedenvironment, the tobacco is taken down and stripped as soon as it is cured,and the bales are stored as briefly as possible before the leaves are threshedand stabilized.

Using tobaccos with low TSNA contents would be effective for reducing thecontent of TSNA in mainstream smoke. The simplest way of reducing TSNAemissions would be to use blends that consist primarily of bright, flue-curedtobacco, as is the tradition in Australia and Canada, where the levels are up

93

to 10-fold lower than those in US blended cigarettes, which contain air-curedburley and other tobaccos. As new curing and other techniques reduce pre-formed (i.e. in the blend) TSNA, however, formation by pyrosynthesis willbecome the predominant source. Nitrosation of nicotine, a tertiary amine,during smoking occurs at a much slower rate than that of nornicotine, a sec-ondary amine. This will favour formation of NNN, N´-nitrosoanatabine andN´-nitrosoanabasine rather than NNK during smoking.

There is little evidence that specialized filters, additives, paper type or poros-ity or other physical parameters significantly affect TSNA in mainstreamsmoke, apart from their general effects on overall smoke concentration.Therefore, no engineering changes to products were identified that wouldspecifically affect TSNA concentration. Irwin (1990), in an internal reviewfor British American Tobacco, reported that filtration efficiency for NNN andNNK varied closely with total particulate matter and nicotine, suggesting noselectivity.

Carbonyl compounds (acetaldehyde, acrolein, formaldehyde): These com-pounds are grouped as they share similar molecular structures and properties.In theory, therefore, the methods for their reduction in smoke should be sim-ilar or related.

Acetaldehyde appears to derive from combustion of polysaccharides, includ-ing cellulose and added sugars. A number of options have been explored bythe tobacco industry to reduce acetaldehyde emissions.

Reducing the concentration of sugars added to tobacco or using blends thatcontain fewer sugars should reduce acetaldehyde levels. Norman (1999)summarized research on tobacco type, reconstituted sheet and expanded to-bacco. On the basis of amount per gram of tobacco burnt, burley tobaccodelivered more than bright; reconstituted tobacco was not significantly dif-ferent from strip tobacco; and expanded tobacco delivered slightly moreacetaldehyde. This would suggest that reducing the amount of burley andexpanded tobacco in a blend could reduce acetaldehyde yields. Rodgman(1977) noted that increasing the amount of G13 expanded tobacco in a blendyielded less acetaldehyde per cigarette but slightly increased the acetaldehydeper puff.

Because acetaldehyde is a component of the vapour phase that has a lowrelative molecular mass, it does not affect cellulose acetate fibres or othermechanical filtration devices. Despite a relatively short contact time(< 100 ms), certain volatile components in cigarette smoke are selectivelyremoved by activated charcoal, which has a porous structure and a high sur-face area (e.g. Tiggelbeck 1967, 1976). In mainstream smoke deliveries,charcoal-filtered cigarettes have been shown to significantly reduce manyvolatile components, including acetaldehyde (e.g. Xue, Thomas, Koller,

94

2002). The amount of volatile compounds removed varies directly with theamount of charcoal present in the filter (Norman, 1999). Polzin et al. (2008)reported on three charcoal filtered brands with different levels of charcoal (0,45, 120 and 180 mg) but similar FTC tar delivery (5–6 mg) and other designparameters. When normalized for nicotine and smoked under Health Canadaintense conditions, the filter with the highest carbon content resulted in agreater reduction in acetaldehyde than the non-charcoal brand (brand 1 inFigure 3.9). The trend in the Canadian intense smoking data shows little dif-ference in levels until brand 5, which may suggest that substantial amountsof charcoal (180 mg) are required to reduce acetaldehyde significantly. Thelarge difference between the ISO and Health Canada levels for the 120-mgcarbon filter (brand 4) suggests that, at least for acetaldehyde, the higherpuffing intensity or vent blocking may overwhelm even 120 mg of charcoalon a filter. These data suggest that a filter containing at least 180 mg of char-coal could reduce the acetaldehyde concentration of smoke by approximately45% under intense smoking conditions.

Figure 3.9Acetaldehyde per milligram of nicotine in different brands machine smoked in theISO and Health Canada intense regimens

Brand

1200.0

1000.0

800.0

600.0

400.0

200.0

0.01 2 3 4 5

Ace

tald

ehyd

e (u

g/m

g ni

cotin

e)

HC ISO

Rodgman (1977) reported the results of an RJ Reynolds study showing thata combination of charcoal and bauxites removed up to 30% of acetaldehydefrom smoke. As noted in this internal memo, however, “Charcoal-filteredcigarettes with substantially reduced aldehydes (as well as other gas-phasecomponents) have generally not fared well in the [US] marketplace. Gener-ally, charcoal filtration does not affect FTC ‘tar’ yield”. Laugeson and Fowles

95

(2005) reported no difference in acetaldehyde levels from two Mild Sevenversions containing charcoal and from a similar brand without carbon underCanadian intense conditions. In fact, when normalized for nicotine, the MildSeven Light actually showed the highest level of acetaldehyde.

Philip Morris USA holds a patent (Rainer, Feins, 1980) for a filter claimedto remove aldehydes selectively from mainstream smoke “comprising a gran-ular carrier containing concentrated hydrogen peroxide, water, and a hy-drophilic stabilizer for said hydrogen peroxide.” A filter constructed in thismanner with silica gel granules was claimed to reduce the acetaldehyde levelby 78% as compared with a standard cellulose acetate filter and by approxi-mately 50% as compared with an activated charcoal filter.

Brown & Williamson made commercial use of Duolite in filters, particularlyon its Fact ‘low-gas’ cigarette introduced in the USA in the 1970s. An em-ployee’s hand-drawing of the resin’s structure is shown in Figure 3.10(Litzinger, 1972). Duolite is a “porous granular resin … reacts with hydrogencyanide and aldehydes removing them from smoke. The resin, partially com-bined with acetic acid to catalyze the removal of adlehydes was made toBrown & Williamson specifications by Diamond Shamrock. … Unlike char-coal, which is indiscriminate in that it also removes flavor compounds,Duolite filters smoke selectively” (Brown & Williamson Tobacco, 1977).Duolite resin contains primary, secondary and tertiary amine groups, andacetaldehyde is removed from smoke by addition to primary and secondaryamines; subsequent to capture, it may undergo further reactions. The authorsof the report on the effectiveness of Duolite noted that the efficiency of fil-tration depended on smoke pH, regardless of the amount of Duolite used.Brown & Williamson did extensive research on the resin’s properties, filtra-tion effectiveness and particle release during manufacture, packaging andsmoking (including inhalation and pyrolisis).

In 2002, Brown & Williamson introduced Advance Lights, featuring a‘Trionic’ filter, which comprised an ion-exchange resin, charcoal and cellu-lose acetate at the mouth end and claimed to provide “All of the taste … Lessof the toxins.” Little information is available about this product; however, itis described in some detail in the comments of Brown & Williamson in re-sponse to petitions from the Campaign for Tobacco Free Kids and the Societyfor Research on Nicotine and Tobacco for advance regulation by the US Foodand Drug Administration (Brown & Williamson, 2002):

“Smoke released from the burning tobacco first moves through the filter seg-ment containing the Ion Exchange Resin. This material interacts primarily withthe semi-volatile constituents of tobacco smoke, particularly … aldehydes (in-cluding formaldehyde) and others like hydrogen cyanide. Some of these mate-rials are trapped within the first segment, reducing their levels as the smokecontinues through the filer.

96

“Smoke next moves into the filter segment containing specialty carbon. Thismaterial adsorbs both semi-volatile constituents and gases. It continues to ad-sorb the aldehydes from the smoke, as well as trapping other constituents likebenzene and acrolein. … The result of smoke passing through the three-stageTrionic filter is a significant reduction in many of the toxins, as compared tothe levels found in smoke delivered by the leading lights cigarette brands.”

The acetaldehyde concentration, normalized for tar, does not appear to varyby cigarette circumference (Seeman, Dixon, Haussmann, 2002). Therefore,reducing the circumference of cigarettes would not appear to be an effectivereduction strategy. More permeable papers reduce volatile phase constituentsgenerally, but the reduction is less than 10% per cigarette across the range of20–100 CORESTA [Cooperation Centre for Scientific Research Relative toTobacco] units (Owens, 1978). Perforated cigarette papers would reducevapour phase constituents by up to 25% per cigarette across the range of 20–200 CORESTA units, with the greatest drop between 20 and 100 units.Rodgman (1977) noted that “increased porosity paper, perforations in thetobacco rod wrapper, and/or perforations in the filter wrapper” can decreasealdehydes generally, including acetaldehyde. This makes sense, given that itis a gas-phase constituent; thus, increasing the paper porosity would allowmore vapour constituents to leave the burning rod. No data could be foundon the effects of paper additives on the acetaldehyde concentration, althoughthe 1977 memo notes that slower-burning papers yield more acetaldehydeper cigarette, probably as a consequence of a higher puff count.

Figure 3.10Brown & Williamson employee’s hand drawing of Duolite structure

97

Acrolein is an aldehyde and is part of the gas phase of cigarette smoke.Therefore, design changes that would reduce acetaldehyde, as outlined in theprevious section, would also tend to reduce acrolein. Indeed, a chart in theRJ Reynolds memo (Rodgman, 1977) shows that, for increasing total air di-lution, reductions in acrolein parallel the reduction in acetaldehyde. Thisincludes filter ventilation effects, which would not be effective underCanadian intense smoking conditions, in which the filter vents are blocked.Reducing the use of humectants such as glycerol might also lower emissionsof acrolein.

Charcoal filters are known to reduce acrolein levels through adsorption.Polzin et al. (2008) show that acrolein can be reduced by as much as 69%under Canadian intense conditions (normalized for nicotine) when a high-carbon filter (120–180 mg charcoal) is used. Laugeson and Fowles (2005)reported a difference of approximately 15% in acrolein levels in Mild Sevenvarieties (which contain far less charcoal than the varieties tested by Polzinet al.) and a non-charcoal brand under Canadian intense condition. This heldtrue whether emissions were normalized for nicotine or not. In the 1970s,Lorillard explored an alternative means for reducing acrolein, by addition ofvarious reactive compounds to the filter or tobacco rod, including hydrazines,amines with high relative molecular mass, ammonium phosphate andammonium carbonate (Ihrig, 1972). By 1976, they had identified some ad-ditives that worked to reduce acrolein selectively, but none were patentable(Ihrig, 1976) and the work appears to have been abandoned.

Figure 3.11 shows the levels of acrolein per milligram of nicotine in differentbrands machine smoked in the ISO and Health Canada intense regimens.

Formaldehyde is known to be generated by both natural and added sugars(Baker, 2006). Therefore, reducing or eliminating exogenous sugars fromtobacco processing (e.g. casings) should reduce the formaldehyde yields ofcigarettes. Carbon filters have been shown to have some efficacy againstformaldehyde in smoke, like other gas-phase constituents. Laugeson andFowles (2005) reported an approximately 33% reduction in formaldehydefrom Mild Seven cigarettes when compared with a non-charcoal-filter com-parison brand under intense smoking conditions (normalized for nicotine).They reported a reduction of 75% in formaldehyde from a 180-mg charcoalmarket brand (Marlboro Ultrasmooth) compared to Marlboro regular whennot normalized for nicotine, but the reduction was only 51% when adjustedfor nicotine.

It has been shown that the presence of ammonium compounds and aminoacids inhibits the generation of formaldehyde from sugars (Baker, 2006).RJ Reynolds explored the reduction in formaldehyde in mainstream smokeby reactions with ammonia. In a memo, Furin and Gentry (1990) described

98

the results of a study of the effects of including ammonia sources (urea andglycine) in the tobacco rod on mainstream yields of various carbonyl com-pounds. Reductions of 70–98% were found, depending on the level of ureaor glycine added. They proposed that formaldehyde is reduced by reactionwith ammonia to form hexamethylenetetramine. The authors noted potentialtradeoffs between urea and glycine as sources of ammonia, as urea has2.5 times more ammonia, yet the reductions were similar to those obtainedwith glycine at equal applications.

A memo from Philip Morris (1998) described the use of silica gels to removeformaldehyde from gas streams. -Aminopropylsilane–silica gels are re-ported to filter formaldehyde from cigarette smoke selectively. In a modelsystem in which formaldehyde-laden gas was puffed through a smoking ma-chine under FTC conditions, the ability of various configurations of silicagels, -aminopropylsilane–silica gels and traditional cellulose acetate filtersto remove formaldehyde were compared. Correlation analysis of the physicalproperties of 15 silica gels showed that pore diameter and pore volume weremost strongly correlated (values of 0.63 and 0.59, respectively) with the per-centage formaldehyde reduction, while particle size showed a negative or nocorrelation (–0.17 to 0.09). Philip Morris holds at least one patent for-aminopropylsilane–silica gels in cigarette filters (Koller et al., 2005).

Benzene is a volatile component of the gas phase with multiple rod precursors(see Ferguson, 2000). Therefore, charcoal filters have been proposed for

Figure 3.11Acrolein per milligram of nicotine in different brands machine smoked in the ISO andHealth Canada intense regimens

Brand

Ace

tald

ehyd

e (u

g/m

g ni

cotin

e)

HC ISO

180.0

160.0

140.0

100.0

120.0

80.0

60.0

40.0

20.0

0.01 2 3 4 5

99

reducing benzene levels in smoke. G.M. Polzin and colleagues (2008) re-ported that, under intense smoking conditions and normalization for nicotine,a filter containing at least 120 mg of carbon can reduce benzene emissionsby nearly 85%. This observation of machine-generated smoke parallels dataon human exposure: Scherer and colleagues (2006) reported that smokingcarbon-filter cigarettes resulted in statistically significant reductions inbiomarkers of exposure to benzene (e.g. S-phenyl mercapturic acid in urine),although the mass-market charcoal-filter cigarettes studied would probablycontain far less than 120 mg charcoal. Searches of tobacco documents andpatents failed to reveal other methods tested for the reduction of benzene inmainstream smoke.

Benzo[a]pyrene is the prototypic PAH. Methods to reduce PAH levels inmainstream smoke have been investigated ever since these compounds wereidentified as significant components of smoke. In the 1950s and early 1960s,several patents were issued for processes for extracting precursor componentsfrom tobacco. Rodgman and colleagues at RJ Reynolds (reviewed byRodgman and Perfetti, 2006) showed that extracting saturated aliphatic hy-drocarbons, phytosterols and terpenoids such as solanesol from tobaccodecreased the PAH levels in the mainstream smoke of cigarettes made fromextracted tobaccos. A study by Tennessee Eastman (1959) examined the ef-fect of extraction on benzo[a]pyrene levels. Extracting these components alsoincreased the per-unit-weight content of carbohydrates (including cellulose)in the tobacco, which, as discussed earlier, are precursors of aldehydes. Bythe 1960s, it was shown that higher levels of nitrates in tobacco led to lowerPAH levels in smoke, probably due to the ability of nitrate to interfere withthe free-radical processes that generate PAHs.

Wynder and Hoffman (1967) postulated that the observed drops inbenzo[a]pyrene yields over time might be due to the introduction of recon-stituted tobacco, particularly that made from stems (with relatively highnitrate levels), in cigarette design. As nitrates are precursors of the TSNA insmoke, however, this may have increased the TSNA content while loweringthe PAH content, in essence trading one class of carcinogen for another.Similarly, given differences in benzo[a]pyrene yield between flue-cured andburley tobaccos, a blend containing more burley would be expected to haveless benzo[a]pyrene. Burley is, however, known to produce more TSNA.

Norman (1999) showed that different types of reconstituted and expandedtobacco have different yields of benzo[a]pyrene (and other constituents), de-pending on how they are processed. Paper process reconstituted tobacco hasless benzo[a]pyrene than slurry process reconstituted tobacco (0.52 vs.0.85 g/g) but more formaldehyde (56.8 vs. 40.5 g/g) (Halter, Ito, 1978).Similarly, different types of expanded tobacco had different levels of

100

benzo[a]pyrene: a standard blend had 0.71 g/g, an RJ Reynolds puffed to-bacco cigarette had 0.76 g/g, a Philip Morris expanded cigarette had0.45 g/g, and a freeze-dried tobacco cigarette had 0.57 g/g (Halter, Ito,1978). Rodgman (2001) summarized the changes in cigarettes at variousstages of production that had been investigated in order to reduce PAH yieldsin smoke, as shown in Table 3.7.

Table 3.7Changes to cigarettes at various stages of production explored to reduce PAH yieldsin smoke

Changes to plants Changes to tobacco rod Changes to design

Tobacco type Combustion modifiers Paper porosityStalk position Casing materials Paper additivesNitrate content Flavours Paper coatingsNicotine content Tobacco weight Filter efficiency or selectivityCuring Reconstituted sheet Filter additivesGrading Stem inclusion Filter ventilationFermentation Expanded laminaExtraction Moisture contentDenicotinization Dilutants (e.g. Cytrel)AmmoniationPesticidesAgricultural chemicals

Adapted from Rodgman (2001), Table 5PAH, polycyclic aromatic hydrocarbon

The Cippilone case, one of the early cigarette product liability cases in theUSA, brought to light the existence of the ‘XA’ cigarette, in which the tobaccohad been treated with nitrate salts containing a palladium catalyst to reducethe biological activity of smoke. In the 1977 patent issued to Liggett for thepalladium additive (Norman, Bryant, 1977), it is mentioned that the additivereduces PAHs in smoke. An ‘Attorney Work Product’ (1992) provides a goodsummary of the testimony given at the trial. Liggett never marketed a cigarettewith the palladium technology, however, allegedly because the companyfeared this would be tantamount to admitting that their other cigarettes causedcancer, and because their claims of reduced biological activity would be basedon tests (e.g. mouse painting) that they had in other instances derided. Thepalladium technology was resurrected in the Omni cigarette sold by VectorTobacco in the early 2000s. Vector claimed 2–42% lower PAH emissions(depending on the measurement method) than conventional cigarettes. Hu-man studies showed, however, that people who switched to Omni did not

101

have reduced PAH exposure, as measured by urinary 1-hydroxypyrene(Hatsukami et al., 2004; Hughes et al., 2004).

More recent attempts to reduce PAHs in mainstream smoke include the ideaof ‘free-radical traps’ to inhibit their formation during pyrolosis or selectivefiltration of PAHs. Fillagent patented a porphyrin additive for filters thatpurports to trap PAHs (Lesser & Von Borstel, 2003), and this appears to havebeen introduced into a new Fact cigarette (no relation to the Brown &Williamson product described earlier) being test-marketed in the USA.

Lodovici et al. (2007) reported that a DNA-based solution (extracted fromsalmon sperm: Maillard, Hada, 2005) could be added to a cellulose acetatefilter to bind with PAHs, including benzo[a]pyrene, as smoke passed throughthe filter. The solution added approximately 10 mg of DNA per cigarette tothe filter. They reported that the smoking machine yield of benzo[a]pyrenewas 40% lower in treated than in control cigarettes. It is not clear whetherthis level of reduction is reproducible, however, given that most PAHs do notreact with DNA until reduced to epoxides (which is not known to occur insmoke). It is also not known whether 10 mg of DNA would be sufficient toreduce PAHs to the reported level.

1,3-Butadiene, like most gas phase constituents, has been proposed to beremoved selectively from smoke by activated charcoal, but little informationis available. RJ Reynolds explored a ‘carbon scrubber’ filter technique, in-volving a 17.8-cm piece of wood-pulp paper impregnated with 40- m parti-cles of carbon placed to provide channels of 0.2–0.7 mm (RJ Reynolds, 1994;Stiles et al., 1994). The technique was supposed to reduce gas-phase con-stituents, including butadiene, acrolein and formaldehyde. An internal report(Rogers, 1993a) noted that the level of butadiene per milligram wet particu-late matter was 24% lower in a modified Camel Light with the carbonscrubber in conjuction with higher filter ventilation. Adding potassium car-bonate to the blend increased the percentage reduction to approximately 48%.Another analysis with ‘straight-grade’ cigarettes showed no significant dif-ference in butadiene content (Rogers, Bluhm, 1993b). The research on carbonscrubber filters went as far as human studies (Rogers, Bluhm, 1993b; Smithet al., 1994a,b). The product evolved into a disposable filter attachment (i.e.an aftermarket accessory) rather than a filter used in manufacture (RJReynolds, 1995).

Another option is to remove potential precursors in tobacco; however,Ferguson (2000) noted that there are many possible precursors in tobacco,and butadiene is probably formed through a number of competing pathways.Thus, lowering one set of precursors in the rod might not necessarily reducethe butadiene yield in smoke.

102

Carbon monoxide has been a target for emission reduction for some time. Asit is a gas-phase component, selective filtration has been considered as astrategy; however, most studies show that activated charcoal is ineffective insignificantly reducing CO in mainstream smoke (e.g. Tigglebeck, 1967),probably because the gas is so abundant. British American Tobacco notedthat CO “is present in smoke at substantially greater concentrations than othersmoke components considered in this review. Hence substantial quantities ofsubstances which combine chemically with it will be required to effectfiltration.” (Irwin, 1990). Polzin et al. (2008) showed that even 120 mg ormore of charcoal, while effective in lowering other gas-phase volatiles (seeearlier sections), had no significant effect on CO.

Tolman (1970) summarized a number of possible techniques for removingCO from smoke and British American Tobacco’s work to date in the area ofselective CO filtration:

chemisorption in a microporous subtrate (molecular sieve), physical ad-sorption,

chemical complex formation,

absorption in a liquid in which CO is soluble,

oxidation to CO2,

catalytic oxidation in which oxygen is consumed,

disproportionation to carbon and CO2 and

reduction to carbon.

British American Tobacco appears to have done substantial work on selectivefiltration of CO in the early 1970s, led by Tolman (1970, 1971, 1973). Inparticular, they explored manganese dioxide and other catalysts for convert-ing CO to CO2 in smoke. In summarizing this work, Irwin (1990) noted thateffective catalysts were obtained by depositing manganese dioxide in thepores of activated carbon, but that the exothermic CO to CO2 conversioncatalysed by manganese oxide heated the charcoal to the point that it glowedred (probably from oxidation of accumulated particulates) and the filter al-most disintegrated. Irwin found that there was “little evidence of effective-ness” of complexing agents and concluded that the best route for futureresearch was molecular sieves.RJ Reynolds developed an internal ‘reactor system’ for rapid assessment offilter additives for reducing CO, relying on tests of the substance’s ability toremove CO from a gas stream (Townsend, 1978). They also performed ex-tensive research in the 1980s on paper porosity and filter dilution to reduce

103

CO. One internal memo mentioned the possibility of building a ‘catalyticconverter’ into the filter to remove CO by converting it to CO2 with hydrogenperoxide in some form (Maselli, 1986). In the 1990s, they explored a groovedfilter that was claimed to reduce CO by up to 50%, although this relied mainlyon filter dilution.In 1997, a Greek cigarette manufacturer began marketing cigarettes featuringa ‘bio-filter’, claiming that haemoglobin embedded in the filter sequesteredCO (Deliconstantinos, Villiotou, Stavridis, 1994; Valavanidis, Haralambous,2001). Given the strong affinity of haemoglobin for CO, this seems to be aplausible strategy; however, a Lorillard memo (Robinson, 1998) reported thatin-house attempts to reproduce the findings reported by Deliconstantinos,Villiotou and Stavridis (1994) had been unsuccessful. The author speculatedthat “failure of the proteins to alter the delivery of the measured componentsmay have resulted from the incorporation of a smaller than adequate amountof each protein. A larger amount of each protein could not be incorporatedusing the Deliconstantinos method since the protein solutions used were atthe saturation limit.” Valavanidis and Haralambous (2001) reported less COremoval than achieved in the original paper (30–35% vs. 90%). Tolman(1971) calculated that 10–20 g of haemoglobin would be required on a filterto remove 15–20 mg of CO from smoke. Philip Morris also exploredhaemoglobin filters over the years (Philip Morris, 1997). Patents exist forhaemoglobin and other absorbent additives to filters (Scheinberg, 1974;Charalambous, Bullin Haines, Morgan, 1987; Kubrina et al., 1989; Stavridis,Deliconstaninos, 1999), but no or only small changes to CO yields areclaimed, suggesting that they are insufficient to remove the large amounts ofCO present.

RJ Reynolds explored palladium salts as a method for reducing CO in smoke(Riggs, 1989). Lewis, Norman and Robinson (1990) summarized their in-vestigations of addition of palladium and copper catalysts to the filter andshowed that they could reduce CO in smoke by as much as 20%. Concernabout cost, availability of sufficient materials and toxicology appear to havecaused them to abandon the work.

3.6.4 Considerations in applying mandatory lowering to multiple toxicants

In considering a recommendation to expand the list of toxicants for mandatedlowering of their levels in cigarette smoke, TobReg examined the effects ofcorrelating the different toxicants across brands, the effects of various levelsof reduction of these toxicants on the number of brands remaining on themarket, and methods for examining and tracking the net effect on toxicantyields of the proposed regulatory strategy. These analyses were conductedwithout considering the effect of brand reduction due to the previously rec-ommended mandated limits for NNN and NNK, as TobReg considered that

104

changes in curing and manufacturing processes would probably allow mostif not all brands to meet those limits.

The lower the regulatory level set for a given toxicant, the greater will be thereduction in the level of that toxicant in the brands remaining on the market,because more brands will either be reformulated to comply with new toxicantlimits or be banned from sale.

A possible effect of setting levels for multiple toxicants is the elimination ofmost brands on the market if the level is set at the median value for severaltoxicants (Figure 3.12). It is impossible to know which brands would have tobe removed from the market rather than simply undergoing product re-designin order to remain on the market, because TobReg does not know the fullextent of the industry’s reformulation capacity. Higher mandated limits allowmore brands to remain on the market unchanged and more toxicants to beregulated simultaneously, but result in higher levels of toxicants in themachine-generated smoke from the remaining brands. The result is a com-plicated trade-off between the number of toxicants regulated, the magnitudeof the reductions in toxicant levels and the number of brands removed fromthe market. TobReg examined a number of considerations relative to thesetrade-offs in arriving at the recommended set of toxicants and their maximumlevels.

Figure 3.12Proportion of brands remaining after constituents removed, by percentile in order oftheir cancer index; international brands, modified intense regimen

0

52

05

57

001

Perc

ent o

f bra

nds

rem

aini

ng

tnacixoT

09yB 08yB 07yB 06yB 05yB

:elitnecreptimillamixaM

105

The analyses presented give the maximum number of brands that would beeliminated on the basis of existing data, on the unlikely assumption that nomodifications would be made to existing products with the introduction ofmandated lowering regulations. As described elsewhere in this report, a spe-cific anticipated outcome of the proposed product regulation strategy is achange in the manufacture of individual brands to reduce the toxicant yieldsmeasured under standardized machine smoking conditions. It is expected thatthe setting of mandated limits will spur further advances in methods to lowertoxicant yields. As a result, TobReg assumed the most draconian potentialimpact on existing brands. In reality, it is anticipated that far fewer brandswould actually be restricted from sale if the mandated limits for toxicantswere implemented.

Correlations among levels of different toxicants

One consideration in examining the list of toxicants is the extent to which thelevels of one toxicant in a brand are correlated with those of other toxicantsin the same brand. This association can be quantified statistically as a corre-lation coefficient. The correlation coefficients for the toxicants under con-sideration, plus NNN and NNK, are presented in Tables 3.8 and 3.9 for thesample of international Philip Morris brands and for the Canadian data minusthose brands high in NNN. The plots of the actual data from which the cor-relations are derived are presented in Figures A4.1 and A4.2 in Annex 3.4,with the r2, which is the correlation coefficient for each toxicant with all ofthe other toxicants on the list recommended for mandated reduction. Alsoincluded in Annex 3.4 is the correlation matrix for all the toxicants reportedto Health Canada and measured by Counts and colleagues (2005). Some tox-icants are highly correlated, others show little correlation, and a few arenegatively correlated.

A positive correlation suggests that removing brands with high levels of onetoxicant by mandating lower levels will also lower the level of the correlatedtoxicant, which is high in the same brands. This might mean that the levelsof one toxicant could serve as a proxy for several other toxicants in a regu-latory strategy. Caution should be exercised in relying on this assumption, asa strong correlation does not mean that the toxicants are generated by thesame mechanism in the smoke. For example, toxicants might be derived fromdifferent flavouring agents in tobacco and therefore subject to independentmanipulation, but they may be correlated because some brands contain mul-tiple flavouring agents and other brands only a few.

106

Table 3.8Correlation coefficients for levels of toxicants in the same brand, internationalbrands, modified intense regimen

(1) (2) (3) (4) (5) (6) (7) (8) (9)

(1) Acetaldehyde 1.000(2) Acrolein 0.949 1.000(3) Benzene 0.760 0.764 1.000(4) Benzo[a]pyrene –0.272 –0.208 –0.266 1.000(5) 1,3 Butadiene 0.852 0.855 0.874 –0.401 1.000(6) Carbon monoxide 0.830 0.761 0.673 –0.369 0.807 1.000(7) Formaldehyde 0.156 0.244 0.081 0.340 0.094 –0.111 1.000(8) NNK 0.010 –0.111 0.002 –0.393 –0.051 0.214 –0.639 1.000(9) NNN 0.103 –0.014 0.042 –0.476 0.062 0.252 –0.717 0.853 1.000

NNK, 4-(N-nitrosomethylamino)-1-(3-pyridyl)-1-butanone; NNN, N´-nitrosonornicotine

Table 3.9Correlation coefficients for levels of toxicants in the same brand, Canadian brands(NNN limited), modified intense regimen

(1) (2) (3) (4) (5) (6) (7) (8) (9)

(1) Acetaldehyde 1.000(2) Acrolein 0.792 1.000(3) Benzene 0.957 0.779 1.000(4) Benzo[a]pyrene 0.810 0.474 0.859 1.000(5) 1,3-Butadiene 0.953 0.836 0.944 0.765 1.000(6) Carbon monoxide 0.949 0.848 0.972 0.810 0.949 1.000(7) Formaldehyde 0.844 0.568 0.867 0.912 0.827 0.810 1.000(8) NNK 0.707 0.426 0.762 0.817 0.627 0.744 0.710 1.000(9) NNN 0.454 0.128 0.462 0.464 0.376 0.447 0.347 0.774 1.000

NNK, 4-(N-nitrosomethylamino)-1-(3-pyridyl)-1-butanone; NNN, N´-nitrosonornicotine

A negative correlation suggests that brands with a high level of one toxicanthave a relatively low level of the other. Mandatory lowering of the levels oftwo toxicants with negative correlation coefficients will result in more brandsbeing eliminated from the market than when two brands have strong positivecorrelation coefficients. The effect on the levels in the remaining brands isless clear, as the regulatory level set for each toxicant might also result inelimination of the brands that are low in the second toxicant, with the resultthat the toxicity indices for the brands remaining on the market might changeless than would be expected from the number of brands eliminated.

107

The uncertain nature of cigarette manufacturers’ responses and the complex-ity of predicting the effect of mandated reductions on toxicant yields understandardized conditions on the brands remaining on the market make a strongcase for continued monitoring of the toxicant yields of cigarettes after im-plementation of regulations to mandate lower levels. After discussion,TobReg considered that regulating a larger number of toxicants is the moreappropriate strategy, rather than allowing closely correlated toxicants to serveas proxies for one another. This approach was considered to offer the greatestprotection against uncertainties about the manufacturing changes likely to beimplemented in response to the proposed regulatory strategy and their con-sequences for toxicant yields.

Effect on number of brands affected by limits

In the absence of a change in cigarette manufacture, the number of brandspotentially eliminated from the market by the proposed regulatory approachwill depend on the level set and the number of toxicants regulated.Figure 3.12 presents the percentage reduction in number of brands on themarket with the regulated levels set at different percentiles of the distributionof the toxicant per milligram of nicotine for the brands on that market, andwith the sequential addition of the toxicants being considered. The numberof brands potentially eliminated as each toxicant is regulated depends on theorder in which the toxicants are added to the analyses. In Figure 3.12, thetoxicants are entered in the order of their animal carcinogenicity index fromTable 3.3. Data for the sample of international Philip Morris brands demon-strate that if all the brands with values for the seven toxicants above themedian were eliminated, almost all the brands would be eliminated from themarket. There is clearly a need to balance the number of toxicants regulatedagainst the levels mandated in order to avoid substantial market disruption.

Monitoring for unintended increases in toxicant yields

Mandating a reduction in the levels of some toxicants does not guarantee thatthe levels of other, non- regulated toxicants will also be reduced. It is thereforepossible that removal from the market of brands high in levels of the regulatedtoxicants will leave brands with high levels of unregulated toxicants. In ad-dition, changes in cigarette design and manufacture implemented to lowerthe regulated toxicants might have the effect of increasing the levels of un-regulated toxicants. TobReg recognized the potential for these unintendedconsequences of mandating the reduction of specific toxicants and consideredhow monitoring might be put in place to detect them.

One approach would be to track the yields, or the percentage change in medianyields, of the entire list of toxicants measured by Health Canada. Different

108

toxicants have different potencies as toxicants, however, and the range oftoxicant per milligram of nicotine (defining the reduction possible with theproposed regulatory approach) is different. As a result, the net effect of amandated reduction of some toxicants on total toxicant yields from the brandsremaining will depend on the toxicant selected and the levels set.

On the basis of these considerations, TobReg recommends using the sum ofthe individual toxicant animal carcinogenicity indices of the brands remain-ing on the market to examine the potential unintended consequences ofregulation on net toxicant yields. TobReg chose this approach as a means ofintegrating data on animal toxicity and data on the levels of toxicants in smokegenerated under standardized conditions for the purpose of monitoring theeffects of implementing the proposed regulatory approach on net toxicantyields. The sum of the toxicant animal carcinogenicity indices generated andchanges in that index are mathematical constructs of toxicant yields and arebased mainly on animal toxicological evidence. They do not account for po-tential interactions between toxicants or the effects of toxicants for which noanimal carcinogenicity index has been generated. They are not quantitativeestimates of the human toxicity of the smoke from different brands, of therisk for cancer or other diseases due to smoking different brands, and shouldnot be applied in a quantitative way for estimating the risk of smoking or theactual toxicity of the smoke generated. They are recommended solely for useas a monitoring tool for regulators.

Figure 3.13 presents the sum of the toxicant animal carcinogenicity indicesfor the brands remaining on the market with mandated limits set at differentpercentiles of the distribution of the toxicant per milligram of nicotine for thebrands on that market. The effect of sequentially adding toxicant mandatedlimits at the same percentile is presented, with the toxicants added in orderof their toxicant animal carcinogenicity index. The sum is calculated for allthe toxicants listed in Annex 3.3 Table A3.1 for which there are toxicantanimal carcinogenicity indices, with the exception of NNN and NNK. Thesum of the indices is not limited to the indices of the seven toxicants underconsideration. The value presented is the brand-specific arithmetic sum ofthe hazard indices for all toxicants for which there are cancer indices, ex-pressed as the mean of the brands remaining on the market. The dashed linerepresents the sum for the brand with the lowest value and therefore representsthe lowest value that can be achieved with mandated lowering based on thedistribution of toxicants in the brands reported by Counts and colleagues(2005). Further lowering would require modification of the existing productsby the manufacturers.

109

Figure 3.13Means of sums of animal carcinogenicity indices for brands remaining as additionaltoxicants are regulated; international brands, modified intense regimen

0

5

01

51

02

52

03

53

Mea

n of

sum

of c

ance

r haz

ard

inde

x

tnacixoT

09yB 08yB 07yB 06yB 05yB muSniM

:elitnecreptimillamixaM

The result of this analysis suggests that the net toxicant yield of the brandsremaining on the market will decrease (at least by this measure), and no netincrease in toxicant yield will result from the proposed regulatory strategy.Ongoing monitoring would be needed to ensure that changes in cigarettemanufacture in response to the proposed mandated lowering do not increasethe yields of other toxicants.

Almost the entire range of values for the existing market is contained in thetop one third of the sum of the toxicant animal carcinogenicity indices, lim-iting the effect of any regulatory strategy on this summary measure of toxicantyields. It is important to note, however, that this range is based on existingbrands and existing technology, and a regulatory strategy that encouragesinnovation in lowering the levels of toxicants in smoke might substantiallyreduce the minimum values for modified brands. This reality supportsTobReg’s recommendation that market variation in yields be used only forsetting the initial mandated lowering levels and that targets based on whatcan be achieved with advancing technology be examined as a second phaseof the mandated lowering approach.

The range of the summary measure presented in Figure 3.13 supports theposition that all brands of cigarettes generate smoke with substantial toxicityand that, while it may be possible to lower the toxicity of smoke generated

110

under standardized conditions, it is unlikely that large reductions in toxicityare achievable for conventional cigarettes without substantial advances incigarette design.

Neither the metric used (the sum of the toxicant animal carcinogenicity in-dices) nor changes in this metric are measures of human toxicity or risk, andthey should not be used as such in estimating changes in human risk with useof different brands or resulting from the regulatory strategy. The data inFigure 3.13 are presented simply to show the effect of regulating some tox-icants on the net toxicant yields of the brands remaining on the market. Thereader should resist the temptation to view these changes as quantitative es-timates of the risk reduction benefit of product regulation, for several reasons.The hazard indices are constructed from potency estimates obtained from theavailable animal data, and such data are not available for all the known tox-icants. The sensitivity of animals and humans to the different toxicants maydiffer. The effect of exposure to the multiple toxic compounds in smoke hasnot been established as additive. These realities limit the use of these measuresas estimates of even total animal toxicity and obviate their direct extrapolationto humans. The regulatory strategy proposed focuses on lowering levels ofknown toxicants in the cigarettes on the market as good manufacturing prac-tice. The strategy is not based on a provable or certain reduction in humantoxicity resulting from the use of those products, the toxicant animal car-cinogenicity index is not a quantitative estimate of human cancer risk, and aregulatory standard that can reliably quantify a reduction in human risk isbeyond current scientific capability.

3.7 Recommended toxicants and recommended mandated limits

The toxicants currently recommended for mandated reductions by TobRegare NNN and NNK, acetaldehyde, acrolein, benzene, benzo[a]pyrene,1,3-butadiene, CO and formaldehyde. The levels recommended in this reportrepresent TobReg’s judgement on the most practical trade-offs, consideringthe need to regulate a range of toxicants, to mandate lowering of those toxi-cants to the greatest extent and yet not to eliminate most brands, in theircurrent form, from the market.As stated in the first report of TobReg on this regulatory approach (WHO,2007), the median values for NNN and NNK are used as recommended levelsbecause there is strong evidence that existing technology can dramaticallylower the amounts of these toxicants in tobacco smoke. An initial level of125% of the median value is recommended for the other toxicants. The higherlevel for the other toxicants reflects the somewhat greater uncertainty aboutthe extent to which these toxicants can be reduced with existing approaches.Substantially lower levels should be the ultimate goal of this regulatorystrategy.

111

An examination of the variation in these toxicants by brand in the data setsavailable (see Annex 3.5) demonstrates that the levels in most brands fallbelow 125% of the median per milligram of nicotine of all the brands in thedata set. The median level of all brands is likely to be a relatively stable valueon which to base the initial regulatory level. It was also viewed as preferableto base the recommended levels on a specific value rather than on a specifiedpercentile of the distribution of brands. TobReg therefore recommends thatthe levels for the proposed regulation of constituents other than NNN andNNK be set at 125% of the median value of the brands reported. This choicereflects a judgement that incorporates the previously described, conflictinggoals of: regulating a number of agents, reducing the levels of toxicants inbrands on the market to the greatest extent and maintaining enough brandsrequiring minimal modification on the market to satisfy consumer demand.

Regulators may of course select different levels that are more appropriate fortheir own circumstances and are encouraged to base levels on the yields ofthe brands sold in their own markets, using the principles set out in this report.Table 3.10 presents the recommended levels, based on the data available onthe international brands (Counts et al., 2005) and the Canadian brands (HealthCanada, 2004) tested with the modified intense machine-smoking regimen.

Table 3.10Toxicants recommended for mandated lowering

Toxicant Level in μg/mg nicotine Value

Internationalbrandsa

Canadianbrandsb

NNK 0.072 0.047 Median of data setNNN 0.114 0.027 Median of data setAcetaldehyde 860 670 125% of median of data setAcrolein 83 97 125% of median of data setBenzene 48 50 125% of median of data setBenzo[a]pyrene 0.011 0.011 125% of median of data set1,3-Butadiene 67 53 125% of median of data setCarbon monoxide 18400 15400 125% of median of data setFormaldehyde 47 97 125% of median of data set

NNK, 4-(N-nitrosomethylamino)-1-(3-pyridyl)-1-butanone; NNN, N´-nitrosonornicotinea Based on data from Counts et al., 2005b Based on the data reported to Health Canada minus brands with levels of NNN per mg nicotine >

0.1 ng, which eliminates most US and Gauloise brands. (http://www.hc-sc.gc.ca/hl-vs/tobac-tabac/legislation/reg/indust/constitu_e.html)

112

The levels in Table 3.1 in the column labelled ‘international brands’ are de-rived from data on brands in an international sample (Counts at al., 2005),which are US-style cigarettes containing a blend of tobaccos. The columnlabelled ‘Canadian brands’ is based on a set reported to Health Canada, whichrepresent unblended cigarettes containing predominantly flue-cured (bright)tobacco. Regulators should use data reported for their own markets to setlevels, if they are available, or select the values derived from the data set thatconforms most closely to the cigarettes available on the market being regu-lated. The differences in the reported levels in individual brands are particu-larly large for the TSNA (NNN, NNK).

In smaller countries, where the style of cigarettes on the market is not readilydefinable, it is recommended that the mandated limits be based on the inter-national sample initially and that alternative standards be considered as moreinformation on local brands becomes available.

Regulation of brands for toxicants other than NNN and NNK at the levelsproposed in Table 3.10 would result in 60% of the brands remaining on themarket without modification of the product for the international set of brandsand 59% of the brands remaining on the Canadian market.

Given the existing technology for lowering NNN and NNK levels in tobacco,it is expected that most brands will be able to comply with the recommendedmandated limits for these compounds. The higher levels present in the inter-national sample of US blend cigarettes than in those from Canada suggestthat countries should establish mandated limits for these toxicants that arebased on measurements made for the brands on their own markets. The extentto which each of the other toxicants listed in Table 3.10 can be lowered withexisting technology is less clear, but, given the variation in levels in the brandscurrently being sold, it is expected that the manufacturers of many brandswith high levels of toxicants on initial reporting would be able to lower thoselevels in order to continue marketing them.

Success in lowering the levels of individual toxicants can indicate to regula-tory authority the potential of existing technology for achieving furtherlowering of toxicant yields and can help in deciding to set goals for toxicantyields to further lower existing mandated limits for individual toxicants. Overtime, it is envisioned that there would be a progressive lowering of the man-dated limits, both in response to advances in technology and to the develop-ment of new approaches to lowering toxicant yields.

3.8 Implementing mandatory reductions in toxicant yields

The data presented in this report reflect a limited set of brands in a few mar-kets, whereas there are substantial differences in the levels of toxicants in

113

brands in different markets. This suggests that a regulatory strategy shouldbe phased, with a period of required reporting of toxicant levels by cigarettemanufacturers, followed by an announcement of the maximum levels fortoxicants, and finally enforcement of those mandated limits. For countrieswithout cigarette manufacturing facilities or with limited laboratory capacity,the values presented in Tables 3.1 and 3.10 represent reasonable initial reg-ulatory levels.

The first phase of the proposed regulatory strategy for mandating reductionsin toxicant yields would be to require reporting of the levels of constituentspresent in cigarette smoke from those brands being regulated. It is expectedthat the tobacco industry would be responsible for the cost of this testing andreporting, either directly or through taxation or licensing. The initial reportingshould cover all the toxicants currently reported to Health Canada. This de-fined list can be used to assess whether there are differences between thetoxicant levels generated by brands on the market to be regulated and thosereported here. Of particular concern are some of the metals, the levels ofwhich can vary substantially with the source of the tobacco used to manu-facture cigarettes. After the initial reporting, a more limited list of toxicants,presented in Table 3.6, is recommended for annual reporting, with the morecomplete list reported every 5 years.

The levels of the smoke constituents listed in Table 3.6 would be reportedannually by brand for 2–3 years to allow analysis of variation in constituentyields of brands on the market. Initial mandated limits for toxicants other thanNNN and NNK would be set on the basis of the level of the toxicant permilligram of nicotine representing 125% of the median value for the reportedbrands. The levels for NNN and NNK would be set at the median value forthe reported brands. Mandated limits for NNN, NNK, acetaldehyde, acrolein,benzene, benzo[a]pyrene, 1,3-butadiene, CO and formaldehyde would thenbe established and announced to cigarette manufacturers and importers at theend of the reporting period. Beginning 2 years after announcement of themandated limits, brands with toxicant yields per milligram of nicotine ex-ceeding the set levels would be excluded from import or sale. Data on thelevels of toxicant yields for the toxicants in Table 3.6 would continue to bereported annually.

3.8.1 Considerations for modified cigarettes and potential reduced exposureproducts

The recommendations in this report are intended to apply to traditional man-ufactured cigarettes in which tobacco is burnt; they should not be applied tocigarettes in which tobacco is heated or in which techniques other than com-bustion are used to deliver nicotine. Assessments of these unconventionaltobacco products and other potential reduced exposure products were

114

discussed in a previous report (Scientific Advisory Committee on TobaccoProduct Regulation, 2003).

It is possible to alter the level of a toxicant per milligram of nicotine incigarette smoke by changing the nicotine yield of the cigarette, as well as byaltering the level of the toxicant. The yield of nicotine in cigarette smoke canbe increased by adding nicotine to the tobacco or the filter or by using high-nicotine varieties of tobacco. While these approaches may theoretically haveindependent use in decreasing exposure to tobacco toxicants, their potentialto do so remains uncertain. The usefulness of increased nicotine yields as amethod of lowering the level of toxicant per milligram of nicotine to a levelbelow the mandated limit value is therefore also unproven for reducing thetoxicity of smoke. Regulatory authorities should neither encourage nor allowincreased nicotine yields to be used as a method for complying with the man-dated values regulation.

Detection of increasing nicotine yields in brands can be facilitated by trackingmachine-delivered nicotine yields per cigarette over time and by examiningthe distribution of tar to nicotine ratios for the brands in a given market. Forthose brands with increasing nicotine yields over time and those brands withtar to nicotine ratios in the bottom one third of the brands on the market,regulators may choose to require that the brand have below a mandated limitvalue per milligram of tar as well as per milligram of nicotine.

Some brands of cigarettes may be offered with reduced nicotine in the tobaccoused to make the cigarette. Nicotine can be removed from the tobacco leaf,and genetically altered tobacco is available with a very low nicotine content.Cigarettes made with these tobaccos have low nicotine deliveries under anytesting method and correspondingly may have very high levels of toxicantsper milligram of nicotine. Regulators may want to evaluate separately brandsfrom manufacturers that appear to be intentionally lowering nicotine in thetobacco. Once the regulatory authority is satisfied that the manufacturer isactually using low-nicotine tobacco in the product, the authority may want touse a mandated limit value per milligram of tar instead of per milligram ofnicotine.

3.8.2 Communication of regulatory values and testing results to the public

Mandated reduction of toxicant levels recommended in this report constitutesa first step towards improved tobacco product regulation. TobReg recognizesthe limitations of machine measurements and of setting levels per milligramof nicotine. Existing science does not allow a definitive conclusion that re-duction of the level of nitrosamines, or any other toxicants in cigarette smoke,will reduce cancer incidence or the rate of any other tobacco-related diseasein smokers who use cigarettes with lower levels of these toxicants. Science

115

has also not demonstrated that the specified changes in regulatory values willresult overall in a meaningful change in the actual exposure of consumers,although that is an anticipated outcome. Mandating levels and banning brandswith higher levels from the market does not equate to a statement that theremaining brands are safe or less hazardous than the brands removed. It alsodoes not represent government recognition of the safety of the products thatremain on the market. The proposed strategy for lowering toxicant yields isbased on sound precautionary approaches similar to those used for other con-sumer products.

Given the limitations of existing science, regulatory authorities have an obli-gation to ensure that the public is not misled by the results of the recom-mended machine testing and mandated lowering regulatory strategy, in theway that the public was misled by the use of machine testing for tar andnicotine yields. TobReg noted that labelling of cigarettes with tar, nicotineand CO levels measured with the ISO regimen persists and continues to beharmful to the public. TobReg recommends that any regulatory approachprohibit the use of the results of the proposed testing in marketing or othercommunications with consumers, including product labelling. It is also rec-ommended that manufacturers be prohibited from making statements such asthat a brand has met government regulatory standards and also from publi-cizing the relative ranking of brands by testing level. Because information isoften transmitted to smokers through the kinds of news stories that accom-pany new regulation implementation, it is the responsibility of the regulatorystructure to monitor:

the accuracy of news reports,

tobacco industry marketing,

smokers’ understanding of the new regulations,

how smokers interpret the new regulations relative to the hazard of theproducts remaining on the market and

whether understanding about the hazard of the remaining products is in-fluencing initiation or cessation rates.

Regulators should then pursue whatever corrective action is necessary toprevent consumers from being misled. These monitoring and surveillanceconcerns are described in more detail in the WHO report on evaluation ofnew or modified tobacco products (Scientific Advisory Committee on To-bacco Product Regulation, 2003).

116

3.9 Issues for regulators mandating lower levels of toxicants

In proposing mandated reductions in toxicant yields, TobReg consideredsome of the ramifications of its recommendations on existing markets and onthe people who control them. It is not TobReg’s obligation to tell the industryhow to achieve the changes, as the expertise for this is within the domain ofthe companies, and there are a number of ways for them to meet the limits.It is, however, obvious, that implementing TobReg’s proposed regulatorystrategy over time will change the market substantially. Tobacco may besourced differently, tobacco-growing is likely to change, manufacturing tech-niques will change, the number of additives is likely to be reduced andultimately the number of brands is likely to be reduced. Manufacturers willbear an analytical burden, which will be part of routine quality control; thisought not to be regarded as an imposition. Regulators will have to validate arandom sample of the tobacco industry analyses for monitoring and qualityassurance, and this will require resources. Taxation or licensing fees for eachbrand to be marketed could provide a source of funds for this activity.As explained elsewhere, the setting of mandated limits is proposed as a pre-cautionary approach and is not an excuse for ‘health’ or advertising claims.There is no reason why any of the changes proposed so far should requirepublic information campaigns, as the cigarettes will remain carcinogenic andtoxic despite reductions in toxicant levels. Thus, there is no reason whycountries that have banned advertising should re-introduce it. Similarly, thereis no reason for allowing advertisements promoting changes as a result ofregulation, as the outcomes of the changes will be uncertain until many yearsof observation have passed.A number of issues for regulators arise from mandated lowering of toxicantyields. In those countries with no tobacco manufacturing industry, the limitsshould pose few difficulties, as the governments will merely place importprohibitions on products that do not meet the regulations. Companies thatmanufacture regulation-compliant products will benefit, while those that donot will be eradicated from the market.Those countries which are home to manufacturing companies and a growingindustry can expect to encounter objections from both manufacturers andgrowers, as both groups will be affected by regulation. Although some in-ternational companies have already embraced the principle of regulation(Philip Morris, 2002), the introduction of mandated limits will require a rea-sonable time. Some changes, such as reduction of nitrosamine levels, arerelatively straightforward and are already occurring (Gray, Boyle, 2004).Others may require different agricultural and manufacturing practices, andthey may take longer. There are, however, few excuses for continuing tomanufacture excessively toxic or carcinogenic products for any longer thannecessary, and an urgent approach is justified.

117

It is important to consider how the effects of the proposed regulation mightbe assessed over time. As this is an attempt to reduce toxicant and carcinogenyields, the most direct evaluation would be to measure a change in the meanyield of the toxicant per milligram of nicotine in the brands on the marketbefore and after the regulatory mandate is implemented. This can be accom-plished by comparing the data on yields reported by tobacco manufacturersin the initial reporting phase of the programme to the yields reported for thefirst year after the ban on brands that are over the mandated limit. Comparisonof the means of toxicant per milligram of nicotine for all brands on the marketbefore and after the regulation is the recommended comparison. TobRegconsidered sales weighting the brands on the market for the comparison butdecided that this would be less appropriate because of the difficulty in ob-taining validated sales data by brand and because the present approach isdirected at changing the toxicity yields of most or all the brands on the market.Sales weighting the brands would suggest that the toxicant yields might beused as a measure of human exposure or risk, a use that TobReg stronglyrejects.

A second concern in the evaluation of regulations mandating lowering oftoxicant yields is the possibility that removing brands with lower levels ofone toxicant would leave on the market brands with higher levels of othertoxicants, resulting in an increase in the net toxicant yield of the brands re-maining on the market. This issue is discussed earlier in this report, whereregular calculation of the toxicant animal carcinogenicity and toxicant non-cancer response indices is recommended for the brands remaining on themarket, with the trend of their sum over time.A principal reason for regulating cigarettes is the harm they cause. It is there-fore tempting to suggest that the best method for evaluating the effect ofregulation would be to evaluate the harm caused by cigarettes with epidemi-ological approaches and disease end-points. Epidemiology can be expectedto provide at best only a partial evaluation. Cigarette composition, and henceemissions, have changed over time, but many of the design changes haveremained industry secrets. If regulations are introduced progressively andregulatory levels are lowered over time, cigarette design and composition canbe expected to continue to change. It has not been possible to conduct a cohortstudy based on detailed, progressive information on brands because of to-bacco industry secrecy about changes in manufacturing practices. As therehave been changes in brands over time, even studies in which the brandsmoked is recorded have limited ability to assess differences in the risks ofusing cigarettes with different designs. The proposed regulatory approach isintended to result in continued, progressive brand changes, in this case in thedirection of lower toxicity. The changes in disease risks that will result fromthese changes will manifest over several decades, and the risks that result willprobably be a product of cumulative exposure over time. The complexity of

118

the brand changes, the complexity of the behavioural responses to brandchanges that influence exposure, and the switching among brands that char-acterizes smokers’ tobacco use suggest that cohort studies by brand withdisease end-points are unlikely to be informative for regulatory monitoringin the future.For specific toxicants, such as NNK, it may be possible to monitor levels ofexposure in the population of smokers before and after mandated loweringof yields is implemented in order to examine the effect of product regulationon population exposure to that toxicant. Population exposure measures arenot direct evaluations of the regulatory approach but are research approachesto examine the effect of changes in cigarette yields of specific toxicants onthe actual exposure of smokers to those toxicants. Obviously, tobacco-associated disease rates are and will continue to be routinely monitored.

It will be useful to monitor industry compliance and published emission levelstogether with regular surveys of smoking topography by brand and levels ofexposure biomarkers by brand. Such periodic evaluation of human exposuremight allow an estimate of trends in exposure resulting from the designchanges implemented to comply with the mandated reduction regulation.

3.10 Recommendations for follow-up and future work

3.10.1Toxicants in tobacco smoke

Tobacco smoke is a complex mixture of chemicals. No or limited toxicolog-ical data are available for most chemicals, and the concentrations of thecompounds in smoke are generally not known by brand. As a result, the haz-ardous effects of tobacco smoke on humans are not well explained by theusual toxicological paradigm of dose multiplied by toxic potency (Fowles,Dybing, 2003) for those toxicants. The present report recommends reductionsin priority toxicants for which data are available as a first step, focusing on afew well-known toxicants that have been quantified in smoke and can beinfluenced by modifying the design of cigarettes.

There is substantially more documentation of the carcinogenic effects thanof the respiratory, cardiovascular or reproductive effects of the toxicants insmoke. Regulatory actions to reduce the risk of cigarette smoke will thereforerequire substantial work to define the toxicity of individual toxicants otherthan cancer, and also to expand the understanding of cancer.

Ongoing research will probably provide new insights into the quantitativeeffects of other smoke constituents, which may be candidates for productregulation. The list of chemicals for regulation should be revised periodicallyin order to reflect this expansion of knowledge. In addition, the mandatedlimit for each toxicant might have to be revised periodically. Toxicological

119

evidence and better understanding of the determinants of smoke chemistrymight allow the setting of reduction goals on the basis of what can be achievedrather than on the levels in existing products.

By setting and mandating reductions in toxicant yields from tobacco products,governments accept their responsibility to the public for regulating tobaccoproducts as consumer products, albeit extremely hazardous ones. This regu-latory responsibility will be discharged by reducing known toxicant yields,rather than the approach used for food and other consumer products (usuallybased on the no-effect level concept) or medical drugs (based on a risk–benefit concept). For tobacco products, the basis for regulating is the ‘bestavailable technology’, in combination with a precautionary approach, whileacknowledging existing harmful effects due to ongoing use.

TobReg recommends that the list of toxicants recommended for regulationand the levels recommended be reviewed and potentially revised every2 years; the new recommendations from this review should be presented tothe Conference of Parties to the WHO Framework Convention on TobaccoControl and the WHO Tobacco Free Initiative. As the hazardous effects oftobacco smoke are still not completely understood, new paradigms for de-scribing the toxicity of tobacco smoke may arise in the near future and mayadd to the scientific foundation for tobacco regulation.

The concept of mandated lowering of toxicant yields in smoke could be ex-tended to other products, including additives to tobacco products. Suchregulatory efforts might work in tandem. For example, regulation of sugarsas additives could influence the levels of acetaldehyde, while acrolein levelsmight be influenced by the levels of humectants added.

Enforcement of product regulation will require capacity-building for regula-tory agencies interested in regulation, which could be facilitated throughTobLabNet, an international cooperative effort of non-tobacco industry to-bacco laboratories. Capacity-building should include investment in testingand control facilities as well as data-handling capacity and expertise, andresources to support these tasks should be included in national tobaccolegislation.

TobReg and TobLabNet, which are scientific advisory bodies to the WHOTobacco Free Initiative, can provide advice about which toxicants are highpriorities for product regulation and suggest the levels to be used by legalbodies as regulatory levels. TobReg will provide advice mainly about thenature of the chemicals and the mandated limits. TobLabNet will provideadvice about the required methods for control and implementation. Periodi-cally, ideally every 2 years, the recommendations of TobReg and TobLabNetwill be reviewed and potentially revised.

120

3.10.2Smokeless tobacco products

Smokeless tobacco products are used widely in North America, parts ofEurope, Asia (particularly South-East Asia) and Africa. The marketing ofmoist snuff was prohibited in the European Union in 1992, but Sweden wasgranted derogation from the ban on its entry into the Union in 1994.

Smokeless tobacco is consumed without burning the product at the time ofuse. It can be used orally or nasally. Oral smokeless tobacco products areplaced in the mouth, cheek or lip and sucked (‘dipped’) or chewed. Tobaccopastes and powders are used in a similar manner and placed on the gums andteeth. Fine tobacco mixtures may also be inhaled and absorbed into the nasalpassages.

Globally, smokeless tobacco products vary widely with respect to their com-position and toxicant content. Many of the Asian products contain thepsychoactive substance arecoline from the areca nut mixed with tobacco. Inaddition, the products contain other flavourings, such as sugars, menthol andlicorice. Toombak is used primarily in Sudan and consists of tobacco andsodium bicarbonate.

Products on the market in the USA also vary with respect to their content ofnicotine (Table 3.11). Widely different levels of unprotonated (non-ionized,free-base) nicotine are available for absorption because of differences in thepH of the products in solution.The levels of TSNA also vary considerably in products sold in different partsof the world. Generally, moist snuff products sold in Europe and the USAcontain less TSNA than products on the Indian market or imported from SouthAsia to the United Kingdom (Tables 3.12–3.14). The levels of NNN andNNK in Sudanese toombak are extremely high, with concentrations of NNNreported to be 141–3085 g/g tobacco and of NNK 188–7870 g/g tobacco(Idris et al., 1991, 1995; Prokopczyk et al., 1995). This product also has avery high pH (8.0–11), resulting in a high concentration of non-ionized (un-protonated) nicotine.

Smokeless tobacco products may contain PAHs when flue-cured (fire-cured)tobacco is used. The levels of benzo[a]pyrene were 0.11–19.3 ng/g in a studyby McNeill et al. (2006; Table 3.12), which also showed low levels of metalsin the products tested. To date, 31 carcinogens have been identified in smoke-less tobacco; the 14 for which there is sufficient evidence of carcinogenicityin humans or animals are listed in Table 3.15 (IARC, in press).

There are no standardized methods for measuring toxicants or emissions fromsmokeless tobacco products. In the methods reported in the scientific litera-ture for analysing TSNA, samples are prepared in aqueous buffer or methanoland then extracted either directly with ethyl acetate or after purification by

121

solid-phase extraction (Österdahl et al., 2004; Stepanov et al., 2006). Ana-lytical identification and quantification of TSNA have been undertaken withliquid chromatography–triple mass spectrometry (Österdahl, Jannson,Paccou, 2004) or gas chromatography–thermal energy analyser instrumen-tation (Stepanov et al., 2006).

In principle, smokeless tobacco products could initially be regulated as in theinterim step described above for cigarettes. This would entail examination ofthe toxicity of the products as a function of the concentration of the examinedtoxicant and its potency, normalized for the non-ionized (unprotonated) con-centration of nicotine. So far, the database on toxic constituents in smokelesstobacco products is much more limited than that on toxicants in cigarettesand cigarette smoke. It is therefore more difficult to characterize the hazardof emissions from smokeless tobacco products. Also, smokeless tobaccoproducts vary much more in their form and composition than industriallyproduced cigarettes. The available published data, however, allow charac-terization of the hazard of NNN and NNK, and some products contain veryhigh levels of these carcinogens. For the same product type sold within acountry (e.g. moist snuff, chewing tobacco), it should be possible to set man-dated limits on the basis of measured variations within that product type andprohibit the marketing of those products that cannot meet the limit. Suchregulation would, however, require the availability of standard methods foranalytical quantification.

Table 3.11Chemical composition of the five leading brands of smokeless tobacco purchasedat various locations in the USA

Toxicant Skoal BanditStraight

(2% of market)

HawkenWintergreen

(1% ofmarket)

Skoal OriginalFine Cut

Wintergreen(39% of market)

CopenhagenSnuff

(42% ofmarket)

Kodiak(11% ofmarket)

pH 5.37 ± 0.12 5.71 ±0.1 7.46 ±0.14 8.00 ± 0.31 8.19 ± 0.11Nicotine(% dry weight)

2.29 ± 0.46 0.46 ± 0.02 2.81 ± 0.34 2.91 ± 0.18 2.5 ± 0.22

Nicotine (mg/g) 10.1 ± 0.8 3.2 ± 0.2 11.9 ± 1.3 12.0 ± 0.7 10.9 ± 0.8Unprotonatednicotine (%)a

0.23 ± 0.05 0.5 ± 0.11 22.0 ± 5.73 49.0 ± 16.7 59.7 ± 6.01

From Djordjevic et al. (1995)a The percentage of unprotonated nicotine, which depends on pH, was calculated according to the

Henderson-Hasselbach equation and by using a pKa value of 8.02 for nicotine (Henningfield,Radzius, Cone, 1995)

122

Tab

le 3

.12

To

xica

nts

in s

mo

kele

ss t

ob

acco

pro

du

cts

avai

lab

le in

th

e U

nit

ed K

ing

do

m a

nd

inte

rnat

ion

ally

Bra

nd

TS

NA

a

(μg

/g)

Ben

zo[a

]pyr

ene

(ng

/g)

ND

MA

(μg

/g)

Ch

rom

ium

(ng

/g)

Nic

kel

(ng

/g)

Ars

enic

(ng

/g)

Lea

d(n

g/g

)N

ico

tin

e(m

g/g

)A

vera

ge

pH

Fre

en

ico

tin

e(m

g/g

)

Pro

du

cts

pu

rch

ased

in t

he

Un

ited

Kin

gd

om

Gut

hka

prod

ucts

Man

ikch

ard

0.28

90.

40N

D0.

261.

220.

040.

153.

19.

193.

0T

ulsi

mix

1.43

61.

28N

D0.

331.

430.

060.

198.

29.

528.

0Z

arda

pro

duct

sH

akim

pur

y29

.705

0.32

ND

2.15

5.35

0.29

1.36

42.7

6.00

0.4

Dal

al m

isti

Zar

da1.

574

8.89

ND

0.87

2.09

0.11

1.14

8.6

6.15

0.1

Bab

a za

rda

(GP

)0.

716

2.04

ND

2.34

5.88

0.24

1.18

48.4

5.32

0.1

Too

th-c

lean

ing

pow

der

A. Q

uard

ir G

ull

5.11

75.

987

3.56

5.31

0.46

1.39

64.0

9.94

63.2

Drie

d to

bacc

o le

aves

Tob

acco

leaf

0.22

30.

11N

D2.

344.

370.

201.

0683

.55.

520.

3P

rod

uct

s p

urc

has

ed o

uts

ide

the

Un

ited

Kin

gd

om

Bab

a 12

0 (I

ndia

)2.

361

2.83

ND

2.08

2.94

0.40

1.56

55.0

4.88

0.04

Snu

s (S

wed

en)

0.47

81.

99N

D1.

542.

590.

300.

5015

.27.

866.

3A

riva

(US

A)

ND

0.40

ND

1.40

2.19

0.12

0.28

9.2

7.57

2.4

Cop

enha

gen

(US

A)

3.50

919

.33

ND

1.69

2.64

0.23

0.45

25.8

7.39

4.9

Fro

m M

cNei

ll et

al.

(200

6)T

SN

A, t

obac

co-s

peci

fic n

itros

amin

es; N

DM

A, N

-nitr

osod

imet

hyla

min

e; N

D, n

ot d

etec

ted,

det

ectio

n lim

it 5

ng/g

a Tot

al T

SN

A: N

NK

(4-

(N-n

itros

omet

hyla

min

o)-1

-(3-

pyrid

yl)-

1-bu

tano

ne),

NN

N (

N´-

nitr

oson

orni

cotin

e) a

nd N

´-ni

tros

oana

basi

ne

123

Table 3.14Tobacco-specific nitrosamines in new and conventional smokeless tobacco products

Product Tobacco-specific nitrosamine(μg/g product wet weight)

NNN NNK NAT NAB Total

Hard snuff Ariva 0.019 0.037 0.12 0.008 0.19Hard snuff Stonewall 0.056 0.043 0.17 0.007 0.28Swedish snus General 0.98 0.18 0.79 0.06 2.0Split-free tobacco Exalt purchased in Sweden 2.3 0.27 0.98 0.13 3.7Split-free tobacco Exalt purchased in USA 2.1 0.24 0.69 0.05 3.1Split-free tobacco Revel mint flavoured 0.62 0.033 0.32 0.018 0.99Split-free tobacco Revel wintergreen flavoured 0.64 0.032 0.31 0.017 1.0Smokeless tobacco Copenhagen snuff 2.2 0.75 1.8 0.12 4.8Smokeless tobacco Copenhagen long cut 3.9 1.6 1.9 0.13 7.5Smokeless tobacco Skoal long cut straight 4.5 0.47 4.1 0.22 9.2Smokeless Skoal bandits 0.9 0.17 0.24 0.014 1.3Smokeless Kodiak ice 2.0 0.29 0.72 0.063 3.1Smokeless Kodiak wintergreen 2.2 0.41 1.8 0.15 4.5

From Stepanov et al. (2006)NNN, N´-nitrosonornicotine; NNK, 4-(N-nitrosomethylamino)-1-(3-pyridyl)-1-butanone; NAT,

N´-nitrosoanatabine; NAB, N´-nitrosoanabasine

Table 3.13Chemical composition of smokeless tobacco products used in India

Toxicant Minimumvalue

Brand Maximumvalue

Brand

pH 5.21 Baba Zarda 120 10.1 Lime Mix–Miraj tobaccoAmmonia (μg/g) 3.8 Goa 1000 Zarda + Supari 5 280 Gai Chhap ZardaTotal carbonate (μg/g) 140 Dabur Red Toothpowder 2 040 Baba Zarda 120Nicotine (mg/g) 1.2 Khaini 10.16 Dentobac Creamy SnuffNNN (μg/g) 1.92 Goa 1000 Zarda + Supari 7.36 Baba Zarda 120NNK (μg/g) 4.38 IPCO Creamy Snuff 11.58 Shimla Jarda + S. SupariBenzo[a]pyrene (μg/g) < 0.0001 ? 0.94 IPCO Creamy SnuffCadmium (μg/g) 0.3 Click Eucalyptus 0.5 Baba Zarda 120Arsenic (μg/g) 0.1 Dabur Red Toothpowder 1.94 Moolchan Super Jarda

From Gupta, 2004 cited in IARC (2007)NNK, 4-(N-nitrosomethylamino)-1-(3-pyridyl)-1-butanone; NNN, N´-nitrosonornicotine

124

Table 3.15Concentrations of classified carcinogenic agents identified in smokeless tobaccoproductsa

Agent Type ofproductb

Concentrationrange (ng/mg)

IARCclassification

Benzo[a]pyrene MS, DS, Z, G < 0.1–90 2AUrethane CT 310–375 2BFormaldehyde MS, DS 1600–7400 1Acetaldehyde MS, DS 1400–27000 2BCrotonaldehyde MS, DS 200–2400 3N-Nitrosodimethylamine MS, CT ND–270 2AN-Nitrosopyrrolidine MS, CT ND–860 2BN-Nitrosopiperidine MS, CT ND–110 2BN-Nitrosomorpholine MS, CT ND–690 2BN´-Nitrososarcosine MS ND–6300 2BNNN MS, CT, Z, G 400–58000 1NNK MS, CT, Z, G ND–7800 1N´-Nitrosoanabasine MS, CT, Z, G Present–1190 3Nickel MS, G 180–2700 1Arsenic Z, G 40–290 1Chromium MS, Z, G 260–2340 1

Adapted from IARC (2007)MS, moist snuff; DS, dry snuff; Z, zarda product; G, gutkha product); CT, chewing tobacco; ND, not

detected; NNN, N´- nitrosonornicotine; NNK, 4-(N-nitrosomethylamino)-1-(3-pyridyl)-1-butanonea In addition, radioactive polonium-210, uranium-235 and -238 are present at picocurie levels in moist

snuff.b Not all carcinogens were measured in each product.

References

Ahrendt SA et al. (2001) Cigarette smoking is strongly associated with mu-tation of the K-ras gene in patients with primary adenocarcinoma of the lung.Cancer, 92:1525–1530.

Attorney Work Product (1992) The ‘palladium cigarette’ and its role in theCipollone trial. Bates No: 2021382482.

Australian Department of Health and Aging (2001) Cigarette emissionsdata, 2001. Available from: http://www.health.gov.au/internet/wcms/publishing.nsf/Content/health-pubhlth-strateg-drugs-tobacco-emis_data.htm.

Baker RR (2006) The generation of formaldehyde in cigarettes—Overviewand recent experiments. Food and Chemical Toxicology, 44:1799–1822.

Baan R et al. (2007) Carcinogenicity of alcoholic beverages. Lancet Oncol-ogy, 8:292–293.

125

Beland FA et al. (2005) High-performance liquid chromatography electro-spray ionization tandem mass spectrometry for the detection and quantitationof benzo[a]pyrene–DNA adducts. Chemical Research in Toxicology,18:1306–1315.

Benowitz NL et al. (1983) Smokers of low-yield cigarettes do not consumeless nicotine. New England Journal of Medicine, 309:139–142.

Boysen G, Hecht SS (2003) Analysis of DNA and protein adducts ofbenzo[a]pyrene in human tissues using structure-specific methods. MutationResearch, 543:17–30.

Brennan P et al. (2004) Pooled analysis of alcohol dehydrogenasegenotypes and head and neck cancer: a HuGE review. American Journal ofEpidemiology, 159:1–16.

Brown & Williamson (1977) Duolite. Internal communication. Bates No:504000094

Brown & Williamson (2002) Comments of Brown & Williamson TobaccoCorporation on the petitions for regulation of Brown & Williamson ‘Ad-vance’ cigarettes submitted by Nat’l Center for Tobacco-Free Kids, et al.,and the Society for Research on Nicotine and Tobacco. Dicket Nos:01P-0571, 02P-0206.

Cao J et al. (2007) Acetaldehyde, a major toxicant of tobacco smoke, en-hances behavioral, endocrine, and neuronal responses to nicotine in adoles-cent and adult rats. Neuropsychopharmacology, 32:2025–2035.

Castelao JE et al. (2001) Gender- and smoking-related bladder cancer risk.Journal of the National Cancer Institute, 93:538–545.

Charalambous J, Bullin Haines LI, Morgan JS (1987) Tobacco smoke filter.Filtrona Ltd. UK Patent GB 2,150,806.

Chen L et al. (2006) Quantitation of an acetaldehyde adduct in human leuko-cyte DNA and the effect of smoking cessation. Chemical Research in Toxi-cology, 20:108–113.

Chung FL, Chen HJC, Nath RG (1996) Lipid peroxidation as a potentialsource for the formation of exocyclic DNA adducts. Carcinogenesis,17:2105–2111.

Chung FL et al. (1999) Endogenous formation and significance of1,N2-propanodeoxyguanosine adducts. Mutation Research Fundamentaland Molecular Mechanisms of Mutagenesis, 424:71–81.

Cogliano V et al. (2004) Smokeless tobacco and tobacco-relatednitrosamines. Lancet Oncology, 5:708.

126

Cohen SM et al. (1992) Acrolein intiates rat urinary bladder carcinogenesis.Cancer Research, 52:3577–3581.

Counts ME et al. (2004) Mainstream smoke toxicant yields and predictingrelationships from a worldwide market sample of cigarette brands: ISOsmoking conditions. Regulatory Toxicology and Pharmacology, 39:111–134.

Counts ME et al. (2005) Smoke composition and predicting relationships forinternational commercial cigarettes smoked with three machine-smokingconditions. Regulatory Toxicology and Pharmacology, 41:185–227.

Deliconstantinos G, Villiotou V, Stavrides JC (1994) Scavenging effects ofhemoglobin and related heme containing compounds on nitric oxide, reactiveoxidants and carcinogenic volatile nitrosocompounds of cigarette smoke. Anew method for protection against the dangerous cigarette constituents. An-ticancer Research, 14:2717–2726.

Denissenko MF et al. (1996) Preferential formation of benzo[a]pyreneadducts at lung cancer mutational hot spots in P53. Science, 274:430–432.

Department of Health and Human Services (2004a) The health consequencesof tobacco use: a report of the Surgeon General. Rockville, Maryland, PublicHealth Service, Centers for Disease Control and Prevention, National Centerfor Chronic Disease Prevention and Health Promotion, Office on Smokingand Health.

Department of Health and Human Services (2004b) Report on carcinogens,11th Ed. Research Triangle Park, North Carolina.

Ding YS et al. (2005) Determination of 14 polycyclic aromatic hydrocarbonsin mainstream smoke from domestic cigarettes. Environmental Science andTechnology, 39:471–478.

Djordjevic MV et al. (1995) US commercial brands of moist snuff, 1994. I.Assessment of nicotine, moisture, and pH. Tobacco Control, 4:62–66.

Djordjevic MV, Stellman SD, Zang E (2000) Doses of nicotine and lungcarcinogens delivered to cigarette smokers. Journal of the National CancerInstitute, 92:106–111.

Dybing E et al. (1997) T25: a simplified carcinogenic potency index. De-scription of the system and study of correlations between carcinogenicpotency and species/site specificity and mutagenicity. Pharmacology andToxicology, 80:272–279.

European Commission (1999) Guidelines for setting specific concentrationlimits for carcinogens in Annex 1 of Directive 67/548/EEC. Inclusion of

127

potency considerations. Brussels, Commission Working Group on the Clas-sification and Labelling of Dangerous Substances.

Feng Z et al. (2006) Acrolein is a major cigarette-related lung cancer agent:preferential binding at p53 mutational hotspots and inhibition of DNA repair.Proceedings of the National Academy of Sciences of the United States,103:15404–15409.

Fennell TR et al. (2000) Hemoglobin adducts from acrylonitrile and ethyleneoxide in cigarette smokers: effects of glutathione S-transferase T1-null andM1-null genotypes. Cancer Epidemiology, Biomarkers and Prevention,9:705–712.

Ferguson R (2000) Reports on filler precursors of selected smoke con-stituents. Consultant evaluation. 10 October 2000. Bates No: 2082742847.

Feron VJ et al. (1991) Aldehydes: occurrence, carcinogenic potential, mech-anism of action and risk assessment. Mutation Research, 259:363–385.

Fowles J, Dybing E (2003) Application of toxicological risk assessmentprinciples to the chemical toxicants of cigarette smoke. Tobacco Control,12:424–430.

Furin OD, Gentry JS (1990) Confirmation of formaldehyde reduction viareaction with ammonia. Bates No: 509495561.

Gold LS et al. (1984) A carcinogenic potency database of the standardizedresults of animal bioassays. Environmental Health Perspectives, 58:9–319.

Gray N, Boyle P (2004) The case of the disappearing nitrosamines: a poten-tially global phenomenon. Tobacco Control, 13:13–16.

Halter HM, Ito TI (1978) Effect of tobacco reconstitution and expansion pro-cesses on smoke composition. Recent Advances in Tobacco Science,4:113–132.

Hatsukami DK et al. (2004) Evaluation of carcinogen exposure in people whoused ‘reduced exposure’ tobacco products. Journal of the National CancerInstitute, 96:844–852.

128

Health Canada (2000) Tobacco industry reporting: tobacco reporting regu-

e.html.

Health Canada (2004) Data for 2004 using the intense smoking protocol.

e.html

lations. http://www.hc-sc.gc.ca/hl-vs/pubs/tobac-tabac/tir-rft/emissions_

http://www.hc-sc.gc.ca/hl-vs/tobac-tabac/legislation/reg/indust/constitu_

Hecht SS (1999) Tobacco smoke carcinogens and lung cancer. Journal of theNational Cancer Institute, 91:1194–1210.

Hecht SS (2002) Human urinary carcinogen metabolites: biomarkers for in-vestigating tobacco and cancer. Carcinogenesis, 23:907–922.

Hecht SS (2003) Tobacco carcinogens, their biomarkers, and tobacco-induced cancer. Nature Reviews Cancer, 3:733–744.

Henningfield JE, Radzius A, Cone EJ (1995) Estimation of available nicotinecontent of six smokeless tobacco products. Tobacco Control, 4:57–61.

Hughes JR et al. (2004) Smoking behaviour and toxin exposure during sixweeks use of a potential reduced exposure product: Omni. Tobacco Con-trol, 13:175–180.

IARC (1986) Tobacco smoking. Lyon, International Agency for Research onCancer (IARC Monographs on the Evaluation of Carcinogenic Risk of Chem-icals to Humans, Vol. 38) p 95, IARC, Lyon, FR.

IARC (1987) Overall evaluations of carcinogenicity: an updating of IARCMonographs Volumes 1 to 42. Lyon, International Agency for Research onCancer (IARC Monographs on the Evaluation of the Carcinogenic Risk ofChemicals to Humans, Suppl. 7).

IARC (1995a) Acrolein. Lyon, International Agency for Research on Cancer(IARC Monographs on the Evaluation of Carcinogenic Risks to Humans,Vol. 63).

IARC (1995b) Dry cleaning, some chlorinated solvents and other industrialchemicals. Lyon, International Agency for Research on Cancer (IARC Mono-graphs on the Evaluation of Carcinogenic Risks to Humans, Vol. 63).

IARC (1999a) Re-evaluation of some organic chemicals, hydrazine and hy-drogen peroxide. Lyon, International Agency for Research on Cancer (IARCMonographs on the Evaluation of Carcinogenic Risks to Humans, Vol. 71).

IARC (1999b) Beryllium, cadmium, mercury, and exposures in the glassmanufacturing industry. Lyon, International Agency for Research on Cancer(IARC Monographs on the Evaluation of Carcinogenic Risks to Humans,Vol. 58).

IARC (2004) Tobacco smoke and involuntary smoking. Lyon, InternationalAgency for Research on Cancer (IARC Monographs on the Evaluation ofCarcinogenic Risks to Humans, Vol. 83).

IARC (2006) Formaldehyde, 2-butoxyethanol and 1-tert-butoxy-2-propanol. Lyon, International Agency for Research on Cancer (IARC Mono-graphs on the Evaluation of Carcinogenic Risks to Humans, Vol. 88).

129

IARC (2007) Smokeless tobacco and some related nitrosamines. Lyon, In-ternational Agency for Research on Cancer (IARC Monographs on theEvaluation of Carcinogenic Risks to Humans, Vol. 89).

IARC (in press) 1,3-Butadiene, ethylene oxide and vinyl halides (vinylfluoride, vinyl chloride and vinyl bromide). Lyon, International Agency forResearch on Cancer (IARC Monographs on the Evaluation of CarcinogenicRisks to Humans, Vol. 97).

Idris AM et al. (1991) Unusually high levels of carcinogenic nitrosamines inSudan snuff (toombak). Carcinogenesis, 12:1115–1118.

Idris AM, Ahmed HM, Malik MO (1995) Toombak dipping and cancer ofthe oral cavity in the Sudan: a case–control study. International Journal ofCancer, 63:477–480.

Ihrig AM (1972) Lorillard research and development project request: selec-tive removal of acrolein. Bates No: 80600686.

Ihrig A (1976) Lorillard monthly management status report: selective re-moval of acrolein. Bates No: 01409538.

Irwin WDE (1990) Filter materials for the reduction of selected minor smokecomponents: a literature review. British American Tobacco. Bates No:570262514.

ISO (2006) Final report on the work of ISO/TC 126/WG 9. Smoking methodsfor cigarettes. Geneva, International Organization for Standardization (ISO/TC 126/WG 9).

Jarvis MJ et al. (2001) Nicotine yield from machine-smoked cigarettes andnicotine intakes in smokers: evidence from a representative population sur-vey. Journal of the National Cancer Institute, 93:134–138.

Kensler CJ, Battista SP (1963) Components of cigarette smoke with ciliary-depressant activity. Their selective removal by filters containing activatedcharcoal granules. New England Journal of Medicine, 269:1161–1166.

Koller KB et al. (2005) Filter for selective removal of a gaseous compo-nent. Philip Morris and Co. US Patent 6,911,189 B1.

Kubrina LN et al. (1989) European Patent application 0,351,252.

Laugesen M, Fowles J (2005) Scope for regulation of cigarette smoke toxi-city: the case for including charcoal filters. New Zealand Medical Journal,118:U1402.

Leikauf GD (2002) Hazardous air pollutants and asthma. EnvironmentalHealth Perspectives, 110 (Suppl. 4):505–526.

130

Lesser C, Von Borstel RW (2003) Cigarette filter containing dry water anda porphyrin. US Patent No. 6530377B1.

Lewis LS, Norman AB, Robinson AL (1990) The evaluation of palladium/copper catalysts for CO removal. Bates No: 508537386.

Litzinger EF (1972) Duolite: properties and effects on smoke. Bates No:508006642.

Lodovici M et al. (2007) DNA solution in cigarette filters reduces polycyclicaromatic hydrocarbon (PAH) levels in mainstream tobacco smoke. Food andChemical Toxicology, 45:1752–1756.

Ludvig J, Miner B, Eisenberg MJ (2005) Smoking cessation in patients withcoronary artery disease. American Heart Journal, 149:565–572.

Maclure M et al. (1990) Decline of the hemoglobin adduct of 4-minobiphenylduring withdrawal from smoking. Cancer Research, 50:181–184.

Maillard F, Hada G (2005) Filtration method and filter consisting of nitrogen-containing cycles of heterocycles such as DNA or RNA. Sun Zero Co. USPatent 6,866,045.

Maselli JA (1986) RJ Reynolds internal communication. Bates No:505491763.

Melikian AA et al. (2007a) Effect of delivered dosage of cigarette smoketoxins on the levels of urinary biomarkers of exposure. Cancer Epidemiology,Biomarkers and Prevention,16:1408–1415.

Melikian AA et al. (2007b) Gender differences relative to smoking behaviorand emissions of toxins from mainstream cigarette smoke. Nicotine and To-bacco Research, 9:377–387.

McNeill A et al. (2006) Levels of toxins in oral tobacco products in the UK.Tobacco Control, 15:64–67.

Mills XE et al. (1995) Increased prevalence of K-ras oncogene mutations inlung adenocarcinma. Cancer Research, 55:1444–1447.

Nath RG et al. (1998) 1,N2-Propanodeoxyguanosine adducts: potential newbiomarkers of smoking-induced DNA damage in human oral tissue. CancerResearch, 58:581–584.

National Cancer Institute (2001) Risk associated with smoking cigarettes withlow machine-measured yields of tar and nicotine. Bethesda, Maryland, De-partment of Health and Human Services, Public Health Service, NationalInstitutes of Health (Smoking and Tobacco Control Monograph No. 13)http://news.findlaw.com/hdocs/docs/tobacco/nihnci112701cigstdy.pdf.

131

Nesnow S et al. (1998) Mechanistic relationships between DNA adducts,oncogene mutations, and lung tumorigenesis in strain A mice. ExperimentalLung Research, 24:395–405.

Norman A (1999) Cigarette design and materials. In: Davis EL, Nielsen MT,eds. Tobacco. Production, chemistry and technology. Oxford, BlackwellScience.

Norman V, Bryant HG Jr (1977) Tobacco composition. Liggett and MyersInc. US Patent 4,055,191.

Österdahl BG, Jannson C, Paccou A (2004) Decreased levels of tobacco-specific N-nitrosamines in moist snuff on the Swedish market. Journal ofAgricultural and Food Chemistry, 52:5085–5088.

Owens WF Jr (1978) Effect of cigarette paper on smoke yield and composi-tion. Recent Advances in Tobacco Science, 4:3–24.

Pappas RS et al. (2006) Cadmium, lead, and thallium in mainstream tobaccosmoke particulate. Food and Chemical Toxicology, 44:714–723.

Park S, Tretyakova N. (2004) Structural characterization of the major DNA–DNA cross-link of 1,2,3,4-diepoxybutane. Chemical Research in Toxicol-ogy, 17:129–136.

Peele DM, Riddick MG, Edwards ME (2001) Formation of tobacco-specificnitrosamines in flue cured tobacco. Recent Advances in Tobacco Science,27:3–12.

Pfeifer GP et al. (2002) Tobacco smoke carcinogens, DNA damage and p53mutations in smoking-associated cancers. Oncogene, 21:7435–7451.

Philip Morris (1997) Internal research on hemoglobin filters from CFile.Bates No: 2060531188.

Philip Morris (1998) An evaluation of formaldehyde reduction by varioussilica gels and -aminopropylsilane-silica gels in a flowing gas stream. BatesNo: 2075879330.

Philip Morris (2002) FDA regulation. Philip Morris USA 2002. Availablefrom: http://www.philipmorrisusa.com/company_news/fda_news_list.asp.

Polzin GM et al. (2008) Effect of charcoal-containing filters on gas phasevolatile organic compounds in mainstream cigarette smoke. Tobacco Con-trol, 17 (Suppl.1):i10–i16.

Prokopczyk B et al. (1995) Improved methodology for the quantitative as-sessment of tobacco-specific N-nitrosamines in tobacco by supercritical fluidextraction. Journal of Agricultural and Food Chemistry, 43:916–922.

132

Rainer NB, Feins IR (1980) Filter material for selective removal of aldehydesfor cigarette smoke. Philip Morris and Co., US Patent 4,182,743.

Riggs DM (1989) CO reduction in premier fuels by catalyst impregnation.Bates No: 514820423.

RJ Reynolds (1994) Carbon scrubber. Bates No: 514723150.

RJ Reynolds (1995) New product concept: carbon scrubber filter attach-ment. Bates No: 510357383.

Robinson EA (1998) Incorporation of proteins into filters of experimentalcigarettes. Memo to LH Gains. Lorillard Tobacco Co. Bates No: 96555438.

Rodgman A (1977) RJR inter-office memorandum: aldehydes. Bates No:503245386.

Rodgman A (2001) Studies of polycyclic aromatic hydrocarbons in cigarettemainstream smoke: identification, tobacco precursors, control of levels: areview. Beitraege zur Tobakforschung, 19:361–379.

Rodgman A, Perfetti TA (2006) The composition of cigarette smoke: achronology of the studies of four polycyclic aromatic hydrocarbons.Beitraege zur Tobakforschung, 22:208–254.

Rogers JC, Bluhm BK (1993a) Determination of 1,3-butadiene and isopreneconcentrations in mainstream cigarette smoke of cigarettes with and withoutcarbon scrubber filters. Bates No: 510636228.

Rogers JC, Bluhm BK (1993b) Determination of 1,3-butadiene and isopreneconcentrations in mainstream cigarette smoke of straight-grade cigaretteswith and without carbon scrubber filters. Bates No: 508441181.

Roux E et al. (1999) Human isolated airway contraction: interaction betweenair pollutants and passive sensitization. American Journal of RespiratoryCritical Care Medicine, 160:439–445.

Scheinberg IH (1974) UK Patent 1,369,067

Scherer G (2006) Carboxyhemoglobin and thiocyanate as biomarkers of ex-posure to carbon monoxide and hydrogen cyanide in tobacco smoke. Exper-imental Toxicology and Pathology, 58:101–124.

Scherer G et al. (2006) Influence of smoking charcoal filter tipped cigaretteson various biomarkers of exposure. Inhalation Toxicology, 18:821–829.

Scientific Advisory Committee on Tobacco Product Regulation (2002) Con-clusions and recommendations on health claims derived from ISO/FTCmethod to measure cigarette yield. Geneva, World Health Organization.

133

Scientific Advisory Committee on Tobacco Product Regulation (2003) State-ment of principles guiding the evaluation of new or modified tobacco prod-ucts. Geneva, World Health Organization.

Seeman JI, Dixon M, Haussmann HJ (2002) Acetaldehyde in mainstreamtobacco smoke: formation and occurrence in smoke and bioavailability in thesmoker. Chemical Research in Toxicology, 15:1331–1350.

Silverstein P (1992) Smoking and wound healing. American Journal ofMedicine, 93:22S–24S.

Skipper PL, Tannenbaum SR (1990) Protein adducts in the molecular dosime-try of chemical carcinogens. Carcinogenesis, 11:507–518.

Smith C et al. (1994a) Comparison study utilizing subject’s usual brandcigarette, (3) carbon scrubber prototypes, and (1) grooved filter prototype todetermine their effects on human smoker’s breath carbon monoxide and %carboxyhemoglobin levels. Bates No: 510636183.

Smith C et al. (1994b) HRRC proposal: comparison study utilizing subject’susual brand cigarette, (3) carbon scrubber prototypes, and (1) grooved filterprototype to determine their effects on human smoker’s breath carbonmonoxide and % carboxyhemoglobin levels. Bates No: 510641328.

Smith LE et al. (2000) Targeting of lung cancer mutational hotspots by poly-cyclic aromatic hydrocarbons. Journal of the National Cancer Institute,92:803–811.

Stavridis I, Deliconstantinos G (1999) Removal of noxious oxidants and car-cinogenic volatile nitrosocompounds from cigarette smoke using biologicalsubstances. US Patent 5,909,736.

Stepanov I et al. (2006) Tobacco-specific nitrosamines in new tobacco prod-ucts. Nicotine and Tobacco Research, 8:309–313.

Stiles M et al. (1994) Human smoking behavior study examining carbonscrubber filter technology. Bates No: 511823912.

Straif K et al. (2005) Carcinogenicity of polycyclic aromatic hydrocarbons.Lancet Oncology, 6:931–932.

Stratton K et al. (2001) Clearing the smoke. Assessing the science base fortobacco harm reduction. Washington DC, National Academy Press.

Tennessee Eastman (1959) Composition of tobacco smoke. III. Effects of theextraction of tobacco on amount of benzo[a]pyrene in cigarette smoke tar.Kingsport, Tennessee (Tennessee Eastman Corp. Research Report No.4-1201-3).

134

Tiggelbeck D (1967) Comments on selective cigarette smoke filtration. In:eds. Wynder EL, Hoffmann D, eds. Toward a Less Harmful Cigarette.Bethesda, Maryland, National Cancer Institute (National Cancer InstituteMonograph 28:249–258).

Tiggelbeck D (1976) Vapor phase modification. An under-utilized technol-ogy. In: Wynder EL, Hoffmann D, Gori GB, eds. Proceedings of the ThirdWorld Conference on Smoking and Health. Bethesda, Maryland, Departmentof Health, Education and Welfare, Public Health Service, National Institutesof Health, National Cancer Institute, pp. 507–514 (DHEW Publication No.(NIH) 76-1221).

Tolman TW (1970) The theoretical and practical aspects of carbon monoxidefiltration. British-American Tobacco Co. Ltd (Report No. RD.687-R). BatesNo: 650315732.

Tolman TW (1971) Further work on the filtration of carbon monoxide andnitric oxide. British-American Tobacco Co. Ltd. (Report No. RD. 802-R).Bates No: 650316111.

Tolman TW (1973) The oxidation of carbon monoxide in cigarette smokecatalysed by manganese dioxide. British-American Tobacco Co. Ltd (ReportNo. RD.1011-R. 1973). Bates No: 570508384.

Townsend DE (1978) A reactor system for the rapid initial evaluation ofprospective filter additives for carbon monoxide removal from cigarettesmoke. RJ Reynolds Tobacco Co. Bates No: 500607965.

Tretyakova NT et al. (2002) Formation of benzo[a]pyrene diol epoxide-DNAadducts at specific guanines within K-ras and p53 gene sequences: stableisotope-labeling mass spectrometry approach. Biochemistry, 41:9535–9544.

Valavanidis A, Haralambous E (2001) A comparative study by electron para-magnetic resonance of free radical species in the mainstream and sidestreamsmoke of cigarettes with conventional acetate filters and ‘bio-filters’. RedoxReport, 6:161–171.

Vleeming W, Rambali B, Opperhuizen A (2002) The role of nitric oxide incigarette smoking and nicotine addiction. Nicotine and Tobacco Research,4:341–348.

Weinberger B et al. (2001) The toxicology of inhaled nitric oxide. Toxico-logical Science, 59:5–16.

Westra WH et al. (1993) K-ras oncogene activation in lung adenocarcinomasfrom former smokers. Cancer, 72:432–438.

135

WHO. Guiding principles for the development of tobacco product researchand testing capacity and proposed protocols for the initiation of tobaccoproduct testing. Geneva, World Health Organization, Study Group on To-bacco Product Regulation. http://www.who.int/tobacco/global_interaction/tobreg/goa_2003_principles/en/index.html.

WHO (2007) Setting maximal limits for toxic constituents in cigarettesmoke. Geneva, World Health Organization, Study Group on Tobacco Prod-uct Regulation. http://www.who.int/tobacco/global_interaction/tobreg/tsr/en/index.html.

Wu W, Ashley D, Watson C (2003) Simultaneous determination of 5 tobacco-specific nitrosamines in mainstream cigarette smoke by isotope dilutionliquid chromatography/electrospray ionization tandem mass spectrometry.Analytical Chemistry, 75:4827–4832.

Wynder EL, Hoffman D (1967) Tobacco and tobacco smoke: studies in ex-perimental carcinogenesis. New York, Academic Press.

Xue L, Thomas CE, Koller KB (2002) Mainstream smoke gas phase filtrationperformance of adsorption materials evaluated with a puff-by-puff multiplexGC-MS method. Beitraege zur Tabakforschung International, 20:251–256.

Zhang S et al. (2006) Analysis of crotonaldehyde- and acetaldehyde-derived1,N2-propanodeoxyguanosine adducts in DNA from human tissues usingliquid chromatography-electrosopray ionization-tandem mass spectrometry.Chemical Research in Toxicology, 19:1386–1392.

136

Annex 3.1 Levels of toxicants per milligram of nicotine forinternational Philip Morris brands, Canadian brands andAustralian brands

137

Tab

le A

1.1.

Yie

lds

per

mill

igra

m o

f n

ico

tin

e o

f to

xica

nts

rep

ort

ed t

o H

ealt

h C

anad

a (2

004)

Bra

nd

Nic

oti

ne

Car

bo

nm

on

oxi

de

Am

mo

nia

1-A

min

on

aph

thal

ene

2-A

min

on

aph

thal

ene

3-A

min

ob

iph

enyl

Exp

ort ‘

A’ U

ltra

Ligh

t Reg

ular

Siz

e2.

499.

078.

5410

.48

6.71

1.60

Van

tage

Slim

s E

xtra

Lig

ht 4

100

’s2.

929.

078.

5011

.16

7.05

1.64

Exp

ort ‘

A’ M

ediu

m K

ing

Siz

e2.

659.

6610

.60

8.45

6.19

1.68

Exp

ort ‘

A’ F

ull F

lavo

ur K

ing

Siz

e2.

7910

.04

11.0

88.

425.

811.

64C

rave

n A

Ext

ra L

ight

Kin

g S

ize

2.38

11.7

79.

1211

.23

6.98

1.70

Exp

ort ‘

A’ M

ediu

m R

egul

ar S

ize

2.51

10.8

39.

0311

.73

7.25

1.72

Exp

ort ‘

A’ E

xtra

Lig

ht R

egul

ar S

ize

2.48

9.05

8.77

10.0

96.

801.

72M

ore

Reg

ular

120

’s2.

7014

.07

13.4

412

.93

8.04

2.06

Exp

ort ‘

A’ L

ight

/ S

peci

al L

ight

Kin

g S

ize

2.60

9.92

11.0

88.

966.

231.

65

Exp

ort ‘

A’ L

ight

/ S

peci

al L

ight

Reg

ular

Siz

e2.

3710

.89

9.77

12.4

67.

732.

01

Mor

e M

enth

ol 1

20’s

2.60

15.6

513

.77

12.2

78.

041.

98V

anta

ge U

ltra

Ligh

t 1 K

ing

Siz

e2.

1011

.32

9.02

12.0

17.

561.

79E

xpor

t ‘A

’ Ext

ra L

ight

Kin

g S

ize

2.27

10.7

011

.10

9.16

6.34

1.68

Exp

ort ‘

A’ M

ild K

ing

Siz

e2.

609.

7711

.69

7.46

5.38

1.56

Exp

ort ‘

A’ M

ild R

egul

ar S

ize

2.60

9.77

11.6

97.

465.

381.

56E

xpor

t ‘A

’ Ultr

a Li

ght K

ing

Siz

e2.

579.

8811

.83

7.55

5.45

1.58

Num

ber

7 K

ing

Siz

e2.

6011

.79

9.97

13.8

48.

332.

06C

rave

n M

ilds

Ultr

a M

ild K

ing

Siz

e1.

5914

.88

9.88

13.8

08.

932.

39

138

Cam

el L

ight

Kin

g S

ize

1.74

14.5

415

.06

15.7

510

.75

2.66

Win

ston

Lig

ht K

ing

Siz

e1.

7414

.54

15.0

615

.75

10.7

52.

66E

xpor

t Pla

in R

egul

ar S

ize

2.60

8.38

12.4

212

.73

7.42

1.62

Rot

hman

s S

peci

al M

ild K

ing

Siz

e2.

5611

.63

9.73

11.9

28.

472.

11W

inst

on L

ight

100

’s2.

1914

.06

15.4

315

.75

10.5

52.

59W

inst

on 1

00’s

2.60

11.3

115

.38

17.1

910

.65

2.57

Gau

lois

es B

lond

es K

ing

Siz

e2.

5511

.58

18.8

212

.15

7.45

2.16

Pla

yer’s

Ext

ra L

ight

Reg

ular

Siz

e2.

1610

.47

8.99

11.8

77.

661.

82C

amel

Kin

g S

ize

2.27

12.5

116

.43

16.3

410

.22

2.48

Sal

em K

ing

Siz

e2.

2712

.51

16.4

316

.34

10.2

22.

48W

inst

on K

ing

Siz

e2.

2712

.51

16.4

316

.34

10.2

22.

48E

xpor

t ‘A

’ Ful

l Fla

vour

Reg

ular

Siz

e2.

6410

.43

10.6

511

.46

7.14

1.82

Du

Mau

rier E

xtra

Lig

ht R

egul

ar S

ize

2.25

12.0

89.

6211

.43

7.15

1.70

Cra

ven

A K

ing

Siz

e2.

6911

.75

10.7

110

.28

6.26

1.58

Sal

em L

ight

Kin

g S

ize

2.32

13.1

515

.17

13.4

99.

312.

23M

atin

ee E

xtra

Lig

ht K

ing

Siz

e1.

7613

.42

9.52

11.3

37.

181.

79C

amel

Pla

in R

egul

ar S

ize

2.45

9.14

18.7

819

.63

11.3

92.

63D

u M

aurie

r Li

ght K

ing

Siz

e2.

5512

.04

9.65

10.9

86.

811.

63V

isco

unt E

xtra

Mild

Kin

g S

ize

1.72

14.6

211

.13

12.2

08.

192.

15S

port

sman

Pla

in R

egul

ar S

ize

2.91

7.81

10.3

39.

985.

161.

14V

atna

ge M

ediu

m 7

Kin

g S

ize

2.42

11.6

49.

9711

.19

7.13

1.74

Ben

son

& H

edge

s D

elux

e M

enth

olU

ltra

Ligh

ts 1

00’s

2.52

10.9

39.

559.

015.

711.

66

Pla

yer’s

Lig

ht R

egul

ar S

ize

2.19

10.5

29.

6911

.41

7.22

1.87

Van

tage

Max

15

Kin

g S

ize

3.10

10.4

58.

559.

354.

771.

18C

rave

n M

enth

ol K

ing

Siz

e2.

2513

.87

8.84

10.4

26.

771.

72M

atin

ee S

ilver

Reg

ular

Siz

e1.

9711

.22

9.85

10.9

66.

141.

40

139

Bel

mon

t Mild

s R

egul

ar S

ize

2.35

10.3

99.

377.

675.

111.

23B

elm

ont M

ilds

Kin

g S

ize

2.34

11.3

19.

839.

836.

411.

50P

laye

r’s L

ight

Kin

g S

ize

2.52

12.3

510

.77

12.6

57.

902.

00G

aulo

ises

Blo

ndes

Lig

hts

Kin

g S

ize

1.97

13.3

618

.29

13.7

28.

642.

59V

anta

ge R

ich

12 K

ing

Siz

e2.

6010

.81

9.69

11.6

26.

041.

39G

aulo

ises

Blo

ndes

Ultr

a Li

ght

Kin

g S

ize

1.68

15.7

319

.00

15.4

49.

502.

85

Mar

k te

n P

lain

Kin

g S

ize

3.21

7.99

10.9

110

.28

5.30

1.22

Du

Mau

rier

Ligh

t Reg

ular

Siz

e1.

9614

.09

10.6

615

.06

9.24

2.45

Du

Mau

rier

Kin

g S

ize

2.49

12.2

79.

8212

.69

7.84

1.97

Pla

yer’s

Reg

ular

Siz

e2.

4611

.22

10.7

010

.31

6.62

1.81

Du

Mau

rier

Ext

ra L

ight

Kin

g S

ize

2.14

12.4

89.

6110

.89

7.17

1.91

Pla

yer’s

Spe

cial

Ble

nd K

ing

Siz

e2.

639.

3211

.60

11.9

46.

621.

44P

laye

r’s S

peci

al B

lend

Reg

ular

Siz

e2.

449.

6312

.30

12.1

77.

011.

62P

laye

r’s P

lain

Reg

ular

Siz

e2.

687.

8711

.19

12.1

35.

751.

31P

laye

r’s E

xtra

Lig

ht K

ing

Siz

e2.

3813

.60

9.70

12.0

97.

621.

98D

u M

aurie

r R

egul

ar S

ize

2.06

13.2

811

.03

11.2

87.

041.

81C

anyo

n R

egul

ar S

ize

1.87

14.7

115

.40

8.34

14.2

21.

63D

akar

Reg

ular

Siz

e1.

8714

.71

15.4

08.

3414

.22

1.63

Dis

cret

ion

Reg

ular

Siz

e1.

8714

.71

15.4

08.

3414

.22

1.63

Fin

e R

egul

ar S

ize

1.87

14.7

115

.40

8.34

14.2

21.

63G

ipsy

Reg

ular

Siz

e1.

8714

.71

15.4

08.

3414

.22

1.63

Sel

esta

Reg

ular

Siz

e1.

8714

.71

15.4

08.

3414

.22

1.63

Sm

okin

g R

egul

ar S

ize

1.87

14.7

115

.40

8.34

14.2

21.

63T

rem

blay

Reg

ular

Siz

e1.

8714

.71

15.4

08.

3414

.22

1.63

Can

yon

Kin

g S

ize

2.13

14.4

614

.13

8.26

14.1

81.

61D

akar

Kin

g S

ize

2.13

14.4

614

.13

8.26

14.1

81.

61

140

Dis

cret

ion

Kin

g S

ize

2.13

14.4

614

.13

8.26

14.1

81.

61F

ine

Kin

g S

ize

2.13

14.4

614

.13

8.26

14.1

81.

61G

ipsy

Kin

g S

ize

2.13

14.4

614

.13

8.26

14.1

81.

61S

eles

ta K

ing

Siz

e2.

1314

.46

14.1

38.

2614

.18

1.61

Sm

okin

g K

ing

Siz

e2.

1314

.46

14.1

38.

2614

.18

1.61

Tre

mbl

ay K

ing

Siz

e2.

1314

.46

14.1

38.

2614

.18

1.61

Dak

ar L

ight

Kin

g S

ize

2.06

14.8

512

.28

8.74

14.4

71.

68F

ine

Ligh

t Kin

g S

ize

2.06

14.8

512

.28

8.74

14.4

71.

68G

ipsy

Lig

ht K

ing

Siz

e2.

0614

.85

12.2

88.

7414

.47

1.68

Rod

eo K

ing

Siz

e2.

0614

.85

12.2

88.

7414

.47

1.68

Sel

esta

Lig

ht K

ing

Siz

e2.

0614

.85

12.2

88.

7414

.47

1.68

Sm

okin

g Li

ght K

ing

Siz

e2.

0614

.85

12.2

88.

7414

.47

1.68

Tre

mbl

ay L

ight

Kin

g S

ize

2.06

14.8

512

.28

8.74

14.4

71.

68D

akar

Lig

ht R

egul

ar S

ize

1.81

14.6

418

.07

8.51

14.1

41.

64F

ine

Ligh

t Reg

ular

Siz

e1.

8114

.64

18.0

78.

5114

.14

1.64

Gip

sy L

ight

Reg

ular

Siz

e1.

8114

.64

18.0

78.

5114

.14

1.64

Rod

eo R

egul

ar S

ize

1.81

14.6

418

.07

8.51

14.1

41.

64S

eles

ta L

ight

Reg

ular

Siz

e1.

8114

.64

18.0

78.

5114

.14

1.64

Sm

okin

g Li

ght R

egul

ar S

ize

1.81

14.6

418

.07

8.51

14.1

41.

64**

* M

eans

***

2.25

12.5

412

.65

10.7

89.

691.

80**

* S

tand

ard

devi

atio

n **

*0.

352.

193.

032.

673.

460.

36**

* C

oeffi

cien

t of v

aria

tion

***

0.15

0.17

0.24

0.25

0.36

0.20

Max

imum

3.21

15.7

319

.00

19.6

314

.47

2.85

75th

per

cent

ile2.

5514

.64

15.3

812

.15

14.1

81.

979t

h pe

rcen

tile

2.64

14.8

518

.07

15.1

314

.22

2.48

Med

ian

2.25

12.5

112

.28

10.3

18.

191.

68M

inim

um1.

597.

818.

507.

464.

771.

14

141

Bra

nd

4-A

min

ob

iph

enyl

Ben

zo[a

]pyr

ene

Fo

rmal

deh

yde

Ace

tald

ehyd

eA

ceto

ne

Acr

ole

in

Exp

ort ‘

A’ U

ltra

Ligh

t Reg

ular

Siz

e1.

205.

1960

.76

431.

1823

1.86

58.4

0V

anta

ge S

lims

Ext

ra L

ight

4 1

00’s

1.22

5.33

46.2

042

2.65

237.

8159

.09

Exp

ort ‘

A’ M

ediu

m K

ing

Siz

e1.

425.

7445

.66

443.

0221

8.11

51.7

0E

xpor

t ‘A

’ Ful

l Fla

vour

Kin

g S

ize

1.36

5.88

50.8

247

4.91

235.

1355

.20

Cra

ven

A E

xtra

Lig

ht K

ing

Siz

e1.

355.

9763

.96

545.

7027

7.99

77.7

9E

xpor

t ‘A

’ Med

ium

Reg

ular

Siz

e1.

336.

0168

.02

489.

8025

5.26

62.9

0E

xpor

t ‘A

’ Ext

ra L

ight

Reg

ular

Siz

e1.

346.

0557

.01

387.

7222

0.95

57.1

5M

ore

Reg

ular

120

’s1.

696.

0739

.19

619.

2631

1.11

62.5

9E

xpor

t ‘A

’ Lig

ht /

Spe

cial

Lig

ht K

ing

Siz

e1.

346.

0846

.62

444.

2322

3.08

51.9

2E

xpor

t ‘A

’ Lig

ht /

Spe

cial

Lig

ht R

egul

ar S

ize

1.52

6.11

65.4

850

0.32

275.

3466

.78

Mor

e M

enth

ol 1

20’s

1.71

6.23

42.4

266

0.77

327.

3167

.69

Van

tage

Ultr

a Li

ght 1

Kin

g S

ize

1.38

6.24

70.1

752

4.96

284.

3971

.70

Exp

ort ‘

A’ E

xtra

Lig

ht K

ing

Siz

e1.

446.

2653

.30

519.

3824

8.02

62.5

6E

xpor

t ‘A

’ Mild

Kin

g S

ize

1.25

6.38

67.6

247

2.69

223.

4657

.31

Exp

ort ‘

A’ M

ild R

egul

ar S

ize

1.25

6.38

67.6

247

2.69

223.

4657

.31

Exp

ort ‘

A’ U

ltra

Ligh

t Kin

g S

ize

1.26

6.46

68.4

047

8.21

226.

0757

.98

Num

ber

7 K

ing

Siz

e1.

576.

4961

.26

517.

5728

2.62

78.6

9C

rave

n M

ilds

Ultr

a M

ild K

ing

Siz

e1.

856.

5882

.37

716.

9638

6.06

99.5

2C

amel

Lig

ht K

ing

Siz

e2.

216.

7249

.31

686.

7835

9.77

73.5

6W

inst

on L

ight

Kin

g S

ize

2.21

6.72

49.3

168

6.78

359.

7773

.56

Exp

ort P

lain

Reg

ular

Siz

e1.

336.

8154

.23

389.

6219

6.15

50.3

8R

othm

ans

Spe

cial

Mild

Kin

g S

ize

1.61

6.86

67.0

750

6.07

280.

5580

.73

142

Win

ston

Lig

ht 1

00’s

2.11

6.94

39.4

565

7.99

337.

4468

.95

Win

ston

100

’s2.

006.

9643

.65

544.

6228

2.69

57.3

1G

aulo

ises

Blo

ndes

Kin

g S

ize

1.84

7.06

44.3

054

8.02

264.

9961

.15

Pla

yer’s

Ext

ra L

ight

Reg

ular

Siz

e1.

377.

1669

.21

505.

9727

9.46

65.8

8C

amel

Kin

g S

ize

2.00

7.18

50.6

658

6.34

303.

9663

.00

Sal

em K

ing

Siz

e2.

007.

1850

.66

586.

3430

3.96

63.0

0W

inst

on K

ing

Siz

e2.

007.

1850

.66

586.

3430

3.96

63.0

0E

xpor

t ‘A

’ Ful

l Fla

vour

Reg

ular

Siz

e1.

377.

2458

.97

437.

3123

8.21

60.7

5D

u M

aurie

r E

xtra

Lig

ht R

egul

ar S

ize

1.25

7.31

60.8

050

9.90

270.

5876

.58

Cra

ven

A K

ing

Siz

e1.

207.

3954

.00

463.

7024

9.95

74.1

6S

alem

Lig

ht K

ing

Siz

e1.

767.

4150

.00

607.

7630

5.17

64.2

2M

atin

ee E

xtra

Lig

ht K

ing

Siz

e1.

347.

4781

.23

615.

5532

3.24

86.6

6C

amel

Pla

in R

egul

ar S

ize

2.00

7.51

52.2

442

4.90

221.

2246

.94

Du

Mau

rier

Ligh

t Kin

g S

ize

1.23

7.54

58.1

151

0.09

271.

3776

.28

Vis

coun

t Ext

ra M

ild K

ing

Siz

e1.

597.

5585

.42

641.

6933

5.48

96.8

4S

port

sman

Pla

in R

egul

ar S

ize

0.86

7.57

63.3

336

4.48

194.

8047

.84

Vat

nage

Med

ium

7 K

ing

Siz

e1.

317.

6964

.68

474.

1127

7.74

65.6

5B

enso

n &

Hed

ges

Del

uxe

Men

thol

Ultr

a Li

ghts

100

’s1.

247.

7072

.60

506.

6226

5.77

77.6

0

Pla

yer’s

Lig

ht R

egul

ar S

ize

1.35

7.74

89.8

351

6.22

289.

6876

.73

Van

tage

Max

15

Kin

g S

ize

0.91

7.77

65.1

043

0.97

222.

5848

.71

Cra

ven

Men

thol

Kin

g S

ize

1.29

7.83

93.5

860

9.68

313.

3985

.97

Mat

inee

Silv

er R

egul

ar S

ize

1.20

7.87

88.3

259

8.98

308.

1271

.07

Bel

mon

t Mild

s R

egul

ar S

ize

0.94

8.09

68.9

947

6.54

242.

7457

.92

Bel

mon

t Mild

s K

ing

Siz

e1.

208.

1264

.57

535.

7825

7.84

61.1

5

143

Pla

yer’s

Lig

ht K

ing

Siz

e1.

538.

1360

.38

528.

5229

9.86

80.6

5G

aulo

ises

Blo

ndes

Lig

hts

Kin

g S

ize

2.29

8.13

44.7

166

0.03

319.

0972

.15

Van

tage

Ric

h 12

Kin

g S

ize

1.05

8.27

77.0

847

9.23

246.

9255

.00

Gau

lois

es B

lond

es U

ltra

Ligh

t Kin

g S

ize

2.61

8.31

43.3

476

6.50

368.

1181

.93

Mar

k te

n P

lain

Kin

g S

ize

0.87

8.41

71.3

739

7.37

205.

7053

.61

Du

Mau

rier

Ligh

t Reg

ular

Siz

e1.

908.

5373

.21

595.

4332

7.51

91.0

8D

u M

aurie

r K

ing

Siz

e1.

418.

5777

.17

517.

5929

1.39

83.6

6P

laye

r’s R

egul

ar S

ize

1.29

8.71

74.0

344

3.81

244.

6470

.16

Du

Mau

rier

Ext

ra L

ight

Kin

g S

ize

1.45

8.93

72.1

954

7.00

314.

8081

.64

Pla

yer’s

Spe

cial

Ble

nd K

ing

Siz

e1.

239.

0159

.32

506.

4625

0.19

54.7

5P

laye

r’s S

peci

al B

lend

Reg

ular

Siz

e1.

369.

1471

.72

523.

3625

2.87

57.3

8P

laye

r’s P

lain

Reg

ular

Siz

e1.

049.

1880

.60

406.

7221

0.07

52.6

1P

laye

r’s E

xtra

Lig

ht K

ing

Siz

e1.

489.

2559

.15

531.

1229

5.11

86.3

5D

u M

aurie

r R

egul

ar S

ize

1.35

9.59

79.8

753

1.02

296.

5583

.78

Can

yon

Reg

ular

Siz

e2.

3416

.79

116.

0465

1.87

318.

1878

.61

Dak

ar R

egul

ar S

ize

2.34

16.7

911

6.04

651.

8731

8.18

78.6

1D

iscr

etio

n R

egul

ar S

ize

2.34

16.7

911

6.04

651.

8731

8.18

78.6

1F

ine

Reg

ular

Siz

e2.

3416

.79

116.

0465

1.87

318.

1878

.61

Gip

sy R

egul

ar S

ize

2.34

16.7

911

6.04

651.

8731

8.18

78.6

1S

eles

ta R

egul

ar S

ize

2.34

16.7

911

6.04

651.

8731

8.18

78.6

1S

mok

ing

Reg

ular

Siz

e2.

3416

.79

116.

0465

1.87

318.

1878

.61

Tre

mbl

ay R

egul

ar S

ize

2.34

16.7

911

6.04

651.

8731

8.18

78.6

1C

anyo

n K

ing

Siz

e2.

2816

.81

97.6

564

6.48

319.

2577

.46

Dak

ar K

ing

Siz

e2.

2816

.81

97.6

564

6.48

319.

2577

.46

Dis

cret

ion

Kin

g S

ize

2.28

16.8

197

.65

646.

4831

9.25

77.4

6F

ine

Kin

g S

ize

2.28

16.8

197

.65

646.

4831

9.25

77.4

6

144

Gip

sy K

ing

Siz

e2.

2816

.81

97.6

564

6.48

319.

2577

.46

Sel

esta

Kin

g S

ize

2.28

16.8

197

.65

646.

4831

9.25

77.4

6S

mok

ing

Kin

g S

ize

2.28

16.8

197

.65

646.

4831

9.25

77.4

6T

rem

blay

Kin

g S

ize

2.28

16.8

197

.65

646.

4831

9.25

77.4

6D

akar

Lig

ht K

ing

Siz

e2.

4717

.38

103.

4065

7.77

322.

3380

.58

Fin

e Li

ght K

ing

Siz

e2.

4717

.38

103.

4065

7.77

322.

3380

.58

Gip

sy L

ight

Kin

g S

ize

2.47

17.3

810

3.40

657.

7732

2.33

80.5

8R

odeo

Kin

g S

ize

2.47

17.3

810

3.40

657.

7732

2.33

80.5

8S

eles

ta L

ight

Kin

g S

ize

2.47

17.3

810

3.40

657.

7732

2.33

80.5

8S

mok

ing

Ligh

t Kin

g S

ize

2.47

17.3

810

3.40

657.

7732

2.33

80.5

8T

rem

blay

Lig

ht K

ing

Siz

e2.

4717

.38

103.

4065

7.77

322.

3380

.58

Dak

ar L

ight

Reg

ular

Siz

e2.

3617

.90

118.

2364

4.20

316.

5779

.56

Fin

e Li

ght R

egul

ar S

ize

2.36

17.9

011

8.23

644.

2031

6.57

79.5

6G

ipsy

Lig

ht R

egul

ar S

ize

2.36

17.9

011

8.23

644.

2031

6.57

79.5

6R

odeo

Reg

ular

Siz

e2.

3617

.90

118.

2364

4.20

316.

5779

.56

Sel

esta

Lig

ht R

egul

ar S

ize

2.36

17.9

011

8.23

644.

2031

6.57

79.5

6S

mok

ing

Ligh

t Reg

ular

Siz

e2.

3617

.90

118.

2364

4.20

316.

5779

.56

***

Mea

ns *

**1.

7710

.52

77.3

556

6.54

289.

0671

.31

***

Sta

ndar

d de

viat

ion

***

0.52

4.73

24.8

492

.69

42.5

111

.71

***

Coe

ffici

ent o

f var

iatio

n **

*0.

290.

450.

320.

160.

150.

16M

axim

um2.

6117

.90

118.

2376

6.50

386.

0699

.52

75th

per

cent

ile2.

2916

.79

97.6

564

6.48

319.

2579

.56

9th

perc

entil

e2.

3617

.38

116.

0465

7.77

322.

5181

.70

Med

ian

1.69

8.09

71.3

758

6.34

303.

9676

.73

Min

imum

0.86

5.19

39.1

936

4.48

194.

8046

.94

145

Bra

nd

Pro

pio

nal

deh

yde

Cro

ton

ald

ehyd

eB

uty

rald

ehyd

eH

ydro

gen

cya

nid

eM

ercu

ryL

ead

Exp

ort ‘

A’ U

ltra

Ligh

t Reg

ular

Siz

e36

.94

25.9

224

.55

87.5

22.

09V

anta

ge S

lims

Ext

ra L

ight

4 1

00’s

39.2

524

.92

25.4

593

.03

1.97

Exp

ort ‘

A’ M

ediu

m K

ing

Siz

e37

.36

21.8

524

.26

112.

452.

54E

xpor

t ‘A

’ Ful

l Fla

vour

Kin

g S

ize

41.1

823

.15

26.1

310

3.23

2.52

Cra

ven

A E

xtra

Lig

ht K

ing

Siz

e43

.73

31.1

631

.62

118.

472.

87E

xpor

t ‘A

’ Med

ium

Reg

ular

Siz

e40

.52

29.2

128

.03

106.

642.

50E

xpor

t ‘A

’ Ext

ra L

ight

Reg

ular

Siz

e35

.67

21.8

522

.43

86.0

72.

08M

ore

Reg

ular

120

’s53

.33

29.0

036

.30

219.

263.

1015

.96

Exp

ort ‘

A’ L

ight

/ S

peci

al L

ight

Kin

g S

ize

38.4

621

.46

24.5

410

8.46

2.66

Exp

ort ‘

A’ L

ight

/ S

peci

al L

ight

Reg

ular

Siz

e45

.90

28.7

629

.08

115.

792.

64

Mor

e M

enth

ol 1

20’s

58.0

831

.42

37.6

923

0.38

3.07

16.5

4V

anta

ge U

ltra

Ligh

t 1 K

ing

Siz

e44

.33

29.7

729

.56

101.

102.

78E

xpor

t ‘A

’ Ext

ra L

ight

Kin

g S

ize

43.6

124

.54

27.7

112

4.23

2.58

Exp

ort ‘

A’ M

ild K

ing

Siz

e41

.54

23.1

524

.15

109.

232.

56E

xpor

t ‘A

’ Mild

Reg

ular

Siz

e41

.54

23.1

524

.15

109.

232.

56E

xpor

t ‘A

’ Ultr

a Li

ght K

ing

Siz

e42

.02

23.4

224

.44

110.

512.

59N

umbe

r 7

Kin

g S

ize

44.9

728

.75

30.3

813

0.39

3.06

Cra

ven

Mild

s U

ltra

Mild

Kin

g S

ize

63.9

039

.62

41.3

016

0.81

3.93

Cam

el L

ight

Kin

g S

ize

57.9

329

.89

40.0

621

1.49

4.05

Win

ston

Lig

ht K

ing

Siz

e57

.93

29.8

940

.06

211.

494.

05E

xpor

t Pla

in R

egul

ar S

ize

34.4

623

.27

23.3

896

.15

2.10

Rot

hman

s S

peci

al M

ild K

ing

Siz

e44

.89

30.4

430

.21

126.

962.

89

146

Win

ston

Lig

ht 1

00’s

54.7

929

.59

38.7

720

0.00

3.83

12.2

8W

inst

on 1

00’s

46.1

525

.23

32.1

916

2.69

2.97

12.5

4G

aulo

ises

Blo

ndes

Kin

g S

ize

46.2

625

.87

29.4

015

7.59

2.43

17.4

0P

laye

r’s E

xtra

Lig

ht R

egul

ar S

ize

43.9

130

.24

30.0

510

3.23

2.77

Cam

el K

ing

Siz

e49

.34

28.6

334

.89

196.

483.

3311

.76

Sal

em K

ing

Siz

e49

.34

28.6

334

.89

196.

483.

3311

.76

Win

ston

Kin

g S

ize

49.3

428

.63

34.8

919

6.48

3.33

11.7

6E

xpor

t ‘A

’ Ful

l Fla

vour

Reg

ular

Siz

e40

.31

24.0

625

.32

116.

802.

319.

76D

u M

aurie

r E

xtra

Lig

ht R

egul

ar S

ize

42.6

728

.61

29.4

911

9.06

3.29

Cra

ven

A K

ing

Siz

e39

.88

25.0

727

.07

123.

823.

17S

alem

Lig

ht K

ing

Siz

e52

.16

29.7

836

.03

200.

433.

3211

.94

Mat

inee

Ext

ra L

ight

Kin

g S

ize

52.5

334

.48

35.2

312

2.18

3.67

Cam

el P

lain

Reg

ular

Siz

e35

.92

25.6

326

.04

149.

802.

5212

.57

Du

Mau

rier

Ligh

t Kin

g S

ize

43.0

029

.10

29.9

912

4.16

3.31

Vis

coun

t Ext

ra M

ild K

ing

Siz

e53

.90

34.9

435

.87

145.

213.

18S

port

sman

Pla

in R

egul

ar S

ize

31.3

226

.85

22.0

375

.37

2.07

Vat

nage

Med

ium

7 K

ing

Siz

e42

.16

26.1

027

.52

121.

332.

7310

.61

Ben

son

& H

edge

s D

elux

e M

enth

olU

ltra

Ligh

ts 1

00’s

44.0

027

.93

28.5

211

4.01

3.11

Pla

yer’s

Lig

ht R

egul

ar S

ize

48.4

030

.33

29.7

010

0.46

2.73

Van

tage

Max

15

Kin

g S

ize

35.8

124

.71

24.5

210

4.84

2.56

11.8

4C

rave

n M

enth

ol K

ing

Siz

e49

.98

35.8

434

.45

133.

703.

42M

atin

ee S

ilver

Reg

ular

Siz

e49

.64

31.3

232

.44

116.

243.

56B

elm

ont M

ilds

Reg

ular

Siz

e38

.33

19.5

923

.85

88.1

52.

34B

elm

ont M

ilds

Kin

g S

ize

42.7

620

.10

25.6

697

.06

2.57

147

Pla

yer’s

Lig

ht K

ing

Siz

e49

.04

30.7

331

.51

133.

893.

14G

aulo

ises

Blo

ndes

Lig

hts

Kin

g S

ize

56.4

029

.47

35.5

717

2.76

3.10

17.2

8V

anta

ge R

ich

12 K

ing

Siz

e40

.00

26.5

426

.92

103.

462.

7712

.65

Gau

lois

es B

lond

es U

ltra

Ligh

t Kin

g S

ize

65.3

133

.25

40.3

720

3.05

3.68

18.7

6M

ark

ten

Pla

in K

ing

Siz

e34

.91

29.6

124

.62

85.0

82.

09D

u M

aurie

r Li

ght R

egul

ar S

ize

53.5

733

.04

34.9

713

8.46

3.62

Du

Mau

rier

Kin

g S

ize

47.4

530

.73

30.5

011

6.93

2.76

Pla

yer’s

Reg

ular

Siz

e39

.72

24.5

825

.27

106.

292.

59D

u M

aurie

r E

xtra

Lig

ht K

ing

Siz

e51

.64

30.9

932

.66

124.

393.

31P

laye

r’s S

peci

al B

lend

Kin

g S

ize

41.4

425

.36

27.6

411

2.17

2.71

Pla

yer’s

Spe

cial

Ble

nd R

egul

ar S

ize

42.6

225

.78

27.8

311

4.75

2.83

Pla

yer’s

Pla

in R

egul

ar S

ize

34.1

429

.51

22.8

094

.78

2.75

13.7

3P

laye

r’s E

xtra

Lig

ht K

ing

Siz

e49

.54

29.1

032

.08

131.

053.

04D

u M

aurie

r R

egul

ar S

ize

47.7

629

.85

31.2

212

8.35

3.30

Can

yon

Reg

ular

Siz

e54

.01

35.9

937

.06

170.

052.

9421

.82

Dak

ar R

egul

ar S

ize

54.0

135

.99

37.0

617

0.05

2.94

21.8

2D

iscr

etio

n R

egul

ar S

ize

54.0

135

.99

37.0

617

0.05

2.94

21.8

2F

ine

Reg

ular

Siz

e54

.01

35.9

937

.06

170.

052.

9421

.82

Gip

sy R

egul

ar S

ize

54.0

135

.99

37.0

617

0.05

2.94

21.8

2S

eles

ta R

egul

ar S

ize

54.0

135

.99

37.0

617

0.05

2.94

21.8

2S

mok

ing

Reg

ular

Siz

e54

.01

35.9

937

.06

170.

052.

9421

.82

Tre

mbl

ay R

egul

ar S

ize

54.0

135

.99

37.0

617

0.05

2.94

21.8

2C

anyo

n K

ing

Siz

e53

.99

34.6

937

.04

160.

092.

7820

.61

Dak

ar K

ing

Siz

e53

.99

34.6

937

.04

160.

092.

7820

.61

Dis

cret

ion

Kin

g S

ize

53.9

934

.69

37.0

416

0.09

2.78

20.6

1F

ine

Kin

g S

ize

53.9

934

.69

37.0

416

0.09

2.78

20.6

1

148

Gip

sy K

ing

Siz

e53

.99

34.6

937

.04

160.

092.

7820

.61

Sel

esta

Kin

g S

ize

53.9

934

.69

37.0

416

0.09

2.78

20.6

1S

mok

ing

Kin

g S

ize

53.9

934

.69

37.0

416

0.09

2.78

20.6

1T

rem

blay

Kin

g S

ize

53.9

934

.69

37.0

416

0.09

2.78

20.6

1D

akar

Lig

ht K

ing

Siz

e54

.85

35.6

837

.38

170.

393.

2523

.25

Fin

e Li

ght K

ing

Siz

e54

.85

35.6

837

.38

170.

393.

2523

.25

Gip

sy L

ight

Kin

g S

ize

54.8

535

.68

37.3

817

0.39

3.25

23.2

5R

odeo

Kin

g S

ize

54.8

535

.68

37.3

817

0.39

3.25

23.2

5S

eles

ta L

ight

Kin

g S

ize

54.8

535

.68

37.3

817

0.39

3.25

23.2

5S

mok

ing

Ligh

t Kin

g S

ize

54.8

535

.68

37.3

817

0.39

3.25

23.2

5T

rem

blay

Lig

ht K

ing

Siz

e54

.85

35.6

837

.38

170.

393.

2523

.25

Dak

ar L

ight

Reg

ular

Siz

e53

.31

35.1

936

.63

163.

543.

0421

.82

Fin

e Li

ght R

egul

ar S

ize

53.3

135

.19

36.6

316

3.54

3.04

21.8

2G

ipsy

Lig

ht R

egul

ar S

ize

53.3

135

.19

36.6

316

3.54

3.04

21.8

2R

odeo

Reg

ular

Siz

e53

.31

35.1

936

.63

163.

543.

0421

.82

Sel

esta

Lig

ht R

egul

ar S

ize

53.3

135

.19

36.6

316

3.54

3.04

21.8

2S

mok

ing

Ligh

t Reg

ular

Siz

e53

.31

35.1

936

.63

163.

543.

0421

.82

***

Mea

ns *

**48

.21

30.3

532

.28

142.

912.

9418

.75

***

Sta

ndar

d de

viat

ion

***

7.35

4.80

5.37

35.9

20.

434.

44**

* C

oeffi

cien

t of v

aria

tion

***

0.15

0.16

0.17

0.25

0.15

0.24

Max

imum

65.3

139

.62

41.3

023

0.38

4.05

23.2

575

th p

erce

ntile

53.9

935

.19

37.0

417

0.05

3.25

21.8

29t

h pe

rcen

tile

54.8

535

.87

37.3

819

6.48

3.35

23.2

5M

edia

n49

.54

29.8

934

.45

145.

212.

9420

.61

Min

imum

31.3

219

.59

22.0

375

.37

1.97

9.76

149

Bra

nd

Cad

miu

mN

itri

c o

xid

eN

itro

gen

oxi

des

NN

NN

NK

NA

T

Exp

ort ‘

A’ U

ltra

Ligh

t Reg

ular

Siz

e72

.54

47.6

451

.45

26.4

050

.87

39.0

1V

anta

ge S

lims

Ext

ra L

ight

4 1

00’s

65.3

941

.62

43.9

420

.69

49.0

338

.38

Exp

ort ‘

A’ M

ediu

m K

ing

Siz

e61

.13

60.7

563

.02

28.6

842

.26

41.1

3E

xpor

t ‘A

’ Ful

l Fla

vour

Kin

g S

ize

64.1

660

.93

63.4

428

.67

44.8

038

.71

Cra

ven

A E

xtra

Lig

ht K

ing

Siz

e79

.97

57.4

161

.44

16.6

433

.68

27.3

2E

xpor

t ‘A

’ Med

ium

Reg

ular

Siz

e75

.75

58.5

862

.97

27.0

050

.03

42.6

6E

xpor

t ‘A

’ Ext

ra L

ight

Reg

ular

Siz

e70

.84

43.7

447

.12

23.7

150

.34

38.6

5M

ore

Reg

ular

120

’s62

.22

130.

3714

1.48

132.

5981

.85

10.7

4E

xpor

t ‘A

’ Lig

ht /

Spe

cial

Lig

ht K

ing

Siz

e61

.54

57.6

959

.62

33.4

646

.58

40.7

7E

xpor

t ‘A

’ Lig

ht /

Spe

cial

Lig

htR

egul

ar S

ize

75.9

657

.56

61.3

726

.77

54.4

044

.78

Mor

e M

enth

ol 1

20’s

59.6

214

5.77

160.

3813

1.92

87.6

911

.15

Van

tage

Ultr

a Li

ght 1

Kin

g S

ize

80.6

859

.06

64.1

926

.06

51.0

442

.40

Exp

ort ‘

A’ E

xtra

Lig

ht K

ing

Siz

e63

.88

65.2

067

.40

28.1

939

.65

37.0

0E

xpor

t ‘A

’ Mild

Kin

g S

ize

58.4

647

.31

49.6

224

.62

41.5

434

.23

Exp

ort ‘

A’ M

ild R

egul

ar S

ize

58.4

647

.31

49.6

224

.62

41.5

434

.23

Exp

ort ‘

A’ U

ltra

Ligh

t Kin

g S

ize

59.1

447

.86

50.1

924

.90

42.0

234

.63

Num

ber

7 K

ing

Siz

e86

.67

59.1

162

.38

20.7

935

.49

27.5

2C

rave

n M

ilds

Ultr

a M

ild K

ing

Siz

e72

.88

63.3

067

.20

31.3

745

.66

35.0

9C

amel

Lig

ht K

ing

Siz

e82

.76

188.

5120

5.17

155.

1797

.70

12.6

4W

inst

on L

ight

Kin

g S

ize

82.7

618

8.51

205.

1715

5.17

97.7

012

.64

Exp

ort P

lain

Reg

ular

Siz

e60

.77

46.5

449

.23

29.6

943

.85

40.0

0R

othm

ans

Spe

cial

Mild

Kin

g S

ize

83.3

056

.92

60.7

021

.63

37.4

629

.12

150

Win

ston

Lig

ht 1

00’s

79.9

117

7.17

193.

6115

1.14

90.8

711

.87

Win

ston

100

’s77

.31

157.

3117

1.15

153.

4696

.54

13.0

8G

aulo

ises

Blo

ndes

Kin

g S

ize

52.1

421

4.43

233.

6312

8.19

45.8

613

8.38

Pla

yer’s

Ext

ra L

ight

Reg

ular

Siz

e76

.43

53.9

557

.79

19.5

937

.62

31.8

7C

amel

Kin

g S

ize

76.6

517

5.33

192.

5116

0.35

100.

4412

.78

Sal

em K

ing

Siz

e76

.65

175.

3319

2.51

160.

3510

0.44

12.7

8W

inst

on K

ing

Siz

e76

.65

175.

3319

2.51

160.

3510

0.44

12.7

8E

xpor

t ‘A

’ Ful

l Fla

vour

Reg

ular

Siz

e79

.93

53.1

657

.42

27.4

758

.37

43.2

8D

u M

aurie

r E

xtra

Lig

ht R

egul

ar S

ize

84.6

561

.83

66.8

417

.16

32.5

930

.90

Cra

ven

A K

ing

Siz

e79

.39

52.6

956

.53

18.1

436

.59

28.3

6S

alem

Lig

ht K

ing

Siz

e77

.59

164.

2217

9.31

142.

2492

.24

11.6

4M

atin

ee E

xtra

Lig

ht K

ing

Siz

e80

.06

68.7

974

.52

19.8

536

.14

35.5

9C

amel

Pla

in R

egul

ar S

ize

74.6

911

6.33

126.

1216

3.27

103.

6713

.47

Du

Mau

rier

Ligh

t Kin

g S

ize

81.9

067

.78

74.0

217

.07

34.9

131

.57

Vis

coun

t Ext

ra M

ild K

ing

Siz

e77

.05

65.2

170

.77

29.0

642

.93

32.5

9S

port

sman

Pla

in R

egul

ar S

ize

64.3

630

.98

31.3

219

.62

25.8

123

.75

Vat

nage

Med

ium

7 K

ing

Siz

e90

.32

58.0

962

.85

26.8

155

.32

48.2

7B

enso

n &

Hed

ges

Del

uxe

Men

thol

Ultr

aLi

ghts

100

’s70

.15

54.2

659

.24

16.3

436

.75

26.9

5

Pla

yer’s

Lig

ht R

egul

ar S

ize

76.8

247

.80

50.9

117

.07

39.5

433

.16

Van

tage

Max

15

Kin

g S

ize

69.5

256

.13

60.6

520

.00

27.1

030

.65

Cra

ven

Men

thol

Kin

g S

ize

79.0

264

.11

69.6

415

.56

34.7

726

.94

Mat

inee

Silv

er R

egul

ar S

ize

66.5

044

.72

46.0

48.

0717

.11

18.3

2B

elm

ont M

ilds

Reg

ular

Siz

e37

.90

40.4

641

.31

14.4

820

.02

19.1

6B

elm

ont M

ilds

Kin

g S

ize

40.1

945

.75

47.0

426

.94

30.7

927

.79

151

Pla

yer’s

Lig

ht K

ing

Siz

e85

.41

61.8

265

.97

18.8

340

.07

34.6

6G

aulo

ises

Blo

ndes

Lig

hts

Kin

g S

ize

53.3

525

4.56

277.

9313

9.73

49.7

915

2.94

Van

tage

Ric

h 12

Kin

g S

ize

71.1

553

.46

56.5

421

.92

37.8

538

.85

Gau

lois

es B

lond

es U

ltra

Ligh

t Kin

g S

ize

58.7

828

4.39

310.

5215

6.74

57.0

016

9.21

Mar

k te

n P

lain

Kin

g S

ize

57.0

328

.36

28.6

78.

7313

.40

14.6

5D

u M

aurie

r Li

ght R

egul

ar S

ize

88.7

573

.60

77.3

517

.54

36.0

632

.41

Du

Mau

rier

Kin

g S

ize

88.5

758

.26

61.1

218

.12

35.6

830

.77

Pla

yer’s

Reg

ular

Siz

e79

.46

51.7

156

.37

15.8

938

.42

29.5

3D

u M

aurie

r E

xtra

Lig

ht K

ing

Siz

e87

.08

61.8

066

.36

17.2

436

.20

30.9

5P

laye

r’s S

peci

al B

lend

Kin

g S

ize

51.3

355

.89

58.5

619

.85

20.9

924

.45

Pla

yer’s

Spe

cial

Ble

nd R

egul

ar S

ize

52.0

551

.64

54.9

220

.00

19.3

424

.47

Pla

yer’s

Pla

in R

egul

ar S

ize

79.8

534

.66

35.3

49.

8922

.43

21.2

7P

laye

r’s E

xtra

Lig

ht K

ing

Siz

e93

.22

68.4

773

.61

17.4

738

.46

31.5

6D

u M

aurie

r R

egul

ar S

ize

90.0

659

.68

63.9

315

.88

36.6

730

.30

Can

yon

Reg

ular

Siz

e70

.59

70.0

578

.61

27.1

770

.05

42.8

3D

akar

Reg

ular

Siz

e70

.59

70.0

578

.61

27.1

770

.05

42.8

3D

iscr

etio

n R

egul

ar S

ize

70.5

970

.05

78.6

127

.17

70.0

542

.83

Fin

e R

egul

ar S

ize

70.5

970

.05

78.6

127

.17

70.0

542

.83

Gip

sy R

egul

ar S

ize

70.5

970

.05

78.6

127

.17

70.0

542

.83

Sel

esta

Reg

ular

Siz

e70

.59

70.0

578

.61

27.1

770

.05

42.8

3S

mok

ing

Reg

ular

Siz

e70

.59

70.0

578

.61

27.1

770

.05

42.8

3T

rem

blay

Reg

ular

Siz

e70

.59

70.0

578

.61

27.1

770

.05

42.8

3C

anyo

n K

ing

Siz

e68

.08

82.6

392

.02

28.1

273

.24

44.9

8D

akar

Kin

g S

ize

68.0

882

.63

92.0

228

.12

73.2

444

.98

Dis

cret

ion

Kin

g S

ize

68.0

882

.63

92.0

228

.12

73.2

444

.98

Fin

e K

ing

Siz

e68

.08

82.6

392

.02

28.1

273

.24

44.9

8

152

Gip

sy K

ing

Siz

e68

.08

82.6

392

.02

28.1

273

.24

44.9

8S

eles

ta K

ing

Siz

e68

.08

82.6

392

.02

28.1

273

.24

44.9

8S

mok

ing

Kin

g S

ize

68.0

882

.63

92.0

228

.12

73.2

444

.98

Tre

mbl

ay K

ing

Siz

e68

.08

82.6

392

.02

28.1

273

.24

44.9

8D

akar

Lig

ht K

ing

Siz

e70

.87

80.1

089

.32

28.8

873

.79

45.2

9F

ine

Ligh

t Kin

g S

ize

70.8

780

.10

89.3

228

.88

73.7

945

.29

Gip

sy L

ight

Kin

g S

ize

70.8

780

.10

89.3

228

.88

73.7

945

.29

Rod

eo K

ing

Siz

e70

.87

80.1

089

.32

28.8

873

.79

45.2

9S

eles

ta L

ight

Kin

g S

ize

70.8

780

.10

89.3

228

.88

73.7

945

.29

Sm

okin

g Li

ght K

ing

Siz

e70

.87

80.1

089

.32

28.8

873

.79

45.2

9T

rem

blay

Lig

ht K

ing

Siz

e70

.87

80.1

089

.32

28.8

873

.79

45.2

9D

akar

Lig

ht R

egul

ar S

ize

70.7

270

.72

79.0

126

.91

63.5

441

.05

Fin

e Li

ght R

egul

ar S

ize

70.7

270

.72

79.0

126

.91

63.5

441

.05

Gip

sy L

ight

Reg

ular

Siz

e70

.72

70.7

279

.01

26.9

163

.54

41.0

5R

odeo

Reg

ular

Siz

e70

.72

70.7

279

.01

26.9

163

.54

41.0

5S

eles

ta L

ight

Reg

ular

Siz

e70

.72

70.7

279

.01

26.9

163

.54

41.0

5S

mok

ing

Ligh

t Reg

ular

Siz

e70

.72

70.7

279

.01

26.9

163

.54

41.0

5**

* M

eans

***

71.3

581

.55

88.8

443

.54

56.0

537

.76

***

Sta

ndar

d de

viat

ion

***

10.3

148

.55

53.4

646

.49

22.3

624

.42

***

Coe

ffici

ent o

f var

iatio

n **

*0.

140.

600.

601.

070.

400.

65M

axim

um93

.22

284.

3931

0.52

163.

2710

3.67

169.

2175

th p

erce

ntile

77.5

980

.10

89.3

228

.88

73.2

442

.83

9th

perc

entil

e83

.57

166.

4518

1.95

144.

0288

.33

45.2

9M

edia

n70

.72

68.7

974

.52

27.1

751

.04

38.3

8M

inim

um37

.90

28.3

628

.67

8.07

13.4

010

.74

153

Bra

nd

NA

BP

yrid

ine

Qu

ino

line

Hyd

roq

uin

on

eR

eso

rcin

ol

Cat

ech

ol

Ph

eno

lm

- an

d p

-Cre

sols

Exp

ort ‘

A’ U

ltra

Ligh

t Reg

ular

Siz

e2.

9314

.64

0.28

62.5

31.

0568

.38

17.0

38.

23V

anta

ge S

lims

Ext

ra L

ight

4 1

00’s

2.84

12.4

00.

2758

.80

0.88

65.6

718

.64

8.76

Exp

ort ‘

A’ M

ediu

m K

ing

Siz

e2.

9414

.42

0.26

44.9

10.

9855

.47

11.3

27.

40E

xpor

t ‘A

’ Ful

l Fla

vour

Kin

g S

ize

3.01

14.0

10.

2847

.31

0.99

59.8

612

.54

7.56

Cra

ven

A E

xtra

Lig

ht K

ing

Siz

e2.

0514

.42

0.26

65.7

21.

0670

.53

14.0

57.

38E

xpor

t ‘A

’ Med

ium

Reg

ular

Siz

e3.

2115

.51

0.29

62.1

80.

9166

.12

16.0

18.

29E

xpor

t ‘A

’ Ext

ra L

ight

Reg

ular

Siz

e2.

7614

.60

0.27

63.4

80.

9168

.42

19.0

58.

62M

ore

Reg

ular

120

’s36

.93

17.0

00.

3958

.89

1.12

69.2

619

.33

12.2

6E

xpor

t ‘A

’ Lig

ht /

Spe

cial

Lig

ht K

ing

Siz

e3.

3515

.08

0.28

47.3

11.

0059

.62

12.6

98.

00E

xpor

t ‘A

’ Lig

ht /

Spe

cial

Lig

htR

egul

ar S

ize

3.30

15.2

00.

2764

.47

1.07

68.2

018

.14

9.63

Mor

e M

enth

ol 1

20’s

37.2

317

.27

0.40

59.2

31.

1971

.54

18.5

412

.12

Van

tage

Ultr

a Li

ght 1

Kin

g S

ize

2.99

14.8

50.

2660

.36

0.80

67.8

913

.93

7.33

Exp

ort ‘

A’ E

xtra

Lig

ht K

ing

Siz

e3.

0015

.51

0.27

47.5

81.

0759

.47

11.7

67.

53E

xpor

t ‘A

’ Mild

Kin

g S

ize

2.38

13.1

50.

3048

.46

0.86

59.2

313

.50

7.77

Exp

ort ‘

A’ M

ild R

egul

ar S

ize

2.38

13.1

50.

3048

.46

0.86

59.2

313

.50

7.77

Exp

ort ‘

A’ U

ltra

Ligh

t Kin

g S

ize

2.41

13.3

10.

3049

.03

0.87

59.9

213

.66

7.86

Num

ber

7 K

ing

Siz

e2.

2914

.72

0.32

71.2

71.

1279

.13

18.5

59.

86C

rave

n M

ilds

Ultr

a M

ild K

ing

Siz

e17

.99

0.30

71.8

71.

3177

.47

14.7

58.

51C

amel

Lig

ht K

ing

Siz

e41

.55

21.6

10.

3259

.77

1.11

64.3

712

.13

8.39

Win

ston

Lig

ht K

ing

Siz

e41

.55

21.6

10.

3259

.77

1.11

64.3

712

.13

8.39

Exp

ort P

lain

Reg

ular

Siz

e3.

1413

.65

0.55

53.4

60.

7672

.69

45.0

021

.58

Rot

hman

s S

peci

al M

ild K

ing

Siz

e2.

3313

.72

0.30

68.1

91.

0874

.10

18.6

89.

97

154

Win

ston

Lig

ht 1

00’s

37.6

720

.55

0.35

61.1

91.

1770

.78

17.1

211

.23

Win

ston

100

’s41

.15

18.0

80.

3960

.00

1.05

71.5

418

.92

12.1

5G

aulo

ises

Blo

ndes

Kin

g S

ize

16.0

718

.82

0.29

44.3

00.

7847

.43

10.9

86.

98P

laye

r’s E

xtra

Lig

ht R

egul

ar S

ize

1.98

15.6

70.

3068

.60

1.07

70.7

916

.86

8.58

Cam

el K

ing

Siz

e43

.35

22.1

10.

4063

.00

1.15

70.9

317

.67

11.5

4S

alem

Kin

g S

ize

43.3

522

.11

0.40

63.0

01.

1570

.93

17.6

711

.54

Win

ston

Kin

g S

ize

43.3

522

.11

0.40

63.0

01.

1570

.93

17.6

711

.54

Exp

ort ‘

A’ F

ull F

lavo

ur R

egul

ar S

ize

2.72

16.5

20.

3369

.46

1.02

72.3

618

.01

9.00

Du

Mau

rier

Ext

ra L

ight

Reg

ular

Siz

e2.

1315

.10

0.26

58.0

11.

0965

.50

12.7

87.

05C

rave

n A

Kin

g S

ize

1.74

13.7

80.

2772

.51

1.04

76.4

315

.50

7.84

Sal

em L

ight

Kin

g S

ize

40.9

921

.38

0.39

64.6

61.

1873

.71

18.8

812

.16

Mat

inee

Ext

ra L

ight

Kin

g S

ize

2.94

16.4

20.

2860

.10

0.89

68.1

714

.01

7.83

Cam

el P

lain

Reg

ular

Siz

e42

.78

20.6

90.

8064

.90

1.13

83.2

753

.88

27.1

4D

u M

aurie

r Li

ght K

ing

Siz

e2.

6614

.68

0.27

61.5

51.

0470

.22

14.4

77.

83V

isco

unt E

xtra

Mild

Kin

g S

ize

17.4

80.

2974

.63

1.14

76.5

015

.23

7.92

Spo

rtsm

an P

lain

Reg

ular

Siz

e1.

7211

.36

0.55

66.0

81.

2093

.96

53.6

923

.20

Vat

nage

Med

ium

7 K

ing

Siz

e3.

4715

.26

0.26

68.6

31.

1471

.46

15.1

47.

83B

enso

n &

Hed

ges

Del

uxe

Men

thol

Ultr

aLi

ghts

100

’s1.

7314

.19

0.28

67.7

01.

1771

.78

16.3

88.

43

Pla

yer’s

Lig

ht R

egul

ar S

ize

1.92

14.7

30.

2765

.31

1.11

70.4

016

.46

9.13

Van

tage

Max

15

Kin

g S

ize

2.35

12.4

20.

3150

.97

0.90

65.4

816

.06

9.29

Cra

ven

Men

thol

Kin

g S

ize

2.13

15.4

00.

2971

.39

1.08

74.0

112

.40

7.49

Mat

inee

Silv

er R

egul

ar S

ize

15.2

30.

2963

.45

1.15

69.0

416

.95

9.19

Bel

mon

t Mild

s R

egul

ar S

ize

10.6

50.

3262

.60

1.32

82.1

915

.76

8.77

Bel

mon

t Mild

s K

ing

Siz

e1.

719.

830.

3064

.57

1.33

79.5

315

.39

9.32

155

Pla

yer’s

Lig

ht K

ing

Siz

e2.

5414

.93

0.30

64.5

90.

9371

.77

18.1

29.

40G

aulo

ises

Blo

ndes

Lig

hts

Kin

g S

ize

18.2

919

.82

0.30

47.2

50.

8650

.30

10.6

77.

06V

anta

ge R

ich

12 K

ing

Siz

e2.

9613

.81

0.33

55.7

71.

1271

.92

19.1

510

.92

Gau

lois

es B

lond

es U

ltra

Ligh

t Kin

g S

ize

19.0

021

.37

0.31

54.0

31.

1355

.81

11.8

77.

96M

ark

ten

Pla

in K

ing

Siz

e12

.78

0.57

68.5

71.

4395

.68

54.5

423

.75

Du

Mau

rier

Ligh

t Reg

ular

Siz

e15

.59

0.24

62.8

71.

2166

.59

12.8

17.

52D

u M

aurie

r K

ing

Siz

e2.

0514

.20

0.25

66.2

31.

2374

.44

15.4

88.

11P

laye

r’s R

egul

ar S

ize

1.86

15.2

10.

2866

.08

1.02

68.4

615

.47

7.83

Du

Mau

rier

Ext

ra L

ight

Kin

g S

ize

17.5

70.

3063

.14

1.20

73.6

815

.47

8.38

Pla

yer’s

Spe

cial

Ble

nd K

ing

Siz

e2.

3415

.21

0.37

65.7

80.

9665

.40

19.7

310

.27

Pla

yer’s

Spe

cial

Ble

nd R

egul

ar S

ize

2.41

14.8

40.

3266

.39

1.19

63.5

215

.74

8.81

Pla

yer’s

Pla

in R

egul

ar S

ize

1.90

12.1

60.

5365

.67

1.16

81.7

254

.10

22.4

6P

laye

r’s E

xtra

Lig

ht K

ing

Siz

e2.

0316

.03

0.29

69.1

01.

1573

.39

14.9

87.

96D

u M

aurie

r R

egul

ar S

ize

2.34

16.3

50.

3065

.88

1.10

71.3

113

.82

7.39

Can

yon

Reg

ular

Siz

e3.

9018

.18

0.39

78.0

71.

4785

.56

16.0

411

.18

Dak

ar R

egul

ar S

ize

3.90

18.1

80.

3978

.07

1.47

85.5

616

.04

11.1

8D

iscr

etio

n R

egul

ar S

ize

3.90

18.1

80.

3978

.07

1.47

85.5

616

.04

11.1

8F

ine

Reg

ular

Siz

e3.

9018

.18

0.39

78.0

71.

4785

.56

16.0

411

.18

Gip

sy R

egul

ar S

ize

3.90

18.1

80.

3978

.07

1.47

85.5

616

.04

11.1

8S

eles

ta R

egul

ar S

ize

3.90

18.1

80.

3978

.07

1.47

85.5

616

.04

11.1

8S

mok

ing

Reg

ular

Siz

e3.

9018

.18

0.39

78.0

71.

4785

.56

16.0

411

.18

Tre

mbl

ay R

egul

ar S

ize

3.90

18.1

80.

3978

.07

1.47

85.5

616

.04

11.1

8C

anyo

n K

ing

Siz

e4.

3417

.09

0.40

75.5

91.

6188

.73

19.3

913

.05

Dak

ar K

ing

Siz

e4.

3417

.09

0.40

75.5

91.

6188

.73

19.3

913

.05

Dis

cret

ion

Kin

g S

ize

4.34

17.0

90.

4075

.59

1.61

88.7

319

.39

13.0

5F

ine

Kin

g S

ize

4.34

17.0

90.

4075

.59

1.61

88.7

319

.39

13.0

5

156

Gip

sy K

ing

Siz

e4.

3417

.09

0.40

75.5

91.

6188

.73

19.3

913

.05

Sel

esta

Kin

g S

ize

4.34

17.0

90.

4075

.59

1.61

88.7

319

.39

13.0

5S

mok

ing

Kin

g S

ize

4.34

17.0

90.

4075

.59

1.61

88.7

319

.39

13.0

5T

rem

blay

Kin

g S

ize

4.34

17.0

90.

4075

.59

1.61

88.7

319

.39

13.0

5D

akar

Lig

ht K

ing

Siz

e4.

1016

.99

0.43

72.8

21.

7887

.38

19.0

812

.91

Fin

e Li

ght K

ing

Siz

e4.

1016

.99

0.43

72.8

21.

7887

.38

19.0

812

.91

Gip

sy L

ight

Kin

g S

ize

4.10

16.9

90.

4372

.82

1.78

87.3

819

.08

12.9

1R

odeo

Kin

g S

ize

4.10

16.9

90.

4372

.82

1.78

87.3

819

.08

12.9

1S

eles

ta L

ight

Kin

g S

ize

4.10

16.9

90.

4372

.82

1.78

87.3

819

.08

12.9

1S

mok

ing

Ligh

t Kin

g S

ize

4.10

16.9

90.

4372

.82

1.78

87.3

819

.08

12.9

1T

rem

blay

Lig

ht K

ing

Siz

e4.

1016

.99

0.43

72.8

21.

7887

.38

19.0

812

.91

Dak

ar L

ight

Reg

ular

Siz

e4.

0118

.56

0.40

74.5

91.

7385

.64

15.5

210

.72

Fin

e Li

ght R

egul

ar S

ize

4.01

18.5

60.

4074

.59

1.73

85.6

415

.52

10.7

2G

ipsy

Lig

ht R

egul

ar S

ize

4.01

18.5

60.

4074

.59

1.73

85.6

415

.52

10.7

2R

odeo

Reg

ular

Siz

e4.

0118

.56

0.40

74.5

91.

7385

.64

15.5

210

.72

Sel

esta

Lig

ht R

egul

ar S

ize

4.01

18.5

60.

4074

.59

1.73

85.6

415

.52

10.7

2S

mok

ing

Ligh

t Reg

ular

Siz

e4.

0118

.56

0.40

74.5

91.

7385

.64

15.5

210

.72

***

Mea

ns *

**8.

7716

.46

0.35

65.9

61.

2575

.30

18.2

610

.70

***

Sta

ndar

d de

viat

ion

***

13.0

82.

620.

099.

110.

3010

.74

8.72

3.76

***

Coe

ffici

ent o

f var

iatio

n **

*1.

490.

160.

240.

140.

240.

140.

480.

35M

axim

um43

.35

22.1

10.

8078

.07

1.78

95.6

854

.54

27.1

475

th p

erce

ntile

4.34

18.1

80.

4074

.59

1.47

85.6

419

.08

12.1

59t

h pe

rcen

tile

37.6

319

.96

0.43

75.5

91.

7388

.73

19.3

913

.05

Med

ian

3.90

16.9

90.

3366

.08

1.15

73.3

916

.04

9.97

Min

imum

1.71

9.83

0.24

44.3

00.

7647

.43

10.6

76.

98

157

Bra

nd

o-C

reso

l1,

3-B

uta

die

ne

Iso

pre

ne

Acr

ylo

nit

rile

Ben

zen

eT

olu

ene

Sty

ren

e

Exp

ort ‘

A’ U

ltra

Ligh

t Reg

ular

Siz

e3.

7330

.24

163.

917.

0830

.35

54.9

29.

00V

anta

ge S

lims

Ext

ra L

ight

4 1

00’s

3.93

29.9

115

6.33

6.86

28.8

052

.17

7.58

Exp

ort ‘

A’ M

ediu

m K

ing

Siz

e2.

7532

.57

215.

478.

7533

.36

54.7

210

.34

Exp

ort ‘

A’ F

ull F

lavo

ur K

ing

Siz

e3.

0032

.62

215.

419.

0334

.55

56.2

79.

75C

rave

n A

Ext

ra L

ight

Kin

g S

ize

3.15

43.2

226

8.65

9.28

39.1

367

.93

11.3

9E

xpor

t ‘A

’ Med

ium

Reg

ular

Siz

e3.

7433

.78

194.

928.

3933

.14

59.6

810

.65

Exp

ort ‘

A’ E

xtra

Lig

ht R

egul

ar S

ize

4.10

30.5

215

9.13

7.45

30.3

855

.02

9.12

Mor

e R

egul

ar 1

20’s

4.26

43.7

032

0.00

9.26

36.4

165

.93

10.0

4E

xpor

t ‘A

’ Lig

ht /

Spe

cial

Lig

ht K

ing

Siz

e3.

1031

.46

208.

858.

3132

.42

54.2

310

.54

Exp

ort ‘

A’ L

ight

/ S

peci

al L

ight

Reg

ular

Siz

e4.

0836

.80

201.

668.

8534

.68

63.1

69.

83

Mor

e M

enth

ol 1

20’s

4.35

44.6

233

0.38

9.27

36.1

965

.38

10.5

4V

anta

ge U

ltra

Ligh

t 1 K

ing

Siz

e3.

3737

.73

222.

319.

0439

.13

66.8

710

.72

Exp

ort ‘

A’ E

xtra

Lig

ht K

ing

Siz

e2.

9635

.24

233.

489.

4735

.90

58.5

911

.19

Exp

ort ‘

A’ M

ild K

ing

Siz

e3.

3735

.58

219.

629.

2331

.81

55.7

79.

42E

xpor

t ‘A

’ Mild

Reg

ular

Siz

e3.

3735

.58

219.

629.

2331

.81

55.7

79.

42E

xpor

t ‘A

’ Ultr

a Li

ght K

ing

Siz

e3.

4035

.99

222.

189.

3432

.18

56.4

29.

53N

umbe

r 7

Kin

g S

ize

4.06

38.1

324

9.66

8.75

35.6

764

.43

10.6

2C

rave

n M

ilds

Ultr

a M

ild K

ing

Siz

e3.

3856

.54

345.

7713

.15

46.0

082

.44

13.8

7C

amel

Lig

ht K

ing

Siz

e3.

1553

.68

398.

8513

.39

51.8

493

.68

13.9

1W

inst

on L

ight

Kin

g S

ize

3.15

53.6

839

8.85

13.3

951

.84

93.6

813

.91

Exp

ort P

lain

Reg

ular

Siz

e9.

2729

.19

186.

926.

3826

.96

46.5

48.

19R

othm

ans

Spe

cial

Mild

Kin

g S

ize

4.10

38.8

924

5.84

9.04

35.6

263

.03

10.1

2

158

Win

ston

Lig

ht 1

00’s

5.04

52.0

538

0.37

12.5

646

.62

82.6

513

.33

Win

ston

100

’s4.

4640

.77

305.

009.

2737

.19

66.5

410

.81

Gau

lois

es B

lond

es K

ing

Siz

e3.

1435

.67

299.

109.

1334

.10

61.5

410

.98

Pla

yer’s

Ext

ra L

ight

Reg

ular

Siz

e3.

8836

.87

228.

757.

7636

.16

64.9

810

.94

Cam

el K

ing

Siz

e4.

4343

.61

319.

3811

.10

40.6

272

.25

12.9

1S

alem

Kin

g S

ize

4.43

43.6

131

9.38

11.1

040

.62

72.2

512

.91

Win

ston

Kin

g S

ize

4.43

43.6

131

9.38

11.1

040

.62

72.2

512

.91

Exp

ort ‘

A’ F

ull F

lavo

ur R

egul

ar S

ize

4.22

33.2

317

5.63

7.83

31.6

057

.06

10.2

4D

u M

aurie

r E

xtra

Lig

ht R

egul

ar S

ize

3.13

40.0

724

8.37

8.29

37.6

663

.77

10.6

6C

rave

n A

Kin

g S

ize

3.38

38.8

623

6.09

9.12

35.3

363

.58

10.1

0S

alem

Lig

ht K

ing

Siz

e5.

0845

.69

326.

2910

.86

42.4

675

.43

12.7

6M

atin

ee E

xtra

Lig

ht K

ing

Siz

e3.

4749

.68

304.

2610

.04

44.9

072

.35

12.8

4C

amel

Pla

in R

egul

ar S

ize

10.5

734

.53

261.

638.

7331

.84

56.7

39.

55D

u M

aurie

r Li

ght K

ing

Siz

e3.

4341

.67

255.

718.

1238

.74

63.3

711

.13

Vis

coun

t Ext

ra M

ild K

ing

Siz

e3.

4150

.54

295.

0011

.29

45.6

779

.38

12.8

9S

port

sman

Pla

in R

egul

ar S

ize

9.98

32.0

117

8.97

6.95

27.1

945

.09

6.88

Vat

nage

Med

ium

7 K

ing

Siz

e3.

5137

.64

201.

649.

6038

.40

66.8

510

.61

Ben

son

& H

edge

s D

elux

e M

enth

ol U

ltra

Ligh

ts 1

00’s

3.87

37.9

923

2.38

8.80

34.3

259

.78

10.8

9

Pla

yer’s

Lig

ht R

egul

ar S

ize

3.92

38.2

721

7.41

8.01

34.7

260

.06

10.0

8V

anta

ge M

ax 1

5 K

ing

Siz

e3.

8731

.61

190.

327.

4533

.19

9.03

56.2

3C

rave

n M

enth

ol K

ing

Siz

e3.

3646

.84

281.

339.

5143

.62

73.2

312

.86

Mat

inee

Silv

er R

egul

ar S

ize

4.31

42.6

926

3.96

7.72

39.7

065

.99

11.2

2B

elm

ont M

ilds

Reg

ular

Siz

e3.

4137

.90

215.

068.

0929

.81

46.8

47.

24B

elm

ont M

ilds

Kin

g S

ize

3.42

46.1

824

5.01

8.55

33.7

851

.31

6.84

159

Pla

yer’s

Lig

ht K

ing

Siz

e4.

0541

.00

247.

009.

5339

.79

70.3

210

.73

Gau

lois

es B

lond

es L

ight

s K

ing

Siz

e3.

0544

.71

375.

4912

.04

42.1

776

.72

12.1

9V

anta

ge R

ich

12 K

ing

Siz

e4.

5034

.62

200.

388.

4237

.15

9.62

61.8

1G

aulo

ises

Blo

ndes

Ultr

a Li

ght K

ing

Siz

e3.

5651

.06

438.

1714

.01

48.0

985

.50

14.2

5M

ark

ten

Pla

in K

ing

Siz

e10

.28

28.9

817

3.91

6.70

26.4

946

.13

8.10

Du

Mau

rier

Ligh

t Reg

ular

Siz

e3.

1545

.75

273.

3410

.20

41.3

371

.25

11.9

1D

u M

aurie

r K

ing

Siz

e3.

5137

.71

211.

088.

7038

.12

65.8

010

.36

Pla

yer’s

Reg

ular

Siz

e3.

6936

.42

210.

058.

2935

.45

61.3

710

.18

Du

Mau

rier

Ext

ra L

ight

Kin

g S

ize

3.73

42.3

925

5.24

9.46

41.2

072

.69

12.5

5P

laye

r’s S

peci

al B

lend

Kin

g S

ize

4.75

33.4

225

8.17

7.03

32.7

058

.17

9.13

Pla

yer’s

Spe

cial

Ble

nd R

egul

ar S

ize

3.93

36.5

227

7.05

8.11

35.0

858

.61

9.51

Pla

yer’s

Pla

in R

egul

ar S

ize

11.0

829

.85

192.

165.

5628

.13

48.1

37.

35P

laye

r’s E

xtra

Lig

ht K

ing

Siz

e3.

5745

.74

267.

4910

.50

42.6

775

.76

11.8

4D

u M

aurie

r R

egul

ar S

ize

3.61

42.3

223

8.40

9.14

40.2

270

.33

11.6

9C

anyo

n R

egul

ar S

ize

4.22

49.8

435

5.08

11.9

348

.56

91.4

412

.67

Dak

ar R

egul

ar S

ize

4.22

49.8

435

5.08

11.9

348

.56

91.4

412

.67

Dis

cret

ion

Reg

ular

Siz

e4.

2249

.84

355.

0811

.93

48.5

691

.44

12.6

7F

ine

Reg

ular

Siz

e4.

2249

.84

355.

0811

.93

48.5

691

.44

12.6

7G

ipsy

Reg

ular

Siz

e4.

2249

.84

355.

0811

.93

48.5

691

.44

12.6

7S

eles

ta R

egul

ar S

ize

4.22

49.8

435

5.08

11.9

348

.56

91.4

412

.67

Sm

okin

g R

egul

ar S

ize

4.22

49.8

435

5.08

11.9

348

.56

91.4

412

.67

Tre

mbl

ay R

egul

ar S

ize

4.22

49.8

435

5.08

11.9

348

.56

91.4

412

.67

Can

yon

Kin

g S

ize

4.84

46.9

534

2.25

11.5

547

.42

88.2

611

.88

Dak

ar K

ing

Siz

e4.

8446

.95

342.

2511

.55

47.4

288

.26

11.8

8D

iscr

etio

n K

ing

Siz

e4.

8446

.95

342.

2511

.55

47.4

288

.26

11.8

8F

ine

Kin

g S

ize

4.84

46.9

534

2.25

11.5

547

.42

88.2

611

.88

160

Gip

sy K

ing

Siz

e4.

8446

.95

342.

2511

.55

47.4

288

.26

11.8

8S

eles

ta K

ing

Siz

e4.

8446

.95

342.

2511

.55

47.4

288

.26

11.8

8S

mok

ing

Kin

g S

ize

4.84

46.9

534

2.25

11.5

547

.42

88.2

611

.88

Tre

mbl

ay K

ing

Siz

e4.

8446

.95

342.

2511

.55

47.4

288

.26

11.8

8D

akar

Lig

ht K

ing

Siz

e4.

7648

.01

364.

5611

.50

47.4

888

.35

11.8

4F

ine

Ligh

t Kin

g S

ize

4.76

48.0

136

4.56

11.5

047

.48

88.3

511

.84

Gip

sy L

ight

Kin

g S

ize

4.76

48.0

136

4.56

11.5

047

.48

88.3

511

.84

Rod

eo K

ing

Siz

e4.

7648

.01

364.

5611

.50

47.4

888

.35

11.8

4S

eles

ta L

ight

Kin

g S

ize

4.76

48.0

136

4.56

11.5

047

.48

88.3

511

.84

Sm

okin

g Li

ght K

ing

Siz

e4.

7648

.01

364.

5611

.50

47.4

888

.35

11.8

4T

rem

blay

Lig

ht K

ing

Siz

e4.

7648

.01

364.

5611

.50

47.4

888

.35

11.8

4D

akar

Lig

ht R

egul

ar S

ize

3.87

51.6

037

7.35

11.9

950

.11

92.2

712

.49

Fin

e Li

ght R

egul

ar S

ize

3.87

51.6

037

7.35

11.9

950

.11

92.2

712

.49

Gip

sy L

ight

Reg

ular

Siz

e3.

8751

.60

377.

3511

.99

50.1

192

.27

12.4

9R

odeo

Reg

ular

Siz

e3.

8751

.60

377.

3511

.99

50.1

192

.27

12.4

9S

eles

ta L

ight

Reg

ular

Siz

e3.

8751

.60

377.

3511

.99

50.1

192

.27

12.4

9S

mok

ing

Ligh

t Reg

ular

Siz

e3.

8751

.60

377.

3511

.99

50.1

192

.27

12.4

9**

* M

eans

***

4.33

42.6

028

8.66

10.0

240

.63

71.5

612

.28

***

Sta

ndar

d de

viat

ion

***

1.57

7.19

71.9

71.

887.

1917

.66

7.32

***

Coe

ffici

ent o

f var

iatio

n **

*0.

360.

170.

250.

190.

180.

250.

60M

axim

um11

.08

56.5

443

8.17

14.0

151

.84

93.6

861

.81

75th

per

cent

ile4.

7548

.01

355.

0811

.55

47.4

888

.35

12.5

59t

h pe

rcen

tile

4.84

51.6

037

7.35

11.9

948

.56

91.4

412

.91

Med

ian

4.05

43.6

129

5.00

9.53

40.6

271

.25

11.8

4M

inim

um2.

7528

.98

156.

335.

5626

.49

9.03

6.84

161

Tab

le A

1.2.

Yie

lds

per

mill

igra

m o

f n

ico

tin

e o

f to

xica

nts

in P

hili

p M

orr

is in

tern

atio

nal

bra

nd

s re

po

rted

by

Co

un

ts e

t al

. (20

04)

Bra

nd

Nic

oti

ne

Tar

Car

bo

n m

on

oxi

de

Ace

tald

ehyd

eA

ceto

ne

Acr

ole

in

Mar

lbor

o Lo

ng S

ize

F h

ard

pack

/Arg

entin

a2.

1215

.19

12.7

458

6.79

316.

5158

.54

Mar

lbor

o Lo

ng S

ize

Filt

er h

ard

pack

/Ven

ezue

la1.

9915

.78

12.9

166

1.81

346.

2361

.11

SG

Ven

til R

egul

ar F

ilter

sof

t pac

k/E

urop

ean

Uni

on1.

4818

.92

13.9

277

7.70

397.

3075

.81

Pet

ra R

egul

ar F

ilter

har

d pa

ck/C

EM

A1.

8517

.51

13.0

863

4.05

334.

0559

.46

Mar

lbor

o K

ing

Siz

e F

ilter

sof

t pac

k/U

SA

2.25

16.7

113

.82

574.

6730

2.67

58.1

3M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

/Nor

way

2.08

15.2

413

.08

660.

1035

0.48

64.4

2L

& M

Kin

g S

ize

Filt

er h

ard

pack

/Eur

opea

n U

nion

1.88

17.6

615

.27

783.

5140

8.51

74.7

3M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

/Mal

aysi

a2.

2116

.74

12.8

165

5.66

317.

6562

.31

Che

ster

field

Orig

inal

s K

ing

Siz

e F

ilter

har

d pa

ck1.

9416

.39

14.5

967

2.68

345.

8864

.64

L &

M K

ing

Siz

e F

ilter

har

d pa

ck/M

alay

sia

2.40

14.5

810

.79

547.

5026

2.08

53.5

0M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

25

s/A

ustr

alia

2.38

14.7

113

.07

584.

0331

0.08

59.2

4M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

/Tai

wan

2.27

15.0

712

.56

555.

5128

1.06

55.0

2M

arlb

oro

100

Filt

er h

ard

pack

/Eur

opea

n U

nion

2.27

15.5

514

.93

662.

5633

0.84

64.5

4M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

Med

ium

/Eur

opea

n U

nion

1.94

15.3

114

.74

635.

0534

0.21

60.3

1P

arlia

men

t 100

Filt

er s

oft p

ack

Ligh

t/US

A2.

4114

.40

13.7

359

6.68

292.

9554

.65

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/J

apan

2.56

12.9

311

.13

451.

1726

1.72

40.6

6L

& M

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/E

urop

ean

Uni

on1.

5016

.93

17.2

787

0.00

448.

0084

.93

Filt

er6

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/E

urop

ean

Uni

on1.

7914

.64

14.3

669

7.21

360.

8973

.85

Che

ster

field

Orig

inal

s K

ing

Siz

e F

ilter

har

d pa

ck1.

6716

.11

15.6

371

8.56

381.

4471

.14

Dia

na K

ing

Siz

e F

ilter

sof

t pac

k S

peci

ally

Mild

/E1.

9415

.52

13.3

568

8.14

350.

5272

.94

Mur

atti

Am

bass

ador

Kin

g S

ize

Filt

er h

ard

pack

/Eur

opea

n U

nion

1.80

16.1

114

.67

766.

1137

8.89

69.1

7M

erit

Kin

g S

ize

Filt

er h

ard

pack

/Eur

opea

n U

nion

1.47

15.5

116

.87

804.

7642

7.21

77.7

6

162

Par

liam

ent 1

00 F

ilter

sof

t pac

k/C

EM

A2.

1714

.15

13.5

965

8.53

344.

7062

.49

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck L

ight

/Ger

man

y/U

nite

d K

ingd

om1.

5815

.32

15.5

774

1.14

373.

4278

.86

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck L

ight

/Jap

an1.

4316

.22

18.6

072

3.08

381.

1266

.78

Mar

lbor

o 10

0 F

ilter

har

d pa

ck L

ight

/Ger

man

y2.

0613

.74

14.4

774

7.57

366.

0274

.81

Mer

it K

ing

Siz

e F

ilter

sof

t pac

k U

ltra-

light

/US

A1.

5814

.37

16.7

175

0.00

403.

8070

.44

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck U

ltra-

light

Men

thol

/US

A1.

5816

.52

17.0

382

7.22

422.

1588

.23

Par

liam

ent K

ing

Siz

e F

ilter

har

d pa

ck L

ight

/Jap

an1.

7915

.03

15.2

065

6.42

347.

4962

.51

Virg

inia

Slim

s 10

0 F

ilter

har

d pa

ck U

ltra-

light

Men

thol

1.88

14.7

314

.26

661.

1733

9.36

67.3

4C

hest

erfie

ld IN

TL

Kin

g S

ize

Filt

er h

ard

pack

Ultr

a-lig

ht/

1.40

14.0

015

.86

782.

8641

3.57

78.4

3P

hilip

Mor

ris 1

00 F

ilter

har

d pa

ck S

uper

L1.

8514

.16

16.0

072

2.16

375.

1469

.84

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck U

ltra-

light

/Eur

opea

n U

nion

1.47

14.7

616

.94

750.

3437

8.23

69.3

2D

iana

Kin

g S

ize

Filt

er h

ard

pack

Ultr

a-lig

ht/E

urop

ean

Uni

on1.

4115

.82

17.0

277

7.30

392.

2084

.40

Virg

inia

Slim

s 10

0 F

ilter

har

d pa

ck U

ltra-

light

Men

thol

1.39

20.0

727

.27

986.

3349

1.37

99.5

0P

hilip

Mor

ris O

ne K

ing

Siz

e F

ilter

har

d pa

ck/E

urop

ean

Uni

on1.

1614

.48

17.8

488

6.21

445.

6988

.36

Mur

atti

Kin

g S

ize

Filt

er h

ard

pack

Ultr

a-lig

ht 1

mg/

CE

MA

1.07

17.3

820

.65

997.

2050

0.93

92.7

1Lo

ngbe

ach

One

Kin

g S

ize

Filt

er h

ard

pack

/Aus

tral

ia1.

1913

.70

16.3

085

7.98

470.

5989

.50

Virg

inia

Slim

s 10

0 F

ilter

har

d pa

ck M

enth

ol 1

1.30

17.8

525

.54

902.

3145

2.31

81.5

4M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

/Mex

ico

2.47

13.4

410

.69

466.

8025

6.28

47.0

0R

affle

s 10

0 F

ilter

har

d pa

ck/E

urop

ean

Uni

on2.

7011

.22

10.9

351

7.04

266.

3049

.70

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/B

razi

l2.

2415

.67

12.6

854

6.43

290.

1852

.10

Pet

er J

acks

on K

ing

Siz

e F

ilter

har

d pa

ck M

enth

ol/A

ustr

alia

1.78

13.0

312

.08

535.

9628

6.52

56.2

4M

arlb

oro

100

Filt

er h

ard

pack

Lig

ht/U

SA

2.15

14.5

114

.65

565.

5830

6.98

54.7

4M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/N

orw

ay1.

8215

.05

14.4

561

2.64

319.

7857

.20

Che

ster

field

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/E

urop

ean

Uni

on1.

5414

.74

15.1

368

7.01

352.

6066

.49

Phi

lip M

orris

Kin

g S

ize

Filt

er h

ard

pack

Sup

er L

ight

1.73

14.6

814

.86

615.

6131

5.61

61.1

0M

erit

Kin

g S

ize

Filt

er s

oft p

ack

Ultr

a-lig

ht/U

SA

1.34

15.6

017

.99

694.

0336

4.93

66.5

7

163

1R4F

ilter

Ken

tuck

y R

efer

ence

1.83

14.3

716

.45

791.

2641

2.57

66.8

91R

4Filt

er K

entu

cky

Ref

eren

ce1.

9314

.20

15.2

370

4.15

355.

9657

.62

** M

ean

**15

.33

15.1

969

4.97

359.

4267

.55

** S

tand

ard

devi

atio

n **

1.55

3.12

121.

7059

.72

12.5

1**

Coe

ffici

ent o

f var

iatio

n **

0.10

0.21

0.18

0.17

0.19

Max

imum

2.70

20.0

727

.27

997.

2050

0.93

99.5

075

th p

erce

ntile

2.14

16.1

116

.41

774.

5139

6.02

74.7

990

th p

erce

ntile

2.38

17.4

017

.86

859.

1844

5.92

85.2

6M

edia

n1.

8415

.13

14.7

068

7.58

351.

5666

.53

Min

imum

1.07

11.2

210

.69

451.

1725

6.28

40.6

6

164

Bra

nd

Bu

tyra

ldeh

yde

Cro

ton

ald

ehyd

eM

eth

ylet

hyl

keto

ne

Pro

pio

nal

deh

yde

Mar

lbor

o Lo

ng S

ize

F h

ard

pack

/Arg

entin

a34

.72

28.5

888

.54

50.0

0M

arlb

oro

Long

Siz

e F

ilter

har

d pa

ck/V

enez

uela

39.9

531

.01

95.2

357

.54

SG

Ven

til R

egul

ar F

ilter

sof

t pac

k/E

urop

ean

Uni

on48

.45

36.3

511

2.97

68.7

8P

etra

Reg

ular

Filt

er h

ard

pack

/CE

MA

42.1

130

.27

97.3

058

.27

Mar

lbor

o K

ing

Siz

e F

ilter

sof

t pac

k/U

SA

36.1

822

.49

76.3

650

.13

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/N

orw

ay41

.15

34.0

410

3.27

56.6

8L

& M

Kin

g S

ize

Filt

er h

ard

pack

/Eur

opea

n U

nion

48.9

434

.41

110.

9667

.50

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/M

alay

sia

41.2

230

.00

82.2

657

.10

Che

ster

field

Orig

inal

s K

ing

Siz

e F

ilter

har

d pa

ck42

.63

29.7

490

.21

58.6

6L

& M

Kin

g S

ize

Filt

er h

ard

pack

/Mal

aysi

a35

.92

23.6

764

.00

48.2

9M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

25

s/A

ustr

alia

37.5

626

.60

79.9

652

.69

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/T

aiw

an33

.00

17.6

765

.86

47.6

2M

arlb

oro

100

Filt

er h

ard

pack

/Eur

opea

n U

nion

40.8

426

.74

88.1

158

.81

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck M

ediu

m/E

urop

ean

Uni

on38

.09

25.9

384

.33

55.5

2P

arlia

men

t 100

Filt

er s

oft p

ack

Ligh

t/US

A34

.36

17.5

168

.76

50.6

6M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

/Jap

an26

.25

15.5

561

.02

39.5

7L

& M

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/E

urop

ean

Uni

on52

.27

37.2

712

3.07

74.3

3F

ilter

6 K

ing

Siz

e F

ilter

har

d pa

ck L

ight

/Eur

opea

n U

nion

44.9

229

.05

95.7

062

.07

Che

ster

field

Orig

inal

s K

ing

Siz

e F

ilter

har

d pa

ck45

.27

31.5

097

.96

64.6

1D

iana

Kin

g S

ize

Filt

er s

oft p

ack

Spe

cial

ly M

ild/E

43.7

633

.66

96.4

960

.82

Mur

atti

Am

bass

ador

Kin

g S

ize

Filt

er h

ard

pack

/Eur

opea

n U

nion

43.0

627

.94

98.1

166

.61

Mer

it K

ing

Siz

e F

ilter

har

d pa

ck/E

urop

ean

Uni

on48

.91

36.9

411

9.93

67.0

1

165

Par

liam

ent 1

00 F

ilter

sof

t pac

k/C

EM

A37

.97

25.9

992

.26

57.9

7M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/G

erm

any/

Uni

ted

Kin

gdom

44.9

427

.59

95.4

464

.30

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck L

ight

/Jap

an38

.74

22.8

785

.10

62.5

2M

arlb

oro

100

Filt

er h

ard

pack

Lig

ht/G

erm

any

45.6

328

.25

95.1

065

.00

Mer

it K

ing

Siz

e F

ilter

sof

t pac

k U

ltra-

light

/US

A47

.66

31.9

699

.56

69.0

5M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

Ultr

a-lig

ht M

enth

ol/U

SA

51.4

636

.58

105.

5775

.38

Par

liam

ent K

ing

Siz

e F

ilter

har

d pa

ck L

ight

/Jap

an36

.82

23.6

385

.47

53.0

7V

irgin

ia S

lims

100

Filt

er h

ard

pack

Ultr

a-lig

ht M

enth

ol41

.49

29.1

082

.34

56.6

5C

hest

erfie

ld IN

TL

Kin

g S

ize

Filt

er h

ard

pack

Ultr

a-lig

ht/

46.1

436

.43

117.

7163

.93

Phi

lip M

orris

100

Filt

er h

ard

pack

Sup

er L

45.7

831

.57

94.8

663

.57

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck U

ltra-

light

/Eur

opea

n U

nion

45.6

530

.00

91.7

764

.56

Dia

na K

ing

Siz

e F

ilter

har

d pa

ck U

ltra-

light

/Eur

opea

n U

nion

47.7

333

.83

100.

2870

.35

Virg

inia

Slim

s 10

0 F

ilter

har

d pa

ck U

ltra-

light

Men

thol

60.0

038

.27

115.

9788

.42

Phi

lip M

orris

One

Kin

g S

ize

Filt

er h

ard

pack

/Eur

opea

n U

nion

56.3

833

.88

106.

5572

.76

Mur

atti

Kin

g S

ize

Filt

er h

ard

pack

Ultr

a-lig

ht 1

mg/

CE

MA

63.5

541

.31

124.

3985

.98

Long

beac

h O

ne K

ing

Siz

e F

ilter

har

d pa

ck/A

ustr

alia

53.8

739

.58

119.

1676

.30

Virg

inia

Slim

s 10

0 F

ilter

har

d pa

ck M

enth

ol 1

54.6

233

.31

104.

9274

.00

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/M

exic

o29

.88

19.9

667

.09

39.9

2R

affle

s 10

0 F

ilter

har

d pa

ck/E

urop

ean

Uni

on36

.85

23.0

778

.11

43.5

6M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

/Bra

zil

35.0

423

.62

79.0

647

.77

Pet

er J

acks

on K

ing

Siz

e F

ilter

har

d pa

ck M

enth

ol/A

ustr

alia

33.3

723

.93

79.1

045

.73

Mar

lbor

o 10

0 F

ilter

har

d pa

ck L

ight

/US

A37

.26

22.7

980

.19

49.0

2M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/N

orw

ay37

.86

24.7

886

.65

51.5

9C

hest

erfie

ld K

ing

Siz

e F

ilter

har

d pa

ck L

ight

/Eur

opea

n U

nion

41.4

327

.21

91.8

857

.60

Phi

lip M

orris

Kin

g S

ize

Filt

er h

ard

pack

Sup

er L

ight

38.2

124

.39

80.5

852

.02

Mer

it K

ing

Siz

e F

ilter

sof

t pac

k U

ltra-

light

/US

A45

.52

27.3

987

.84

59.0

3

166

1R4F

ilter

Ken

tuck

y R

efer

ence

51.0

428

.52

114.

2170

.44

1R4F

ilter

Ken

tuck

y R

efer

ence

43.0

625

.44

98.2

959

.90

** M

ean

**42

.95

28.8

493

.20

60.2

7**

Sta

ndar

d de

viat

ion

**7.

555.

8915

.78

10.7

3**

Coe

ffici

ent o

f var

iatio

n **

0.18

0.20

0.17

0.18

Max

imum

63.5

541

.31

124.

3988

.42

75th

per

cent

ile47

.28

33.5

710

2.52

66.9

190

th p

erce

ntile

52.4

336

.62

116.

1574

.03

Med

ian

42.3

728

.55

93.5

658

.92

Min

imum

26.2

515

.55

61.0

239

.57

167

Bra

nd

Fo

rmal

deh

yde

Acr

ylo

nit

rile

Ben

zen

e1,

3-B

uta

die

ne

Iso

pre

ne

Mar

lbor

o Lo

ng S

ize

F h

ard

pack

/Arg

entin

a45

.57

9.34

34.3

446

.79

443.

87M

arlb

oro

Long

Siz

e F

ilter

har

d pa

ck/V

enez

uela

44.8

710

.00

34.5

749

.30

440.

20S

G V

entil

Reg

ular

Filt

er s

oft p

ack/

Eur

opea

n U

nion

73.7

210

.88

37.9

152

.03

343.

92P

etra

Reg

ular

Filt

er h

ard

pack

/CE

MA

57.7

310

.27

39.1

945

.73

323.

78M

arlb

oro

Kin

g S

ize

Filt

er s

oft p

ack/

US

A29

.16

13.3

837

.24

50.1

346

8.89

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/N

orw

ay49

.04

10.1

939

.57

53.0

843

3.17

L &

M K

ing

Siz

e F

ilter

har

d pa

ck/E

urop

ean

Uni

on52

.98

11.4

943

.88

56.8

642

3.94

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/M

alay

sia

28.0

111

.72

34.3

043

.39

352.

49C

hest

erfie

ld O

rigin

als

Kin

g S

ize

Filt

er h

ard

pack

46.4

49.

2334

.69

47.2

736

9.59

L &

M K

ing

Siz

e F

ilter

har

d pa

ck/M

alay

sia

33.7

510

.25

33.5

844

.83

406.

67M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

25

s/A

ustr

alia

28.1

99.

9233

.15

41.4

335

7.14

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/T

aiw

an28

.77

10.9

730

.70

44.0

538

9.87

Mar

lbor

o 10

0 F

ilter

har

d pa

ck/E

urop

ean

Uni

on36

.61

12.0

336

.87

49.6

942

4.67

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck M

ediu

m/E

urop

ean

Uni

on44

.38

10.8

837

.22

48.3

540

0.00

Par

liam

ent 1

00 F

ilter

sof

t pac

k Li

ght/U

SA

25.3

911

.12

31.6

646

.60

421.

99M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

/Jap

an29

.41

6.72

25.7

035

.78

283.

20L

& M

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/E

urop

ean

Uni

on53

.00

12.0

047

.00

62.4

046

4.67

Filt

er6

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/E

urop

ean

Uni

on71

.79

11.6

241

.01

54.8

040

0.00

Che

ster

field

Orig

inal

s K

ing

Siz

e F

ilter

har

d pa

ck44

.07

12.5

142

.81

55.6

944

4.31

Dia

na K

ing

Siz

e F

ilter

sof

t pac

k S

peci

ally

Mild

/E49

.23

10.5

238

.09

49.2

335

2.58

Mur

atti

Am

bass

ador

Kin

g S

ize

Filt

er h

ard

pack

/Eur

opea

n U

nion

59.3

910

.00

34.4

453

.28

364.

44M

erit

Kin

g S

ize

Filt

er h

ard

pack

/Eur

opea

n U

nion

43.8

112

.65

43.8

860

.48

482.

99

168

Par

liam

ent 1

00 F

ilter

sof

t pac

k/C

EM

A30

.78

9.08

31.2

448

.11

383.

41M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/G

erm

any/

Uni

ted

Kin

gdom

46.5

813

.61

39.8

161

.65

517.

09M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/J

apan

52.3

19.

9334

.76

56.0

837

7.62

Mar

lbor

o 10

0 F

ilter

har

d pa

ck L

ight

/Ger

man

y40

.05

12.3

336

.26

54.0

347

3.30

Mer

it K

ing

Siz

e F

ilter

sof

t pac

k U

ltra-

light

/US

A24

.18

14.9

447

.66

59.2

451

0.13

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck U

ltra-

light

Men

thol

/US

A36

.77

13.6

744

.37

59.1

148

7.97

Par

liam

ent K

ing

Siz

e F

ilter

har

d pa

ck L

ight

/Jap

an24

.80

9.66

32.0

150

.67

443.

58V

irgin

ia S

lims

100

Filt

er h

ard

pack

Ultr

a-lig

ht M

enth

ol32

.23

12.0

236

.54

50.4

844

5.21

Che

ster

field

INT

L K

ing

Siz

e F

ilter

har

d pa

ck U

ltra-

light

/46

.86

12.5

745

.57

65.7

950

3.57

Phi

lip M

orris

100

Filt

er h

ard

pack

Sup

er L

35.4

111

.89

41.4

154

.11

478.

38M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

Ultr

a-lig

ht/E

urop

ean

Uni

on45

.85

11.0

937

.62

53.4

046

3.27

Dia

na K

ing

Siz

e F

ilter

har

d pa

ck U

ltra-

light

/Eur

opea

n U

nion

40.2

813

.05

43.5

561

.35

468.

79V

irgin

ia S

lims

100

Filt

er h

ard

pack

Ultr

a-lig

ht M

enth

ol36

.62

18.8

551

.08

74.8

969

3.53

Phi

lip M

orris

One

Kin

g S

ize

Filt

er h

ard

pack

/Eur

opea

n U

nion

26.2

117

.33

50.0

075

.52

745.

69M

urat

ti K

ing

Siz

e F

ilter

har

d pa

ck U

ltra-

light

1 m

g/C

EM

A28

.79

17.0

147

.85

75.1

464

4.86

Long

beac

h O

ne K

ing

Siz

e F

ilter

har

d pa

ck/A

ustr

alia

90.5

010

.17

44.2

965

.46

499.

16V

irgin

ia S

lims

100

Filt

er h

ard

pack

Men

thol

126

.15

17.3

141

.62

67.6

270

0.77

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/M

exic

o35

.51

9.92

30.1

239

.39

363.

16R

affle

s 10

0 F

ilter

har

d pa

ck/E

urop

ean

Uni

on35

.89

10.1

932

.26

43.2

239

7.78

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/B

razi

l38

.97

12.6

337

.86

52.0

542

1.43

Pet

er J

acks

on K

ing

Siz

e F

ilter

har

d pa

ck M

enth

ol/A

ustr

alia

73.2

010

.67

37.1

955

.73

424.

72M

arlb

oro

100

Filt

er h

ard

pack

Lig

ht/U

SA

24.5

615

.95

45.7

255

.02

539.

53M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/N

orw

ay38

.57

12.3

135

.44

47.7

541

0.44

Che

ster

field

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/E

urop

ean

Uni

on47

.01

15.2

643

.12

59.6

152

2.08

Phi

lip M

orris

Kin

g S

ize

Filt

er h

ard

pack

Sup

er L

ight

33.5

314

.97

39.7

756

.71

540.

46M

erit

Kin

g S

ize

Filt

er s

oft p

ack

Ultr

a-lig

ht/U

SA

21.8

719

.48

44.7

064

.18

701.

49

169

1R4F

ilter

Ken

tuck

y R

efer

ence

33.0

616

.01

45.5

257

.38

520.

221R

4Filt

er K

entu

cky

Ref

eren

ce31

.09

15.4

439

.43

48.6

548

1.35

** M

ean

**41

.06

12.3

038

.97

54.0

745

9.03

** S

tand

ard

devi

atio

n **

14.4

82.

735.

698.

8599

.48

** C

oeffi

cien

t of v

aria

tion

**0.

350.

220.

150.

160.

22M

axim

um90

.50

19.4

851

.08

75.5

274

5.69

75th

per

cent

ile46

.79

13.5

543

.79

59.2

149

6.36

90th

per

cent

ile57

.90

16.1

145

.85

65.4

955

0.90

Med

ian

37.6

711

.81

38.0

053

.34

443.

72M

inim

um21

.87

6.72

25.7

035

.78

283.

20

170

Bra

nd

Sty

ren

eT

olu

ene

Am

mo

nia

To

tal h

ydro

gen

cya

nid

e

Mar

lbor

o Lo

ng S

ize

F h

ard

pack

/Arg

entin

a12

.31

59.9

516

.51

150.

61M

arlb

oro

Long

Siz

e F

ilter

har

d pa

ck/V

enez

uela

12.7

664

.92

22.9

616

9.70

SG

Ven

til R

egul

ar F

ilter

sof

t pac

k/E

urop

ean

Uni

on15

.07

68.2

422

.43

180.

07P

etra

Reg

ular

Filt

er h

ard

pack

/CE

MA

15.0

873

.62

21.5

115

5.24

Mar

lbor

o K

ing

Siz

e F

ilter

sof

t pac

k/U

SA

13.5

171

.87

27.7

321

4.89

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/N

orw

ay13

.85

71.8

320

.67

145.

24L

& M

Kin

g S

ize

Filt

er h

ard

pack

/Eur

opea

n U

nion

16.9

181

.54

21.2

220

2.55

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/M

alay

sia

14.8

969

.00

40.7

220

1.27

Che

ster

field

Orig

inal

s K

ing

Siz

e F

ilter

har

d pa

ck14

.02

66.5

521

.29

182.

37L

& M

Kin

g S

ize

Filt

er h

ard

pack

/Mal

aysi

a13

.17

64.1

323

.04

198.

42M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

25

s/A

ustr

alia

13.0

762

.48

25.1

320

3.28

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/T

aiw

an8.

9056

.78

26.1

219

2.38

Mar

lbor

o 10

0 F

ilter

har

d pa

ck/E

urop

ean

Uni

on14

.67

72.6

926

.48

200.

48M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

Med

ium

/Eur

opea

n U

nion

13.8

169

.95

17.8

917

0.36

Par

liam

ent 1

00 F

ilter

sof

t pac

k Li

ght/U

SA

8.88

54.9

026

.72

201.

54M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

/Jap

an8.

3645

.59

16.9

511

5.70

L &

M K

ing

Siz

e F

ilter

har

d pa

ck L

ight

/Eur

opea

n U

nion

18.5

384

.13

20.2

019

9.53

Filt

er6

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/E

urop

ean

Uni

on14

.13

73.7

414

.97

177.

54C

hest

erfie

ld O

rigin

als

Kin

g S

ize

Filt

er h

ard

pack

14.1

377

.13

20.3

619

2.57

Dia

na K

ing

Siz

e F

ilter

sof

t pac

k S

peci

ally

Mild

/E14

.85

70.5

718

.30

180.

26M

urat

ti A

mba

ssad

or K

ing

Siz

e F

ilter

har

d pa

ck/E

urop

ean

Uni

on10

.61

58.8

319

.17

143.

17M

erit

Kin

g S

ize

Filt

er h

ard

pack

/Eur

opea

n U

nion

15.7

181

.02

22.4

520

3.06

171

Par

liam

ent 1

00 F

ilter

sof

t pac

k/C

EM

A8.

7653

.41

23.3

217

3.69

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck L

ight

/Ger

man

y/U

nite

d K

ingd

om14

.94

77.2

220

.06

206.

71M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/J

apan

10.7

761

.12

19.7

918

7.83

Mar

lbor

o 10

0 F

ilter

har

d pa

ck L

ight

/Ger

man

y14

.22

72.0

417

.82

169.

22M

erit

Kin

g S

ize

Filt

er s

oft p

ack

Ultr

a-lig

ht/U

SA

16.0

187

.97

26.2

726

7.53

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck U

ltra-

light

Men

thol

/US

A15

.13

79.4

327

.53

276.

39P

arlia

men

t Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/J

apan

8.77

53.6

327

.32

196.

15V

irgin

ia S

lims

100

Filt

er h

ard

pack

Ultr

a-lig

ht M

enth

ol12

.61

69.2

028

.78

229.

89C

hest

erfie

ld IN

TL

Kin

g S

ize

Filt

er h

ard

pack

Ultr

a-lig

ht/

15.4

380

.50

21.1

420

2.21

Phi

lip M

orris

100

Filt

er h

ard

pack

Sup

er L

14.4

975

.62

19.5

121

8.59

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck U

ltra-

light

/Eur

opea

n U

nion

14.3

571

.84

16.7

321

1.70

Dia

na K

ing

Siz

e F

ilter

har

d pa

ck U

ltra-

light

/Eur

opea

n U

nion

16.1

777

.87

18.3

023

7.73

Virg

inia

Slim

s 10

0 F

ilter

har

d pa

ck U

ltra-

light

Men

thol

16.6

290

.86

24.9

639

0.22

Phi

lip M

orris

One

Kin

g S

ize

Filt

er h

ard

pack

/Eur

opea

n U

nion

16.7

293

.02

21.8

128

7.33

Mur

atti

Kin

g S

ize

Filt

er h

ard

pack

Ultr

a-lig

ht 1

mg/

CE

MA

16.3

683

.18

24.8

632

2.80

Long

beac

h O

ne K

ing

Siz

e F

ilter

har

d pa

ck/A

ustr

alia

16.6

472

.44

9.24

157.

90V

irgin

ia S

lims

100

Filt

er h

ard

pack

Men

thol

115

.15

75.3

122

.69

374.

31M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

/Mex

ico

10.4

552

.06

19.2

714

8.70

Raf

fles

100

Filt

er h

ard

pack

/Eur

opea

n U

nion

11.0

056

.93

11.1

913

5.07

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/B

razi

l12

.81

65.0

017

.46

148.

17P

eter

Jac

kson

Kin

g S

ize

Filt

er h

ard

pack

Men

thol

/Aus

tral

ia11

.40

57.5

310

.06

114.

66M

arlb

oro

100

Filt

er h

ard

pack

Lig

ht/U

SA

11.6

381

.07

24.5

122

0.51

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck L

ight

/Nor

way

13.2

465

.82

20.0

517

8.08

Che

ster

field

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/E

urop

ean

Uni

on13

.25

75.8

418

.12

172.

99P

hilip

Mor

ris K

ing

Siz

e F

ilter

har

d pa

ck S

uper

Lig

ht12

.83

71.9

719

.42

201.

04M

erit

Kin

g S

ize

Filt

er s

oft p

ack

Ultr

a-lig

ht/U

SA

13.4

378

.36

20.8

231

9.33

172

1R4F

ilter

Ken

tuck

y R

efer

ence

16.4

596

.28

16.9

923

3.28

1R4F

ilter

Ken

tuck

y R

efer

ence

13.0

680

.98

17.2

021

8.76

** M

ean

**13

.60

71.1

221

.16

203.

62**

Sta

ndar

d de

viat

ion

**2.

4111

.11

5.19

57.3

4**

Coe

ffici

ent o

f var

iatio

n **

0.18

0.16

0.25

0.28

Max

imum

18.5

396

.28

40.7

239

0.22

75th

per

cent

ile15

.12

78.2

424

.21

217.

6790

th p

erce

ntile

16.4

783

.27

26.7

827

7.49

Med

ian

13.9

371

.92

20.7

519

8.98

Min

imum

8.36

45.5

99.

2411

4.66

173

Bra

nd

Imp

ing

erh

ydro

gen

cyan

ide

Pad

hyd

rog

ency

anid

e

Nit

ric

oxi

de

Nit

rog

eno

xid

es1

Am

ino

nap

hth

alen

e

Mar

lbor

o Lo

ng S

ize

F h

ard

pack

/Arg

entin

a91

.98

58.6

314

4.81

159.

9116

.75

Mar

lbor

o Lo

ng S

ize

Filt

er h

ard

pack

/Ven

ezue

la10

6.18

63.5

214

4.72

158.

7918

.94

SG

Ven

til R

egul

ar F

ilter

sof

t pac

k/E

urop

ean

Uni

on11

3.18

66.8

913

1.76

156.

7617

.50

Pet

ra R

egul

ar F

ilter

har

d pa

ck/C

EM

A10

0.97

54.2

799

.46

111.

8919

.89

Mar

lbor

o K

ing

Siz

e F

ilter

sof

t pac

k/U

SA

139.

6975

.20

212.

4423

4.67

16.8

0M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

/Nor

way

88.8

956

.35

153.

3716

7.31

17.2

1L

& M

Kin

g S

ize

Filt

er h

ard

pack

/Eur

opea

n U

nion

131.

7070

.85

153.

1916

8.09

18.8

3M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

/Mal

aysi

a12

3.39

77.8

321

4.93

247.

0624

.80

Che

ster

field

Orig

inal

s K

ing

Siz

e F

ilter

har

d pa

ck11

2.32

70.0

513

4.54

153.

0917

.84

L &

M K

ing

Siz

e F

ilter

har

d pa

ck/M

alay

sia

126.

1772

.25

166.

2518

7.08

20.6

3M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

25

s/A

ustr

alia

135.

7667

.52

183.

6120

7.98

17.5

2M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

/Tai

wan

127.

6764

.71

208.

3722

3.35

16.7

0M

arlb

oro

100

Filt

er h

ard

pack

/Eur

opea

n U

nion

126.

5673

.92

168.

7218

4.58

17.2

7M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

Med

ium

/Eur

opea

n U

nion

107.

8962

.47

134.

0215

0.52

17.5

3P

arlia

men

t 100

Filt

er s

oft p

ack

Ligh

t/US

A13

2.66

68.8

822

2.41

238.

5916

.39

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/J

apan

68.7

146

.95

144.

5315

8.59

18.7

9L

& M

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/E

urop

ean

Uni

on12

6.27

73.2

716

1.33

177.

3320

.07

Filt

er6

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/E

urop

ean

Uni

on11

7.93

59.6

183

.80

91.0

610

.11

Che

ster

field

Orig

inal

s K

ing

Siz

e F

ilter

har

d pa

ck12

4.19

68.3

212

8.74

148.

5016

.53

Dia

na K

ing

Siz

e F

ilter

sof

t pac

k S

peci

ally

Mild

/E11

6.44

63.8

113

4.02

152.

0616

.29

Mur

atti

Am

bass

ador

Kin

g S

ize

Filt

er h

ard

pack

/Eur

opea

n U

nion

87.6

755

.50

110.

0012

7.78

16.1

7M

erit

Kin

g S

ize

Filt

er h

ard

pack

/Eur

opea

n U

nion

132.

3870

.68

191.

1620

8.16

15.7

8

174

Par

liam

ent 1

00 F

ilter

sof

t pac

k/C

EM

A11

3.36

60.2

817

3.27

186.

6418

.76

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck L

ight

/Ger

man

y/U

nite

d K

ingd

om13

8.67

68.0

413

2.91

148.

1012

.85

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck L

ight

/Jap

an11

8.18

69.6

517

9.72

196.

5016

.85

Mar

lbor

o 10

0 F

ilter

har

d pa

ck L

ight

/Ger

man

y10

5.15

64.1

312

3.30

135.

4412

.72

Mer

it K

ing

Siz

e F

ilter

sof

t pac

k U

ltra-

light

/US

A18

2.09

85.4

427

8.48

312.

0317

.53

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck U

ltra-

light

Men

thol

/US

A18

5.38

91.0

129

7.47

331.

6516

.52

Par

liam

ent K

ing

Siz

e F

ilter

har

d pa

ck L

ight

/Jap

an13

1.28

64.8

624

3.58

262.

5717

.37

Virg

inia

Slim

s 10

0 F

ilter

har

d pa

ck U

ltra-

light

Men

thol

157.

2372

.61

243.

0927

6.60

18.5

1C

hest

erfie

ld IN

TL

Kin

g S

ize

Filt

er h

ard

pack

Ultr

a-lig

ht/

133.

2169

.00

140.

7115

2.86

14.2

1P

hilip

Mor

ris 1

00 F

ilter

har

d pa

ck S

uper

L14

5.51

73.1

415

8.92

176.

2218

.49

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck U

ltra-

light

/Eur

opea

n U

nion

134.

8376

.87

142.

8616

0.54

15.5

1D

iana

Kin

g S

ize

Filt

er h

ard

pack

Ultr

a-lig

ht/E

urop

ean

Uni

on15

6.10

81.6

317

7.30

204.

9616

.81

Virg

inia

Slim

s 10

0 F

ilter

har

d pa

ck U

ltra-

light

Men

thol

253.

6713

6.55

348.

9238

9.93

13.8

8P

hilip

Mor

ris O

ne K

ing

Siz

e F

ilter

har

d pa

ck/E

urop

ean

Uni

on19

2.07

95.2

624

1.38

262.

9316

.12

Mur

atti

Kin

g S

ize

Filt

er h

ard

pack

Ultr

a-lig

ht 1

mg/

CE

MA

213.

9310

8.88

273.

8330

2.80

16.7

3Lo

ngbe

ach

One

Kin

g S

ize

Filt

er h

ard

pack

/Aus

tral

ia97

.48

60.3

485

.71

98.3

29.

83V

irgin

ia S

lims

100

Filt

er h

ard

pack

Men

thol

123

3.77

140.

4632

0.77

352.

3114

.00

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/M

exic

o94

.45

54.2

512

1.46

131.

5813

.16

Raf

fles

100

Filt

er h

ard

pack

/Eur

opea

n U

nion

86.0

049

.07

118.

5212

9.26

11.5

9M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

/Bra

zil

93.0

855

.04

118.

3012

9.46

13.7

9P

eter

Jac

kson

Kin

g S

ize

Filt

er h

ard

pack

Men

thol

/Aus

tral

ia69

.04

45.6

264

.61

71.9

19.

10M

arlb

oro

100

Filt

er h

ard

pack

Lig

ht/U

SA

152.

5667

.95

246.

5126

7.44

15.8

6M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/N

orw

ay11

7.97

60.1

116

5.93

180.

2214

.45

Che

ster

field

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/E

urop

ean

Uni

on11

2.40

60.5

812

3.38

137.

6614

.94

Phi

lip M

orris

Kin

g S

ize

Filt

er h

ard

pack

Sup

er L

ight

133.

9967

.11

152.

0216

8.21

12.4

9M

erit

Kin

g S

ize

Filt

er s

oft p

ack

Ultr

a-lig

ht/U

SA

219.

4899

.85

251.

4927

2.39

19.3

3

175

1R4F

ilter

Ken

tuck

y R

efer

ence

163.

4469

.84

334.

4335

4.64

15.7

91R

4Filt

er K

entu

cky

Ref

eren

ce15

3.63

65.1

330

1.55

325.

9114

.35

** M

ean

**13

2.53

71.0

817

9.73

198.

6416

.20

** S

tand

ard

devi

atio

n **

39.6

418

.60

68.4

574

.36

2.89

** C

oeffi

cien

t of v

aria

tion

**0.

300.

260.

380.

370.

18M

axim

um25

3.67

140.

4634

8.92

389.

9324

.80

75th

per

cent

ile14

4.06

73.2

322

0.54

244.

9417

.53

90th

per

cent

ile18

6.05

91.4

428

0.38

313.

4118

.98

Med

ian

126.

4268

.00

160.

1317

6.77

16.6

1M

inim

um68

.71

45.6

264

.61

71.9

19.

10

176

Bra

nd

2-A

min

on

aph

thal

ene

3-A

min

ob

iph

enyl

4-A

min

ob

iph

enyl

Mar

lbor

o Lo

ng S

ize

F h

ard

pack

/Arg

entin

a10

.19

3.11

2.23

Mar

lbor

o Lo

ng S

ize

Filt

er h

ard

pack

/Ven

ezue

la11

.46

3.41

2.49

SG

Ven

til R

egul

ar F

ilter

sof

t pac

k/E

urop

ean

Uni

on10

.74

3.29

2.57

Pet

ra R

egul

ar F

ilter

har

d pa

ck/C

EM

A10

.92

3.21

2.31

Mar

lbor

o K

ing

Siz

e F

ilter

sof

t pac

k/U

SA

10.3

12.

942.

21M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

/Nor

way

10.5

33.

082.

26L

& M

Kin

g S

ize

Filt

er h

ard

pack

/Eur

opea

n U

nion

11.9

73.

522.

66M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

/Mal

aysi

a14

.34

4.14

3.18

Che

ster

field

Orig

inal

s K

ing

Siz

e F

ilter

har

d pa

ck10

.67

2.69

2.08

L &

M K

ing

Siz

e F

ilter

har

d pa

ck/M

alay

sia

11.7

92.

922.

17M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

25

s/A

ustr

alia

11.0

13.

002.

30M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

/Tai

wan

10.2

22.

752.

09M

arlb

oro

100

Filt

er h

ard

pack

/Eur

opea

n U

nion

9.91

2.82

2.16

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck M

ediu

m/E

urop

ean

Uni

on10

.82

2.68

2.10

Par

liam

ent 1

00 F

ilter

sof

t pac

k Li

ght/U

SA

9.54

2.71

2.00

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/J

apan

10.5

52.

862.

10L

& M

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/E

urop

ean

Uni

on12

.80

3.94

3.09

Filt

er6

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/E

urop

ean

Uni

on6.

031.

681.

31C

hest

erfie

ld O

rigin

als

Kin

g S

ize

Filt

er h

ard

pack

10.9

62.

772.

16D

iana

Kin

g S

ize

Filt

er s

oft p

ack

Spe

cial

ly M

ild/E

10.1

52.

662.

15M

urat

ti A

mba

ssad

or K

ing

Siz

e F

ilter

har

d pa

ck/E

urop

ean

Uni

on9.

562.

742.

06M

erit

Kin

g S

ize

Filt

er h

ard

pack

/Eur

opea

n U

nion

11.1

63.

402.

78

177

Par

liam

ent 1

00 F

ilter

sof

t pac

k/C

EM

A11

.61

3.36

2.59

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck L

ight

/Ger

man

y/U

nite

d K

ingd

om8.

102.

441.

87M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/J

apan

10.9

82.

922.

17M

arlb

oro

100

Filt

er h

ard

pack

Lig

ht/G

erm

any

8.01

2.17

1.70

Mer

it K

ing

Siz

e F

ilter

sof

t pac

k U

ltra-

light

/US

A11

.14

3.19

2.56

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck U

ltra-

light

Men

thol

/US

A11

.20

3.37

2.65

Par

liam

ent K

ing

Siz

e F

ilter

har

d pa

ck L

ight

/Jap

an11

.34

3.29

2.59

Virg

inia

Slim

s 10

0 F

ilter

har

d pa

ck U

ltra-

light

Men

thol

12.3

92.

882.

24C

hest

erfie

ld IN

TL

Kin

g S

ize

Filt

er h

ard

pack

Ultr

a-lig

ht/

10.0

03.

042.

49P

hilip

Mor

ris 1

00 F

ilter

har

d pa

ck S

uper

L11

.19

2.62

2.08

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck U

ltra-

light

/Eur

opea

n U

nion

10.2

72.

522.

09D

iana

Kin

g S

ize

Filt

er h

ard

pack

Ultr

a-lig

ht/E

urop

ean

Uni

on10

.92

3.22

2.50

Virg

inia

Slim

s 10

0 F

ilter

har

d pa

ck U

ltra-

light

Men

thol

8.78

2.55

2.10

Phi

lip M

orris

One

Kin

g S

ize

Filt

er h

ard

pack

/Eur

opea

n U

nion

11.4

73.

713.

09M

urat

ti K

ing

Siz

e F

ilter

har

d pa

ck U

ltra-

light

1 m

g/C

EM

A10

.56

3.37

2.86

Long

beac

h O

ne K

ing

Siz

e F

ilter

har

d pa

ck/A

ustr

alia

6.55

1.78

1.37

Virg

inia

Slim

s 10

0 F

ilter

har

d pa

ck M

enth

ol 1

8.38

2.55

2.11

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/M

exic

o7.

812.

301.

79R

affle

s 10

0 F

ilter

har

d pa

ck/E

urop

ean

Uni

on6.

561.

791.

30M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

/Bra

zil

8.08

2.46

1.84

Pet

er J

acks

on K

ing

Siz

e F

ilter

har

d pa

ck M

enth

ol/A

ustr

alia

5.00

1.53

1.16

Mar

lbor

o 10

0 F

ilter

har

d pa

ck L

ight

/US

A10

.65

2.78

2.17

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck L

ight

/Nor

way

9.18

2.70

2.14

Che

ster

field

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/E

urop

ean

Uni

on9.

092.

492.

01P

hilip

Mor

ris K

ing

Siz

e F

ilter

har

d pa

ck S

uper

Lig

ht7.

982.

361.

86M

erit

Kin

g S

ize

Filt

er s

oft p

ack

Ultr

a-lig

ht/U

SA

10.4

52.

672.

33

178

1R4F

ilter

Ken

tuck

y R

efer

ence

10.2

22.

952.

491R

4Filt

er K

entu

cky

Ref

eren

ce9.

432.

962.

28**

Mea

n **

10.0

62.

852.

22**

Sta

ndar

d de

viat

ion

**1.

760.

530.

43**

Coe

ffici

ent o

f var

iatio

n **

0.18

0.19

0.19

Max

imum

14.3

44.

143.

1875

th p

erce

ntile

11.1

13.

212.

4990

th p

erce

ntile

11.6

33.

402.

67M

edia

n10

.49

2.84

2.17

Min

imum

5.00

1.53

1.16

179

Bra

nd

Ben

zo[a

]pyr

ene

Cat

ech

ol

m-

and

p-C

reso

lo

-Cre

sol

Mar

lbor

o Lo

ng S

ize

F h

ard

pack

/Arg

entin

a7.

7048

.73

9.15

3.65

Mar

lbor

o Lo

ng S

ize

Filt

er h

ard

pack

/Ven

ezue

la8.

6053

.07

10.5

54.

51S

G V

entil

Reg

ular

Filt

er s

oft p

ack/

Eur

opea

n U

nion

8.85

58.1

811

.42

4.08

Pet

ra R

egul

ar F

ilter

har

d pa

ck/C

EM

A13

.82

62.1

613

.68

5.00

Mar

lbor

o K

ing

Siz

e F

ilter

sof

t pac

k/U

SA

9.86

44.8

47.

913.

19M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

/Nor

way

7.10

53.4

19.

384.

11L

& M

Kin

g S

ize

Filt

er h

ard

pack

/Eur

opea

n U

nion

9.99

63.7

811

.33

4.84

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/M

alay

sia

11.7

754

.30

12.5

34.

45C

hest

erfie

ld O

rigin

als

Kin

g S

ize

Filt

er h

ard

pack

10.3

553

.45

9.79

3.81

L &

M K

ing

Siz

e F

ilter

har

d pa

ck/M

alay

sia

12.0

454

.83

9.79

3.65

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck 2

5 s/

Aus

tral

ia9.

0841

.89

8.61

2.74

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/T

aiw

an9.

1547

.18

9.74

3.76

Mar

lbor

o 10

0 F

ilter

har

d pa

ck/E

urop

ean

Uni

on10

.44

51.8

98.

553.

35M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

Med

ium

/Eur

opea

n U

nion

9.79

47.6

87.

632.

98P

arlia

men

t 100

Filt

er s

oft p

ack

Ligh

t/US

A9.

7646

.64

8.17

3.07

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/J

apan

8.21

47.3

09.

413.

65L

& M

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/E

urop

ean

Uni

on9.

6161

.60

8.60

3.47

Filt

er6

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/E

urop

ean

Uni

on13

.84

65.3

66.

032.

39C

hest

erfie

ld O

rigin

als

Kin

g S

ize

Filt

er h

ard

pack

9.98

49.6

46.

652.

48D

iana

Kin

g S

ize

Filt

er s

oft p

ack

Spe

cial

ly M

ild/E

9.77

56.6

510

.82

3.94

Mur

atti

Am

bass

ador

Kin

g S

ize

Filt

er h

ard

pack

/Eur

opea

n U

nion

12.5

565

.11

12.9

44.

59M

erit

Kin

g S

ize

Filt

er h

ard

pack

/Eur

opea

n U

nion

9.18

49.2

56.

462.

62

180

Par

liam

ent 1

00 F

ilter

sof

t pac

k/C

EM

A7.

7157

.93

10.8

33.

79M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/G

erm

any/

Uni

ted

Kin

gdom

9.97

49.4

95.

512.

29M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/J

apan

9.57

57.2

07.

692.

80M

arlb

oro

100

Filt

er h

ard

pack

Lig

ht/G

erm

any

9.44

48.2

07.

092.

78M

erit

Kin

g S

ize

Filt

er s

oft p

ack

Ultr

a-lig

ht/U

SA

7.69

38.8

67.

592.

66M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

Ultr

a-lig

ht M

enth

ol/U

SA

8.85

38.2

35.

762.

10P

arlia

men

t Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/J

apan

7.00

43.5

27.

042.

59V

irgin

ia S

lims

100

Filt

er h

ard

pack

Ultr

a-lig

ht M

enth

ol7.

0339

.10

6.06

2.00

Che

ster

field

INT

L K

ing

Siz

e F

ilter

har

d pa

ck U

ltra-

light

/7.

9649

.57

6.29

2.61

Phi

lip M

orris

100

Filt

er h

ard

pack

Sup

er L

8.16

47.3

58.

052.

89M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

Ultr

a-lig

ht/E

urop

ean

Uni

on7.

7242

.18

5.51

2.08

Dia

na K

ing

Siz

e F

ilter

har

d pa

ck U

ltra-

light

/Eur

opea

n U

nion

9.35

47.8

76.

672.

42V

irgin

ia S

lims

100

Filt

er h

ard

pack

Ultr

a-lig

ht M

enth

ol8.

3736

.91

4.46

1.76

Phi

lip M

orris

One

Kin

g S

ize

Filt

er h

ard

pack

/Eur

opea

n U

nion

5.67

38.3

66.

032.

03M

urat

ti K

ing

Siz

e F

ilter

har

d pa

ck U

ltra-

light

1 m

g/C

EM

A7.

1642

.43

7.20

2.74

Long

beac

h O

ne K

ing

Siz

e F

ilter

har

d pa

ck/A

ustr

alia

8.08

46.8

16.

552.

69V

irgin

ia S

lims

100

Filt

er h

ard

pack

Men

thol

16.

6734

.00

4.69

1.65

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/M

exic

o9.

2641

.46

6.92

2.98

Raf

fles

100

Filt

er h

ard

pack

/Eur

opea

n U

nion

11.0

961

.96

9.81

4.32

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/B

razi

l9.

9449

.33

7.99

3.54

Pet

er J

acks

on K

ing

Siz

e F

ilter

har

d pa

ck M

enth

ol/A

ustr

alia

8.23

48.7

65.

282.

39M

arlb

oro

100

Filt

er h

ard

pack

Lig

ht/U

SA

9.47

43.0

25.

532.

24M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/N

orw

ay9.

0846

.10

6.04

2.58

Che

ster

field

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/E

urop

ean

Uni

on8.

3245

.45

4.68

1.90

Phi

lip M

orris

Kin

g S

ize

Filt

er h

ard

pack

Sup

er L

ight

7.68

40.9

85.

092.

08M

erit

Kin

g S

ize

Filt

er s

oft p

ack

Ultr

a-lig

ht/U

SA

6.21

34.1

07.

012.

83

181

1R4F

ilter

Ken

tuck

y R

efer

ence

7.44

47.6

05.

522.

311R

4Filt

er K

entu

cky

Ref

eren

ce6.

9946

.32

5.39

2.36

** M

ean

**9.

0348

.80

7.83

3.03

** S

tand

ard

devi

atio

n **

1.76

7.93

2.30

0.86

** C

oeffi

cien

t of v

aria

tion

**0.

190.

160.

290.

28M

axim

um13

.84

65.3

613

.68

5.00

75th

per

cent

ile9.

8453

.44

9.40

3.65

90th

per

cent

ile11

.15

61.6

410

.88

4.33

Med

ian

9.08

47.7

87.

402.

79M

inim

um5.

6734

.00

4.46

1.65

182

Bra

nd

Hyd

roq

uin

on

eP

hen

ol

Res

orc

ino

lP

yrid

ine

Qu

ino

line

Mar

lbor

o Lo

ng S

ize

F h

ard

pack

/Arg

entin

a49

.72

12.5

50.

8824

.81

0.36

Mar

lbor

o Lo

ng S

ize

Filt

er h

ard

pack

/Ven

ezue

la67

.79

16.9

80.

6625

.83

0.44

SG

Ven

til R

egul

ar F

ilter

sof

t pac

k/E

urop

ean

Uni

on85

.00

15.6

11.

0523

.78

0.41

Pet

ra R

egul

ar F

ilter

har

d pa

ck/C

EM

A75

.73

19.3

51.

4723

.30

0.46

Mar

lbor

o K

ing

Siz

e F

ilter

sof

t pac

k/U

SA

48.3

611

.64

1.04

24.8

00.

34M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

/Nor

way

56.6

314

.04

0.76

24.3

80.

37L

& M

Kin

g S

ize

Filt

er h

ard

pack

/Eur

opea

n U

nion

80.4

817

.98

1.35

28.1

40.

44M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

/Mal

aysi

a71

.04

18.4

61.

2427

.24

0.44

Che

ster

field

Orig

inal

s K

ing

Siz

e F

ilter

har

d pa

ck61

.44

13.0

41.

1322

.94

0.39

L &

M K

ing

Siz

e F

ilter

har

d pa

ck/M

alay

sia

59.2

114

.33

0.90

21.3

80.

42M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

25

s/A

ustr

alia

46.0

911

.01

0.78

24.2

00.

33M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

/Tai

wan

50.6

215

.02

1.40

17.3

60.

39M

arlb

oro

100

Filt

er h

ard

pack

/Eur

opea

n U

nion

61.9

812

.11

1.28

22.5

60.

36M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

Med

ium

/Eur

opea

n U

nion

50.5

710

.21

0.68

19.9

50.

29P

arlia

men

t 100

Filt

er s

oft p

ack

Ligh

t/US

A49

.63

11.5

40.

9815

.64

0.35

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/J

apan

45.9

413

.98

0.75

13.3

20.

35L

& M

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/E

urop

ean

Uni

on79

.87

12.4

01.

3927

.33

0.33

Filt

er6

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/E

urop

ean

Uni

on79

.33

8.66

1.49

17.3

70.

27C

hest

erfie

ld O

rigin

als

Kin

g S

ize

Filt

er h

ard

pack

59.7

68.

200.

7920

.72

0.29

Dia

na K

ing

Siz

e F

ilter

sof

t pac

k S

peci

ally

Mild

/E63

.35

15.7

70.

7122

.78

0.38

Mur

atti

Am

bass

ador

Kin

g S

ize

Filt

er h

ard

pack

/Eur

opea

n U

nion

83.3

318

.28

1.06

16.6

10.

42M

erit

Kin

g S

ize

Filt

er h

ard

pack

/Eur

opea

n U

nion

53.3

39.

120.

9023

.20

0.28

183

Par

liam

ent 1

00 F

ilter

sof

t pac

k/C

EM

A58

.34

15.1

61.

0314

.47

0.35

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck L

ight

/Ger

man

y/U

nite

d K

ingd

om65

.70

7.85

1.60

21.1

40.

25M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/J

apan

58.1

19.

091.

2314

.97

0.32

Mar

lbor

o 10

0 F

ilter

har

d pa

ck L

ight

/Ger

man

y60

.19

10.3

91.

1820

.34

0.32

Mer

it K

ing

Siz

e F

ilter

sof

t pac

k U

ltra-

light

/US

A45

.06

9.30

1.12

25.8

20.

28M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

Ultr

a-lig

ht M

enth

ol/U

SA

48.2

97.

660.

8424

.18

0.23

Par

liam

ent K

ing

Siz

e F

ilter

har

d pa

ck L

ight

/Jap

an45

.98

10.0

00.

7417

.09

0.28

Virg

inia

Slim

s 10

0 F

ilter

har

d pa

ck U

ltra-

light

Men

thol

38.5

17.

820.

7019

.47

0.28

Che

ster

field

INT

L K

ing

Siz

e F

ilter

har

d pa

ck U

ltra-

light

/62

.29

9.43

1.25

22.9

30.

28P

hilip

Mor

ris 1

00 F

ilter

har

d pa

ck S

uper

L49

.08

10.7

61.

2320

.16

0.30

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck U

ltra-

light

/Eur

opea

n U

nion

49.5

96.

940.

9021

.56

0.25

Dia

na K

ing

Siz

e F

ilter

har

d pa

ck U

ltra-

light

/Eur

opea

n U

nion

59.5

77.

590.

9422

.20

0.26

Virg

inia

Slim

s 10

0 F

ilter

har

d pa

ck U

ltra-

light

Men

thol

44.9

65.

901.

1524

.24

0.24

Phi

lip M

orris

One

Kin

g S

ize

Filt

er h

ard

pack

/Eur

opea

n U

nion

45.7

88.

021.

1423

.71

0.22

Mur

atti

Kin

g S

ize

Filt

er h

ard

pack

Ultr

a-lig

ht 1

mg/

CE

MA

54.2

110

.65

1.23

24.7

70.

27Lo

ngbe

ach

One

Kin

g S

ize

Filt

er h

ard

pack

/Aus

tral

ia51

.60

7.73

1.11

20.0

00.

24V

irgin

ia S

lims

100

Filt

er h

ard

pack

Men

thol

139

.46

6.23

1.02

21.9

20.

25M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

/Mex

ico

43.6

012

.19

0.71

18.8

70.

36R

affle

s 10

0 F

ilter

har

d pa

ck/E

urop

ean

Uni

on63

.07

19.7

80.

9417

.22

0.39

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/B

razi

l62

.68

14.2

40.

9922

.32

0.39

Pet

er J

acks

on K

ing

Siz

e F

ilter

har

d pa

ck M

enth

ol/A

ustr

alia

57.1

39.

550.

9215

.62

0.25

Mar

lbor

o 10

0 F

ilter

har

d pa

ck L

ight

/US

A44

.05

8.56

0.78

18.7

90.

28M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/N

orw

ay49

.12

9.34

0.87

21.3

70.

32C

hest

erfie

ld K

ing

Siz

e F

ilter

har

d pa

ck L

ight

/Eur

opea

n U

nion

59.8

76.

950.

8620

.06

0.28

Phi

lip M

orris

Kin

g S

ize

Filt

er h

ard

pack

Sup

er L

ight

46.7

18.

030.

8720

.81

0.28

Mer

it K

ing

Siz

e F

ilter

sof

t pac

k U

ltra-

light

/US

A35

.37

12.6

10.

9925

.37

0.40

184

1R4F

ilter

Ken

tuck

y R

efer

ence

57.8

17.

540.

7222

.95

0.29

1R4F

ilter

Ken

tuck

y R

efer

ence

52.2

88.

240.

8820

.00

0.30

** M

ean

**56

.55

11.3

61.

0121

.40

0.33

** S

tand

ard

devi

atio

n **

11.9

03.

720.

243.

520.

06**

Coe

ffici

ent o

f var

iatio

n **

0.21

0.33

0.23

0.16

0.20

Max

imum

85.0

019

.78

1.60

28.1

40.

4675

th p

erce

ntile

62.2

114

.02

1.18

24.0

80.

3790

th p

erce

ntile

76.0

917

.08

1.35

25.4

20.

42M

edia

n55

.42

10.5

20.

9821

.74

0.32

Min

imum

35.3

75.

900.

6613

.32

0.22

185

Bra

nd

NN

NN

NK

NA

TN

AB

Mer

cury

Cad

miu

mL

ead

Mar

lbor

o Lo

ng S

ize

F h

ard

pack

/Arg

entin

a10

3.30

79.1

010

4.81

15.1

92.

6927

.17

19.6

7M

arlb

oro

Long

Siz

e F

ilter

har

d pa

ck/V

enez

uela

118.

3473

.27

85.7

316

.08

2.86

47.4

925

.18

SG

Ven

til R

egul

ar F

ilter

sof

t pac

k/E

urop

ean

Uni

on80

.74

49.3

271

.15

9.86

3.65

41.8

222

.30

Pet

ra R

egul

ar F

ilter

har

d pa

ck/C

EM

A61

.89

43.5

157

.73

7.46

2.92

48.0

521

.19

Mar

lbor

o K

ing

Siz

e F

ilter

sof

t pac

k/U

SA

151.

3311

1.11

130.

6217

.11

2.93

63.6

026

.31

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/N

orw

ay18

9.04

101.

9715

1.11

15.8

72.

5050

.48

18.6

1L

& M

Kin

g S

ize

Filt

er h

ard

pack

/Eur

opea

n U

nion

92.7

157

.02

78.7

810

.80

3.51

53.5

624

.52

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/M

alay

sia

185.

7980

.54

140.

5920

.90

2.90

70.7

232

.08

Che

ster

field

Orig

inal

s K

ing

Siz

e F

ilter

har

d pa

ck75

.00

54.8

569

.18

8.87

2.89

44.6

423

.71

L &

M K

ing

Siz

e F

ilter

har

d pa

ck/M

alay

sia

100.

0852

.33

76.4

211

.63

2.50

41.0

426

.04

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck 2

5 s/

Aus

tral

ia13

3.78

97.5

212

2.31

15.5

02.

8657

.18

21.3

0M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

/Tai

wan

169.

7497

.53

143.

3917

.53

2.51

37.5

324

.32

Mar

lbor

o 10

0 F

ilter

har

d pa

ck/E

urop

ean

Uni

on11

0.26

83.0

895

.02

11.3

22.

9150

.88

25.3

3M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

Med

ium

/Eur

opea

n U

nion

83.8

770

.88

76.6

010

.41

3.09

47.8

919

.38

Par

liam

ent 1

00 F

ilter

sof

t pac

k Li

ght/U

SA

162.

4910

9.13

143.

2020

.79

2.57

31.8

323

.98

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/J

apan

125.

2771

.02

121.

4514

.06

2.85

45.0

022

.66

L &

M K

ing

Siz

e F

ilter

har

d pa

ck L

ight

/Eur

opea

n U

nion

94.6

059

.13

83.0

013

.87

4.13

55.8

725

.13

Filt

er6

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/E

urop

ean

Uni

on17

.09

24.9

724

.30

2.96

3.07

35.8

118

.66

Che

ster

field

Orig

inal

s K

ing

Siz

e F

ilter

har

d pa

ck86

.29

60.5

477

.54

10.2

42.

5141

.20

24.8

5D

iana

Kin

g S

ize

Filt

er s

oft p

ack

Spe

cial

ly M

ild/E

51.1

941

.13

49.6

97.

633.

3542

.89

27.1

6M

urat

ti A

mba

ssad

or K

ing

Siz

e F

ilter

har

d pa

ck/E

urop

ean

Uni

on54

.72

52.5

657

.39

6.61

3.06

24.4

420

.67

Mer

it K

ing

Siz

e F

ilter

har

d pa

ck/E

urop

ean

Uni

on12

8.44

63.4

010

8.91

14.4

23.

8861

.09

23.8

1

186

Par

liam

ent 1

00 F

ilter

sof

t pac

k/C

EM

A16

5.48

76.9

613

0.83

19.7

22.

7628

.66

22.0

3M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/G

erm

any/

Uni

ted

Kin

gdom

63.5

450

.32

57.8

56.

843.

4843

.67

16.2

7M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/J

apan

116.

9282

.66

101.

4711

.26

3.43

38.4

632

.87

Mar

lbor

o 10

0 F

ilter

har

d pa

ck L

ight

/Ger

man

y65

.68

51.8

060

.39

7.33

3.11

36.7

015

.49

Mer

it K

ing

Siz

e F

ilter

sof

t pac

k U

ltra-

light

/US

A17

5.82

99.6

815

2.09

20.9

53.

9264

.62

23.7

3M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

Ultr

a-lig

ht M

enth

ol/U

SA

157.

8594

.75

143.

9920

.13

4.18

66.9

625

.13

Par

liam

ent K

ing

Siz

e F

ilter

har

d pa

ck L

ight

/Jap

an14

9.55

94.3

013

5.47

23.4

63.

1329

.89

22.4

0V

irgin

ia S

lims

100

Filt

er h

ard

pack

Ultr

a-lig

ht M

enth

ol14

5.96

91.0

112

9.41

16.8

13.

0960

.11

24.8

9C

hest

erfie

ld IN

TL

Kin

g S

ize

Filt

er h

ard

pack

Ultr

a-lig

ht/

75.7

138

.71

65.8

610

.57

3.86

50.1

422

.93

Phi

lip M

orris

100

Filt

er h

ard

pack

Sup

er L

117.

8479

.30

107.

2412

.65

3.57

47.8

916

.11

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck U

ltra-

light

/Eur

opea

n U

nion

74.3

551

.90

82.1

18.

983.

6738

.23

17.4

8D

iana

Kin

g S

ize

Filt

er h

ard

pack

Ultr

a-lig

ht/E

urop

ean

Uni

on56

.17

42.2

757

.52

8.37

4.11

42.9

826

.31

Virg

inia

Slim

s 10

0 F

ilter

har

d pa

ck U

ltra-

light

Men

thol

120.

7294

.24

106.

1912

.88

4.68

60.9

433

.17

Phi

lip M

orris

One

Kin

g S

ize

Filt

er h

ard

pack

/Eur

opea

n U

nion

161.

2962

.24

147.

8421

.38

5.09

47.0

722

.16

Mur

atti

Kin

g S

ize

Filt

er h

ard

pack

Ultr

a-lig

ht 1

mg/

CE

MA

182.

9985

.98

161.

5018

.79

5.51

40.6

524

.02

Long

beac

h O

ne K

ing

Siz

e F

ilter

har

d pa

ck/A

ustr

alia

17.3

132

.86

41.4

33.

613.

7038

.32

21.6

0V

irgin

ia S

lims

100

Filt

er h

ard

pack

Men

thol

117

4.38

77.5

416

4.62

19.3

14.

9246

.46

19.7

7M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

/Mex

ico

89.5

142

.83

75.7

110

.24

3.24

79.8

018

.66

Raf

fles

100

Filt

er h

ard

pack

/Eur

opea

n U

nion

17.2

223

.33

29.0

74.

372.

5924

.93

14.1

1M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

/Bra

zil

60.4

548

.71

48.6

26.

962.

5046

.12

19.7

8P

eter

Jac

kson

Kin

g S

ize

Filt

er h

ard

pack

Men

thol

/Aus

tral

ia16

.29

28.2

033

.54

3.15

2.75

36.1

814

.44

Mar

lbor

o 10

0 F

ilter

har

d pa

ck L

ight

/US

A13

2.79

98.0

012

1.67

18.5

13.

5360

.37

24.1

4M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/N

orw

ay17

4.01

108.

1914

0.60

17.3

63.

0855

.38

21.8

7C

hest

erfie

ld K

ing

Siz

e F

ilter

har

d pa

ck L

ight

/Eur

opea

n U

nion

61.1

035

.58

57.8

67.

923.

4448

.64

23.5

1P

hilip

Mor

ris K

ing

Siz

e F

ilter

har

d pa

ck S

uper

Lig

ht10

5.09

79.0

210

2.20

11.1

03.

1843

.35

20.2

3M

erit

Kin

g S

ize

Filt

er s

oft p

ack

Ultr

a-lig

ht/U

SA

185.

8286

.34

182.

9123

.21

3.96

47.3

119

.18

187

1R4F

ilter

Ken

tuck

y R

efer

ence

126.

2810

9.23

136.

1220

.98

5.52

87.4

948

.63

1R4F

ilter

Ken

tuck

y R

efer

ence

133.

1110

3.01

142.

8523

.94

5.54

78.1

348

.29

** M

ean

**10

9.98

70.0

699

.72

13.4

03.

4348

.19

23.5

2**

Sta

ndar

d de

viat

ion

**49

.38

25.1

040

.64

5.80

0.82

13.8

26.

58**

Coe

ffici

ent o

f var

iatio

n **

0.45

0.36

0.41

0.43

0.24

0.29

0.28

Max

imum

189.

0411

1.11

182.

9123

.94

5.54

87.4

948

.63

75th

per

cent

ile15

0.89

93.4

413

5.96

18.2

73.

8255

.75

25.0

790

th p

erce

ntile

174.

5310

2.07

148.

1720

.95

4.70

64.8

527

.66

Med

ian

113.

5972

.14

101.

8312

.76

3.15

46.7

722

.79

Min

imum

16.2

923

.33

24.3

02.

962.

5024

.44

14.1

1

188

Bra

nd

Ch

rom

ium

Nic

kel

Ars

enic

Sel

eniu

m

Mar

lbor

o Lo

ng S

ize

F h

ard

pack

/Arg

entin

a4.

06M

arlb

oro

Long

Siz

e F

ilter

har

d pa

ck/V

enez

uela

SG

Ven

til R

egul

ar F

ilter

sof

t pac

k/E

urop

ean

Uni

onP

etra

Reg

ular

Filt

er h

ard

pack

/CE

MA

4.92

Mar

lbor

o K

ing

Siz

e F

ilter

sof

t pac

k/U

SA

3.78

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/N

orw

ay3.

70L

& M

Kin

g S

ize

Filt

er h

ard

pack

/Eur

opea

n U

nion

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/M

alay

sia

5.38

Che

ster

field

Orig

inal

s K

ing

Siz

e F

ilter

har

d pa

ckL

& M

Kin

g S

ize

Filt

er h

ard

pack

/Mal

aysi

a4.

71M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

25

s/A

ustr

alia

4.41

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/T

aiw

an4.

45M

arlb

oro

100

Filt

er h

ard

pack

/Eur

opea

n U

nion

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck M

ediu

m/E

urop

ean

Uni

onP

arlia

men

t 100

Filt

er s

oft p

ack

Ligh

t/US

A3.

86M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

/Jap

anL

& M

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/E

urop

ean

Uni

onF

ilter

6 K

ing

Siz

e F

ilter

har

d pa

ck L

ight

/Eur

opea

n U

nion

Che

ster

field

Orig

inal

s K

ing

Siz

e F

ilter

har

d pa

ckD

iana

Kin

g S

ize

Filt

er s

oft p

ack

Spe

cial

ly M

ild/E

4.38

Mur

atti

Am

bass

ador

Kin

g S

ize

Filt

er h

ard

pack

/Eur

opea

n U

nion

4.89

Mer

it K

ing

Siz

e F

ilter

har

d pa

ck/E

urop

ean

Uni

on

189

Par

liam

ent 1

00 F

ilter

sof

t pac

k/C

EM

A4.

70M

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/G

erm

any/

Uni

ted

Kin

gdom

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck L

ight

/Jap

anM

arlb

oro

100

Filt

er h

ard

pack

Lig

ht/G

erm

any

Mer

it K

ing

Siz

e F

ilter

sof

t pac

k U

ltra-

light

/US

AM

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

Ultr

a-lig

ht M

enth

ol/U

SA

4.75

Par

liam

ent K

ing

Siz

e F

ilter

har

d pa

ck L

ight

/Jap

an4.

25V

irgin

ia S

lims

100

Filt

er h

ard

pack

Ultr

a-lig

ht M

enth

ol5.

16C

hest

erfie

ld IN

TL

Kin

g S

ize

Filt

er h

ard

pack

Ultr

a-lig

ht/

Phi

lip M

orris

100

Filt

er h

ard

pack

Sup

er L

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck U

ltra-

light

/Eur

opea

n U

nion

Dia

na K

ing

Siz

e F

ilter

har

d pa

ck U

ltra-

light

/Eur

opea

n U

nion

Virg

inia

Slim

s 10

0 F

ilter

har

d pa

ck U

ltra-

light

Men

thol

Phi

lip M

orris

One

Kin

g S

ize

Filt

er h

ard

pack

/Eur

opea

n U

nion

Mur

atti

Kin

g S

ize

Filt

er h

ard

pack

Ultr

a-lig

ht 1

mg/

CE

MA

Long

beac

h O

ne K

ing

Siz

e F

ilter

har

d pa

ck/A

ustr

alia

Virg

inia

Slim

s 10

0 F

ilter

har

d pa

ck M

enth

ol 1

Mar

lbor

o K

ing

Siz

e F

ilter

har

d pa

ck/M

exic

o5.

87R

affle

s 10

0 F

ilter

har

d pa

ck/E

urop

ean

Uni

onM

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

/Bra

zil

Pet

er J

acks

on K

ing

Siz

e F

ilter

har

d pa

ck M

enth

ol/A

ustr

alia

Mar

lbor

o 10

0 F

ilter

har

d pa

ck L

ight

/US

AM

arlb

oro

Kin

g S

ize

Filt

er h

ard

pack

Lig

ht/N

orw

ayC

hest

erfie

ld K

ing

Siz

e F

ilter

har

d pa

ck L

ight

/Eur

opea

n U

nion

Phi

lip M

orris

Kin

g S

ize

Filt

er h

ard

pack

Sup

er L

ight

Mer

it K

ing

Siz

e F

ilter

sof

t pac

k U

ltra-

light

/US

A

190

1R4F

ilter

Ken

tuck

y R

efer

ence

6.50

1R4F

ilter

Ken

tuck

y R

efer

ence

6.37

** M

ean

**4.

79**

Sta

ndar

d de

viat

ion

**0.

82**

Coe

ffici

ent o

f var

iatio

n **

0.17

Max

imum

0.00

0.00

6.50

0.00

75th

per

cent

ile#N

UM

!#N

UM

!5.

10#N

UM

!90

th p

erce

ntile

#NU

M!

#NU

M!

6.02

#NU

M!

Med

ian

#NU

M!

#NU

M!

4.70

#NU

M!

Min

imum

0.00

0.00

3.70

0.00

NN

K, 4

-(N

-nitr

osom

ethy

lam

ino)

-1-(

3-py

ridyl

)-1-

buta

none

; NN

N, N

´-ni

tros

onor

nico

tine;

NA

T, N

´-ni

tros

oana

tabi

ne; N

AB

, N´-

nitr

osoa

naba

sine

; CE

MA

,ce

ntra

l Eur

ope,

Mid

dle

Eas

t and

Afr

ica

191

Tab

le A

1.3.

Co

nst

itu

ents

per

mill

igra

m o

f n

ico

tin

e fo

r A

ust

ralia

n b

ran

ds,

mo

dif

ied

inte

nse

mac

hin

e re

gim

en

Bra

nd

Nic

oti

ne

Ace

tald

ehyd

eA

ceto

ne

Acr

ole

inA

cryl

on

itri

leA

mm

on

ia

Long

beac

h M

ild2.

5042

9.32

189.

3245

.88

6.80

10.2

4Lo

ngbe

ach

Sup

er M

ild2.

0049

4.15

223.

9052

.70

7.90

10.2

0Lo

ngbe

ach

Ultr

a M

ild1.

6057

4.50

258.

6361

.94

8.56

9.63

Pet

er J

acks

on E

xtra

Mild

2.40

546.

9624

2.25

58.8

36.

929.

21P

eter

Jac

kson

Sup

er M

ild2.

0058

6.30

268.

8563

.55

8.50

8.90

Pet

er J

acks

on U

ltra

Mild

1.60

677.

5629

5.50

71.4

410

.63

10.3

1H

oriz

on M

ild2.

1061

2.38

265.

2467

.62

9.91

10.9

5H

oriz

on S

uper

Mild

2.20

562.

2724

5.00

62.7

39.

4611

.64

Hor

izon

Ultr

a M

ild1.

5468

3.12

292.

2172

.73

11.0

412

.40

Ben

son

& H

edge

s E

xtra

Mild

81.

9756

4.47

277.

6761

.42

10.1

011

.32

Ben

son

& H

edge

s S

peci

al F

ilter

Kin

g S

ize

hard

pac

k2.

4946

6.67

232.

9350

.20

8.80

11.4

5H

olid

ay 8

Sup

er M

ild h

ard

pack

1.48

508.

1126

0.14

53.1

88.

1813

.11

Win

field

Ext

ra M

ild 2

5 ha

rd p

ack

2.06

512.

6226

0.19

54.8

58.

4511

.60

Win

field

Filt

er K

ing

Siz

e ha

rd p

ack

2.27

486.

7824

4.93

51.1

08.

1510

.93

Win

field

Sup

er M

ild K

ing

Siz

e ha

rd p

ack

2.09

538.

2827

6.08

56.9

47.

9910

.86

Mea

n2.

0254

9.57

255.

5259

.01

8.76

10.8

5S

tand

ard

devi

atio

n0.

3471

.75

27.3

57.

901.

251.

15C

oeffi

cien

t of v

aria

tion

0.17

0.13

0.11

0.13

0.14

0.11

Max

imum

2.50

683.

1229

5.50

72.7

311

.04

13.1

175

th p

erce

ntile

2.24

580.

4027

2.46

63.1

49.

6811

.52

90th

per

cent

ile2.

4565

1.49

286.

3969

.91

10.4

212

.10

Med

ian

2.06

546.

9626

0.14

58.8

38.

5010

.93

192

Bra

nd

Ben

zen

eB

enzo

[a]p

yren

eB

uty

rald

ehyd

eC

adm

ium

Car

bo

nm

on

oxi

de

Cat

ech

ol

Cro

ton

ald

ehyd

e

Long

beac

h M

ild24

.20

8.40

23.2

425

.16

8.72

51.0

816

.12

Long

beac

h S

uper

Mild

30.3

09.

1024

.75

27.9

510

.10

51.9

017

.80

Long

beac

h U

ltra

Mild

34.1

910

.00

31.1

934

.38

11.9

445

.56

20.0

0P

eter

Jac

kson

Ext

ra M

ild30

.75

10.5

428

.79

30.1

710

.13

51.2

521

.08

Pet

er J

acks

on S

uper

Mild

35.8

010

.05

32.2

033

.65

10.9

544

.60

21.8

5P

eter

Jac

kson

Ultr

a M

ild45

.31

9.00

35.8

129

.56

13.5

644

.56

23.0

6H

oriz

on M

ild34

.86

9.24

33.3

339

.52

11.0

550

.48

23.0

5H

oriz

on S

uper

Mild

33.3

69.

4631

.64

44.8

610

.86

47.7

320

.55

Hor

izon

Ultr

a M

ild39

.55

12.2

738

.77

39.5

512

.40

48.7

023

.51

Ben

son

& H

edge

s E

xtra

Mild

837

.41

6.90

38.2

743

.86

10.9

652

.79

20.7

1B

enso

n &

Hed

ges

Spe

cial

Filt

er K

ing

Siz

eha

rd p

ack

32.8

17.

3932

.65

44.5

89.

3654

.62

17.2

3

Hol

iday

8 S

uper

Mild

har

d pa

ck35

.95

8.24

32.9

142

.91

11.0

851

.69

19.3

9W

infie

ld E

xtra

Mild

25

hard

pac

k35

.49

7.38

35.4

939

.13

10.4

451

.46

19.6

6W

infie

ld F

ilter

Kin

g S

ize

hard

pac

k31

.45

7.49

33.6

136

.61

9.69

54.6

316

.96

Win

field

Sup

er M

ild K

ing

Siz

e ha

rd p

ack

33.8

87.

3737

.70

35.2

610

.57

50.2

421

.15

Mea

n34

.35

8.86

32.6

936

.48

10.7

950

.09

20.1

4S

tand

ard

devi

atio

n4.

681.

484.

506.

321.

213.

242.

31C

oeffi

cien

t of v

aria

tion

0.14

0.17

0.14

0.17

0.11

0.06

0.11

Max

imum

45.3

112

.27

38.7

744

.86

13.5

654

.63

23.5

175

th p

erce

ntile

35.8

79.

7335

.65

41.2

311

.06

51.7

921

.50

90th

per

cent

ile38

.69

10.3

538

.05

44.2

912

.22

53.8

923

.06

Med

ian

34.1

99.

0032

.91

36.6

110

.86

51.0

820

.55

193

Bra

nd

Fo

rmal

deh

yde

Hyd

rog

en c

yan

ide

Hyd

roq

uin

on

eIs

op

ren

eL

ead

Mer

cury

Met

hyl

eth

yl k

eto

ne

Long

beac

h M

ild54

.60

88.4

462

.56

293.

08.

1.84

45.9

6Lo

ngbe

ach

Sup

er M

ild63

.95

101.

5068

.90

328.

45.

2.10

51.8

5Lo

ngbe

ach

Ultr

a M

ild62

.31

112.

5661

.94

378.

00.

2.63

58.8

1P

eter

Jac

kson

Ext

ra M

ild71

.17

94.4

267

.83

290.

3811

.04

1.92

57.2

9P

eter

Jac

kson

Sup

er M

ild70

.80

103.

9556

.35

363.

50.

2.35

65.6

5P

eter

Jac

kson

Ultr

a M

ild64

.81

142.

0660

.81

438.

44.

2.75

69.7

5H

oriz

on M

ild75

.24

123.

8160

.48

357.

14.

2.61

62.8

6H

oriz

on S

uper

Mild

58.6

412

1.82

55.9

133

3.18

13.5

52.

4673

.64

Hor

izon

Ultr

a M

ild55

.97

142.

8660

.07

411.

04.

3.25

67.5

3B

enso

n &

Hed

ges

Ext

ra M

ild 8

55.3

312

7.92

57.8

741

1.68

.3.

1277

.16

Ben

son

& H

edge

s S

peci

al F

ilter

Kin

gS

ize

hard

pac

k49

.80

111.

6560

.24

359.

8411

.93

2.54

67.4

7

Hol

iday

8 S

uper

Mild

har

d pa

ck58

.65

138.

5162

.70

343.

92.

2.96

71.6

2W

infie

ld E

xtra

Mild

25

hard

pac

k55

.83

116.

5159

.71

396.

12.

2.71

74.7

6W

infie

ld F

ilter

Kin

g S

ize

hard

pac

k49

.78

112.

7864

.32

355.

5112

.12

2.41

70.4

9W

infie

ld S

uper

Mild

Kin

g S

ize

hard

pac

k53

.59

118.

1860

.77

382.

78.

2.57

80.3

8

Mea

n60

.03

117.

1361

.36

362.

8712

.16

2.55

66.3

5S

tand

ard

devi

atio

n7.

8216

.37

3.64

42.0

81.

040.

409.

54C

oeffi

cien

t of v

aria

tion

0.13

0.14

0.06

0.12

0.09

0.16

0.14

Max

imum

75.2

414

2.86

68.9

043

8.44

13.5

53.

2580

.38

75th

per

cent

ile64

.38

125.

8662

.63

389.

4512

.47

2.73

72.6

390

th p

erce

ntile

71.0

214

0.64

66.4

341

1.42

13.1

23.

0676

.20

Med

ian

58.6

411

6.51

60.7

735

9.84

12.0

22.

5767

.53

194

Bra

nd

NN

NN

NK

NA

TN

AB

Nit

rog

en o

xid

esN

itri

c o

xid

eP

hen

ol

Long

beac

h M

ild19

.12

22.8

44.

0433

.08

68.8

061

.68

13.5

6Lo

ngbe

ach

Sup

er M

ild16

.00

27.1

55.

2527

.95

74.5

068

.00

10.2

0Lo

ngbe

ach

Ultr

a M

ild15

.31

24.4

44.

0636

.94

77.1

369

.63

7.31

Pet

er J

acks

on E

xtra

Mild

14.0

818

.83

4.96

27.6

376

.00

68.9

212

.58

Pet

er J

acks

on S

uper

Mild

10.9

520

.30

5.10

26.3

086

.30

77.7

58.

70P

eter

Jac

kson

Ultr

a M

ild20

.06

26.2

57.

6935

.88

105.

1395

.06

6.25

Hor

izon

Mild

15.0

019

.86

4.66

31.5

710

1.43

91.4

311

.38

Hor

izon

Sup

er M

ild18

.59

24.8

67.

6435

.86

88.6

479

.09

11.1

4H

oriz

on U

ltra

Mild

19.5

523

.31

4.64

31.6

210

3.25

92.8

68.

70B

enso

n &

Hed

ges

Ext

ra M

ild 8

23.9

132

.64

6.50

41.7

385

.28

77.6

711

.83

Ben

son

& H

edge

s S

peci

al F

ilter

Kin

g S

ize

hard

pac

k25

.02

37.6

36.

1045

.78

87.5

580

.32

14.1

8H

olid

ay 8

Sup

er M

ild h

ard

pack

33.2

425

.61

7.77

49.7

393

.24

85.8

113

.31

Win

field

Ext

ra M

ild 2

5 ha

rd p

ack

28.5

938

.20

8.69

48.5

489

.32

82.0

413

.50

Win

field

Filt

er K

ing

Siz

e ha

rd p

ack

25.3

333

.70

6.83

44.9

373

.57

67.4

014

.67

Win

field

Sup

er M

ild K

ing

Siz

e ha

rd p

ack

27.1

334

.59

8.13

48.3

372

.25

66.0

313

.16

Mea

n20

.79

27.3

56.

1437

.72

85.4

977

.58

11.3

6S

tand

ard

devi

atio

n6.

226.

431.

588.

1911

.71

10.5

22.

61C

oeffi

cien

t of v

aria

tion

0.30

0.24

0.26

0.22

0.14

0.14

0.23

Max

imum

33.2

438

.20

8.69

49.7

310

5.13

95.0

614

.67

75th

per

cent

ile25

.18

33.1

77.

6645

.36

91.2

883

.93

13.4

090

th p

erce

ntile

28.0

136

.42

7.99

48.4

610

2.52

92.2

913

.93

Med

ian

19.5

525

.61

6.10

35.8

886

.30

77.7

511

.83

195

Bra

nd

Pro

pio

nal

deh

yde

Pyr

idin

eQ

uin

olin

eR

eso

rcin

ol

Sty

ren

eT

olu

ene

1,3-

Bu

tad

ien

e

Long

beac

h M

ild34

.88

12.5

60.

281.

007.

1239

.16

33.7

2Lo

ngbe

ach

Sup

er M

ild40

.90

11.4

50.

201.

207.

7545

.65

39.4

0Lo

ngbe

ach

Ultr

a M

ild46

.06

13.5

60.

191.

449.

4451

.69

45.5

6P

eter

Jac

kson

Ext

ra M

ild46

.17

13.5

80.

250.

928.

5445

.42

38.0

4P

eter

Jac

kson

Sup

er M

ild47

.20

12.9

00.

200.

909.

9054

.85

44.5

5P

eter

Jac

kson

Ultr

a M

ild57

.63

13.5

60.

191.

3110

.50

65.2

553

.63

Hor

izon

Mild

50.0

015

.33

0.29

1.13

10.3

853

.81

44.1

9H

oriz

on S

uper

Mild

47.7

314

.46

0.28

0.86

9.68

52.7

343

.50

Hor

izon

Ultr

a M

ild58

.12

16.1

70.

24.

12.2

160

.00

51.4

3B

enso

n &

Hed

ges

Ext

ra M

ild 8

47.8

215

.43

0.29

1.22

11.4

759

.90

52.2

8B

enso

n &

Hed

ges

Spe

cial

Filt

er K

ing

Siz

e ha

rd p

ack

40.1

615

.34

0.32

1.57

10.1

252

.21

43.7

8H

olid

ay 8

Sup

er M

ild h

ard

pack

42.3

015

.54

0.29

1.78

10.7

457

.23

49.5

3W

infie

ld E

xtra

Mild

25

hard

pac

k43

.93

17.6

70.

311.

5611

.31

56.3

147

.33

Win

field

Filt

er K

ing

Siz

e ha

rd p

ack

41.4

516

.74

0.33

1.68

10.6

648

.90

42.5

1W

infie

ld S

uper

Mild

Kin

g S

ize

hard

pac

k46

.51

16.8

40.

301.

3410

.81

55.0

247

.85

Mea

n46

.06

14.7

40.

261.

2810

.04

53.2

145

.15

Sta

ndar

d de

viat

ion

6.14

1.77

0.05

0.30

1.38

6.58

5.48

Coe

ffici

ent o

f var

iatio

n0.

130.

120.

190.

230.

140.

120.

12M

axim

um58

.12

17.6

70.

331.

7812

.21

65.2

553

.63

75th

per

cent

ile47

.77

15.8

60.

301.

5310

.78

56.7

748

.69

90th

per

cent

ile54

.58

16.8

00.

321.

6511

.41

59.9

651

.94

Med

ian

46.1

715

.33

0.28

1.27

10.3

853

.81

44.5

5

196

Bra

nd

1-A

min

on

aph

thal

ene

2-A

min

on

aph

thal

ene

3-A

min

ob

iph

enyl

Long

beac

h M

ild8.

805.

361.

16Lo

ngbe

ach

Sup

er M

ild7.

554.

601.

15Lo

ngbe

ach

Ultr

a M

ild8.

135.

191.

31P

eter

Jac

kson

Ext

ra M

ild7.

504.

631.

08P

eter

Jac

kson

Sup

er M

ild7.

054.

351.

15P

eter

Jac

kson

Ultr

a M

ild7.

755.

131.

31H

oriz

on M

ild9.

625.

811.

50H

oriz

on S

uper

Mild

9.23

5.55

1.46

Hor

izon

Ultr

a M

ild10

.00

7.14

1.65

Ben

son

& H

edge

s E

xtra

Mild

810

.86

6.90

1.71

Ben

son

& H

edge

s S

peci

al F

ilter

Kin

g S

ize

hard

pac

k10

.36

6.43

1.65

Hol

iday

8 S

uper

Mild

har

d pa

ck11

.42

7.30

2.03

Win

field

Ext

ra M

ild 2

5 ha

rd p

ack

10.7

36.

941.

73W

infie

ld F

ilter

Kin

g S

ize

hard

pac

k11

.67

7.23

1.82

Win

field

Sup

er M

ild K

ing

Siz

e ha

rd p

ack

10.1

46.

561.

72M

ean

9.39

5.94

1.50

Sta

ndar

d de

viat

ion

1.52

1.04

0.29

Coe

ffici

ent o

f var

iatio

n0.

160.

180.

19M

axim

um11

.67

7.30

2.03

75th

per

cent

ile10

.54

6.92

1.71

90th

per

cent

ile11

.20

7.19

1.78

Med

ian

9.62

5.81

1.50

197

Bra

nd

4-A

min

ob

iph

enyl

m-

and

p-C

reso

lo

-Cre

sol

Long

beac

h M

ild0.

888.

003.

12Lo

ngbe

ach

Sup

er M

ild0.

906.

302.

40Lo

ngbe

ach

Ultr

a M

ild1.

065.

191.

88P

eter

Jac

kson

Ext

ra M

ild0.

837.

923.

08P

eter

Jac

kson

Sup

er M

ild0.

906.

252.

45P

eter

Jac

kson

Ultr

a M

ild1.

004.

561.

69H

oriz

on M

ild1.

147.

102.

67H

oriz

on S

uper

Mild

1.06

6.64

2.54

Hor

izon

Ultr

a M

ild1.

306.

242.

31B

enso

n &

Hed

ges

Ext

ra M

ild 8

1.39

7.11

2.75

Ben

son

& H

edge

s S

peci

al F

ilter

Kin

g S

ize

hard

pac

k1.

269.

243.

28H

olid

ay 8

Sup

er M

ild h

ard

pack

1.63

8.99

3.54

Win

field

Ext

ra M

ild 2

5 ha

rd p

ack

1.33

8.35

3.03

Win

field

Filt

er K

ing

Siz

e ha

rd p

ack

1.39

9.34

3.54

Win

field

Sup

er M

ild K

ing

Siz

e ha

rd p

ack

1.33

7.94

3.23

Mea

n1.

167.

282.

77S

tand

ard

devi

atio

n0.

241.

430.

56C

oeffi

cien

t of v

aria

tion

0.20

0.20

0.20

Max

imum

1.63

9.34

3.54

75th

per

cent

ile1.

338.

183.

1890

th p

erce

ntile

1.39

9.14

3.43

Med

ian

1.14

7.11

2.75

NN

K, 4

-(N

-nitr

osom

ethy

lam

ino)

-1-(

3-py

ridyl

)-1-

buta

none

; NN

N, N

´-ni

tros

onor

nico

tine;

NA

T, N

´-ni

tros

oana

tabi

ne; N

AB

, N´-

nitr

osoa

naba

sine

198

Annex 3.2 Basis for calculation of toxicant animal carcinogenicityindex and toxicant non-cancer response index

The T25 value is the long-term daily dose that will cause tumours in 25% ofthe animals above the background level at a specific tissue site. The T25 isdetermined by linear extrapolation from the lowest dose that gives a statisti-cally significant increase in tumours (referred to herein as the critical dose)(Dybing et al., 1997). T25 values are derived from the following equations:

C = [(B/100 A/100)/(1 A/100)] × 100

T25 = (25/C) × critical dose

where A is the proportion (%) of animals with the tumour in the control group,B is the proportion (%) of animals with the tumour in an exposed group andC is the net increase in tumour frequency (%).

The default values used to convert feed and drinking-water concentrationsinto doses per kilogram bodyweight are given below. Most of the T25 valueswere derived from oral feeding studies. Concentrations used in inhalationbioassays were converted to oral doses by using the default values for in-halation volumes reported in Table A2.1.

Table A2.1Default values for dose calculation, weight and intake by diet, water and inhalation,for a standard experimental period of 2 years

Experimentalanimal

Sex Weight(g) Food/day(g) Water/day(ml) Inhalationvolume(l/h)

Mouse Male 30 3.60 5 1.8Female 25 3.25 5 1.8

Rat Male 500 20.00 25 6.0Female 350 17.50 20 6.0

Hamster Male 125 11.50 15 3.6Female 110 11.50 15 3.6

From Gold et al. (1984)

Acetaldehyde

Groups of 105 male and 105 female 6-week-old Wistar rats were exposed byinhalation to acetaldehyde at a concentration of 0 ppm (control), 750 ppm,1500 ppm or 3000 ppm, 6 h/day, 5 days/week for 28 months. Control mea-surements showed that the mean concentrations of acetaldehyde in the lowand intermediate doses were 735 and 1412 ppm, respectively, while that in

199

the high dose was 3033 ppm from day 0 to day 141, 2167 ppm from day 142to day 210, 2039 ppm from day 211 to day 238, 1433 ppm from day 239 today 300, 1695 ppm from day 301 to day 312, 1472 ppm from day 313 to day359, and 977 ppm for the rest of the study. The reason for the gradual loweringof the high dose was its considerable toxicity, causing growth retardation,respiratory problems and death. All the animals at the high dose died after100 weeks. The incidences of squamous epithelial carcinomas in the nasalcavity were: control males 1/49 (2%), low-dose males 1/52 (2%), interme-diate-dose males 5/53 (9%) and high-dose males 15/49 (31%); controlfemales 0/50 (0%), low-dose females 0/48 (0%), intermediate-dose females5/5 (9%) and high-dose females 17/53 (32%). The incidences of adenocar-cinomas in the nasal cavity were: control males 0/59 (0%), low-dose males16/52 (31%), intermediate-dose males 31/53 (59%) and high-dose males21/49 (43%); control females 0/50 (0%), low-dose females 6/48 (13%),intermediate-dose females 26/53 (49%) and high-dose females 21/53 (40%)(Woutersen et al., 1986).

Remarks on the study

Species, strain, sex: Rat, Wistar, maleRoute: InhalationCritical end-point: Nasal adenocarcinomasDuration: 104 weeks (default)Critical dose: 735 ppm = 735 × 44.1/24.45 (conversion factor) mg/

m3 = 1326 mg/m3; inhalation volume per day 6 l/h ×6 h for 5 days/week = 25.7 l/day; inhalation dose perday for specified body weight 450 g = 1350 mg/m3 ×25.7 l/day × 1000/450 kg = 75.8 mg/kg bw per day;daily dose for 24 months with treatment and observa-tion time of 28 months: 75.8 mg/kg bw per day ×28/24 × 28/24 = 102.1 mg/kg bw per day

Lowest dose causing significantly increased tumour incidence

Control: 0%Dose 102.1 mg/kg bw per day: 31%Net increase: 31%

T25 calculationT25 = 25/31 × 82.1 mg/kg bw per day = 82.4 mg/kg bw per day

Acrylonitrile

A study was conducted by Biodynamics Inc. (1980) in which acrylonitrilewas administered in drinking-water to groups of 100 Fischer 344 rats of eachsex at doses of 1, 3, 10, 30 (stated to correspond to 2.5 mg/kg bw per day)

200

and 100 ppm, with a control group of 200 rats of each sex for 2 years. Interimnecropsies were performed at 6, 12 and 18 months (10 per sex per exposedgroup and 20 per sex in the control group). The study was terminated earlybecause of the low survival rate. An increased incidence of tumours (astro-cytomas of the brain and spinal cord and carcinomas of the Zymbal gland)was seen in the groups at 3 ppm (3/200 in controls, 10/99 in 30 ppm group)or higher, and the incidence was dose-dependent. An increased incidence ofmammary gland tumours was seen in females at 100 ppm.

Remarks on the study

Species, strain, sex: Rat, Fischer 344, maleRoute: Drinking-waterCritical end-point: Brain and spinal cord astrocytomaDuration: 104 weeksCritical dose: 2.5 mg/kg bw per day

Lowest dose causing significantly increased tumour incidence

Control: 3/200 (0.2%)Dose 0.04 mg/kg bw per day: 10/99 (10%)Net increase: [(10/99 3/200) × 100]/(1 3/200) = 9%

T25 calculationT25 = 25/9 × 2.5 mg/kg bw per day = 6.9 mg/kg bw per day

1-Aminonaphthalene

Solutions of 0.01% 1-naphthylamine hydrochloride (freed from 2-naphthy-lamine by multiple fractional recrystallization) were added to the drinking-water of 61 male and female stock mice for 84 weeks. In males, the incidenceof hepatomas was 4/18 (22%) in treated animals and 4/24 (17%) in the con-trols. In females, the corresponding incidences were 5/43 (11%) and 0/36(0%) (Clayson, Ashton, 1963).

Remarks on the study

Species, strain, sex: Mouse, stock (strain not specified), femaleRoute: Drinking-waterCritical end-point: HepatomaDuration: 84 weeksCritical dose: 0.01% = 0.1 mg/ml, female mice weighing 25 g,

drinking 5 ml per day: 0.1 × 5 × 1000/25 mg/kg bwper day = 20 mg/kg bw per day

201

Lowest dose causing significantly increased tumour incidence

Control: 0%Dose 20 mg/kg bw per day: 11%Net increase: 11%

T25 calculationT25 = 25/11 × 20 mg/kg bw per day × 84/104 × 84/104 = 29.7 mg/kg bwper day

2-Aminonaphthalene

In groups of 23 DBA and 25 IF mice given 2-naphthylamine by stomach tubeat a dose of 240 or 400 mg/kg bw per week in arachis oil, respectively, for90 weeks, liver tumours developed in 50% of the animals. In two similargroups of control mice given arachis oil alone, no hepatomas were found(Bonser et al., 1952).

Remarks on the study

Species, strain, sex: Mice, DBA, sex not specifiedRoute: GavageCritical end-point: Liver tumoursDuration: 90 weeksCritical dose: 240 mg/kg bw once a week, 240/7 mg/kg bw per day

= 34.3 mg/kg bw per day dose with administration andobservation for 104 weeks = 34.3 × 90/104 × 90/104mg/kg bw per day = 25.7 mg/kg bw per day

Lowest dose causing significantly increased tumour incidence

Control: 0%Dose 25.7 mg/kg bw per day: 50%Net increase: 50%

T25 calculationT25 = 25/50 × 25.7 mg/kg bw per day = 12.8 mg/kg bw per day

Benzene

Benzene in corn oil was administered to groups of 50 B6C3F1 mice of eachsex by gavage on 5 days per week for 103 weeks. Both sexes were exposedto a dose of 0, 25, 50 or 100 mg/kg bw per day. Benzene mainly causedsignificantly increased incidences of malignant lymphomas in both sexes atdoses 25 mg/kg bw per day, Zymbal gland carcinomas in males at doses 50 mg/kg bw per day and females at doses 100 mg/kg bw per day, lungalveolar and bronchiolar adenomas and carcinomas in males at 100 mg/kg

202

bw per day and females at 50 mg/kg bw per day, Harderian gland adenomasin males at 25 mg/kg bw per day, preputial gland squamous cell carcinomasin males at incidences of 0/21, 5/28, 19/29 (66%) and 31/35 at 0, 25, 50 and100 mg/kg bw per day, respectively, and mammary gland carcinomas in fe-males at 50 mg/kg bw per day (National Toxicology Program, 1986).

Remarks on the study

Species, strain, sex: Mouse, B6C3F1, maleRoute: GavageCritical end-point: Preputial gland, carcinoma (all types)Duration: 104 weeks, dosing for 103 weeksCritical dose: 50 mg/kg bw per day, 5 days/week; daily dose 50 mg/

kg bw × 5/7 = 35.7 mg/kg bw per day; daily dose for104 weeks = 35.7 mg/kg bw × 103/104 = 35.4 mg/kgbw per day

Lowest dose causing significantly increased tumour incidence

Control: 0/21 (0%)Dose 35.4 mg/kg bw per day: 19/29 (66%)Net increase: 66%

T25 calculationT25 = 25/66 × 35.4 mg/kg bw per day = 13.4 mg/kg bw per day

Benzo[a]pyrene

Groups of 24 male Syrian golden hamsters were exposed by inhalation to aircontaining benzo[a]pyrene at 0, 2.2, 9.5 and 46.5 mg/m3, corresponding tototal average doses of 0, 29, 127 and 383 mg benzo[a]pyrene per animal. Theaverage survival time was 96.4 weeks, with the exception of animals at thehigh dose (59 weeks). Of the animals at the intermediate dose, 34.6% devel-oped respiratory tract tumours and 26.9% developed tumours of the upperdigestive tract. None of the controls or low-dose animals developed suchtumours (Thyssen et al., 1981).

203

Remarks on the study

Species, strain, sex: Hamster, Syrian golden, maleRoute: InhalationCritical end-point: Resipiratory tract tumoursDuration: 96.4 weeks average survival time for control, low and

intermediate doseCritical dose: 127 mg total dose in 657 days (96.4 weeks) = 0.19 mg

per animal per day; default body weight of male ham-sters 125 g; daily dose = 0.19 mg × 1000/125 = 1.5mg/kg bw per day

Lowest dose causing significantly increased tumour incidence

Control: 0/24 (0%)Dose 1.5 mg/kg bw per day: 9/26 (34.6%)Net increase: 34.6%

T25 calculationT25 = 25/34.6 × 1.5 mg/kg bw per day = 1.1 mg/kg bw per day

1,3-Butadiene

Groups of 70 male and 70 female B6C3F1 mice aged 8–9 weeks were exposedto air containing 0, 625 or 1250 ppm 1,3-butadiene, 6 h/day, 5 times per weekfor 2 years (Melnick et al., 1990). The results with the two lowest concen-trations used are shown in Table A2.2.

Table A2.2Tumours induced by aerosol exposure to 1,3-butadiene in B6C3F1 mice

Tumour type 0 ppm 6.25 ppm 20 ppm

Males Females Males Females Males Females

Alveolar or bronchiolar adenomaor carcinoma

22/70(31%)

4/70(6%)

23/60(38%)

15/60*(25%)

20/60(33%)

19/60*(32%)

Lymphoma, all malignant 4/70(6%)

10/70(14%)

3/70(4%)

14/70(20%)

8/70(11%)

18/49*(26%)

Haemangiosarcoma, heart 0/70(0%)

0/70(0%)

0/70(0%)

0/70(0%)

1/70(1%)

0/70(0%)

Papilloma or carcinoma of the fore-stomach

1/70(1%)

2/70(3%)

0/700%)

2/70(3%)

1/70(1%)

3/70(4%)

* p < 0.01, Fisher exact test

204

Remarks on the study

Species, strain, sex: Mouse, B6C3F1, femaleRoute: InhalationCritical end-point: Alveolar or bronchiolar adenoma or carcinomaDuration: 104 weeksCritical dose: 6.25 ppm = (6.25 × 2.21)/1000 mg/m3 = 13.8 μg/m3;

inhalation volume per day 2.5 l/h × 6 h for 5 days perweek = 15 l × 5/7 = 10.7 l/day; inhalation dose per dayfor specified body weight 39 g = 13.8 μg/m3 × 10.7 l/day × 1000/39kg = 3.8 mg/kg bw per day

Lowest dose causing significantly increased tumour incidence

Control: 4/7 (6%)Dose 3.8 mg/kg bw per day: 15/60 (25%)Net increase: [(25/100 6/100)/(1 6/100)] × 100 = 20%

T25 calculationT25 = 25/20 × 3.8 mg/kg bw per day = 4.8 mg/kg bw per day

Cadmium

Four groups of 40 male SPF Wistar (TNO/W75) rats, 6 weeks old, wereexposed to 12.5, 25 or 50 g/m3 cadmium as cadmium chloride aerosol (massmedian aerodynamic diameter, 0.55 μm; geometric standard deviation, 1.8)for 23 h/day on 7 days/week for 18 months. Animals were observed for anadditional 13 months, at which time the experiment was ended. A group of41 rats exposed to filtered air served as controls. Histological examinationwas performed on all animals. Body weights and survival were not affectedby cadmium treatment. A dose-related increase in the incidence of malignantpulmonary tumours (mostly adenocarcinomas) was observed in cadmiumchloride-treated rats (12.5 g/m3, 6/39 [15%]; 25 g/m3, 20/38 [53%]; 50

g/m3, 25/35 [71%]) compared with controls (0/38). Multiple pulmonary tu-mours were observed frequently; several tumours showed metastases or wereregionally invasive. The incidence of adenomatous hyperplasia was also in-creased by cadmium treatment (Takenaka et al., 1983).

205

Remarks on the study

Species, strain, sex: Rats, SPF Wistar, maleRoute: InhalationDuration: 23 h/day, 7 days/week for 18 months, observation for

additional 13 months before termination of experimentCritical dose: 25 μg/m3; inhalation volume per 23 h = 0.06 m3 ×

23 = 0.138 m3; male rats weigh 500 g; daily dose:0.138 × 25 μg/0.5 kg = 6.9 μg/kg

Lowest dose causing significantly increased tumour incidence

Control: 0%Dose 6.9 μg/kg bw per day: 53%Net increase: 53%

T25 calculationT25 = 25/53 × 6.9 × 18/24 × 31/24 = 3 g/kg bw per day

Catechol

Groups of 20 or 30 male Wistar, WKY, Lewis and Sprague-Dawley rats(6 weeks of age) were given catechol in the diet at 0 or 0.8% for 104 weeks.The incidence of hyperplasia in the forestomach was significantly increasedin exposed WKY and Sprague-Dawley rats compared with controls. Papil-lomas occurred in 6/30 Sprague-Dawley rats, 2/30 Wistar rats and 1/30 WKYrats and carcinomas in 1/30 Sprague-Dawley and 1/30 Wistar rats, with nonein controls. All strains developed 97–100% incidence of adenomas in theglandular stomach, with none in controls, and adenocarcinomas in theglandular stomach occurred in 23/30 Sprague-Dawley, 22/30 Lewis, 20/30Wistar and 3/30 Wistar rats, with none in controls (Tanaka et al., 1995).

Remarks on the study

Species, strain, sex: Rats, Sprague-Dawley, maleRoute: DietCritical end-point: Glandular stomach adenocarcinomasDuration: 104 weeksCritical dose: 0.8% in diet = 8000 mg/kg diet, male rats eat 20 g per

day and weigh 500 g, daily dose: (8000 mg/kg × 0.02kg)/0.5 kg = 320 mg/kg bw per day

206

Lowest dose causing significantly increased tumour incidence

Control: 0/20 (0%)Dose 320 mg/kg bw per day: 23/30 (77%)Net increase: 77%

T25 calculationT25 = 25/77 × 320 mg/kg bw per day = 104 mg/kg bw per day

Formaldehyde

Groups of 119–120 male and 120 female Fischer 344 rats, 7 weeks of age,were exposed to 0, 2.0, 5.6 or 14.3 ppm (0, 2.5, 6.9 or 17.6 mg/m3) formalde-hyde (> 97.5% pure) vapour by whole-body exposure for 6 h/day on 5 days/week for up to 24 months and were then observed for 6 months with no furtherexposure. While no nasal cavity malignancies were found in rats exposed to0 or 2.0 ppm formaldehyde, two squamous-cell carcinomas (one among 119males and one among 116 females examined) occurred in the group exposedto 5.6 ppm and 107 (51 among 117 males and 52 among 115 females exam-ined) in those exposed to 14.3 ppm (p < 0.001). Five additional nasal cavitytumours (classified as carcinoma, undifferentiated carcinoma or sarcoma andcarcinosarcoma) were identified in rats exposed to 14.3 ppm; two of thesetumours were found in rats that also had squamous-cell carcinomas of thenasal cavity. There was a significant overall increase in the incidence ofpolypoid adenomas in treated animals (males and females combined) whencompared with controls (p = 0.02, Fisher exact test). The incidences of poly-poid adenomas were marginally significantly elevated in females at the lowdose and in males at the intermediate dose (Swenberg et al., 1980; Kernset al., 1983).

Because of the highly nonlinear dose–response relation for formaldehyde-induced nasal tumours, no T25 value was calculated for this compound.

Hydroquinone

Groups of 65 male and 65 female Fischer 344 rats (7–9 weeks of age) weregiven hydroquinone by gavage at 0, 25 or 50 mg/kg bw, 5 days/week for 103weeks. The animals were sacrificed when they were 111–113 weeks old. Inthe males, renal adenomas were found in 0/55 (0%) controls, 4/55 (p = 0.069)(7%) at the low dose and 8/55 (p = 0.003) (14.5%) at the high dose. In thefemales mononuclear leukaemia was found in 9/55 (16%) controls, 15/55(p = 0.048) (27%) at the low dose and 22/55 (p = 0.003) (40%) at the highdose (National Toxicology Program, 1989).

207

Remarks on the study

Species, strain, sex: Rats, Fischer 344, maleRoute: GavageCritical end-point: Renal tubular adenomaDuration: 103 weeksCritical dose: 50 mg/kg bw per day and 5 days per week, daily dose

for 104 weeks with observation for 112 weeks: 50 mg/kg bw per day × 5/7 × 103/104 = 35.4 mg/kg bwper day

Lowest dose causing significantly increased tumour incidence

Control: 0/55 (0%)Dose 35.4 mg/kg bw per day: 8/55 (14.5%)Net increase: 14.5%

T25 calculationT25 = 25/14.5 × 35.4 mg/kg bw per day = 61.0 mg/kg bw per day

Isoprene

Groups of 50 male and 50 female B6C3F1 mice (age unspecified) were ex-posed to isoprene by whole- body inhalation at a concentration of 0, 10, 70,140, 280, 700 or 2200 ppm (0, 28, 200, 400, 800, 2000 or 6160 mg/m3) for 4or 8 h/day on 5 days/week for 20, 40 or 80 weeks, followed by holding periodsuntil termination of the experiment at 96 or 104 weeks. Increases in tumourincidence were found in the lung, liver, heart, spleen and Harderian gland inmales at an exposure of 140 ppm for 40 weeks. In female mice, increasedincidences of Harderian gland adenomas (2/49 controls, 3/49 at 10 ppm and8/49 (p < 0.005) at 70 ppm) and pituitary adenomas (1/49 controls, 6/46 at10 ppm and 9/49 (p < 0.05) at 70 ppm) were seen after exposure for 80 weeks(Cox, Bird, Griffis, 1996; Placke et al., 1996).

Remarks on the study

Species, strain, sex: Mice, B6C3F1, femaleRoute: InhalationCritical end-point: Pituitary adenomasDuration: 80 weeksCritical dose: 200 mg/m3; inhalation volume per 8 h/day = 0.0018

m3/h 8 h 5/7 = 0.0103 m3; female mice weigh 25 g;daily dose: (0.103 × 200 mg/0.025 kg = 82.3 mg/kgbw per day; daily dose for 104 weeks = 82.3 mg/kgbw per day 80/104 = 63.3 mg/kg bw per day

208

Lowest dose causing significantly increased tumour incidence

Control: 1/49 (2%)Dose 63.3 mg/kg bw per day: 9/49 (18%)Net increase: [(18/100 2/100)/(1 2/100)] 100 = 16.7%

T25 calculationT25 = 25/16.7 × 63.3 mg/kg bw per day = 94.8 mg/kg bw per day

Lead

Basic lead acetate has been shown to induce kidney tumours in several ex-periments with rats. In one experiment, Wistar rats received feed containing0.1% and 1% basic lead acetate for 29 and 24 months, respectively. Twocontrol groups consisted of 24–30 rats. In the low-dose group, 11/32 (34%)developed kidney tumours (three carcinomas), and in the high dose 13/24(54%) developed kidney tumours (six carcinomas). No kidney tumours werefound among the control animals (van Esch, van Genderen, Vink, 1962).

Remarks on the study

Species, strain, sex: Rat, Wistar, sex not specifiedRoute: FeedCritical end-point: Renal tumoursDuration: 29 monthsCritical dose: 1000 mg/kg diet, default feed intake 18.8 g/day (mean

of male and female), default body weight 425 g (meanof male and female), daily dose = 1000 mg/kg 0.0188kg × 1000/425 = 44.2 mg/kg bw per day; daily dosefor 2 years = 44.2 mg/kg bw per day × 29/24 = 53.4mg/kg bw per day

Lowest dose causing significantly increased tumour incidence

Control: No renal tumours found (number of controlanimals not reported)

Dose 53.4 mg/kg bw per day: 11/32 (34%)Net increase: 34%

T25 calculationT25 = 25/34 × 53.4 mg/kg bw per day = 39.3 mg/kg bw per day

NNK

Male Fischer rats were treated with NNK dissolved in trioctanoin at 0.03, 0.1,0.3, 1.0, 10 or 50 mg/kg bw or trioactanoin alone subcutaneously three timesper week for 20 weeks. Animals were killed over the course of 104 weeks

209

after cessation of treatment. The following lung tumour incidences werefound: 2.5% (control), 6.7% (0.03 mg/kg bw), 10.0% (0.1 mg/kg bw), 13.3%(0.3 mg/kg bw), 53.3% (1.0 mg/kg bw), 73.3% (10 mg/kg bw), 73.3%(10 mg/kg bw) and 87.1% (50 mg/kg bw). Tumour incidence increased withdose, and the trend was strongly significant. The respective incidences ofalveolar hyperplasia were 2.5%, 16.4%, 16.0%, 40.0%, 73.3%, 93.3% and93.5% (Belinsky et al., 1990).

Remarks on the study

Species, strain, sex: Rats, Fischer, maleRoute: SubcutaneousCritical end-point: Lung adenomas and carcinomasDuration: 20 weeksCritical dose: 0.03 mg/kg bw 3 times per week, daily dose: 0.03 mg/

kg bw × 3/7 = 00.013 mg/kg bw per day, daily dosefor 2 years: 12.9 g/kg bw per day × 20/104 = 2.5 g/kg bw per day

Lowest dose causing significantly increased tumour incidence

Control: 1/40 (2.5%)Dose 2.5 g/kg bw per day: 4/60 (6.7%)%Net increase: [(6.7/100 2.5/100)/(1 2.5/100)] × 100 = 4.3

T25 calculationT25 = 25/4.3 × 2.5 g/kg bw per day = 0.015 mg/kg bw per day

NNN

A group of 20 male Fischer rats, 7 weeks old, was given 200 mg/l NNN inthe drinking-water on 5 days/week for 30 weeks (estimated total, 630 mg).Animals were killed when moribund or after 11 months, and all animals,except those lost to cannibalism or autolysis, were necropsied. All 12 rats inthe treated groups that were necropsied had developed oesophageal tumours(11 papillomas and three carcinomas); in addition, one pharyngeal papillomaand three carcinomas of the nasal cavity with invasion of the brain wereobserved. No tumours were observed in 19 untreated controls (Hoffmannet al., 1975).

210

Remarks on the study

Species, strain, sex: Rat, Fischer, maleRoute: Drinking-waterCritical end-point: Oesophageal papillomas and carcinomasDuration: 30-week treatment period, necropsied after 11 monthsCritical dose: 630 mg total for 210 days, male rats weigh 500 g =

daily dose 630/210 × 1000/500 kg = 6.0 mg/kg bw perday; daily dose for 24 months = 6.0 mg/kg bw per day× 210/728 × 11/24 = 0.8 mg/kg bw per day

Lowest dose causing significantly increased tumour incidence

Control: 0%Dose 0.4 mg/kg bw per day: 12/12 (100%)Net increase: 100%

T25 calculationT25 = 25/100 × 0.4 mg/kg bw per day = 0.2 mg/kg bw per day

References

Belinsky SA et al. (1990) Dose–response relationships beteen O6-methyl-guanine formation in Clara cells and induction of pulmonary neoplasia in therat by 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone. Cancer Research,50:3772–3780.

Biodynamics, Inc. (1990) A twenty-four month oral toxicity/carcinogenicitystudy of arcrylonitrile administered in the drinking water to Fischer 344rats. Division of Biology and Safety Evaluation, East Millstone, New Jersey,under project No 77-1746 for Monsanto Company, St Louis, Missouri.

Bonser GM et al. (1952) The carcinogenic properties of 2-amino-1-naphtholhydrochloride and its parent amine 2- naphthylamine. British Journal ofCancer, 6:412–424.

Clayson DB, Ashton MJ (1963) The metabolism of 1-naphthylamine and itsbearing on the mode of carcinogenesis of the aromatic amines. Acta UnioInternational Contra Cancrum, 19:539–542.

Cox LA, Bird MG, Griffis L (1996) Isoprene cancer risk and the time patternof dose administration. Toxicology, 113:263–272.

Dybing E et al. (1997) T25: a simplified carcinogenic potency index. De-scription of the system and study of correlations between carcinogenicpotency and species/site specificity and mutagenicity. Pharmacology andToxicology, 80:272–279.

211

van Esch GJ, van Genderen H, Vink HH (1962) The induction of renal tu-mours by feeding of basic lead acetate to rats. British Journal of Cancer,16:289–297.

Gold LS et al. (1984) A carcinogenic potency database of the standardizedresults of animal bioassays. Environmental Health Perspectives, 58:9–319.

Hoffmann D et al. (1975) A study of tobacco carcinogenesis. XIV. Effects ofN´-nitrosonornicotine and N´- nitrosoanabasine in rats. Journal of the Na-tional Cancer Institute, 55:977–981.

Kerns WD et al. (1983) Carcinogenicity of formaldehyde in rats and miceafter long-term inhalation exposure. Cancer Research, 43:4382–4392.

Melnick Rl, Huff J, Chou BJ, Miller RA. Carcinogenicity of 1,3-butadienein C57Bl/6 x C3HF1 mice at low exposure concentrations. Cancer Res 1990;50: 6592-6599.

National Toxicology Program (1986) Toxicology and carcinogenesis studiesof benzene (CAS No. 71-43-2) in F344/N rats and B6C3F1 mice (gavagestudies). Research Triangle Park, North Carolina. (NTP TR 289; NationalInstitutes of Health Publication No. 86-2545).

National Toxicology Program (1989) NTP toxicology and carcinogenesisstudies of hydroquinone (CAS No. 123- 31-9) in F344/N rats and B6C3F1mice (gavage studies). Research Triangle Park, North Carolina. (NTP TR366).

Placke ME et al. (1996) Chronic inhalation oncogenicity study of isoprene inB6C3F1 mice. Toxicology, 110:253262.

Swenberg JA et al. (1980) Induction of squamous cell carcinomas of the ratnasal cavity by inhalation exposure to formaldehyde vapor. Cancer Re-search, 40:3398–3402.

Takenaka S et al. (1983) Carcinogenicity of cadmium chloride aerosols in Wrats. Journal of the National Cancer Institute, 70:367–373.

Tanaka H et al. (1995) Rat strain differences in catechol carcinogenicity tothe stomach. Food and Chemical Toxicology, 33:93–98.

Thyssen J et al. (1981) Inhalation studies with benzo(a)pyrene in Syriangolden hamsters. Journal of the National Cancer Institute, 66:575–577.

Woutersen RA et al. (1986) Inhalation toxicity of acetaldehyde in rats. III.Carcinogenicity study. Toxicology, 41:213–231.

212

Annex 3.3 Calculation of toxicant animal carcinogenicity index andtoxicant non-cancer response index for internationalPhilip Morris brands, Canadian brands and Australianbrands

In this report, the T25 carcinogenic potency method (Dybing et al., 1997) wasused, rather than cancer potency factors published by the US EnvironmentalProtection Agency (www.epa.gov/iris) or the California Environmental Pro-tection Agency (www.oehha.ca.gov) and used in the study of Fowles andDybing (2003). The latter potency factors are derived from the lower 95%confidence limit of a modelled benchmark dose for 10% tumour incidence.The authors examined the usefulness of the T25 carcinogenic potency methodextensively (Dybing et al., 1997; Sanner et al., 2001; Sanner, Dybing,2005a,b). Furthermore, the T25 method has been used in European chemicalproduct regulation (European Commission, 1999; Scientific Committee onCosmetic and Non-food Products, 2003). The choice of method for estimatingcarcinogenic potency does not have much effect on calculations of the toxi-cant animal carcinogenicity index, as the various methods for estimatingpotency factors show good correlation. For instance, Dybing et al. (1997)found a correlation coefficient of 0.96 between the T25 values for 110 USNational Cancer Institute/National Toxicology Program carcinogens and theTD50 carcinogenic potency indices derived with the method of Gold et al.(1984). Interestingly, using the T25 method, Sanner and Dybing (2005b)found good agreement between the hazard characterization based on epi-demiological data and that based on animal experiments, although the dataavailable for such comparisons were limited.

Tables A3.1–A3.3 show the cancer and toxicant non-cancer response indicesper milligram of nicotine of toxicants in smoke generated by the modifiedintense smoking regimen based on data from Counts et al. (2005), the Cana-dian brands and the Australian brands. Tables A3.4–A3.9 show the rankingof toxicants in smoke with respect to toxicant animal carcinogenicity andtoxicant non-cancer response indices for the three data sets. Table A3.10 andA3.11 show comparisons of the three data sets for the two indices.

213

Tab

le A

3.1

Can

cer

and

to

xica

nt

no

n-c

ance

r re

spo

nse

ind

ices

per

mill

igra

m o

f n

ico

tin

e o

f to

xica

nts

in s

mo

ke g

ener

ated

by

the

mo

dif

ied

inte

nse

smo

kin

g r

egim

en b

ased

on

dat

a fr

om

Co

un

ts e

t al

. (20

05)

on

sm

oke

to

xica

nt

leve

ls

To

xica

nt

Lev

el in

sm

oke

(μ

g/m

gn

ico

tin

e)T

25a

(mg

/kg

bw

per

day

)

Po

ten

cyp

er m

g/

kg b

wp

er d

ay

To

xica

nt

anim

alca

rcin

og

enic

ity

ind

exT

ole

rab

lele

vel (

μg

/m

)3

To

xica

nt

no

n-c

ance

rre

spo

nse

ind

ex

Mea

n90

%ile

Max

imu

mM

ean

90%

ileM

axim

um

Mea

n90

%ile

Max

imu

m

Ace

tald

ehyd

e69

585

999

782

.40.

017.

08.

610

.09

77.2

95.4

111

Ace

tone

359

446

501

ND

––

––

Non

e–

––

Acr

olei

n67

.685

.399

.5I

––

––

0.06

1127

1422

1658

Acr

ylon

itrile

12.3

16.1

19.5

6.9

0.14

1.7

2.3

2.7

52.

53.

23.

91-

Am

inon

apht

hale

ne16

.2b

19.0

b24

.8b

29.7

0.03

0.00

049

0.00

057

0.00

074

Non

e–

––

2-A

min

onap

htha

lene

10.1

b11

.6b

14.3

b12

.80.

080.

0008

10.

0009

30.

0011

Non

e–

––

3-A

min

obip

heny

l2.

9b3.

4b4.

1bN

D–

––

–N

one

––

–4-

Am

inob

iphe

nyl

2.2b

2.7b

3.2b

ND

c–

––

–N

one

––

–A

mm

onia

21.2

26.8

40.7

ND

––

––

200

0.11

0.13

0.20

Ars

enic

4.8b

6.0b

6.5b

NQ

––

––

0.03

0.16

0.20

0.22

Ben

zene

39.0

45.8

51.1

13.4

0.07

2.7

3.2

3.6

600.

660.

760.

85B

enzo

[a]p

yren

e9.

0b11

.2b

13.8

b1.

1d0.

910.

0082

0.01

020.

0126

Non

e–

––

1,3-

But

adie

ne54

.165

.575

.54.

80.

2111

.413

.815

.920

2.7

3.3

3.8

But

yral

dehy

de43

.052

.463

.6N

D–

––

–N

one

––

–C

adm

ium

48.2

b64

.9b

87.5

b0.

0333

1.6

2.1

2.2

0.02

2.4

3.2

4.4

Car

bon

mon

oxid

e15

.2e

17.9

e27

.3e

ND

––

––

10 0

001.

51.

82.

7

214

Cat

echo

l48

.861

.665

.410

40.

010.

490.

620.

65N

one

––

–C

hrom

ium

NQ

NQ

NQ

NQ

––

––

0.2f

––

–m

– an

d p-

Cre

sol

7.8

10.9

13.7

ND

––

––

600g

0.01

0.02

0.02

o-C

reso

l3.

04.

35.

0N

D–

––

–60

0g0.

010.

010.

01C

roto

nald

ehyd

e28

.836

.641

.3I

––

––

Non

e–

––

For

mal

dehy

de41

.157

.990

.5N

Qh

––

––

313

.719

.330

.2H

ydro

gen

cyan

ide

204

277

390

ND

––

––

922

.730

.843

.3H

ydro

quin

one

56.6

76.1

85.0

61.0

0.02

1.1

1.5

1.7

Non

e–

––

Isop

rene

459

551

746

94.8

d0.

014.

65.

57.

5N

one

––

–Le

ad23

.5b

27.7

b48

.6b

39.3

i0.

030.

000.

000.

00N

one

––

–M

ercu

ry3.

4b4.

7b5.

5bL

––

––

0.09

0.04

0.05

0.06

Met

hyl e

thyl

ket

one

93.2

116

124

ND

––

––

1000

0.09

0.12

0.12

N´-

Nitr

osoa

naba

sine

13.4

b21

.0b

23.9

bL

––

––

Non

e–

––

N´-

Nitr

osoa

nata

bine

99.7

b14

8b18

3bI

––

––

Non

e–

––

Nic

kel

NQ

NQ

NQ

NQ

––

––

0.05

––

–N

itric

oxi

de18

028

034

9N

D–

––

–N

one

––

–N

itrog

en o

xide

s19

931

339

0N

D–

––

–40

j5.

07.

89.

8N

NK

70.1

b10

2b11

1b0.

015k

674.

76.

87.

4N

one

––

–N

NN

110b

175b

189b

0.2

5.0

0.55

0.88

0.95

Non

e–

––

Phe

nol

11.4

17.1

19.8

I–

––

–20

00.

060.

090.

10P

ropi

onal

dehy

de60

.374

.088

.4N

D–

––

–N

one

––

–P

yrid

ine

21.4

25.4

28.1

L–

––

–N

one

––

–Q

uino

line

0.33

0.42

0.46

ND

––

––

Non

e–

––

Res

orci

nol

1.0

1.4

1.6

I–

––

–N

one

––

–

215

Sel

eniu

mN

QN

QN

QN

D–

––

–20

––

–S

tyre

ne13

.616

.518

.5L

––

––

900

0.02

0.02

0.02

Tol

uene

71.1

83.3

96.3

E–

––

–30

00.

240.

280.

32

ND

, no

data

; I, i

nsuf

ficie

nt e

vide

nce

of c

arci

noge

nici

ty; N

Q, n

ot q

uant

ifiab

le; L

, lim

ited

evid

ence

of c

arci

noge

nici

ty; N

NN

, N´-

nitr

oson

orni

cotin

e; N

NK

, 4-(

N-

nitr

osom

ethy

lam

ino)

-1-(

3-py

ridyl

)-1-

buta

none

; E, e

vide

nce

of n

onca

rcin

ogen

icity

a V

alue

s ar

e fr

om s

tudi

es b

y or

al a

dmin

istr

atio

n, u

nles

s ot

herw

ise

stat

ed.

b ng

/mg

nico

tine

c No

prop

er d

ata

for

T25

cal

cula

tion

d In

hala

tion

adm

inis

trat

ion

e m

g/m

g ni

cotin

ef H

exav

alen

t chr

omiu

mg

Cre

sol m

ixtu

reh

Hig

hly

non-

linea

r do

se-r

espo

nse

rela

tion

I Lea

d su

bace

tate

j WH

O g

uide

line

for

nitr

ogen

dio

xide

, not

list

ed b

y th

e C

alifr

onia

Env

ironm

enta

l Pro

tect

ion

Age

ncy

in 2

005

k Sub

cuta

neou

s ad

min

istr

atio

n

216

Tab

le A

3.2

Can

cer

and

to

xica

nt

no

n-c

ance

r re

spo

nse

ind

ices

per

mill

igra

m o

f n

ico

tin

e o

f to

xica

nts

in s

mo

ke g

ener

ated

by

the

mo

dif

ied

inte

nse

smo

kin

g r

egim

en b

ased

on

Can

adia

n d

ata

on

sm

oke

to

xica

nt

leve

ls

To

xica

nt

Lev

el in

sm

oke

(μ

g/m

gn

ico

tin

e)T

25a

(mg

/kg

bw

per

day

)

Po

ten

cyp

er m

g/

kg b

wp

er d

ay

To

xica

nt

anim

alca

rcin

og

enic

ity

ind

exT

ole

rab

lele

vel (

μg

/m

3 )

To

xica

nt

no

n-c

ance

rre

spo

nse

ind

ex

Mea

n90

%ile

Max

imu

mM

ean

90%

ileM

axim

um

Mea

n90

%ile

Max

imu

m

Ace

tald

ehyd

e56

765

876

682

0.01

5.7

6.6

7.7

962

.973

.185

.2A

ceto

ne28

932

338

6N

Db

––

––

Non

e–

––

Acr

olei

n71

.381

.799

.5I

––

––

0.06

1188

1362

1659

Acr

ylon

itrile

10.0

12.0

14.0

6.9

0.14

1.4

1.7

2.0

52.

02.

42.

81-

Am

inon

apht

hale

ne10

.8b

15.1

b19

.6b

29.7

0.03

0.00

032

0.00

045

0.00

059

Non

e–

––

2-A

min

onap

htha

lene

9.7

b14

.2b

14.5

b12

.80.

080.

0007

70.

0011

0.00

12N

one

––

–3-

Am

inob

iphe

nyl

1.8

b2.

5 b

2.8b

ND

––

––

Non

e–

––

4-A

min

obip

heny

l1.

8 b

2.4

b2.

6bN

Dc

––

––

Non

e–

––

Am

mon

ia12

.718

.119

.0N

D–

––

–20

00.

060.

080.

09A

rsen

icN

Db

ND

bN

Db

NQ

––

––

0.03

Ben

zene

40.6

48.6

51.8

13.4

0.07

2.8

3.4

3.6

600.

680.

810.

86B

enzo

[a]p

yren

e10

.5b

17.4

b17

.9b

1.1d

0.91

0.00

960.

0158

0.01

63N

one

––

–1,

3-B

utad

iene

42.6

51.6

56.5

4.8

0.21

8.9

0.01

0811

.920

2.1

2.6

2.8

But

yral

dehy

de32

.337

.441

.3N

D–

––

–N

one

––

Cad

miu

m71

.4b

83.6

b93

.2b

0.03

332.

42.

83.

10.

023.

64.

24.

7C

arbo

n m

onox

ide

11.8

e14

.6e

15.7

eN

D–

––

–10

000

1.2

1.5

1.6

217

Cat

echo

l75

.388

.795

.710

40.

010.

750.

890.

96N

one

––

–C

hrom

ium

ND

ND

ND

NQ

––

––

0.2f

––

–m

- an

d p-

Cre

sol

10.7

13.1

27.1

ND

––

––

600g

0.02

0.02

0.05

o-C

reso

l4.

34.

811

.1N

D–

––

–60

0g0.

010.

010.

02C

roto

nald

ehyd

e30

.335

.939

.6I

––

––

Non

e–

––

For

mal

dehy

de77

.311

611

8N

Qh

––

––

325

.838

.739

.4H

ydro

gen

cyan

ide

143

196

230

ND

––

––

915

.921

.825

.6H

ydro

quin

one

66.0

75.6

78.1

61.0

0.02

1.3

1.5

1.6

Non

e–

––

Isop

rene

289

377

438

94.8

d0.

012.

93.

84.

4N

one

––

Lead

18.7

b23

.3b

23.3

b39

.3i

0.03

0.00

0.00

0.00

Non

e–

––

Mer

cury

2.9b

3.3b

4.0b

L–

––

–0.

090.

030.

040.

04M

ethy

l eth

yl k

eton

eN

DN

DN

DN

D–

––

–10

000.

090.

120.

12N

´-N

itros

oana

basi

ne8.

8b37

.6b

43.3

bL

––

––

Non

e–

––

N´-

Nitr

osoa

nata

bine

37.8

b45

.3b

169b

I–

––

–N

one

––

–N

icke

lN

DN

DN

DN

D–

––

–0.

05–

––

Nitr

ic o

xide

81.5

166

284

ND

––

––

Non

e–

––

Nitr

ogen

oxi

des

88.8

182

311

ND

––

––

20j

2.2

4.6

7.8

NN

K56

.1b

88.3

b10

4b0.

015k

673.

85.

97.

0N

one

––

–N

NN

43.5

b14

4b16

3b0.

25.

00.

220.

720.

82N

one

––

–P

heno

l18

.319

.454

.5I

––

––

200

0.09

0.10

0.27

Pro

pion

alde

hyde

48.2

54.9

65.3

ND

––

––

Non

e–

––

Pyr

idin

e16

.520

.022

.1L

––

––

Non

e–

––

Qui

nolin

e0.

350.

430.

80N

D–

––

–N

one

––

–R

esor

cino

l1.

31.

71.

8I

––

––

Non

e–

––

218

Sel

eniu

mN

DN

DN

DN

D–

––

–20

––

–S

tyre

ne12

.312

.961

.8L

––

––

900

0.01

0.01

0.07

Tol

uene

71.6

91.4

93.7

E–

––

–30

00.

240.

300.

31

ND

, no

data

; I, i

nsuf

ficie

nt e

vide

nce

of c

arci

noge

nici

ty; N

Q, n

ot q

uant

ifiab

le; L

, lim

ited

evid

ence

of c

arci

noge

nici

ty; N

NN

, N´-

nitr

oson

orni

cotin

e; N

NK

, 4-(

N-

nitr

osom

ethy

lam

ino)

-1-(

3-py

ridyl

)-1-

buta

none

; E, e

vide

nce

of n

onca

rcin

ogen

icity

a Val

ues

are

from

stu

dies

by

oral

adm

inis

trat

ion,

unl

ess

othe

rwis

e st

ated

.b n

g/m

g ni

cotin

ec N

o pr

oper

dat

a fo

r T

25 c

alcu

latio

nd I

nhal

atio

n ad

min

istr

atio

ne m

g/m

g ni

cotin

ef H

exav

alen

t chr

omiu

mg C

reso

l mix

ture

h Hig

hly

non-

linea

r do

se-r

espo

nse

rela

tion

I Lea

d su

bace

tate

j WH

O g

uide

line

for

nitr

ogen

dio

xide

, not

list

ed b

y th

e C

alifr

onia

Env

ironm

enta

l Pro

tect

ion

Age

ncy

in 2

005

k Sub

cuta

neou

s ad

min

istr

atio

n

219

Tab

le A

3.3

Can

cer

and

to

xica

nt

no

n-c

ance

r re

spo

nse

ind

ices

per

mill

igra

m n

ico

tin

e o

f to

xica

nts

in s

mo

ke g

ener

ated

by

the

mo

dif

ied

inte

nse

smo

kin

g r

egim

en b

ased

on

Au

stra

lian

dat

a o

n s

mo

ke t

oxi

can

t le

vels

To

xica

nt

Lev

el in

sm

oke

(μ

g/m

gn

ico

tin

e)T

25a

(mg

/kg

bw

per

day

)

Po

ten

cyp

er m

g/

kg b

wp

er d

ay

To

xica

nt

anim

alca

rcin

og

enic

ity

ind

exT

ole

rab

lele

vel (

μg

/m

3 )

To

xica

nt

no

n-c

ance

rre

spo

nse

ind

ex

Mea

n90

%ile

Max

imu

mM

ean

90%

ileM

axim

um

Mea

n90

%ile

Max

imu

m

Ace

tald

ehyd

e55

065

168

382

0.01

5.5

6.5

6.8

961

.172

.375

.9A

ceto

ne25

328

629

6N

D–

––

–N

one

––

–A

crol

ein

59.0

69.9

72.7

I–

––

–0.

0698

311

6512

12A

cryl

onitr

ile8.

810

.411

.06.

90.

141.

21.

51.

55

1.8

2.1

2.2

1-A

min

onap

htha

lene

9.4b

11.2

b11

.7b

29.7

0.03

0.00

0028

0.00

034

0.00

035

Non

e–

––

2-A

min

onap

htha

lene

5.9b

7.2b

7.3b

12.8

0.08

0.00

047

0.00

058

0.00

058

Non

e–

––

3-A

min

obip

heny

l1.

5b1.

9b2.

0bN

D–

––

–N

one

––

–4-

Am

inob

iphe

nyl

1.2b

1.4b

1.6b

ND

c–

––

–N

one

––

–A

mm

onia

10.8

12.1

13.1

ND

––

––

200

0.05

0.06

0.07

Ars

enic

ND

bN

Db

ND

bN

Q–

––

–0.

03–

––

Ben

zene

34.4

38.7

45.3

13.4

0.07

2.4

2.7

3.2

600.

570.

650.

76B

enzo

[a]p

yren

e8.

9b10

.3b

12.3

b1.

1d0.

910.

0081

0.00

940.

0112

Non

e–

––

1,3-

But

adie

ne45

.251

.953

.64.

80.

219.

510

.911

.320

2.3

2.6

2.7

But

yral

dehy

de32

.738

.038

.8N

D–

––

–N

one

––

–C

adm

ium

36.5

b44

.3b

44.9

b0.

0333

1.2

1.5

1.5

0.02

1.8

2.2

2.2

Car

bon

mon

oxid

e10

.8e

12.2

e13

.6e

ND

––

––

1000

01.

11.

21.

4

220

Cat

echo

l50

.153

.954

.610

40.

010.

500.

540.

55N

one

––

–C

hrom

ium

ND

ND

ND

NQ

––

––

0.2f

––

–m

- an

d p-

Cre

sol

7.3

9.1

9.3

ND

––

––

600g

0.01

0.02

0.02

o-C

reso

l2.

83.

43.

5N

D–

––

–60

0g0.

000.

010.

01C

roto

nald

ehyd

e20

.123

.123

.5I

––

––

Non

e–

––

For

mal

dehy

de60

.071

.075

.2N

Qh

––

––

320

.023

.725

.1H

ydro

gen

cyan

ide

117

141

143

ND

––

––

913

.015

.715

.9H

ydro

quin

one

61.4

66.4

68.9

61.0

0.02

1.2

1.3

1.4

Non

e–

––

Isop

rene

363

411

438

94.8

d0.

013.

64.

14.

4N

one

––

–Le

ad12

.2b

13.1

b13

.5b

39.3

i0.

030.

000.

000.

00N

one

––

–M

ercu

ry2.

5b3.

1b3.

3bL

––

––

0.09

0.00

0.00

0.00

Met

hyl e

thyl

ket

one

66.3

76.2

80.4

ND

––

––

1000

0.07

0.08

0.08

N´-

Nitr

osoa

naba

sine

6.1b

8.0b

8.7b

L–

––

–N

one

––

–N

´-N

itros

oana

tabi

ne37

.7b

48.5

b49

.7b

I–

––

–N

one

––

–N

icke

lN

DN

DN

DN

Q–

––

–0.

05–

––

Nitr

ic o

xide

77.6

92.3

95.1

ND

––

––

Non

e–

––

Nitr

ogen

oxi

des

85.5

103

105

ND

––

––

20j

2.1

2.6

2.6

NN

K27

.3b

36.4

b38

.2b

0.01

5k67

1.8

2.4

2.6

Non

e–

––

NN

N20

.8b

28.0

b33

.2b

0.2

5.0

0.10

0.14

0.17

Non

e–

––

Phe

nol

11.4

13.9

14.7

I–

––

–20

00.

060.

070.

07P

ropi

onal

dehy

de46

.154

.658

.1N

D–

––

–N

one

––

–P

yrid

ine

14.7

16.8

17.7

L–

––

–N

one

––

–Q

uino

line

0.26

0.32

0.33

ND

––

––

Non

e–

––

Res

orci

nol

1.33

1.6

1.8

I–

––

–N

one

––

–

221

Sel

eniu

mN

DN

DN

DN

D–

––

–20

g–

––

Sty

rene

10.0

11.4

12.2

L–

––

–90

00.

010.

010.

01T

olue

ne53

.260

.065

.3E

––

––

300

0.18

0.20

0.22

ND

, no

data

; I, i

nsuf

ficie

nt e

vide

nce

of c

arci

noge

nici

ty; N

Q, n

ot q

uant

ifiab

le; L

, lim

ited

evid

ence

of c

arci

noge

nici

ty; N

NN

, N´-

nitr

oson

orni

cotin

e; N

NK

, 4-(

N-

nitr

osom

ethy

lam

ino)

-1-(

3-py

ridyl

)-1-

buta

none

; E, e

vide

nce

of n

onca

rcin

ogen

icity

a Val

ues

are

from

stu

dies

by

oral

adm

inis

trat

ion,

unl

ess

othe

rwis

e st

ated

.b n

g/m

g ni

cotin

ec N

o pr

oper

dat

a fo

r T

25 c

alcu

latio

nd I

nhal

atio

n ad

min

istr

atio

ne m

g/m

g ni

cotin

ef H

exav

alen

t chr

omiu

mg C

reso

l mix

ture

h Hig

hly

non-

linea

r do

se-r

espo

nse

rela

tion

I Lea

d su

bace

tate

j WH

O g

uide

line

for

nitr

ogen

dio

xide

, not

list

ed b

y th

e C

alifr

onia

Env

ironm

enta

l Pro

tect

ion

Age

ncy

in 2

005

k Sub

cuta

neou

s ad

min

istr

atio

n

222

Table A3.4Ranking of toxicants in smoke with respect to toxicant animal carcinogenicityindices: data from Counts et al. (2005)

Toxicant Mean 90th percentile Maximum

1,3-Butadiene 11.4 13.8 15.9Acetaldehyde 7.0 8.6 10.0NNK 4.7 6.8 7.4Isoprene 4.6 5.5 7.5Benzene 2.7 3.2 3.6Acrylonitrile 1.7 2.3 2.7Cadmium 1.6 2.1 2.2Hydroquinone 1.1 1.5 1.7NNN 0.55 0.88 0.95Catechol 0.49 0.62 0.65Benzo[a]pyrene 0.0082 0.0102 0.01262-Aminonaphthalene 0.00081 0.00093 0.0011Lead 0.00 0.00 0.001-Aminonaphthalene 0.00049 0.00057 0.00074

NNK, 4-(N-nitrosomethylamino)-1-(3-pyridyl)-1-butanone; NNN, N´-nitrosonornicotine

Table A3.5Ranking of toxicants in smoke with respect to toxicant non-cancer response indices:data from Counts et al. (2005)

Toxicant Mean 90th percentile Maximum

Acrolein 1127 1422 1658Acetaldehyde 77.2 95.4 111Hydrogen cyanide 22.7 30.8 43.3Formaldehyde 13.7 19.3 30.2Nitrogen oxides 5.0 7.8 9.8Cadmium 2.4 3.2 4.41,3-Butadiene 2.7 3.3 3.8Acrylonitrile 2.5 3.3 3.9Carbon monoxide 1.5 1.8 2.7Benzene 0.66 0.76 0.85Toluene 0.24 0.28 0.32Arsenic 0.16 0.20 0.22Ammonia 0.11 0.13 0.20Methyl ethyl ketone 0.09 0.12 0.12Phenol 0.06 0.09 0.10Mercury 0.04 0.05 0.06Styrene 0.02 0.02 0.02m- and p-Cresol 0.01 0.02 0.02o-Cresol 0.01 0.01 0.01

223

Table A3.6Ranking of toxicants in smoke with respect to toxicant animal carcinogenicityindices: Canadian data

Toxicant Mean 90th percentile Maximum

1,3-Butadiene 8.9 10.8 11.9Acetaldehyde 5.7 6.6 7.7NNK 3.8 5.9 7.0Isoprene 2.9 3.8 4.4Benzene 2.8 3.4 3.6Cadmium 2.4 2.8 3.1Acrylonitrile 1.4 1.7 2.0Hydroquinone 1.3 1.5 1.6Catechol 0.75 0.89 0.96NNN 0.22 0.72 0.82Benzo[a]pyrene 0.0096 0.0158 0.01632-Aminonapthalene 0.00077 0.0011 0.0012Lead 0.00 0.00 0.001-Aminonaphthalene 0.00032 0.00045 0.00059

NNK, 4-(N-nitrosomethylamino)-1-(3-pyridyl)-1-butanone; NNN, N´-nitrosonornicotine

Table A3.7Ranking of toxicants in smoke with respect to toxicant non-cancer response indices:Canadian data

Toxicant Mean 90th percentile Maximum

Acrolein 1188 1362 1659Acetaldehyde 62.9 73.1 85.2Formaldehyde 25.8 38.7 39.4Hydrogen cyanide 15.9 21.8 25.6Cadmium 3.6 4.2 4.7Nitrogen oxides 2.2 4.6 4.61,3-Butadiene 2.1 2.6 2.8Acrylonitrile 2.0 2.4 2.8Carbon monoxide 1.2 1.5 1.6Benzene 0.68 0.81 0.86Toluene 0.24 0.30 0.31Phenol 0.09 0.10 0.27Ammonia 0.06 0.08 0.09Styrene 0.01 0.01 0.07m- and p-Cresol 0.02 0.02 0.05Mercury 0.03 0.04 0.04o-Cresol 0.01 0.01 0.02

224

Table A3.8Ranking of toxicants in smoke with respect to toxicant animal carcinogenicityindices: Australian data

Toxicant Mean 90 percentile Max

1,3-Butadiene 9.5 10.9 11.3Acetaldehyde 5.5 6.5 6.8Isoprene 3.6 4.1 4.4Benzene 2.4 2.7 3.2NNK 1.8 2.4 2.6Acrylonitrile 1.2 1.5 1.5Cadmium 1.2 1.5 1.5Hydroquinone 1.2 1.3 1.4Catechol 0.50 0.54 0.55NNN 0.10 0.14 0.17Benzo[a]pyrene 0.0081 0.0094 0.01122-Aminonaphthalene 0.00047 0.00058 0.00058Lead 0.00 0.00 0.001-Aminonaphthalene 0.00028 0.00034 0.00035

NNK, 4-(N-nitrosomethylamino)-1-(3-pyridyl)-1-butanone; NNN, N´-nitrosonornicotine

Table A3.9Ranking of toxicants in smoke with respect to toxicant non-cancer response indices:Australian data

Toxicant Mean 90th percentile Max

Acrolein 983 1165 1212Acetaldehyde 61.1 72.3 75.9Formaldehyde 20.0 23.7 25.1Hydrogen cyanide 13.0 15.7 15.91,3-Butadiene 2.3 2.6 2.7Nitrogen oxides 2.1 2.6 2.6Cadmium 1.8 2.2 2.2Acrylonitrile 1.8 2.1 2.2Carbon monoxide 1.1 1.2 1.4Benzene 0.57 0.65 0.76Toluene 0.18 0.20 0.22Methyl ethyl ketone 0.07 0.08 0.08Phenol 0.06 0.07 0.07Ammonia 0.05 0.06 0.07m- and p-Cresol 0.01 0.02 0.02Styrene 0.01 0.01 0.01o-Cresol 0.00 0.01 0.01Mercury 0.00 0.00 0.00

225

Table A3.10Ranking of toxicants in smoke with respect to toxicant animal carcinogenicity indices

Toxicant Mean

Counts et al. (2005) Canadian data Australian data Mean

1,3-Butadiene 11.4 8.9 9.5 9.9Acetaldehyde 7.0 5.7 5.5 6.1Isoprene 4.6 2.9 3.6 3.7NNK 4.7 3.8 1.8 3.4Benzene 2.7 2.8 2.4 2.6Cadmium 1.6 2.4 1.2 1.7Acrylonitrile 1.7 1.4 1.2 1.4Hydroquinone 1.1 1.3 1.2 1.2Catechol 0.49 0.75 0.50 0.58NNN 0.55 0.22 0.10 0.29Benzo[a]pyrene 0.0082 0.0096 0.0081 0.00862-Aminonaphthalene 0.00081 0.00077 0.00047 0.00068Lead 0.00 0.00 0.00 0.001-Aminonaphthalene 0.00049 0.00032 0.00028 0.00036

NNK, 4-(N-nitrosomethylamino)-1-(3-pyridyl)-1-butanone; NNN, N´-nitrosonornicotine

Table A3.11Ranking of toxicants in smoke with respect to toxicant non-cancer response indices

Toxicant Mean

Counts et al. (2005) Canadian data Australian data Mean

Acrolein 1127 1188 983 1099Acetaldehyde 77.2 62.9 61.1 67.1Formaldehyde 13.7 25.8 20.0 19.8Hydrogen cyanide 22.7 15.9 13.0 17.2Nitrogen oxides 5.0 2.2 2.1 3.1Cadmium 2.4 3.6 1.8 2.61,3-Butadiene 2.7 2.1 2.3 2.4Acrylonitrile 2.5 2.0 1.8 2.1Carbon monoxide 1.5 1.2 1.1 1.3Benzene 0.66 0.68 0.57 0.64Toluene 0.24 0.24 0.18 0.22Arsenic 0.16 – – 0.16Methyl ethyl ketone 0.09 – 0.07 0.08Ammonia 0.11 0.06 0.05 0.07Phenol 0.06 0.09 0.06 0.07Mercury 0.04 0.03 0.00 0.02Styrene 0.02 0.01 – 0.02m- and p-Cresol 0.01 0.02 0.01 0.01o-Cresol 0.01 0.01 0.0 00.01

226

Some of the calculated T25 values are less certain than others, for instance,when data from older experiments that did not adhere to modern bioassaytesting conditions were used (e.g. for 1- and 2-aminonaphthalene). The valuefor NNN is highly uncertain, as only one dose resulted in a 100% tumourincidence. The reported T25 value for NNK is also rather uncertain becausethe value was derived from an experiment by subcutaneous injection threetimes a week for 20 weeks and normalization of this regime to daily dosagefor 104 weeks, assuming dose and time linearity. Mandated limits for NNKand NNN were recommended on the basis of their known toxicity, their vari-ation across brands and information that the levels of these toxicants can bereduced by altering tobacco curing (WHO, 2007). 4-Aminobiphenyl is a hu-man carcinogen (IARC Group 1; IARC, 2006); however, the experimentaldata on this toxicant do not allow proper calculation of a T25 value.

References

Counts ME et al. (2005) Smoke composition and predicting relationships forinternational commercial cigarettes smoked with three machine-smokingconditions. Regulatory Toxicology and Pharmacology, 41:185–227.

Dybing E et al. (1997) T25: a simplified carcinogenic potency index. Descriptionof the system and study of correlations between carcinogenic potency andspecies/site specificity and mutagenicity. Pharmacology and Toxicology,80:272–279.

European Commission (1999) Guidelines for setting specific concentrationlimits for carcinogens in Annex 1 of Directive 67/548/EEC. Inclusion of potencyconsiderations. Brussels, Commission Working Group on the Classification andLabelling of Dangerous Substances.

Fowles J, Dybing E (2003) Application of toxicological risk assessmentprinciples to the chemical toxicants of cigarette smoke. Tobacco Control,12:424–430.

Gold LS et al. (1984) A carcinogenic potency database of the standardizedresults of animal bioassays. Environmental Health Perspectives, 58:9–319.

IARC (2006) Formaldehyde, 2-butoxyethanol and 1-tert-butoxy-2-propanol.Lyon, International Agency for Research on Cancer (IARC Monographs on theEvaluation of Carcinogenic Risks to Humans, Vol. 88).

Sanner T et al. (2001) A simple method for quantitative risk assessment of non-threshold carcinogens based on the dose descriptor T25. Pharmacology andToxicology, 88:331–341.

Sanner T, Dybing E (2005a) Comparison of carcinogen hazard characterisationbased on animal studies and epidemiology. Basic and Clinical Pharmacologyand Toxicology, 96:66–70.

Sanner T, Dybing E (2005b) Comparison of carcinogenic and in vivo genotoxicpotency estimates. Basic and Clinical Pharmacology and Toxicology, 96:131–139.

227

Scientific Committee on Cosmetic Products and Non-food Products (2003) TheSCCNFP’s notes of guidance for the testing of cosmetic ingredients and theirsafety evaluation. 5th revision. Brussels, European Commission.

WHO (2007) Setting maximal limits for toxic constituents in cigarette smoke.Geneva, World Health Organization, Study Group on Tobacco ProductRegulation. http://www.who.int/tobacco/global_interaction/tobreg/tsr/en/index.html.

228

Table A4.1.Correlation coefficients: international brands, modified intense machine regimen

(1) (2) (3) (4) (5) (6) (7) (8) (9)

(1) Acetaldehyde 1.000(2) Acrolein 0.949 1.000(3) Benzene 0.760 0.764 1.000(4) Benzo[a]pyrene –0.272 –0.208 –0.266 1.000(5) 1,3-Butadiene 0.852 0.855 0.874 –0.401 1.000(6) Carbon monoxide 0.830 0.761 0.673 –0.369 0.807 1.000(7) Formaldehyde 0.156 0.244 0.081 0.340 0.094 –0.111 1.000(8) NNK 0.010 –0.111 0.002 –0.393 –0.051 0.214 –0.639 1.000(9) NNN 0.103 –0.014 0.042 –0.476 0.062 0.252 –0.717 0.853 1.000

NNN, N´-nitrosonornicotine; NNK, 4-(N-nitrosomethylamino)-1-(3-pyridyl)-1-butanone

Table A4.2.Multiple r2: international brands, modified intense machine regimen

Constituent Multiple r2

Benzo[a]pyrene 0.44Formaldehyde 0.63NNK 0.79Benzene 0.81Carbon monoxide 0.82NNN 0.821,3-Butadiene 0.91Acrolein 0.92Acetaldehyde 0.93

NNK, 4-(N-nitrosomethylamino)-1-(3-pyridyl)-1-butanone; NNN, N´-nitrosonornicotine

229

Annex 3.4 Correlation of toxicant yields from international brands,Canadian brands and Canadian brands with less than 100ng/mg nicotine in the modified intense smoking regimenwith machine measurement

Table A4.3.Correlation coefficients: Canadian data (NNN limited), modified intense machineregimen

(1) (2) (3) (4) (5) (6) (7) (8) (9)

(1) Acetaldehyde 1.000(2) Acrolein 0.792 1.000(3) Benzene 0.957 0.779 1.000(4) Benzo[a]pyrene 0.810 0.474 0.859 1.000(5) 1,3-Butadiene 0.953 0.836 0.944 0.765 1.000(6) Carbon monoxide 0.949 0.848 0.972 0.810 0.949 1.000(7) Formaldehyde 0.844 0.568 0.867 0.912 0.827 0.810 1.000(8) NNK 0.707 0.426 0.762 0.817 0.627 0.744 0.710 1.000(9) NNN 0.454 0.128 0.462 0.464 0.376 0.447 0.347 0.774 1.000

NNK, 4-(N-nitrosomethylamino)-1-(3-pyridyl)-1-butanone; NNN, N´-nitrosonornicotine

Table A4.4.Multiple r2: Canadian data (NNN limited), modified intense machine regimen

Constituent Multiple R2

NNN 0.82Formaldehyde 0.90Acrolein 0.91NNK 0.92Benzo[a]pyrene 0.93Acetaldehyde 0.941,3-Butadiene 0.96Benzene 0.97Carbon monoxide 0.98

NNN, N´-nitrosonornicotine; NNK, 4-(N-nitrosomethylamino)-1-(3-pyridyl)-1-butanone

230

Table A4.5.Correlation coefficients: Canadian data (all), modified intense machine regimen

(1) (2) (3) (4) (5) (6) (7) (8) (9)

(1) Acetaldehyde 1.000(2) Acrolein 0.720 1.000(3) Benzene 0.921 0.752 1.000(4) Benzo[a]pyrene 0.630 0.490 0.753 1.000(5) 1,3-Butadiene 0.941 0.776 0.940 0.633 1.000(6) Carbon monoxide 0.945 0.798 0.934 0.680 0.936 1.000(7) Formaldehyde 0.511 0.555 0.658 0.895 0.565 0.559 1.000(8) NNK 0.574 0.137 0.567 0.385 0.532 0.569 0.138 1.000(9) NNN 0.270 –0.197 0.109 –0.258 0.192 0.166 –0.498 0.657 1.000

NNN, N´-nitrosonornicotine; NNK, 4-(N-nitrosomethylamino)-1-(3-pyridyl)-1-butanone

Table A4.6.Multiple r2: Canadian data (all), modified intense machine regimen

Constituent Multiple r2

NNK 0.81Acrolein 0.86NNN 0.88Benzo[a]pyrene 0.91Formaldehyde 0.91Acetaldehyde 0.941,3-Butadiene 0.94Benzene 0.94Carbon monoxide 0.96

NNK, 4-(N-nitrosomethylamino)-1-(3-pyridyl)-1-butanone; NNN, N´-nitrosonornicotine

231

Tab

le A

4.7.

Co

rrel

atio

n m

atri

x fo

r all

con

stit

uen

ts p

er m

illig

ram

of n

ico

tin

e: P

hili

p M

orr

is in

tern

atio

nal

bra

nd

s (C

ou

nts

et a

l., 2

004)

, mo

dif

ied

inte

nse

mac

hin

e re

gim

en

Con

stitu

ent

Tpm

Wat

erC

arbo

n m

onox

ide

Ace

tald

ehyd

eA

ceto

neA

crol

ein

Tar

puf

f cou

nt.

..

..

.T

pm1.

000

..

..

.W

ater

0.95

81.

000

..

..

Car

bon

mon

oxid

e0.

656

0.70

71.

000

..

.A

ceta

ldeh

yde

0.59

70.

611

0.83

01.

000

..

Ace

tone

0.56

30.

585

0.81

70.

983

1.00

0.

Acr

olei

n0.

565

0.58

50.

761

0.94

90.

936

1.00

0B

utyr

alde

hyde

0.53

30.

524

0.78

10.

967

0.95

10.

931

Cro

tona

ldeh

yde

0.48

20.

427

0.55

80.

844

0.86

00.

869

Met

hyl e

thyl

ket

one

0.43

80.

397

0.60

90.

889

0.91

90.

844

Pro

pion

alde

hyde

0.61

70.

622

0.80

40.

984

0.97

00.

948

For

mal

dehy

de0.

065

0.02

8-0

.111

0.15

60.

204

0.24

4A

cryl

onitr

ite0.

321

0.32

20.

694

0.54

50.

511

0.48

7B

enze

ne0.

365

0.34

60.

673

0.76

00.

785

0.76

41,

3-B

utad

iene

0.42

10.

457

0.80

70.

852

0.85

70.

855

Isop

rene

0.26

60.

318

0.75

50.

624

0.60

90.

588

Sty

rene

0.42

90.

372

0.50

10.

723

0.72

30.

742

Tol

uene

0.35

90.

335

0.61

00.

719

0.71

90.

683

Am

mon

ia0.

410

0.34

90.

115

0.07

40.

008

0.05

0T

otal

hyd

roge

n cy

anid

e0.

567

0.59

70.

852

0.67

20.

627

0.62

0Im

ping

er h

ydro

gen

cyan

ide

0.51

70.

549

0.81

90.

646

0.60

60.

594

Pad

hyd

roge

n cy

anid

e0.

647

0.66

90.

882

0.69

30.

641

0.64

4

232

Nitr

ic o

xide

0.36

40.

403

0.60

40.

413

0.38

10.

288

Nitr

ogen

oxi

des

0.39

20.

429

0.61

50.

429

0.39

50.

310

1-A

min

onap

htha

lene

0.29

20.

196

-0.1

05-0

.037

-0.0

65-0

.117

2-A

min

onap

htha

lene

0.33

00.

269

0.02

00.

140

0.12

70.

069

3-A

min

obip

heny

l0.

393

0.30

40.

101

0.28

00.

267

0.19

64-

Am

inob

iphe

nyl

0.42

00.

353

0.24

20.

411

0.39

80.

319

Ben

zo[a

]pyr

ene

0.03

2-0

.067

-0.3

69-0

.272

-0.3

20-0

.208

Cat

echo

l-0

.088

-0.2

18-0

.431

-0.1

42-0

.156

-0.1

44m

- an

d p-

Cre

sol

0.03

0-0

.137

-0.4

81-0

.260

-0.2

82-0

.291

o-C

reso

l-0

.049

-0.2

25-0

.539

-0.3

25-0

.334

-0.3

59H

ydro

quin

one

0.15

30.

008

-0.2

450.

115

0.07

80.

113

Phe

nol

-0.1

29-0

.309

-0.5

61-0

.388

-0.4

13-0

.419

Res

orci

nol

0.25

80.

203

0.22

00.

338

0.29

70.

366

Pyr

idin

e0.

462

0.33

80.

263

0.42

40.

416

0.42

2Q

uino

line

-0.0

15-0

.196

-0.5

34-0

.446

-0.4

80-0

.503

NN

N0.

230

0.23

00.

252

0.10

30.

089

-0.0

14N

NK

0.20

90.

222

0.21

40.

010

0.00

5-0

.111

NA

T0.

194

0.22

00.

318

0.14

40.

140

0.01

2N

AB

0.13

50.

150

0.22

50.

111

0.10

4-0

.030

Mer

cury

0.30

80.

342

0.70

40.

743

0.74

00.

621

Cad

miu

m0.

110

0.07

20.

130

0.09

00.

112

0.02

9Le

ad0.

234

0.22

50.

220

0.22

90.

211

0.08

1S

mok

e pH

0.44

60.

507

0.73

70.

819

0.84

20.

809

..

..

..

.T

pm, t

otal

par

ticul

ate

mat

ter

..

..

..

233

Con

stitu

ent

But

yral

dehy

deC

roto

nald

ehyd

eM

ethy

l eth

yl k

eton

eP

ropi

onal

dehy

deF

orm

alde

hyde

Acr

ylon

itrite

Tar

puf

f cou

nt.

..

..

.T

pm.

..

..

.W

ater

..

..

..

Car

bon

mon

oxid

e.

..

..

.A

ceta

ldeh

yde

..

..

..

Ace

tone

..

..

..

Acr

olei

n.

..

..

.B

utyr

alde

hyde

1.00

0.

..

..

Cro

tona

ldeh

yde

0.87

61.

000

..

..

Met

hyl e

thyl

ket

one

0.88

50.

914

1.00

0.

..

Pro

pion

alde

hyde

0.96

50.

853

0.87

81.

000

..

For

mal

dehy

de0.

137

0.34

80.

354

0.18

81.

000

.A

cryl

onitr

ite0.

610

0.32

40.

385

0.51

4-0

.369

1.00

0B

enze

ne0.

825

0.73

10.

762

0.76

00.

081

0.76

21,

3-B

utad

iene

0.84

60.

692

0.74

00.

810

0.09

40.

734

Isop

rene

0.65

70.

408

0.44

10.

559

-0.3

070.

889

Sty

rene

0.81

00.

831

0.77

30.

746

0.24

50.

472

Tol

uene

0.79

80.

665

0.71

60.

728

-0.0

420.

763

Am

mon

ia0.

049

0.02

6-0

.093

0.07

8-0

.554

0.15

8T

otal

hyd

roge

n cy

anid

e0.

695

0.43

60.

400

0.65

5-0

.455

0.81

8Im

ping

er h

ydro

gen

cyan

ide

0.67

80.

407

0.38

80.

632

-0.4

830.

841

Pad

hyd

roge

n cy

anid

e0.

697

0.47

60.

405

0.67

2-0

.372

0.72

9N

itric

oxi

de0.

413

0.14

40.

204

0.39

9-0

.662

0.66

2N

itrog

en o

xide

s0.

429

0.17

00.

212

0.41

9-0

.651

0.66

1

234

1-A

min

onap

htha

lene

-0.0

520.

036

-0.0

81-0

.012

-0.3

79-0

.141

2-A

min

onap

htha

lene

0.10

50.

189

0.09

90.

149

-0.3

87-0

.053

3-A

min

obip

heny

l0.

234

0.31

40.

294

0.26

8-0

.333

0.04

94-

Am

inob

iphe

nyl

0.37

70.

403

0.39

70.

388

-0.3

780.

217

Ben

zo[a

]pyr

ene

-0.2

37-0

.164

-0.2

15-0

.197

0.34

0-0

.401

Cat

echo

l-0

.159

0.01

00.

072

-0.1

110.

550

-0.5

81m

- an

d p-

Cre

sol

-0.2

57-0

.059

-0.1

40-0

.213

0.15

2-0

.587

o-C

reso

l-0

.309

-0.0

91-0

.140

-0.2

890.

207

-0.5

91H

ydro

quin

one

0.10

30.

242

0.29

80.

148

0.60

2-0

.341

Phe

nol

-0.3

51-0

.173

-0.2

35-0

.364

0.08

9-0

.519

Res

orci

nol

0.32

90.

222

0.27

70.

346

0.24

60.

103

Pyr

idin

e0.

514

0.63

10.

505

0.43

9-0

.055

0.37

3Q

uino

line

-0.4

18-0

.262

-0.3

22-0

.417

0.03

0-0

.450

NN

N0.

076

-0.0

44-0

.047

0.06

9-0

.717

0.33

4N

NK

-0.0

30-0

.180

-0.1

110.

013

-0.6

390.

285

NA

T0.

125

-0.0

45-0

.015

0.10

7-0

.714

0.41

6N

AB

0.08

8-0

.064

-0.0

040.

071

-0.7

350.

380

Mer

cury

0.75

80.

518

0.65

30.

710

-0.1

520.

711

Cad

miu

m0.

161

0.11

50.

150

0.11

8-0

.293

0.38

6Le

ad0.

205

0.05

80.

205

0.25

4-0

.205

0.27

9S

mok

e pH

0.79

80.

696

0.70

10.

775

0.20

50.

550

..

..

..

.T

pm, t

otal

par

ticul

ate

mat

ter

..

..

..

235

Con

stitu

ent

Ben

zene

1,3-

But

adie

neIs

opre

neS

tyre

neT

olue

neA

mm

onia

Tar

puf

f cou

nt.

..

..

.T

pm.

..

..

.W

ater

..

..

..

Car

bon

mon

oxid

e.

..

..

.A

ceta

ldeh

yde

..

..

..

Ace

tone

..

..

..

Acr

olei

n.

..

..

But

yral

dehy

de.

..

..

.C

roto

nald

ehyd

e.

..

..

.M

ethy

l eth

yl k

eton

e.

..

..

.P

ropi

onal

dehy

de.

..

..

.F

orm

alde

hyde

..

..

..

Acr

ylon

itrite

..

..

..

Ben

zene

1.00

0.

..

..

1,3-

But

adie

ne0.

874

1.00

0.

..

.Is

opre

ne0.

737

0.84

51.

000

..

.S

tyre

ne0.

800

0.63

50.

439

1.00

0.

.T

olue

ne0.

931

0.74

10.

669

0.84

91.

000

.A

mm

onia

-0.0

29-0

.066

0.03

80.

001

0.07

91.

000

Tot

al h

ydro

gen

cyan

ide

0.62

80.

683

0.81

10.

451

0.63

00.

394

Impi

nger

hyd

roge

n cy

anid

e0.

643

0.67

20.

813

0.43

50.

654

0.38

9P

ad h

ydro

gen

cyan

ide

0.56

60.

672

0.76

80.

464

0.55

10.

385

Nitr

ic o

xide

0.38

10.

363

0.58

60.

164

0.46

80.

513

Nitr

ogen

oxi

des

0.38

90.

368

0.58

10.

185

0.47

50.

532

236

1-A

min

onap

htha

lene

-0.1

35-0

.262

-0.2

120.

033

-0.0

090.

670

2-A

min

onap

htha

lene

0.04

0-0

.091

-0.0

940.

150

0.16

40.

684

3-A

min

obip

heny

l0.

159

0.06

90.

020

0.25

50.

252

0.65

74-

Am

inob

iphe

nyl

0.30

20.

226

0.18

50.

351

0.38

80.

640

Ben

zo[a

]pyr

ene

-0.2

66-0

.401

-0.5

81-0

.034

-0.2

140.

029

Cat

echo

l-0

.266

-0.3

32-0

.640

-0.0

34-0

.241

-0.2

09m

- an

d p-

Cre

sol

-0.4

51-0

.515

-0.6

27-0

.170

-0.3

990.

220

o-C

reso

l-0

.445

-0.5

23-0

.621

-0.1

69-0

.401

0.08

7H

ydro

quin

one

-0.0

22-0

.088

-0.4

520.

233

0.00

6-0

.110

Phe

nol

-0.4

81-0

.541

-0.5

72-0

.252

-0.4

480.

123

Res

orci

nol

0.23

90.

298

0.08

50.

324

0.23

60.

089

Pyr

idin

e0.

534

0.33

30.

317

0.76

80.

620

0.33

8Q

uino

line

-0.5

07-0

.600

-0.5

62-0

.254

-0.4

230.

195

NN

N0.

042

0.06

20.

317

-0.0

960.

118

0.69

9N

NK

0.00

2-0

.051

0.20

4-0

.150

0.12

10.

568

NA

T0.

100

0.12

20.

408

-0.0

830.

170

0.58

6N

AB

0.08

70.

062

0.35

4-0

.099

0.18

00.

615

Mer

cury

0.69

10.

691

0.68

20.

561

0.72

5-0

.020

Cad

miu

m0.

321

0.03

20.

148

0.37

20.

493

0.32

8Le

ad0.

198

0.05

30.

071

0.17

60.

378

0.24

2S

mok

e pH

0.71

40.

881

0.68

60.

576

0.56

8-0

.098

..

..

..

.T

pm, t

otal

par

ticul

ate

mat

ter

..

..

..

237

Con

stitu

ent

Tot

al h

ydro

gen

cyan

ide

Impi

nger

hyd

roge

ncy

anid

eP

ad h

ydro

gen

cyan

ide

Nitr

ic o

xide

Nitr

ogen

oxi

des

1-A

min

onap

htha

lene

Tar

puf

f cou

nt.

..

..

.T

pm.

..

..

.W

ater

..

..

..

Car

bon

mon

oxid

e.

..

..

.A

ceta

ldeh

yde

..

..

..

Ace

tone

..

..

..

Acr

olei

n.

..

..

.B

utyr

alde

hyde

..

..

..

Cro

tona

ldeh

yde

..

..

..

Met

hyl e

thyl

ket

one

..

..

..

Pro

pion

alde

hyde

..

..

..

For

mal

dehy

de.

..

..

.A

cryl

onitr

ite.

..

..

.B

enze

ne.

..

..

.1,

3-B

utad

iene

..

..

..

Isop

rene

..

..

..

Sty

rene

..

..

..

Tol

uene

..

..

..

Am

mon

ia.

..

..

.T

otal

hyd

roge

n cy

anid

e1.

000

..

..

.Im

ping

er h

ydro

gen

cyan

ide

0.99

31.

000

..

..

Pad

hyd

roge

n cy

anid

e0.

967

0.93

01.

000

..

.N

itric

oxi

de0.

813

0.83

50.

728

1.00

0.

.N

itrog

en o

xide

s0.

825

0.84

40.

745

0.99

81.

000

.

238

1-A

min

onap

htha

lene

0.10

20.

095

0.11

20.

230

0.24

71.

000

2-A

min

onap

htha

lene

0.19

60.

200

0.17

70.

337

0.35

40.

931

3-A

min

obip

heny

l0.

239

0.23

80.

229

0.38

60.

397

0.80

44-

Am

inob

iphe

nyl

0.38

70.

392

0.36

00.

484

0.49

40.

741

Ben

zo[a

]pyr

ene

-0.4

09-0

.426

-0.3

54-0

.482

-0.4

710.

084

Cat

echo

l-0

.609

-0.6

23-0

.549

-0.5

98-0

.600

0.12

9m

- an

d p-

Cre

sol

-0.4

31-0

.457

-0.3

56-0

.354

-0.3

430.

583

o-C

reso

l-0

.518

-0.5

43-0

.441

-0.4

24-0

.422

0.47

0H

ydro

quin

one

-0.4

16-0

.437

-0.3

50-0

.511

-0.5

020.

112

Phe

nol

-0.4

71-0

.489

-0.4

09-0

.380

-0.3

790.

445

Res

orci

nol

0.13

50.

118

0.16

3-0

.128

-0.1

27-0

.010

Pyr

idin

e0.

400

0.38

50.

413

0.24

60.

266

0.37

5Q

uino

line

-0.4

19-0

.437

-0.3

60-0

.316

-0.3

160.

520

NN

N0.

507

0.52

00.

456

0.72

30.

716

0.51

8N

NK

0.40

30.

434

0.31

70.

731

0.72

10.

363

NA

T0.

563

0.58

40.

492

0.78

80.

779

0.42

2N

AB

0.50

40.

538

0.40

70.

806

0.79

30.

480

Mer

cury

0.69

90.

716

0.62

80.

654

0.65

1-0

.103

Cad

miu

m0.

298

0.33

40.

209

0.50

40.

504

0.17

8Le

ad0.

271

0.30

30.

192

0.58

70.

581

0.24

4S

mok

e pH

0.58

40.

566

0.59

50.

234

0.24

7-0

.164

..

..

..

.T

pm, t

otal

par

ticul

ate

mat

ter

..

..

..

239

Con

stitu

ent

2-A

min

onap

htha

lene

3-A

min

obip

heny

l4-

Am

inob

iphe

nyl

Ben

zo[a

]pyr

ene

Cat

echo

lm

- an

d p-

Cre

sol

Tar

puf

f cou

nt.

..

..

.T

pm.

..

..

.W

ater

..

..

..

Car

bon

mon

oxid

e.

..

..

.A

ceta

ldeh

yde

..

..

..

Ace

tone

..

..

..

Acr

olei

n.

..

..

.B

utyr

alde

hyde

..

..

..

Cro

tona

ldeh

yde

..

..

..

Met

hyl e

thyl

ket

one

..

..

..

Pro

pion

alde

hyde

..

..

..

For

mal

dehy

de.

..

..

.A

cryl

onitr

ite.

..

..

.B

enze

ne.

..

..

.1,

3-B

utad

iene

..

..

..

Isop

rene

..

..

..

Sty

rene

..

..

..

Tol

uene

..

..

..

Am

mon

ia.

..

..

.T

otal

hyd

roge

n cy

anid

e.

..

..

.Im

ping

er h

ydro

gen

cyan

ide

..

..

..

Pad

hyd

roge

n cy

anid

e.

..

..

.N

itric

oxi

de.

..

..

.N

itrog

en o

xide

s.

..

..

.1-

Am

inon

apht

hale

ne.

..

..

.2-

Am

inon

apht

hale

ne1.

000

..

..

.3-

Am

inob

iphe

nyl

0.89

71.

000

..

..

240

4-A

min

obip

heny

l0.

865

0.97

11.

000

..

.B

enzo

[a]p

yren

e-0

.067

-0.1

25-0

.239

1.00

0.

.C

atec

hol

0.03

40.

058

-0.0

690.

702

1.00

0.

m-

and

p-C

reso

l0.

403

0.37

40.

229

0.50

80.

680

1.00

0o-

Cre

sol

0.27

70.

284

0.12

90.

500

0.72

40.

957

Hyd

roqu

inon

e0.

044

0.15

50.

069

0.65

00.

886

0.57

8P

heno

l0.

231

0.23

60.

099

0.48

50.

659

0.93

3R

esor

cino

l-0

.058

0.04

50.

041

0.37

20.

353

0.18

1P

yrid

ine

0.41

70.

528

0.56

2-0

.082

-0.0

980.

113

Qui

nolin

e0.

281

0.23

20.

090

0.48

70.

570

0.86

9N

NN

0.55

60.

576

0.60

8-0

.476

-0.4

91-0

.038

NN

K0.

415

0.39

00.

393

-0.3

93-0

.464

-0.1

29N

AT

0.46

90.

491

0.54

7-0

.573

-0.5

82-0

.147

NA

B0.

549

0.59

80.

641

-0.5

32-0

.498

-0.1

03M

ercu

ry0.

060

0.23

20.

393

-0.4

80-0

.389

-0.4

68C

adm

ium

0.26

00.

297

0.35

0-0

.156

-0.3

38-0

.258

Lead

0.32

80.

369

0.38

8-0

.107

-0.0

36-0

.084

Sm

oke

pH-0

.009

0.11

50.

268

-0.4

12-0

.322

-0.4

09.

..

..

..

Tpm

, tot

al p

artic

ulat

e m

atte

r.

..

..

.

241

Con

stitu

ent

o-C

reso

lH

ydro

quin

one

Phe

nol

Res

orci

nol

Pyr

idin

eQ

uino

line

NN

NN

NK

NA

TN

AB

Tar

puf

f cou

nt.

..

..

..

..

.T

pm.

..

..

..

..

.W

ater

..

..

..

..

..

Car

bon

mon

oxid

e.

..

..

..

..

.A

ceta

ldeh

yde

..

..

..

..

..

Ace

tone

..

..

..

..

..

Acr

olei

n.

..

..

..

..

.B

utyr

alde

hyde

..

..

..

..

..

Cro

tona

ldeh

yde

..

..

..

..

..

Met

hyl e

thyl

ket

one

..

..

..

..

..

Pro

pion

alde

hyde

..

..

..

..

..

For

mal

dehy

de.

..

..

..

..

.A

cryl

onitr

ite.

..

..

..

..

.B

enze

ne.

..

..

..

..

.1,

3-B

utad

iene

..

..

..

..

..

Isop

rene

..

..

..

..

..

Sty

rene

..

..

..

..

..

Tol

uene

..

..

..

..

..

Am

mon

ia.

..

..

..

..

.T

otal

hyd

roge

n cy

anid

e.

..

..

..

..

.Im

ping

er h

ydro

gen

cyan

ide

..

..

..

..

..

Pad

hyd

roge

n cy

anid

e.

..

..

..

..

.N

itric

oxi

de.

..

..

..

..

.N

itrog

en o

xide

s.

..

..

..

..

.1-

Am

inon

apht

hale

ne.

..

..

..

..

.2-

Am

inon

apht

hale

ne.

..

..

..

..

.

242

3-A

min

obip

heny

l.

..

..

..

..

.4-

Am

inob

iphe

nyl

..

..

..

..

..

Ben

zo[a

]pyr

ene

..

..

..

..

..

Cat

echo

l.

..

..

..

..

.m

- an

d p-

Cre

sol

..

..

..

..

..

o-C

reso

l1.

000

..

..

..

..

.H

ydro

quin

one

0.60

71.

000

..

..

..

..

Phe

nol

0.96

70.

542

1.00

0.

..

..

..

Res

orci

nol

0.16

20.

458

0.13

11.

000

..

..

..

Pyr

idin

e0.

138

0.14

90.

083

0.14

31.

000

..

..

.Q

uino

line

0.90

80.

475

0.91

60.

073

0.15

81.

000

..

..

NN

N-0

.110

-0.5

05-0

.079

-0.1

380.

229

-0.0

031.

000

..

.N

NK

-0.1

76-0

.514

-0.1

90-0

.227

0.13

6-0

.060

0.85

31.

000

..

NA

T-0

.220

-0.5

94-0

.184

-0.1

750.

187

-0.1

110.

971

0.86

41.

000

.N

AB

-0.1

69-0

.498

-0.1

29-0

.233

0.20

8-0

.066

0.91

70.

831

0.93

91.

000

Mer

cury

-0.4

96-0

.178

-0.4

960.

082

0.30

7-0

.521

0.22

80.

183

0.35

00.

370

Cad

miu

m-0

.235

-0.2

15-0

.245

-0.1

760.

434

-0.1

110.

347

0.40

10.

356

0.40

4Le

ad-0

.102

0.00

8-0

.158

-0.1

220.

160

-0.0

210.

274

0.44

30.

318

0.44

5S

mok

e pH

-0.4

63-0

.092

-0.4

940.

255

0.27

5-0

.548

0.04

9-0

.140

0.11

90.

033

..

..

..

..

..

.T

pm, t

otal

par

ticul

ate

mat

ter

..

..

..

..

..

243

Con

stitu

ent

Mer

cury

Cad

miu

mLe

adS

mok

e pH

Met

hyl e

thyl

ket

one

Chr

omiu

mN

icke

lA

rsen

icS

elen

ium

Tar

puf

f cou

nt.

..

..

..

..

Tpm

..

..

..

..

.W

ater

..

..

..

..

.C

arbo

n m

onox

ide

..

..

..

..

.A

ceta

ldeh

yde

..

..

..

..

.A

ceto

ne.

..

..

..

..

Acr

olei

n.

..

..

..

..

But

yral

dehy

de.

..

..

..

..

Cro

tona

ldeh

yde

..

..

..

..

.M

ethy

l eth

yl k

eton

e.

..

..

..

..

Pro

pion

alde

hyde

..

..

..

..

.F

orm

alde

hyde

..

..

..

..

.A

cryl

onitr

ite.

..

..

..

..

Ben

zene

..

..

..

..

.1,

3-B

utad

iene

..

..

..

..

.Is

opre

ne.

..

..

..

..

Sty

rene

..

..

..

..

.T

olue

ne.

..

..

..

..

Am

mon

ia.

..

..

..

..

Tot

al h

ydro

gen

cyan

ide

..

..

..

..

.Im

ping

er h

ydro

gen

cyan

ide

..

..

..

..

.P

ad h

ydro

gen

cyan

ide

..

..

..

..

.N

itric

oxi

de.

..

..

..

..

Nitr

ogen

oxi

des

..

..

..

..

.1-

Am

inon

apht

hale

ne.

..

..

..

..

2-A

min

onap

htha

lene

..

..

..

..

.3-

Am

inob

iphe

nyl

..

..

..

..

.

244

4-A

min

obip

heny

l.

..

..

..

..

Ben

zo[a

]pyr

ene

..

..

..

..

.C

atec

hol

..

..

..

..

.m

- an

d p-

Cre

sol

..

..

..

..

.o-

Cre

sol

..

..

..

..

.H

ydro

quin

one

..

..

..

..

.P

heno

l.

..

..

..

..

Res

orci

nol

..

..

..

..

.P

yrid

ine

..

..

..

..

.Q

uino

line

..

..

..

..

.N

NN

..

..

..

..

.N

NK

..

..

..

..

.N

AT

..

..

..

..

.N

AB

..

..

..

..

.M

ercu

ry1.

000

..

..

..

..

Cad

miu

m0.

449

1.00

0.

..

..

..

Lead

0.50

60.

592

1.00

0.

..

..

.S

mok

e pH

0.66

2-0

.056

-0.0

081.

000

..

..

..

..

..

..

..

.T

pm, t

otal

par

ticul

ate

mat

ter

..

..

..

..

.

245

Tab

le A

4.8.

Co

rrel

atio

n m

atri

x fo

r al

l co

nst

itu

ents

per

mill

igra

m o

f n

ico

tin

e re

po

rted

to

Hea

lth

Can

ada

for

cig

aret

te b

ran

ds

wit

h N

NN

leve

ls le

ssth

an 1

00 n

g/m

g n

ico

tin

e (H

ealt

h C

anad

a, 2

004)

, mo

dif

ied

inte

nse

mac

hin

e re

gim

en

Con

stitu

ent

Car

bon

Mon

oxid

eA

mm

onia

1-A

min

onap

htha

lene

2-A

min

onap

htha

lene

3-A

min

obip

heny

l4-

Am

inob

iphe

nyl

Car

bon

Mon

oxid

e1.

000

Am

mon

ia0.

625

1.00

01-

Am

inon

apht

hale

ne-0

.338

-0.5

941.

000

2-A

min

onap

htha

lene

0.85

90.

804

-0.5

141.

000

3-A

min

obip

heny

l0.

307

-0.1

550.

550

0.04

71.

000

4-A

min

obip

heny

l0.

877

0.79

0-0

.500

0.98

50.

154

1.00

0B

enzo

[a]p

yren

e0.

810

0.84

7-0

.613

0.95

9-0

.155

0.92

6F

orm

alde

hyde

0.81

00.

817

-0.5

160.

878

-0.0

710.

852

Ace

tald

ehyd

e0.

949

0.64

5-0

.362

0.85

80.

239

0.87

9A

ceto

ne0.

935

0.47

7-0

.119

0.75

60.

452

0.78

0A

crol

ein

0.84

80.

306

0.06

50.

568

0.62

40.

607

Pro

pion

alde

yde

0.94

40.

579

-0.2

570.

816

0.39

70.

853

Cro

tona

ldeh

yde

0.86

30.

578

-0.1

790.

817

0.25

00.

807

But

yral

dehy

de0.

959

0.62

0-0

.267

0.87

10.

307

0.88

3H

ydro

gen

cyan

ide

0.94

50.

746

-0.4

410.

929

0.20

20.

951

Mer

cury

0.67

40.

161

0.13

30.

341

0.46

60.

405

Lead

0.89

20.

657

-0.5

350.

919

0.72

80.

927

Cad

miu

m0.

249

-0.1

780.

502

0.01

50.

603

0.03

5N

itrog

en o

xide

s0.

890

0.55

2-0

.374

0.83

80.

264

0.86

4N

iroge

n ox

ides

0.90

20.

585

-0.4

040.

868

0.23

20.

886

246

NN

N0.

447

0.48

0-0

.438

0.58

80.

138

0.63

4N

NK

0.74

40.

666

-0.5

730.

902

0.07

80.

904

N´-

Nitr

osoa

naba

sine

0.57

30.

500

-0.4

330.

715

0.15

20.

734

N´-

Nitr

osoa

nata

bine

0.73

70.

678

-0.4

830.

854

0.15

10.

870

Pyr

idin

e0.

830

0.71

2-0

.286

0.81

40.

310

0.83

9Q

uino

line

0.20

50.

538

-0.3

620.

527

-0.4

630.

453

Hyd

roqu

inon

e0.

733

0.57

7-0

.187

0.77

20.

103

0.71

5R

esor

cino

l0.

765

0.72

4-0

.531

0.88

5-0

.107

0.86

1C

atec

hol

0.61

90.

606

-0.3

880.

766

-0.2

110.

685

Phe

nol

-0.3

89-0

.033

0.13

0-0

.114

-0.4

19-0

.199

m-

and

p-C

reso

ls-0

.055

0.28

2-0

.144

0.27

1-0

.457

0.18

4o-

Cre

sol

-0.2

730.

079

0.04

60.

025

-0.4

42-0

.063

1,3-

But

adie

ne0.

949

0.65

2-0

.323

0.79

60.

266

0.81

5Is

opre

ne0.

918

0.78

5-0

.483

0.91

50.

090

0.92

6A

cryl

onitr

ile0.

938

0.72

5-0

.489

0.86

40.

253

0.90

1B

enze

ne0.

972

0.71

9-0

.401

0.89

20.

206

0.90

1T

olue

ne0.

864

0.71

1-0

.389

0.87

20.

280

0.89

3S

tyre

ne0.

045

-0.0

670.

012

-0.0

51-0

.204

-0.0

80P

uff c

ount

0.84

70.

453

-0.2

020.

777

0.37

50.

790

247

Con

stitu

ent

Ben

zo[a

]pyr

ene

For

mal

dehy

deA

ceta

ldeh

yde

Ace

tone

Acr

olei

nP

ropi

onal

deyd

e

Car

bon

Mon

oxid

eA

mm

onia

1-A

min

onap

htha

lene

2-A

min

onap

htha

lene

3-A

min

obip

heny

l4-

Am

inob

iphe

nyl

Ben

zo[a

]pyr

ene

1.00

0F

orm

alde

hyde

0.91

21.

000

Ace

tald

ehyd

e0.

810

0.84

41.

000

Ace

tone

0.67

50.

740

0.94

41.

000

Acr

olei

n0.

474

0.56

80.

792

0.91

21.

000

Pro

pion

alde

yde

0.74

60.

788

0.97

20.

974

0.86

01.

000

Cro

tona

ldeh

yde

0.76

20.

844

0.88

70.

894

0.80

10.

886

But

yral

dehy

de0.

805

0.82

80.

976

0.96

50.

843

0.97

4H

ydro

gen

cyan

ide

0.88

50.

844

0.93

20.

855

0.72

00.

910

Mer

cury

0.31

50.

419

0.67

50.

759

0.79

70.

708

Lead

0.91

10.

852

0.87

90.

888

0.88

50.

896

Cad

miu

m-0

.081

-0.0

080.

067

0.31

30.

526

0.19

5N

itrog

en o

xide

s0.

766

0.63

80.

849

0.80

00.

670

0.83

5N

iroge

n ox

ides

0.80

10.

679

0.86

00.

805

0.67

10.

841

NN

N0.

464

0.34

70.

454

0.32

20.

128

0.42

4N

NK

0.81

70.

710

0.70

70.

609

0.42

60.

690

248

N´-

Nitr

osoa

naba

sine

0.59

60.

476

0.52

00.

449

0.25

80.

523

N´-

Nitr

osoa

nata

bine

0.77

00.

668

0.75

60.

689

0.45

50.

749

Pyr

idin

e0.

746

0.77

10.

817

0.80

30.

682

0.83

0Q

uino

line

0.62

10.

516

0.25

90.

099

-0.0

510.

182

Hyd

roqu

inon

e0.

757

0.77

30.

713

0.72

70.

649

0.69

1R

esor

cino

l0.

925

0.83

50.

781

0.67

20.

493

0.72

3C

atec

hol

0.81

10.

750

0.60

70.

542

0.41

80.

554

Phe

nol

-0.0

25-0

.073

-0.3

48-0

.384

-0.3

79-0

.372

m-

and

p-C

reso

ls0.

356

0.26

1-0

.002

-0.1

07-0

.203

-0.0

54o-

Cre

sol

0.11

90.

062

-0.2

21-0

.282

-0.3

17-0

.254

1,3-

But

adie

ne0.

765

0.82

70.

953

0.92

50.

836

0.93

5Is

opre

ne0.

905

0.88

40.

954

0.84

80.

689

0.90

0A

cryl

onitr

ile0.

811

0.80

60.

924

0.85

50.

729

0.91

4B

enze

ne0.

859

0.86

70.

957

0.91

50.

779

0.93

4T

olue

ne0.

802

0.78

50.

871

0.83

40.

749

0.87

1S

tyre

ne0.

017

0.05

3-0

.011

-0.0

21-0

.118

-0.0

43P

uff c

ount

0.69

50.

679

0.85

10.

896

0.84

70.

885

249

Con

stitu

ent

Cro

tona

ldeh

yde

But

yral

dehy

deH

ydro

gen

Cya

nide

Mer

cury

Lead

Cad

miu

m

Car

bon

Mon

oxid

eA

mm

onia

1-A

min

onap

htha

lene

2-A

min

onap

htha

lene

3-A

min

obip

heny

l4-

Am

inob

iphe

nyl

Ben

zo[a

]pyr

ene

For

mal

dehy

deA

ceta

ldeh

yde

Ace

tone

Acr

olei

nP

ropi

onal

deyd

eC

roto

nald

ehyd

e1.

000

But

yral

dehy

de0.

944

1.00

0H

ydro

gen

cyan

ide

0.85

50.

936

1.00

0M

ercu

ry0.

637

0.68

30.

584

1.00

0Le

ad0.

906

0.92

00.

951

0.86

81.

000

Cad

miu

m0.

315

0.21

70.

133

0.43

10.

523

1.00

0N

itrog

en o

xide

s0.

734

0.86

30.

886

0.53

60.

878

0.19

9N

iroge

n ox

ides

0.75

70.

876

0.90

30.

521

0.89

10.

193

NN

N0.

274

0.41

10.

541

-0.0

300.

706

-0.2

13N

NK

0.64

30.

717

0.81

80.

157

0.89

40.

040

250

N´-

Nitr

osoa

naba

sine

0.43

70.

532

0.63

80.

057

0.82

20.

086

N´-

Nitr

osoa

nata

bine

0.67

60.

761

0.81

40.

342

0.91

80.

078

Pyr

idin

e0.

814

0.85

90.

866

0.52

00.

895

0.29

0Q

uino

line

0.40

70.

286

0.36

8-0

.109

0.01

4-0

.287

Hyd

roqu

inon

e0.

792

0.76

50.

730

0.35

70.

673

0.18

8R

esor

cino

l0.

734

0.77

50.

814

0.34

70.

679

-0.1

10C

atec

hol

0.70

00.

648

0.64

60.

171

0.25

2-0

.055

Phe

nol

-0.0

53-0

.280

-0.2

80-0

.382

-0.5

14-0

.145

m-

and

p-C

reso

ls0.

226

0.05

10.

087

-0.2

54-0

.295

-0.2

34o-

Cre

sol

0.06

1-0

.158

-0.1

43-0

.316

-0.4

02-0

.152

1,3-

But

adie

ne0.

850

0.93

60.

889

0.70

30.

780

0.13

2Is

opre

ne0.

852

0.93

30.

954

0.59

80.

885

-0.0

13A

cryl

onitr

ile0.

799

0.90

70.

940

0.54

40.

859

0.08

0B

enze

ne0.

884

0.96

20.

949

0.63

40.

921

0.19

7T

olue

ne0.

822

0.88

40.

887

0.50

80.

775

0.16

5S

tyre

ne0.

005

-0.0

040.

003

0.05

9-0

.049

0.06

4P

uff c

ount

0.82

80.

890

0.79

30.

604

0.79

50.

278

251

Con

stitu

ent

Nitr

ic o

xide

Nitr

ogen

oxi

des

NN

NN

NK

N´-

Nitr

osoa

nata

bine

N´-

Nitr

osoa

naba

sine

Car

bon

Mon

oxid

eA

mm

onia

1-A

min

onap

htha

lene

2-A

min

onap

htha

lene

3-A

min

obip

heny

l4-

Am

inob

iphe

nyl

Ben

zo[a

]pyr

ene

For

mal

dehy

deA

ceta

ldeh

yde

Ace

tone

Acr

olei

nP

ropi

onal

deyd

eC

roto

nald

ehyd

eB

utyr

alde

hyde

Hyd

roge

n cy

anid

eM

ercu

ryLe

adC

adm

ium

Nitr

ogen

oxi

des

1.00

0N

iroge

n ox

ides

0.99

71.

000

NN

N0.

585

0.58

81.

000

NN

K0.

804

0.83

10.

774

1.00

0N

´-N

itros

oana

basi

ne0.

723

0.73

70.

856

0.91

81.

000

N´-

Nitr

osoa

nata

bine

0.84

40.

856

0.82

80.

923

0.91

71.

000

252

Pyr

idin

e0.

782

0.80

10.

498

0.74

30.

637

0.80

6Q

uino

line

0.19

60.

233

0.20

40.

384

0.19

00.

324

Hyd

roqu

inon

e0.

549

0.59

00.

187

0.59

30.

316

0.47

7R

esor

cino

l0.

700

0.72

60.

405

0.71

60.

476

0.63

9C

atec

hol

0.48

20.

523

0.23

00.

590

0.30

50.

443

Phe

nol

-0.3

96-0

.374

-0.2

62-0

.204

-0.2

91-0

.247

m-

and

p-C

reso

ls-0

.042

-0.0

130.

011

0.15

20.

005

0.10

0o-

Cre

sol

-0.2

68-0

.243

-0.1

94-0

.079

-0.1

90-0

.111

1,3-

But

adie

ne0.

763

0.77

80.

376

0.62

70.

429

0.65

2Is

opre

ne0.

837

0.85

60.

478

0.73

60.

526

0.75

3A

cryl

onitr

ile0.

847

0.86

00.

606

0.80

00.

650

0.78

8B

enze

ne0.

881

0.89

80.

462

0.76

20.

597

0.79

0T

olue

ne0.

781

0.80

10.

457

0.76

70.

563

0.72

0S

tyre

ne0.

062

0.06

2-0

.006

-0.0

410.

064

0.06

9P

uff c

ount

0.76

20.

774

0.28

60.

654

0.45

60.

627

253

Con

stitu

ent

Pyr

idin

eQ

uino

line

Hyd

roqu

inon

eR

esor

cino

lC

atec

hol

Phe

nol

Car

bon

Mon

oxid

eA

mm

onia

1-A

min

onap

htha

lene

2-A

min

onap

htha

lene

3-A

min

obip

heny

l4-

Am

inob

iphe

nyl

Ben

zo[a

]pyr

ene

For

mal

dehy

deA

ceta

ldeh

yde

Ace

tone

Acr

olei

nP

ropi

onal

deyd

eC

roto

nald

ehyd

eB

utyr

alde

hyde

Hyd

roge

n cy

anid

eM

ercu

ryLe

adC

adm

ium

Nitr

ogen

oxi

des

Niro

gen

oxid

esN

NN

NN

KN

´-N

itros

oana

basi

neN

´-N

itros

oana

tabi

neP

yrid

ine

1.00

0Q

uino

line

0.23

61.

000

Hyd

roqu

inon

e0.

667

0.45

61.

000

254

Res

orci

nol

0.64

50.

614

0.74

41.

000

Cat

echo

l0.

476

0.77

10.

843

0.85

21.

000

Phe

nol

-0.3

200.

727

0.05

50.

033

0.37

21.

000

m-

and

p-C

reso

ls-0

.031

0.92

70.

293

0.38

30.

644

0.91

2o-

Cre

sol

-0.2

000.

811

0.15

00.

144

0.45

90.

983

1,3-

But

adie

ne0.

779

0.21

80.

731

0.74

40.

618

-0.3

41Is

opre

ne0.

820

0.42

10.

725

0.85

70.

683

-0.2

36A

cryl

onitr

ile0.

829

0.24

50.

655

0.76

30.

587

-0.3

80B

enze

ne0.

882

0.26

50.

730

0.79

00.

626

-0.3

51T

olue

ne0.

851

0.30

70.

765

0.75

00.

623

-0.2

43S

tyre

ne-0

.025

-0.0

50-0

.143

-0.0

39-0

.060

-0.1

08P

uff c

ount

0.74

60.

196

0.72

70.

694

0.59

2-0

.260

255

Con

stitu

ent

m-

and

p-C

reso

lso-

Cre

sol

1,3-

But

adie

neIs

opre

neA

cryl

onitr

ileB

enze

neT

olue

neS

tyre

neP

uff c

ount

Car

bon

Mon

oxid

eA

mm

onia

1-A

min

onap

htha

lene

2-A

min

onap

htha

lene

3-A

min

obip

heny

l4-

Am

inob

iphe

nyl

Ben

zo[a

]pyr

ene

For

mal

dehy

deA

ceta

ldeh

yde

Ace

tone

Acr

olei

nP

ropi

onal

deyd

eC

roto

nald

ehyd

eB

utyr

alde

hyde

Hyd

roge

n cy

anid

eM

ercu

ryLe

adC

adm

ium

Nitr

ogen

oxi

des

Niro

gen

oxid

esN

NN

NN

K

256

N´-

Nitr

osoa

naba

sine

N´-

Nitr

osoa

nata

bine

Pyr

idin

eQ

uino

line

Hyd

roqu

inon

eR

esor

cino

lC

atec

hol

Phe

nol

m-

and

p-C

reso

ls1.

000

o-C

reso

l0.

956

1.00

01,

3-B

utad

iene

-0.0

34-0

.239

1.00

0Is

opre

ne0.

132

-0.1

040.

930

1.00

0A

cryl

onitr

ile-0

.031

-0.2

710.

919

0.92

11.

000

Ben

zene

-0.0

04-0

.225

0.94

40.

952

0.94

11.

000

Tol

uene

0.06

1-0

.131

0.85

60.

884

0.86

60.

881

1.00

0S

tyre

ne-0

.064

-0.0

86-0

.035

-0.0

340.

002

0.06

3-0

.395

1.00

0P

uff c

ount

0.00

5-0

.163

0.82

10.

796

0.76

30.

833

0.89

0-0

.279

1.00

0

NN

K, 4

-(N

-nitr

osom

ethy

lam

ino)

-1-(

3-py

ridyl

)-1-

buta

none

; NN

N, N

´-ni

tros

onor

nico

tine

257

Figure A4.1Correlation scatterplot matrix: international brands, modified intense machineregimen

Figure A4.2Correlation scatterplot matrix: Canadian data (NNN limited), modified intensemachine regimenNNN, N´-nitrosonornicotine

258

Figure A4.3Correlation scatterplot matrix: Canadian data (all), modified intense machine regimen

259

Annex 3.5 Variation in yield of selected toxicants by brand

0%

25%

50%

75%

100%

125%

150%

EXPORT 'A' U

LTRA LI

GHT REGULA

R SIZE

EXPORT 'A' M

EDIUM K

ING S

IZE

CRAVEN A E

XTRA LIGHT K

ING S

IZE

EXPORT 'A' E

XTRA LIGHT R

EGULAR S

IZE

EXPORT 'A' L

IGHT / S

PECIAL L

IGHT K

ING S

IZE

MORE MENTHOL 1

20's

EXPORT 'A' E

XTRA LIGHT K

ING S

IZE

EXPORT 'A' M

ILD R

EGULAR S

IZE

NUMBER 7 KIN

G SIZE

CAMEL LIG

HT KIN

G SIZE

EXPORT PLA

IN R

EGULAR S

IZE

WIN

STON LIGHT 10

0's

GAULOIS

ES BLO

NDES KIN

G SIZE

CAMEL KIN

G SIZE

WIN

STON KIN

G SIZE

DU MAURIE

R EXTRA LI

GHT REGULA

R SIZE

SALEM LI

GHT KIN

G SIZE

CAMEL PLA

IN R

EGULAR S

IZE

VISCOUNT E

XTRA MILD

KIN

G SIZE

VANTAGE MEDIU

M 7 KIN

G SIZE

PLAYER'S LI

GHT REGULA

R SIZE

CRAVEN MENTHOL K

ING S

IZE

BELMONT M

ILDS R

EGULAR S

IZE

PLAYER'S LI

GHT KIN

G SIZE

VANTAGE RIC

H 12 K

ING S

IZE

MARK TEN PLA

IN K

ING S

IZE

DU MAURIE

R KIN

G SIZE

DU MAURIE

R EXTRA LI

GHT KIN

G SIZE

PLAYER'S S

PECIAL B

LEND R

EGULAR S

IZE

PLAYER'S E

XTRA LIGHT K

ING S

IZE

CANYON REGULA

R SIZE

DISCRETIO

N REGULA

R SIZE

GIPSY R

EGULAR S

IZE

SMOKING R

EGULAR S

IZE

CANYON KIN

G SIZE

DISCRETIO

N KIN

G SIZE

GIPSY K

ING S

IZE

SMOKING K

ING S

IZE

DAKAR LIGHT K

ING S

IZE

GIPSY LI

GHT KIN

G SIZE

SELESTA LI

GHT KIN

G SIZE

TREMBLAY LI

GHT KIN

G SIZE

FINE LI

GHT REGULA

R SIZE

RODEO REGULA

R SIZE

SMOKING LI

GHT REGULA

R SIZE

Brand

Perc

enta

ge o

f med

ian

valu

e of

bra

nds

Figure A5.1Carbon monoxide per milligram of nicotine yield by brand as a percentage of themedian yield from brands reported to Heath Canada (2004)

0%

25%

50%

75%

100%

125%

150%

175%

200%

225%

250%

EXPORT 'A' U

LTRA LI

GHT REGULA

R SIZE

EXPORT 'A' M

EDIUM K

ING S

IZE

CRAVEN A E

XTRA LI

GHT KIN

G SIZE

EXPORT 'A' E

XTRA LI

GHT REGULA

R SIZE

EXPORT 'A' L

IGHT /

SPECIA

L LIG

HT KIN

G SIZE

MORE MENTH

OL 120

's

EXPORT 'A' E

XTRA LI

GHT KIN

G SIZE

EXPORT 'A' M

ILD R

EGULAR S

IZE

NUMBER 7 KIN

G SIZE

CAMEL LIG

HT KIN

G SIZE

EXPORT PLA

IN R

EGULAR S

IZE

WIN

STON LI

GHT 100

's

GAULOIS

ES BLO

NDES KIN

G SIZE

CAMEL KIN

G SIZE

WIN

STON K

ING S

IZE

DU MAURIE

R EXTR

A LIGHT R

EGULAR S

IZE

SALEM LI

GHT KIN

G SIZE

CAMEL PLA

IN R

EGULAR S

IZE

VISCOUNT E

XTRA M

ILD K

ING S

IZE

VANTAGE M

EDIUM 7

KING S

IZE

PLAYER'S LI

GHT REGULA

R SIZE

CRAVEN MENTH

OL KIN

G SIZE

BELMONT M

ILDS R

EGULAR S

IZE

PLAYER'S LI

GHT KIN

G SIZE

VANTAGE R

ICH 12

KIN

G SIZE

MARK TEN P

LAIN

KIN

G SIZE

DU MAURIE

R KIN

G SIZE

DU MAURIE

R EXTR

A LIGHT K

ING S

IZE

PLAYER'S S

PECIAL B

LEND R

EGULAR S

IZE

PLAYER'S E

XTRA LI

GHT KIN

G SIZE

CANYON REGULA

R SIZE

DISCRETIO

N REGULA

R SIZE

GIPSY R

EGULAR S

IZE

SMOKING R

EGULAR S

IZE

CANYON KIN

G SIZE

DISCRETIO

N KIN

G SIZE

GIPSY K

ING S

IZE

SMOKING K

ING S

IZE

DAKAR LIGHT K

ING S

IZE

GIPSY LI

GHT KIN

G SIZE

SELESTA

LIGHT K

ING S

IZE

TREMBLA

Y LIGHT K

ING S

IZE

FINE LI

GHT REGULA

R SIZE

RODEO REGULA

R SIZE

SMOKING LI

GHT REGULA

R SIZE

Brand

Per

cent

age

of m

edia

n va

lue

for

bran

ds

Figure A5.2Benzo[a]pyrene per milligram of nicotine yield by brand as a percentage of themedian yield from brands reported to Heath Canada (2004)

260

0%

25%

50%

75%

100%

125%

150%

175%

EXPORT 'A' U

LTRA LI

GHT REGULA

R SIZE

EXPORT 'A' M

EDIUM K

ING S

IZE

CRAVEN A E

XTRA LIGHT K

ING S

IZE

EXPORT 'A' E

XTRA LIGHT R

EGULAR S

IZE

EXPORT 'A' L

IGHT / S

PECIAL L

IGHT K

ING S

IZE

MORE MENTHOL 1

20's

EXPORT 'A' E

XTRA LIGHT K

ING S

IZE

EXPORT 'A' M

ILD R

EGULAR S

IZE

NUMBER 7 KIN

G SIZE

CAMEL LIG

HT KIN

G SIZE

EXPORT PLA

IN R

EGULAR S

IZE

WIN

STON LIGHT 10

0's

GAULOIS

ES BLO

NDES KIN

G SIZE

CAMEL KIN

G SIZE

WIN

STON KIN

G SIZE

DU MAURIE

R EXTRA LI

GHT REGULA

R SIZE

SALEM LI

GHT KIN

G SIZE

CAMEL PLA

IN R

EGULAR S

IZE

VISCOUNT E

XTRA MILD

KIN

G SIZE

VANTAGE MEDIU

M 7 KIN

G SIZE

PLAYER'S LI

GHT REGULA

R SIZE

CRAVEN MENTHOL K

ING S

IZE

BELMONT M

ILDS R

EGULAR S

IZE

PLAYER'S LI

GHT KIN

G SIZE

VANTAGE RIC

H 12 K

ING S

IZE

MARK TEN PLA

IN K

ING S

IZE

DU MAURIE

R KIN

G SIZE

DU MAURIE

R EXTRA LI

GHT KIN

G SIZE

PLAYER'S S

PECIAL B

LEND R

EGULAR S

IZE

PLAYER'S E

XTRA LIGHT K

ING S

IZE

CANYON REGULA

R SIZE

DISCRETIO

N REGULA

R SIZE

GIPSY R

EGULAR S

IZE

SMOKING R

EGULAR S

IZE

CANYON KIN

G SIZE

DISCRETIO

N KIN

G SIZE

GIPSY K

ING S

IZE

SMOKING K

ING S

IZE

DAKAR LIGHT K

ING S

IZE

GIPSY LI

GHT KIN

G SIZE

SELESTA LI

GHT KIN

G SIZE

TREMBLAY LI

GHT KIN

G SIZE

FINE LI

GHT REGULA

R SIZE

RODEO REGULA

R SIZE

SMOKING LI

GHT REGULA

R SIZE

Brand

Perc

enta

ge o

f med

ian

valu

e fo

r bra

nds

Figure A5.3Formaldehyde per milligram of nicotine yield by brand as a percentage of the medianyield for brands reported to Health Canada (2004)

0%

25%

50%

75%

100%

125%

150%

EXPORT 'A' U

LTRA LI

GHT REGULA

R SIZE

EXPORT 'A' M

EDIUM K

ING S

IZE

CRAVEN A E

XTRA LIGHT K

ING S

IZE

EXPORT 'A' E

XTRA LIGHT R

EGULAR S

IZE

EXPORT 'A' L

IGHT / S

PECIAL L

IGHT K

ING S

IZE

MORE MENTHOL 1

20's

EXPORT 'A' E

XTRA LIGHT K

ING S

IZE

EXPORT 'A' M

ILD R

EGULAR S

IZE

NUMBER 7 KIN

G SIZE

CAMEL LIG

HT KIN

G SIZE

EXPORT PLA

IN R

EGULAR S

IZE

WIN

STON LIGHT 10

0's

GAULOIS

ES BLO

NDES KIN

G SIZE

CAMEL KIN

G SIZE

WIN

STON KIN

G SIZE

DU MAURIE

R EXTRA LI

GHT REGULA

R SIZE

SALEM LI

GHT KIN

G SIZE

CAMEL PLA

IN R

EGULAR S

IZE

VISCOUNT E

XTRA MILD

KIN

G SIZE

VANTAGE MEDIU

M 7 KIN

G SIZE

PLAYER'S LI

GHT REGULA

R SIZE

CRAVEN MENTHOL K

ING S

IZE

BELMONT M

ILDS R

EGULAR S

IZE

PLAYER'S LI

GHT KIN

G SIZE

VANTAGE RIC

H 12 K

ING S

IZE

MARK TEN PLA

IN K

ING S

IZE

DU MAURIE

R KIN

G SIZE

DU MAURIE

R EXTRA LI

GHT KIN

G SIZE

PLAYER'S S

PECIAL B

LEND R

EGULAR S

IZE

PLAYER'S E

XTRA LIGHT K

ING S

IZE

CANYON REGULA

R SIZE

DISCRETIO

N REGULA

R SIZE

GIPSY R

EGULAR S

IZE

SMOKING R

EGULAR S

IZE

CANYON KIN

G SIZE

DISCRETIO

N KIN

G SIZE

GIPSY K

ING S

IZE

SMOKING K

ING S

IZE

DAKAR LIGHT K

ING S

IZE

GIPSY LI

GHT KIN

G SIZE

SELESTA LI

GHT KIN

G SIZE

TREMBLAY LI

GHT KIN

G SIZE

FINE LI

GHT REGULA

R SIZE

RODEO REGULA

R SIZE

SMOKING LI

GHT REGULA

R SIZE

Brand

Perc

enta

ge o

f med

ian

valu

e of

bra

nds

Figure A5.4Acetaldehyde yield per milligram of nicotine by brand as a percentage of the medianyield for brands reported to Health Canada (2004)

261

0%

25%

50%

75%

100%

125%

150%

EXPORT 'A' U

LTRA LI

GHT REGULA

R SIZE

EXPORT 'A' M

EDIUM K

ING S

IZE

CRAVEN A EXTRA LI

GHT KIN

G SIZE

EXPORT 'A' E

XTRA LIGHT REGULA

R SIZE

EXPORT 'A' L

IGHT / S

PECIAL L

IGHT KIN

G SIZE

MORE MENTHOL 1

20's

EXPORT 'A' E

XTRA LIGHT KIN

GSIZE

EXPORT 'A' M

ILD R

EGULAR SIZE

NUMBER7 K

ING SIZE

CAMEL LIG

HT KIN

G SIZE

EXPORT PLAIN

REGULA

RSIZE

WINSTON

LIGHT 10

0's

GAULOISES BLO

NDESKIN

G SIZE

CAMEL KIN

GSIZE

WINSTON KIN

G SIZE

DU MAURIE

R EXTRA LIGHT REGULA

R SIZE

SALEM LI

GHT KIN

GSIZE

CAMEL PLA

IN R

EGULAR SIZE

VISCOUNT E

XTRA MILD

KING S

IZE

VANTAGE MEDIUM 7

KING S

IZE

PLAYER'S

LIGHT REGULA

R SIZE

CRAVENMENTHOL K

ING

SIZE

BELMONT M

ILDS REGULA

R SIZE

PLAYER'S LI

GHT KING SIZE

VANTAGE RIC

H12

KIN

G SIZE

MARK TENPLA

IN KIN

G SIZE

DU MAURIE

R KIN

G SIZE

DUMAURIE

R EXTRA LI

GHT KIN

G SIZE

PLAYER'S S

PECIAL BLEND R

EGULAR SIZE

PLAYER'S

EXTRALIG

HT KIN

G SIZE

CANYON REGULA

RSIZE

DISCRETIO

NREGULA

R SIZE

GIPSY REGULAR

SIZE

SMOKING R

EGULAR SIZE

CANYON KIN

G SIZE

DISCRETION K

ING S

IZE

GIPSY KING S

IZE

SMOKING

KING SIZE

DAKAR LIGHT K

ING SIZE

GIPSY LIGHT KIN

G SIZE

SELESTA LIG

HT KING S

IZE

TREMBLAY

LIGHT KIN

G SIZE

FINE LI

GHT REGULAR S

IZE

RODEO REGULA

R SIZE

SMOKING LI

GHT REGULAR SIZE

Brands

Perc

enta

ge o

f med

ian

valu

e of

bra

nds

Figure A5.5Acrolein yield by brand per milligram of nicotine as a percentage of the median yieldfor brands reported to Health Canada (2004)

0%

50%

100%

150%

200%

250%

300%

350%

400%

450%

500%

550%

600%

650%

EXPORT 'A' U

LTRA LI

GHT REGULA

R SIZE

EXPORT 'A' M

EDIUM K

ING S

IZE

CRAVEN A E

XTRA LIGHT K

ING S

IZE

EXPORT 'A' E

XTRA LIGHT R

EGULAR S

IZE

EXPORT 'A' L

IGHT / S

PECIAL L

IGHT K

ING S

IZE

MORE MENTHOL 1

20's

EXPORT 'A' E

XTRA LIGHT K

ING S

IZE

EXPORT 'A' M

ILD R

EGULAR S

IZE

NUMBER 7 KIN

G SIZE

CAMEL LIG

HT KIN

G SIZE

EXPORT PLA

IN R

EGULAR S

IZE

WIN

STON LIGHT 10

0's

GAULOIS

ES BLO

NDES KIN

G SIZE

CAMEL KIN

G SIZE

WIN

STON KIN

G SIZE

DU MAURIE

R EXTRA LI

GHT REGULA

R SIZE

SALEM LI

GHT KIN

G SIZE

CAMEL PLA

IN R

EGULAR S

IZE

VISCOUNT E

XTRA MILD

KIN

G SIZE

VANTAGE MEDIU

M 7 KIN

G SIZE

PLAYER'S LI

GHT REGULA

R SIZE

CRAVEN MENTHOL K

ING S

IZE

BELMONT M

ILDS R

EGULAR S

IZE

PLAYER'S LI

GHT KIN

G SIZE

VANTAGE RIC

H 12 K

ING S

IZE

MARK TEN PLA

IN K

ING S

IZE

DU MAURIE

R KIN

G SIZE

DU MAURIE

R EXTRA LI

GHT KIN

G SIZE

PLAYER'S S

PECIAL B

LEND R

EGULAR S

IZE

PLAYER'S E

XTRA LIGHT K

ING S

IZE

CANYON REGULA

R SIZE

DISCRETIO

N REGULA

R SIZE

GIPSY R

EGULAR S

IZE

SMOKING R

EGULAR S

IZE

CANYON KIN

G SIZE

DISCRETIO

N KIN

G SIZE

GIPSY K

ING S

IZE

SMOKING K

ING S

IZE

DAKAR LIGHT K

ING S

IZE

GIPSY LI

GHT KIN

G SIZE

SELESTA LI

GHT KIN

G SIZE

TREMBLAY LI

GHT KIN

G SIZE

FINE LI

GHT REGULA

R SIZE

RODEO REGULA

R SIZE

SMOKING LI

GHT REGULA

R SIZE

Brand

Perc

enta

ge o

f med

ian

valu

e fo

r bra

nds

Figure A5.6NNN yield by brand per milligram of nicotine as a percentage of the median yield forbrands reported to Health Canada (2004)NNN, N´-nitrosonornicotine

262

0,000

0,250

0,500

0,750

1,000

1,250

1,500

1,750

2,000

2,250

EXPORT 'A' U

LTRA LI

GHT REGULA

R SIZE

EXPORT 'A' M

EDIUM K

ING S

IZE

CRAVEN A E

XTRA LIGHT K

ING S

IZE

EXPORT 'A' E

XTRA LIGHT R

EGULAR S

IZE

EXPORT 'A' L

IGHT / S

PECIAL L

IGHT K

ING S

IZE

MORE MENTHOL 1

20's

EXPORT 'A' E

XTRA LIGHT K

ING S

IZE

EXPORT 'A' M

ILD R

EGULAR S

IZE

NUMBER 7 KIN

G SIZE

CAMEL LIG

HT KIN

G SIZE

EXPORT PLA

IN R

EGULAR S

IZE

WIN

STON LIGHT 10

0's

GAULOIS

ES BLO

NDES KIN

G SIZE

CAMEL KIN

G SIZE

WIN

STON KIN

G SIZE

DU MAURIE

R EXTRA LI

GHT REGULA

R SIZE

SALEM LI

GHT KIN

G SIZE

CAMEL PLA

IN R

EGULAR S

IZE

VISCOUNT E

XTRA MILD

KIN

G SIZE

VANTAGE MEDIU

M 7 KIN

G SIZE

PLAYER'S LI

GHT REGULA

R SIZE

CRAVEN MENTHOL K

ING S

IZE

BELMONT M

ILDS R

EGULAR S

IZE

PLAYER'S LI

GHT KIN

G SIZE

VANTAGE RIC

H 12 K

ING S

IZE

MARK TEN PLA

IN K

ING S

IZE

DU MAURIE

R KIN

G SIZE

DU MAURIE

R EXTRA LI

GHT KIN

G SIZE

PLAYER'S S

PECIAL B

LEND R

EGULAR S

IZE

PLAYER'S E

XTRA LIGHT K

ING S

IZE

CANYON REGULA

R SIZE

DISCRETIO

N REGULA

R SIZE

GIPSY R

EGULAR S

IZE

SMOKING R

EGULAR S

IZE

CANYON KIN

G SIZE

DISCRETIO

N KIN

G SIZE

GIPSY K

ING S

IZE

SMOKING K

ING S

IZE

DAKAR LIGHT K

ING S

IZE

GIPSY LI

GHT KIN

G SIZE

SELESTA LI

GHT KIN

G SIZE

TREMBLAY LI

GHT KIN

G SIZE

FINE LI

GHT REGULA

R SIZE

RODEO REGULA

R SIZE

SMOKING LI

GHT REGULA

R SIZE

Brand

Perc

enta

ge o

f med

ian

valu

e of

bra

nds

Figure A5.7NNK yield by brand per milligram of nicotine as a percentage of the median yield forbrands reported to Health Canada (2004)NNK, 4-(N-nitrosomethylamino)-1-(3-pyridyl)-1-butanone

0%

25%

50%

75%

100%

125%

150%

EXPORT 'A' U

LTRA LI

GHT REGULA

R SIZE

EXPORT 'A' M

EDIUM K

ING S

IZE

CRAVEN A E

XTRA LIGHT K

ING S

IZE

EXPORT 'A' E

XTRA LIGHT R

EGULAR S

IZE

EXPORT 'A' L

IGHT / S

PECIAL L

IGHT K

ING S

IZE

MORE MENTHOL 1

20's

EXPORT 'A' E

XTRA LIGHT K

ING S

IZE

EXPORT 'A' M

ILD R

EGULAR S

IZE

NUMBER 7 KIN

G SIZE

CAMEL LIG

HT KIN

G SIZE

EXPORT PLA

IN R

EGULAR S

IZE

WIN

STON LIGHT 10

0's

GAULOIS

ES BLO

NDES KIN

G SIZE

CAMEL KIN

G SIZE

WIN

STON KIN

G SIZE

DU MAURIE

R EXTRA LI

GHT REGULA

R SIZE

SALEM LI

GHT KIN

G SIZE

CAMEL PLA

IN R

EGULAR S

IZE

VISCOUNT E

XTRA MILD

KIN

G SIZE

VANTAGE MEDIU

M 7 KIN

G SIZE

PLAYER'S LI

GHT REGULA

R SIZE

CRAVEN MENTHOL K

ING S

IZE

BELMONT M

ILDS R

EGULAR S

IZE

PLAYER'S LI

GHT KIN

G SIZE

VANTAGE RIC

H 12 K

ING S

IZE

MARK TEN PLA

IN K

ING S

IZE

DU MAURIE

R KIN

G SIZE

DU MAURIE

R EXTRA LI

GHT KIN

G SIZE

PLAYER'S S

PECIAL B

LEND R

EGULAR S

IZE

PLAYER'S E

XTRA LIGHT K

ING S

IZE

CANYON REGULA

R SIZE

DISCRETIO

N REGULA

R SIZE

GIPSY R

EGULAR S

IZE

SMOKING R

EGULAR S

IZE

CANYON KIN

G SIZE

DISCRETIO

N KIN

G SIZE

GIPSY K

ING S

IZE

SMOKING K

ING S

IZE

DAKAR LIGHT K

ING S

IZE

GIPSY LI

GHT KIN

G SIZE

SELESTA LI

GHT KIN

G SIZE

TREMBLAY LI

GHT KIN

G SIZE

FINE LI

GHT REGULA

R SIZE

RODEO REGULA

R SIZE

SMOKING LI

GHT REGULA

R SIZE

Brand

Perc

enta

ge o

f med

ian

valu

e of

bra

nds

Figure A5.81,3-Butadiene yield by brand per milligram of nicotine as a percentage of the medianyield for brands reported to Health Canada (2004)

263

0%

25%

50%

75%

100%

125%

150%

EXPORT 'A' U

LTRA LI

GHT REGULA

R SIZE

EXPORT 'A' M

EDIUM K

ING S

IZE

CRAVEN A E

XTRA LIGHT K

ING S

IZE

EXPORT 'A' E

XTRA LIGHT R

EGULAR S

IZE

EXPORT 'A' L

IGHT / S

PECIAL L

IGHT K

ING S

IZE

MORE MENTHOL 1

20's

EXPORT 'A' E

XTRA LIGHT K

ING S

IZE

EXPORT 'A' M

ILD R

EGULAR S

IZE

NUMBER 7 KIN

G SIZE

CAMEL LIG

HT KIN

G SIZE

EXPORT PLA

IN R

EGULAR S

IZE

WIN

STON LIGHT 10

0's

GAULOIS

ES BLO

NDES KIN

G SIZE

CAMEL KIN

G SIZE

WIN

STON KIN

G SIZE

DU MAURIE

R EXTRA LI

GHT REGULA

R SIZE

SALEM LI

GHT KIN

G SIZE

CAMEL PLA

IN R

EGULAR S

IZE

VISCOUNT E

XTRA MILD

KIN

G SIZE

VANTAGE MEDIU

M 7 KIN

G SIZE

PLAYER'S LI

GHT REGULA

R SIZE

CRAVEN MENTHOL K

ING S

IZE

BELMONT M

ILDS R

EGULAR S

IZE

PLAYER'S LI

GHT KIN

G SIZE

VANTAGE RIC

H 12 K

ING S

IZE

MARK TEN PLA

IN K

ING S

IZE

DU MAURIE

R KIN

G SIZE

DU MAURIE

R EXTRA LI

GHT KIN

G SIZE

PLAYER'S S

PECIAL B

LEND R

EGULAR S

IZE

PLAYER'S E

XTRA LIGHT K

ING S

IZE

CANYON REGULA

R SIZE

DISCRETIO

N REGULA

R SIZE

GIPSY R

EGULAR S

IZE

SMOKING R

EGULAR S

IZE

CANYON KIN

G SIZE

DISCRETIO

N KIN

G SIZE

GIPSY K

ING S

IZE

SMOKING K

ING S

IZE

DAKAR LIGHT K

ING S

IZE

GIPSY LI

GHT KIN

G SIZE

SELESTA LI

GHT KIN

G SIZE

TREMBLAY LI

GHT KIN

G SIZE

FINE LI

GHT REGULA

R SIZE

RODEO REGULA

R SIZE

SMOKING LI

GHT REGULA

R SIZE

Brand

Perc

enta

ge o

f med

ian

valu

e of

bra

nds

Figure A5.9Benzene yield by brand per milligram of nicotine as a percentage of the median yieldfor brands reported to Health Canada (2004)

0

0,25

0,5

0,75

1

1,25

1,5

1,75

2

Mar

lbor

o LS

F H

P/A

rgen

tina

Mar

lbor

o LS

F H

P/V

enez

uela

SG

Ven

til R

egul

ar F

SP

/EU

Pet

ra R

egul

ar F

HP

/CE

MA

Mar

lbor

o K

S F

SP

/US

Mar

lbor

o K

S F

HP

/Nor

way

L &

M K

S F

HP

/EU

Mar

lbor

o K

S F

HP

/Mal

aysi

aC

hest

erfie

ld O

rigin

als

KS

F H

PL

& M

KS

F H

P/M

alay

sia

Mar

lbor

o K

S F

HP

25_

s/A

ustra

liM

arlb

oro

KS

F H

P/T

aiw

anM

arlb

oro

100

F H

P/E

UM

arlb

oro

KS

F H

P M

ediu

m/E

UP

arlia

men

t 100

F S

P L

t/US

Mar

lbor

o K

S F

HP

/Jap

anL

& M

KS

F H

P L

t/EU

F6 K

S F

HP

Lt/E

UC

hest

erfie

ld O

rigin

als

KS

F H

PD

iana

KS

F S

P S

peci

ally

Mild

/EM

urat

ti A

mba

ssad

or K

S F

HP

/EU

Mer

it K

S F

HP

/EU

Par

liam

ent 1

00 F

SP

/CE

MA

Mar

lbor

o K

S F

HP

Lt/G

erm

any/

UK

Mar

lbor

o K

S F

HP

Lt/J

apan

Mar

lbor

o 10

0 F

HP

Lt/G

erm

any

Mer

it K

S F

SP

Ult/

US

Mar

lbor

o K

S F

HP

Ult

Men

/US

Par

liam

ent K

S F

HP

Lt/J

apan

Virg

inia

Slim

s 10

0 F

HP

Ult

Me

Che

ster

field

INTL

KS

F H

P U

lt/P

hilip

Mor

ris 1

00 F

HP

Sup

er L

Mar

lbor

o K

S F

HP

Ult/

EU

Dia

na K

S F

HP

Ult/

EU

Virg

inia

Slim

s 10

0 F

HP

Ult

Me

Phi

lip M

orris

One

KS

F H

P/E

UM

urat

ti K

S F

HP

Ult

1 m

g/C

EM

ALo

ngbe

ach

One

KS

F H

P/A

ustra

liV

irgin

ia S

lims

100

F H

P M

en 1

Mar

lbor

o K

S F

HP

/Mex

ico

Raf

fles

100

F H

P/E

UM

arlb

oro

KS

F H

P/B

razi

lP

eter

Jac

kson

KS

F H

P M

en/A

ust

Mar

lbor

o 10

0 F

HP

Lt/U

SM

arlb

oro

KS

F H

P L

t/Nor

way

Che

ster

field

KS

F H

P L

t/EU

Phi

lip M

orris

KS

F H

P S

uper

Li

Mer

it K

S F

SP

Ulti

ma/

US

1R4F

Ken

tuck

y R

efer

ence

1R4F

Ken

tuck

y R

efer

ence

Brand

Perc

enta

ge o

f med

ian

valu

e of

bra

nds

Figure A5.10Carbon monoxide per milligrams of nictone yield as a percentage of the median yieldfor Philip Morris International Brands

264

Mar

lbor

o LS

F H

P/Ar

gent

ina

Mar

lbor

o LS

F H

P/Ve

nezu

ela

SG V

entil

Reg

ular

F S

P/E

UPe

tra R

egul

ar F

HP/

CEM

AM

arlb

oro

KS F

SP/

US

Mar

lbor

o KS

F H

P/N

orw

ayL

& M

KS

F H

P/EU

Mar

lbor

o KS

F H

P/M

alay

sia

Che

ster

field

Orig

inal

s K

S F

HP

L &

M K

S F

HP/

Mal

aysi

aM

arlb

oro

KS F

HP

25_

s/Au

stra

liM

arlb

oro

KS F

HP/

Taiw

anM

arlb

oro

100

F H

P/EU

Mar

lbor

o KS

F H

P M

ediu

m/E

UPa

rliam

ent 1

00 F

SP

Lt/U

SM

arlb

oro

KS F

HP/

Japa

nL

& M

KS

F H

P Lt

/EU

F6 K

S F

HP

Lt/E

UC

hest

erfie

ld O

rigin

als

KS F

HP

Dia

na K

S F

SP S

peci

ally

Mild

/EM

urat

ti Am

bass

ador

KS

F H

P/EU

Mer

it KS

F H

P/EU

Parli

amen

t 100

F S

P/C

EMA

Mar

lbor

o KS

F H

P L

t/Ger

man

y/U

KM

arlb

oro

KS F

HP

Lt/J

apan

Mar

lbor

o 10

0 F

HP

Lt/G

erm

any

Mer

it KS

F S

P U

lt/U

SM

arlb

oro

KS F

HP

Ult

Men

/US

Parli

amen

t KS

F H

P Lt

/Jap

anVi

rgin

ia S

lims

100

F H

P U

lt M

eC

hest

erfie

ld IN

TL K

S F

HP

Ult/

Philip

Mor

ris 1

00 F

HP

Sup

er L

Mar

lbor

o KS

F H

P U

lt/EU

Dia

na K

S F

HP

Ult/

EU

Virg

inia

Slim

s 10

0 F

HP

Ult

Me

Philip

Mor

ris O

ne K

S F

HP/

EUM

urat

ti KS

F H

P U

lt 1

mg/

CEM

ALo

ngbe

ach

One

KS

F H

P/Au

stra

liVi

rgin

ia S

lims

100

F H

P M

en 1

Mar

lbor

o KS

F H

P/M

exic

oR

affle

s 10

0 F

HP

/EU

Mar

lbor

o KS

F H

P/B

razi

lPe

ter J

acks

on K

S F

HP

Men

/Aus

tM

arlb

oro

100

F H

P Lt

/US

Mar

lbor

o KS

F H

P L

t/Nor

way

Che

ster

field

KS

F H

P Lt

/EU

Philip

Mor

ris K

S F

HP

Supe

r Li

Mer

it KS

F S

P U

ltim

a/U

S1R

4F K

entu

cky

Ref

eren

ce1R

4F K

entu

cky

Ref

eren

ce

Brand

0

0,25

0,5

0,75

1

1,25

1,5

1,75

Perc

enta

ge o

f med

ian

valu

e of

bra

nds

Figure A5.11Acetaldehyde per milligrams of nictone yield as a percentage of the median yield forPhilip Morris international brands

Mar

lbor

o LS

F H

P/A

rgen

tina

Mar

lbor

o LS

F H

P/V

enez

uela

SG

Ven

til R

egul

ar F

SP

/EU

Pet

ra R

egul

ar F

HP

/CE

MA

Mar

lbor

o K

S F

SP

/US

Mar

lbor

o K

S F

HP

/Nor

way

L &

M K

S F

HP

/EU

Mar

lbor

o K

S F

HP

/Mal

aysi

aC

hest

erfie

ld O

rigin

als

KS

F H

PL

& M

KS

F H

P/M

alay

sia

Mar

lbor

o K

S F

HP

25_

s/A

ustra

liM

arlb

oro

KS

F H

P/T

aiw

anM

arlb

oro

100

F H

P/E

UM

arlb

oro

KS

F H

P M

ediu

m/E

UP

arlia

men

t 100

F S

P L

t/US

Mar

lbor

o K

S F

HP

/Jap

anL

& M

KS

F H

P L

t/EU

F6 K

S F

HP

Lt/E

UC

hest

erfie

ld O

rigin

als

KS

F H

PD

iana

KS

F S

P S

peci

ally

Mild

/EM

urat

ti A

mba

ssad

or K

S F

HP

/EU

Mer

it K

S F

HP

/EU

Par

liam

ent 1

00 F

SP

/CE

MA

Mar

lbor

o K

S F

HP

Lt/G

erm

any/

UK

Mar

lbor

o K

S F

HP

Lt/J

apan

Mar

lbor

o 10

0 F

HP

Lt/G

erm

any

Mer

it K

S F

SP

Ult/

US

Mar

lbor

o K

S F

HP

Ult

Men

/US

Par

liam

ent K

S F

HP

Lt/J

apan

Virg

inia

Slim

s 10

0 F

HP

Ult

Me

Che

ster

field

INTL

KS

F H

P U

lt/P

hilip

Mor

ris 1

00 F

HP

Sup

er L

Mar

lbor

o K

S F

HP

Ult/

EU

Dia

na K

S F

HP

Ult/

EU

Virg

inia

Slim

s 10

0 F

HP

Ult

Me

Phi

lip M

orris

One

KS

F H

P/E

UM

urat

ti K

S F

HP

Ult

1 m

g/C

EM

ALo

ngbe

ach

One

KS

F H

P/A

ustra

liV

irgin

ia S

lims

100

F H

P M

en 1

Mar

lbor

o K

S F

HP

/Mex

ico

Raf

fles

100

F H

P/E

UM

arlb

oro

KS

F H

P/B

razi

lP

eter

Jac

kson

KS

F H

P M

en/A

ust

Mar

lbor

o 10

0 F

HP

Lt/U

SM

arlb

oro

KS

F H

P L

t/Nor

way

Che

ster

field

KS

F H

P L

t/EU

Phi

lip M

orris

KS

F H

P S

uper

Li

Mer

it K

S F

SP

Ulti

ma/

US

1R4F

Ken

tuck

y R

efer

ence

1R4F

Ken

tuck

y R

efer

ence

Brand

0

0,25

0,5

0,75

1

1,25

1,5

1,75

Perc

enta

ge o

f med

ian

valu

e of

bra

nds

Figure A5.12Acrolein per milligrams of nictone yield as a percentage of the median yield for PhilipMorris international brands

265

Mar

lbor

o LS

F H

P/A

rgen

tina

Mar

lbor

o LS

F H

P/V

enez

uela

SG

Ven

til R

egul

ar F

SP

/EU

Pet

ra R

egul

ar F

HP

/CE

MA

Mar

lbor

o K

S F

SP

/US

Mar

lbor

o K

S F

HP

/Nor

way

L &

M K

S F

HP

/EU

Mar

lbor

o K

S F

HP

/Mal

aysi

aC

hest

erfie

ld O

rigin

als

KS

F H

PL

& M

KS

F H

P/M

alay

sia

Mar

lbor

o K

S F

HP

25_

s/A

ustra

liM

arlb

oro

KS

F H

P/T

aiw

anM

arlb

oro

100

F H

P/E

UM

arlb

oro

KS

F H

P M

ediu

m/E

UP

arlia

men

t 100

F S

P L

t/US

Mar

lbor

o K

S F

HP

/Jap

anL

& M

KS

F H

P L

t/EU

F6 K

S F

HP

Lt/E

UC

hest

erfie

ld O

rigin

als

KS

F H

PD

iana

KS

F S

P S

peci

ally

Mild

/EM

urat

ti A

mba

ssad

or K

S F

HP

/EU

Mer

it K

S F

HP

/EU

Par

liam

ent 1

00 F

SP

/CE

MA

Mar

lbor

o K

S F

HP

Lt/G

erm

any/

UK

Mar

lbor

o K

S F

HP

Lt/J

apan

Mar

lbor

o 10

0 F

HP

Lt/G

erm

any

Mer

it K

S F

SP

Ult/

US

Mar

lbor

o K

S F

HP

Ult

Men

/US

Par

liam

ent K

S F

HP

Lt/J

apan

Virg

inia

Slim

s 10

0 F

HP

Ult

Me

Che

ster

field

INTL

KS

F H

P U

lt/P

hilip

Mor

ris 1

00 F

HP

Sup

er L

Mar

lbor

o K

S F

HP

Ult/

EU

Dia

na K

S F

HP

Ult/

EU

Virg

inia

Slim

s 10

0 F

HP

Ult

Me

Phi

lip M

orris

One

KS

F H

P/E

UM

urat

ti K

S F

HP

Ult

1 m

g/C

EM

ALo

ngbe

ach

One

KS

F H

P/A

ustra

liV

irgin

ia S

lims

100

F H

P M

en 1

Mar

lbor

o K

S F

HP

/Mex

ico

Raf

fles

100

F H

P/E

UM

arlb

oro

KS

F H

P/B

razi

lP

eter

Jac

kson

KS

F H

P M

en/A

ust

Mar

lbor

o 10

0 F

HP

Lt/U

SM

arlb

oro

KS

F H

P L

t/Nor

way

Che

ster

field

KS

F H

P L

t/EU

Phi

lip M

orris

KS

F H

P S

uper

Li

Mer

it K

S F

SP

Ulti

ma/

US

1R4F

Ken

tuck

y R

efer

ence

1R4F

Ken

tuck

y R

efer

ence

Brand

0

0,25

0,5

0,75

1

1,25

1,5

1,75

2

2,25

2,5

2,75

Perc

enta

ge o

f med

ian

valu

e of

bra

nds

Figure A5.13Formaldehyde per milligrams of nictone yield as a percentage of the median yield forPhilip Morris international brands

Mar

lbor

o LS

F H

P/Ar

gent

ina

Mar

lbor

o LS

F H

P/V

enez

uela

SG

Ven

til R

egul

ar F

SP

/EU

Pet

ra R

egul

ar F

HP

/CE

MA

Mar

lbor

o KS

F S

P/U

SM

arlb

oro

KS

F H

P/N

orw

ayL

& M

KS

F H

P/EU

Mar

lbor

o K

S F

HP

/Mal

aysi

aC

hest

erfie

ld O

rigin

als

KS F

HP

L &

M K

S F

HP/

Mal

aysi

aM

arlb

oro

KS F

HP

25_s

/Aus

trali

Mar

lbor

o KS

F H

P/Ta

iwan

Mar

lbor

o 10

0 F

HP/

EUM

arlb

oro

KS F

HP

Med

ium

/EU

Parli

amen

t 100

F S

P Lt

/US

Mar

lbor

o KS

F H

P/Ja

pan

L &

M K

S F

HP

Lt/E

UF6

KS

F H

P Lt

/EU

Che

ster

field

Orig

inal

s KS

F H

PD

iana

KS

F SP

Spe

cial

ly M

ild/E

Mur

atti

Amba

ssad

or K

S F

HP/

EUM

erit

KS F

HP/

EUPa

rliam

ent 1

00 F

SP/

CEM

AM

arlb

oro

KS F

HP

Lt/G

erm

any/

UK

Mar

lbor

o KS

F H

P Lt

/Jap

anM

arlb

oro

100

F H

P Lt

/Ger

man

yM

erit

KS F

SP

Ult/

US

Mar

lbor

o KS

F H

P U

lt M

en/U

SPa

rliam

ent K

S F

HP

Lt/J

apan

Virg

inia

Slim

s 10

0 F

HP

Ult

Me

Che

ster

field

INTL

KS

F H

P U

lt/Ph

ilip M

orris

100

F H

P Su

per L

Mar

lbor

o KS

F H

P U

lt/EU

Dia

na K

S F

HP

Ult/

EUVi

rgin

ia S

lims

100

F H

P U

lt M

ePh

ilip M

orris

One

KS

F H

P/EU

Mur

atti

KS F

HP

Ult

1 m

g/C

EMA

Long

beac

h O

ne K

S F

HP/

Aust

rali

Virg

inia

Slim

s 10

0 F

HP

Men

1M

arlb

oro

KS F

HP/

Mex

ico

Raf

fles

100

F H

P/EU

Mar

lbor

o KS

F H

P/Br

azil

Pete

r Jac

kson

KS

F H

P M

en/A

ust

Mar

lbor

o 10

0 F

HP

Lt/U

SM

arlb

oro

KS F

HP

Lt/N

orw

ayC

hest

erfie

ld K

S F

HP

Lt/E

UPh

ilip M

orris

KS

F H

P Su

per L

iM

erit

KS F

SP

Ulti

ma/

US

1R4F

Ken

tuck

y R

efer

ence

1R4F

Ken

tuck

y R

efer

ence

Brand

0

0,25

0,5

0,75

1

1,25

1,5

Perc

enta

ge o

f med

ian

valu

e of

bra

nds

Figure A5.14Benzene per milligrams of nictone yield as a percentage of the median yield for PhilipMorris international brands

266

Mar

lbor

o LS

F H

P/Ar

gent

ina

Mar

lbor

o LS

F H

P/Ve

nezu

ela

SG V

entil

Reg

ular

F S

P/E

UPe

tra R

egul

ar F

HP/

CEM

AM

arlb

oro

KS F

SP/

US

Mar

lbor

o KS

F H

P/N

orw

ayL

& M

KS

F H

P/EU

Mar

lbor

o KS

F H

P/M

alay

sia

Che

ster

field

Orig

inal

s KS

F H

PL

& M

KS

F H

P/M

alay

sia

Mar

lbor

o KS

F H

P 25

_s/A

ustra

liM

arlb

oro

KS F

HP/

Taiw

anM

arlb

oro

100

F H

P/EU

Mar

lbor

o KS

F H

P M

ediu

m/E

UPa

rliam

ent 1

00 F

SP

Lt/U

SM

arlb

oro

KS F

HP

/Jap

anL

& M

KS

F H

P Lt

/EU

F6 K

S F

HP

Lt/E

UC

hest

erfie

ld O

rigin

als

KS F

HP

Dia

na K

S F

SP S

peci

ally

Mild

/EM

urat

ti Am

bass

ador

KS

F H

P/EU

Mer

it KS

F H

P/EU

Parli

amen

t 100

F S

P/C

EMA

Mar

lbor

o KS

F H

P L

t/Ger

man

y/U

KM

arlb

oro

KS F

HP

Lt/J

apan

Mar

lbor

o 10

0 F

HP

Lt/G

erm

any

Mer

it KS

F S

P U

lt/U

SM

arlb

oro

KS F

HP

Ult

Men

/US

Parli

amen

t KS

F H

P Lt

/Jap

anVi

rgin

ia S

lims

100

F H

P U

lt M

eC

hest

erfie

ld IN

TL K

S F

HP

Ult/

Philip

Mor

ris 1

00 F

HP

Sup

er L

Mar

lbor

o KS

F H

P U

lt/EU

Dia

na K

S F

HP

Ult/

EU

Virg

inia

Slim

s 10

0 F

HP

Ult

Me

Philip

Mor

ris O

ne K

S F

HP/

EUM

urat

ti KS

F H

P U

lt 1

mg/

CEM

ALo

ngbe

ach

One

KS

F H

P/Au

stra

liVi

rgin

ia S

lims

100

F H

P M

en 1

Mar

lbor

o KS

F H

P/M

exic

oR

affle

s 10

0 F

HP

/EU

Mar

lbor

o KS

F H

P/Br

azil

Pete

r Jac

kson

KS

F H

P M

en/A

ust

Mar

lbor

o 10

0 F

HP

Lt/U

SM

arlb

oro

KS F

HP

Lt/N

orw

ayC

hest

erfie

ld K

S F

HP

Lt/E

UPh

ilip M

orris

KS

F H

P Su

per L

iM

erit

KS F

SP

Ulti

ma/

US

1R4F

Ken

tuck

y R

efer

ence

1R4F

Ken

tuck

y R

efer

ence

Brand

0

0,25

0,5

0,75

1

1,25

1,5

Perc

enta

ge o

f med

ian

valu

e of

bra

nds

Figure A5.151,3-Butadiene per milligrams of nictone yield as a percentage of the median yield forPhilip Morris international brands

Mar

lbor

o LS

F H

P/Ar

gent

ina

Mar

lbor

o LS

F H

P/Ve

nezu

ela

SG V

entil

Reg

ular

F S

P/E

UPe

tra R

egul

ar F

HP/

CEM

AM

arlb

oro

KS F

SP/

US

Mar

lbor

o KS

F H

P/N

orw

ayL

& M

KS

F H

P/EU

Mar

lbor

o KS

F H

P/M

alay

sia

Che

ster

field

Orig

inal

s KS

F H

PL

& M

KS

F H

P/M

alay

sia

Mar

lbor

o KS

F H

P 2

5_s/

Aust

rali

Mar

lbor

o KS

F H

P/T

aiw

anM

arlb

oro

100

F H

P/EU

Mar

lbor

o KS

F H

P M

ediu

m/E

UPa

rliam

ent 1

00 F

SP

Lt/U

SM

arlb

oro

KS F

HP/

Japa

nL

& M

KS

F H

P Lt

/EU

F6 K

S F

HP

Lt/E

UC

hest

erfie

ld O

rigin

als

KS F

HP

Dia

na K

S F

SP S

peci

ally

Mild

/EM

urat

ti Am

bass

ador

KS

F H

P/EU

Mer

it KS

F H

P/EU

Parli

amen

t 100

F S

P/C

EMA

Mar

lbor

o KS

F H

P L

t/Ger

man

y/U

KM

arlb

oro

KS F

HP

Lt/J

apan

Mar

lbor

o 10

0 F

HP

Lt/G

erm

any

Mer

it KS

F S

P U

lt/U

SM

arlb

oro

KS F

HP

Ult

Men

/US

Parli

amen

t KS

F H

P Lt

/Jap

anVi

rgin

ia S

lims

100

F H

P U

lt M

eC

hest

erfie

ld IN

TL K

S F

HP

Ult/

Philip

Mor

ris 1

00 F

HP

Supe

r LM

arlb

oro

KS F

HP

Ult/

EUD

iana

KS

F H

P U

lt/E

UVi

rgin

ia S

lims

100

F H

P U

lt M

ePh

ilip M

orris

One

KS

F H

P/EU

Mur

atti

KS F

HP

Ult

1 m

g/C

EMA

Long

beac

h O

ne K

S F

HP/

Aust

rali

Virg

inia

Slim

s 10

0 F

HP

Men

1M

arlb

oro

KS F

HP/

Mex

ico

Raf

fles

100

F H

P/E

UM

arlb

oro

KS F

HP

/Bra

zil

Pete

r Jac

kson

KS

F H

P M

en/A

ust

Mar

lbor

o 10

0 F

HP

Lt/U

SM

arlb

oro

KS F

HP

Lt/N

orw

ayC

hest

erfie

ld K

S F

HP

Lt/E

UPh

ilip M

orris

KS

F H

P Su

per L

iM

erit

KS F

SP

Ulti

ma/

US

1R4F

Ken

tuck

y R

efer

ence

1R4F

Ken

tuck

y R

efer

ence

Brand

0

0,25

0,5

0,75

1

1,25

1,5

1,75

Perc

enta

ge o

f med

ian

valu

e of

bra

nds

Figure A5.16NNN per milligrams of nictone yield as a percentage of the median yield for PhilipMorris international brandsNNN, N´-nitrosonornicotine

267

Mar

lbor

o LS

F H

P/Ar

gent

ina

Mar

lbor

o LS

F H

P/Ve

nezu

ela

SG V

entil

Reg

ular

F S

P/E

UPe

tra R

egul

ar F

HP/

CEM

AM

arlb

oro

KS F

SP/

US

Mar

lbor

o KS

F H

P/N

orw

ayL

& M

KS

F H

P/EU

Mar

lbor

o KS

F H

P/M

alay

sia

Che

ster

field

Orig

inal

s KS

F H

PL

& M

KS

F H

P/M

alay

sia

Mar

lbor

o KS

F H

P 2

5_s/

Aust

rali

Mar

lbor

o KS

F H

P/T

aiw

anM

arlb

oro

100

F H

P/EU

Mar

lbor

o KS

F H

P M

ediu

m/E

UPa

rliam

ent 1

00 F

SP

Lt/U

SM

arlb

oro

KS F

HP/

Japa

nL

& M

KS

F H

P Lt

/EU

F6 K

S F

HP

Lt/E

UC

hest

erfie

ld O

rigin

als

KS F

HP

Dia

na K

S F

SP S

peci

ally

Mild

/EM

urat

ti Am

bass

ador

KS

F H

P/EU

Mer

it KS

F H

P/EU

Parli

amen

t 100

F S

P/C

EMA

Mar

lbor

o KS

F H

P L

t/Ger

man

y/U

KM

arlb

oro

KS F

HP

Lt/J

apan

Mar

lbor

o 10

0 F

HP

Lt/G

erm

any

Mer

it KS

F S

P U

lt/U

SM

arlb

oro

KS F

HP

Ult

Men

/US

Parli

amen

t KS

F H

P Lt

/Jap

anVi

rgin

ia S

lims

100

F H

P U

lt M

eC

hest

erfie

ld IN

TL K

S F

HP

Ult/

Philip

Mor

ris 1

00 F

HP

Sup

er L

Mar

lbor

o KS

F H

P U

lt/EU

Dia

na K

S F

HP

Ult/

EU

Virg

inia

Slim

s 10

0 F

HP

Ult

Me

Philip

Mor

ris O

ne K

S F

HP/

EUM

urat

ti KS

F H

P U

lt 1

mg/

CEM

ALo

ngbe

ach

One

KS

F H

P/Au

stra

liVi

rgin

ia S

lims

100

F H

P M

en 1

Mar

lbor

o KS

F H

P/M

exic

oR

affle

s 10

0 F

HP

/EU

Mar

lbor

o KS

F H

P/B

razi

lPe

ter J

acks

on K

S F

HP

Men

/Aus

tM

arlb

oro

100

F H

P Lt

/US

Mar

lbor

o KS

F H

P L

t/Nor

way

Che

ster

field

KS

F H

P Lt

/EU

Philip

Mor

ris K

S F

HP

Supe

r Li

Mer

it KS

F S

P U

ltim

a/U

S1R

4F K

entu

cky

Ref

eren

ce1R

4F K

entu

cky

Ref

eren

ce

Brand

0

0,25

0,5

0,75

1

1,25

1,5

1,75

Perc

enta

ge o

f med

ian

valu

e of

bra

nds

Figure A5.17NNK per milligrams of nictone yield as a percentage of the median yield for PhilipMorris international brandsNNK, 4-(N-nitrosomethylamino)-1-(3-pyridyl)-1-butanone

Perc

enta

ge o

f med

ian

valu

e of

bra

nds

0

0,2

0,4

0,6

0,8

1

1,2

1,4

1,6

1,8

Brand

Mar

lbor

o LS

F H

P/Ar

gent

ina

Mar

lbor

o LS

F H

P/Ve

nezu

ela

SG V

entil

Reg

ular

F S

P/E

UPe

tra R

egul

ar F

HP/

CEM

AM

arlb

oro

KS F

SP/

US

Mar

lbor

o KS

F H

P/N

orw

ayL

& M

KS

F H

P/EU

Mar

lbor

o KS

F H

P/M

alay

sia

Che

ster

field

Orig

inal

s K

S F

HP

L &

M K

S F

HP/

Mal

aysi

aM

arlb

oro

KS F

HP

25_s

/Aus

trali

Mar

lbor

o KS

F H

P/T

aiw

anM

arlb

oro

100

F H

P/EU

Mar

lbor

o KS

F H

P M

ediu

m/E

UPa

rliam

ent 1

00 F

SP

Lt/U

SM

arlb

oro

KS F

HP/

Japa

nL

& M

KS

F H

P Lt

/EU

F6 K

S F

HP

Lt/E

UC

hest

erfie

ld O

rigin

als

KS

F H

PD

iana

KS

F SP

Spe

cial

ly M

ild/E

Mur

atti

Amba

ssad

or K

S F

HP/

EUM

erit

KS F

HP/

EUPa

rliam

ent 1

00 F

SP/

CEM

AM

arlb

oro

KS F

HP

Lt/G

erm

any/

UK

Mar

lbor

o KS

F H

P L

t/Jap

anM

arlb

oro

100

F H

P Lt

/Ger

man

yM

erit

KS F

SP

Ult/

US

Mar

lbor

o KS

F H

P U

lt M

en/U

SPa

rliam

ent K

S F

HP

Lt/J

apan

Virg

inia

Slim

s 10

0 F

HP

Ult

Me

Che

ster

field

INTL

KS

F H

P U

lt/Ph

ilip M

orris

100

F H

P S

uper

LM

arlb

oro

KS F

HP

Ult/

EUD

iana

KS

F H

P U

lt/E

UVi

rgin

ia S

lims

100

F H

P U

lt M

ePh

ilip M

orris

One

KS

F H

P/EU

Mur

atti

KS F

HP

Ult

1 m

g/C

EMA

Long

beac

h O

ne K

S F

HP/

Aust

rali

Virg

inia

Slim

s 10

0 F

HP

Men

1M

arlb

oro

KS F

HP/

Mex

ico

Raf

fles

100

F H

P/E

UM

arlb

oro

KS F

HP/

Bra

zil

Pete

r Jac

kson

KS

F H

P M

en/A

ust

Mar

lbor

o 10

0 F

HP

Lt/U

SM

arlb

oro

KS F

HP

Lt/N

orw

ayC

hest

erfie

ld K

S F

HP

Lt/E

UPh

ilip M

orris

KS

F H

P Su

per L

iM

erit

KS F

SP

Ulti

ma/

US

1R4F

Ken

tuck

y R

efer

ence

1R4F

Ken

tuck

y R

efer

ence

Figure A5.18Benzo[a]pyrene per milligrams of nictone yield as a percentage of the median yieldfor Philip Morris international brands

268

4. Recommendation on cigarettemachine smoking regimens

Tobacco product characterization is an essential element of tobacco productcontrol and should be an integral part of comprehensive tobacco control pro-grammes. Assessment of the toxicity of tobacco products allows applicationof regulatory strategies similar to those used for other consumer products,including informing regulatory agencies about the levels of toxicants intobacco product content and emissions and reducing the levels of knowntoxicants present in the emissions, while not requiring measurement ofexposure of human populations.

The Scientific Advisory Committee on Tobacco Product Regulation estab-lished by WHO held its first meeting in October 2000. In November 2003,the Director-General of WHO changed its status from an advisory committeeto a study group, and it is now known as the WHO Study Group on TobaccoProduct Regulation (TobReg). The Director-General reports to the ExecutiveBoard on the results and recommendations of the Study Group in order todraw the attention of Member States to WHO’s efforts in tobacco productregulation. The objective of TobReg is to make scientifically sound recom-mendations to WHO Member States on the most effective, evidence-basedmeans for filling regulatory gaps in tobacco control and for establishing acoordinated regulatory framework for tobacco products. TobReg’s memberscomprise national and international experts on product regulation, treatmentof tobacco dependence, laboratory analysis of tobacco contents and emis-sions, and design features.

An advisory note from the WHO Scientific Advisory Committee on TobaccoProduct Regulation in 2003, entitled Recommendation on health claimsderived from ISO/FTC method to measure cigarette yield, concluded that thecurrent standard method for measuring tar, nicotine and carbon monoxide incigarette smoke with the single International Organization for Standardiza-tion (ISO)/US Federal Trade Commission (FTC) machine smoking regimenis unacceptable for public health. TobReg examined the scientific evidenceand found that the tar, nicotine and carbon monoxide yields per cigarettemeasured by machine under the existing ISO regimen are not valid estimatesof human exposure or of the exposure deriving from smoking different brands

269

of cigarettes. Communication of such measures to smokers creates harm bymisleading them into believing that their exposure and risk can be reduced ifthey switch to cigarette brands with lower machine-measured yields, and byoffering the products as an alternative to cessation.

The meeting of the ISO Technical Committee (TC) 126 in Las Vegas, USA,on 22–23 May 2006, which discussed harmonized, international standardsfor tobacco and tobacco products, recognized that misuse of machine-measured yields can result in false information to consumers. It passed aformal resolution, adopting the following statements as the rationale for allmachine smoking regimens:

“No machine smoking regime can represent all human smoking behaviors;

“Methods are recommended which test the product under conditions of differ-ent intensities of machine smoking testing in order to collect main stream smoke;

“Machine smoking testing is useful to characterize cigarette emissions for designand regulatory purposes, but communication of machine measurements tosmokers can result in misunderstanding about differences in exposure and riskacross brands; and

“Smoke emission data from machine measurements may be used as inputs forproduct hazard assessment, but they are not intended to be nor are they validmeasures of human exposure or risks. Communicating differences betweenproducts in machine measurements as differences in exposure or risk is a misuseof testing using ISO standards.”

In order to address Articles 9 and 10 of the WHO Framework Convention onTobacco Control on regulation of the contents, emissions and design featuresof tobacco products and of tobacco product disclosures, the Conference ofParties at its first session in Geneva, 6–17 February 2006, created a workinggroup to draft guidelines for implementation of those articles. At its secondmeeting in Ottawa, Canada, 26–28 October 2006, the working group re-quested WHO’s Tobacco Laboratory Network (TobLabNet) to evaluate thetechnical advantages and disadvantages of the current ISO method, theCanadian ‘intense’ method, the method of the State of Massachusetts (USA),and the ‘compensatory regime’. TobLabNet, a global network of government,academic institutions and independent laboratories administered by theWHO Tobacco Free Initiative, met in Beijing, China, 20–22 November 2006to fulfill that task. TobReg considers that the discussion clearly favoured theCanadian ‘intense’ method, which includes a more intense machine smokingregimen and is the most useful single regimen for assessing emissions. Nev-ertheless, TobReg and TobLabNet recognize that more than one machineregimen is needed to characterize smoke emissions adequately.

At the second session of the Conference of Parties, in Bangkok, Thailand,30 June–6 July 2007, the possibility of working with ISO TC126 to validatemethods for measuring the contents of tobacco products was discussed as a

270

problematic but potentially valuable approach. ISO TC126 formed WorkingGroup 10, which is a forum for the exchange of information between ISO andWHO, including issues related to deliberations by WHO or the Conferenceof Parties on standard setting for product regulation in general and the designand validation of test methods in particular. ISO TC126 exchanges informa-tion with WHO in order to conform to the needs of the Conference of Parties,if that body chooses to exercise its competence in standards development fortobacco products, specifically in the design and validation of tobacco testingand measuring methods.

On the basis of existing scientific evidence, TobReg at its meeting in Stanford,California, 25–27 July 2007, made the following recommendation throughthe Head of the Convention Secretariat of the WHO Framework Conventionon Tobacco Control, the Tobacco Free Initiative, to the Conference of Partiesworking group mandated to prepare guidelines for Articles 9 and 10:

TobReg recommends that the Conference of Parties Working Group finalize itsdecision on a machine smoking regimen and recommends that the regimen se-lected be the Canadian ‘intense’ regimen. To assist Parties in strengthening theregulation of the content of tobacco products, the working group on Articles 9and 10 established by the Conference of Parties at its first session to developprepare for implementation of Articles 9 and 10 of the Framework Convention,charged the Tobacco Free Initiative to work with TobLabNet to identify themethod of choice for each of the three groups of chemicals in cigarette tobaccoand for the emissions of four groups of chemicals in mainstream cigarettesmoke. TobLabNet is prepared to carry out the mandated method validationbut requires a decision from the Conference of Parties concerning finalizationof the machine smoking regimen and any need for interaction with ISO.

In this regard, TobReg considers that use of the Canadian ‘intense’ regimenwould reflect the yields resulting under intense smoking conditions and wouldalso form a basis for preventive public health strategies. The yields derivedfrom this regimen could be used, for example, in setting product performancestandards.

After this recommendation was written, the working group on Artices 9 and10 included in its progress report to the third session of the Conference ofParties in Durban, South Africa (17–22 November 2008), a recommendationfor use of two smoking regimens—the ISO regimen and an intense regimen—for validation of the test methods outlined in their report. The working groupidentified three sets of contents (ammonia, nicotine and humectants) and foursets of emissions (tobacco-specific nitrosamines, benzo[a]pyrene, aldehydesand volatile organic compounds) for method validation. The report recom-mended that the Conference of Parties, through the Convention Secretariat,request WHO’s Tobacco Free Initiative to validate the analytical chemicalmethods for testing and measuring the cigarette contents and emissions iden-tified as priorities in its report, using the two smoking regimens set out inparagraph 18 of the report and to inform the Conference of Parties throughthe Convention Secretariat of the progress made on a regular basis. The first

271

smoking regimen is ISO 3308:2000, which defines a routine analyticalcigarette smoking machine used under standard conditions, with a 35-ml puffvolume, a puff frequency of one every 60 s and no modification to ventilationholes. The second is similar but with a 55-ml puff volume, a puff frequencyof one every 30 s and blocking of all ventilation holes with Mylar adhesivetape.

The Conference of Parties at its third session deliberated on the progressreport and mandated the working group to continue its work, elaboratingguidelines step-by-step, and to submit a first set of draft guidelines to theConference of Parties for consideration at its fourth session. In addition, theConference of Parties requested the Convention Secretariat to invite WHO’sTobacco Free Initiative to undertake the following work:

(1) submit a report for consideration by the Conference of Parties at its fourthsession, which:

(a) identifies best practices in reporting to regulators as regards contents,emissions and product characteristics, including electronic systems;

(b) identifies best practices in informing the public; and

(c) collects information on legal cases and analyses the legal issuesrelated to tobacco product disclosures;

(2) validate, within 5 years, the analytical chemical methods for testing andmeasuring the cigarette contents and emissions identified as priorities inthe progress report of the working group (FCTC/COP/3/6), using the twosmoking regimens set out in paragraph 18 of that report, and to informthe Conference of Parties through the Convention Secretariat on a regularbasis of the progress made.

272

5. Overall recommendations

5.1 Harm reduction and smokeless tobacco products: regulatoryrecommendations and research needs

Main recommendations

The objective of harm reduction in relation to tobacco is to decrease morbidityand mortality among continuing tobacco and nicotine users who are unwillingor unable to quit, with due consideration of effects at the population level.Cigarette smoke is the most hazardous form of nicotine intake, and medicinalnicotine is the least hazardous. Among the smokeless tobacco products onthe market, products with low levels of nitrosamines, such as Swedish snus,are considerably less hazardous than cigarettes, while the risks associatedwith some products used in Africa and Asia approach those of smoking. Thefinding in Sweden that long-term use of smokeless tobacco by men reducescigarette smoking has not been replicated in other countries where smokelesstobacco has been widely available. The Swedish experience may not thereforebe generalizable, and it would be premature to use it as a basis for publichealth recommendations.

It is recommended that research be conducted to determine whether and underwhat conditions smokeless tobacco might be used as an aid in smoking ces-sation and whether marketing smokeless tobacco as a method for harmreduction would encourage initiation of smoking and smokeless tobacco use.The evidence that use of smokeless tobacco leads to cessation of cigarettesmoking is inconclusive, although survey data from Sweden suggest that thismay have occurred. The evidence that initiation of smokeless tobacco useleads to a higher prevalence of use of combustible tobacco products is con-flicting. In view of the wide diversity in the composition, toxicity and patternsof use of smokeless tobacco products by geographical region, it is inappro-priate to consider smokeless tobacco as a single product.

Significance for public health policies

All smokeless tobacco products should be subjected to comprehensive reg-ulatory control by an independent, scientific government agency. The control

273

must include and require disclosure of ingredients by manufacturers. Asclaims for reduced exposure might be interpreted as claims for harm reduc-tion, the former must be based on evidence of reduced risk.

In view of the wide variety of smokeless products with respect to composition,patterns of use, history of use and user characteristics, public health policymust cater for different populations. Continued testing and measurement ofthe contents and emissions of smokeless tobacco products must be conductedin order to identify regional variations. Careful attention should be paid toproduct characteristics and risks, the pattern of tobacco use in the population,social and cultural differences and marketing messages in order to evaluatethe potential of specific smokeless tobacco products to reduce harm.

Implications for WHO programmes

The wide range of smokeless tobacco products and use characteristics meansthat WHO should support individual and population-based research on spe-cific products. Better knowledge about the effects and mechanisms of actionof smokeless tobacco products and about what modifications can be made toalter the effects is needed so that governments can implement the WHOFramework Convention on Tobacco Control. WHO should continue researchon the health hazards and risks to individuals and populations due to use ofsmokeless tobacco products.

5.2 ‘Fire-safer’ cigarettes: approaches to reducing ignitionpropensity

Main recommendations

Fires due to cigarettes and the related deaths are a major global public healthproblem. In view of the role of cigarettes as the source of residential fires andresulting deaths, public health policy is required to reduce the number of suchincidents over time.

Techniques for reducing ignition propensity are available and are used. Anignition propensity performance standard has been established for cigaretteswhich makes them significantly less likely to cause fires if left unattended. Itis recommended that standards such as that of the National Institute of Stan-dards and Technology in the USA be implemented in Member States.

Research is needed to ensure the effectiveness of regulations for reducedignition propensity and on the effects of changes in cigarette design to providethe basis for further policies. As in countries that already have reduced igni-tion propensity policies, others should require tobacco manufacturers to testignition strength, report to the appropriate authority and cover the costs ofresearch and implementation.

274

Significance for public health policies

Adequate, appropriate monitoring, reporting and archiving is needed of theeffectiveness of techniques for reducing ignition propensity for reducingdeaths, injuries and property damage due to cigarette-induced fires. Suchmonitoring will increase public assurance and lead to more effective policies.

Claims that use of products with reduced ignition propensity will reduce riskshould not be allowed, as they could lead consumers to perceive a loweredoverall health risk. Public education programmes should be continued, toinform consumers that tobacco products are lethal and that smokers shouldquit. Such programmes should also include education campaigns to teach thepublic how to prevent fires.

Implications for WHO programmes

As techniques for reducing ignition potential are available and may be ben-eficial, Member States should require cigarettes with reduced ignition po-tential based on the National Institute of Standards and Technology standardor any other standard that has been shown to be effective. Countries andjurisdictions within countries should retain the right to alter the standard onthe basis of population-based data on the effectiveness of the standard. Poli-cies should require tobacco manufacturers to commission testing by inde-pendent laboratories that have been accredited in accordance with ISOstandard 17025, General requirements for the competence of calibration andtesting laboratories. WHO should facilitate implementation of such policiesand support the development of more effective means of redcuing the damagedue to cigarette-ignited fires.

5.3 Mandated lowering of toxicants in cigarette smoke: tobacco-specific nitrosamines and selected other constituents

Main recommendations

Communication of machine-measured tar, nicotine and carbon monoxideyields per cigarette as a measure of human exposure or difference in riskbetween products is currently misleading smokers to believe that low yieldcigarettes have less risk and are a reasonable alternative to cessation. Thescientific capacity to assess human exposure and harm from use of differentproducts for use in regulation remains incomplete. In the interim, a newstrategy for regulation is proposed which establishes product performancestandards, requires disclosure of emissions, and mandates a lowering of thetoxicant levels generated under standardized conditions by prohibiting thesale of brands that do not meet established levels of these standards. Thisapproach is similar to the regulation of most consumer products where

275

toxicant levels present in the product are reduced to the extent possible aspart of good manufacturing processes. An essential component of this rec-ommendation is the regulatory prohibition of the communication of thesemeasures to the public as measures of human exposure or risk, and of anyranking of products by their toxicant yields.

Toxicant levels would be compared using units per mg nicotine in order tofocus on the toxicity of cigarettes under standardized conditions and avoidtheir use as measures of exposure. Toxicants were selected based on multiplecriteria, the most important of which was evidence of toxicity.

The primary goal of the proposed regulatory strategy is to reduce the levelsof toxic constituents measured under standardized conditions in the smokeof cigarettes allowed on the market. A secondary goal is to prevent the in-troduction into a market of cigarettes with higher levels of smoke toxicantsthan are present in brands already on the market.

Significance for public health policies

Regulatory bodies should consider adoption of the new regulatory strategyin order to avoid the continuing harm resulting from communication of tar,nicotine and carbon monoxide values per cigarette and as a means of reduc-ing the toxicants known to be present in smoke in a manner similar to thatused to regulate toxicant levels in other consumer products. The recom-mended regulatory strategy should be implemented in phases beginning witha period of required annual reporting of toxicant levels by cigarette manu-facturers to the regulatory authority. This should be followed by the promul-gation of the levels of toxicants above which brands cannot be offered forsale. Finally, the established levels would be enforced and violating brandsbanned from the market.

Any regulatory approach based on yields under standardized conditionsshould prohibit the use of testing results, ranking of brands by testing levels,or statements that the brand has met governmental regulatory standards asindicators of risk or exposure. Regulatory authorities have an obligation toensure that the testing results are not used to mislead the public, which hasoccurred previously.

Implications for WHO programmes

In view of the harmful effects of the present approach, which allows com-munication of measurements of emissions per cigarette, WHO should pro-mote the prompt replacement of that approach with the recommendedregulatory strategy. Mandated lowering of levels of toxicants per mg nicotinein cigarette smoke will make regulation of cigarettes consistent with other

276

regulatory approaches which mandate reduction of known toxicants in prod-ucts used by humans. The WHO Framework Convention on Tobacco Controlrecognizes the need for tobacco product regulation in Articles 9 and 10.

5.4 Cigarette machine smoking regimens

Main recommendations

The Conference of the Parties at its second session acknowledged the poten-tially value of working with the International Standards Organizations (ISO)Technical Committee 126 to establish an ISO standard for a machine smokingregimen. The recommendation for a new regimen is based on a scientificconsensus that a new, more intense smoking regimen is needed to characterizebetter the composition of tobacco smoke produced by cigarettes under con-ditions reflecting the upper range of intensities of the patterns of their use.After evaluation of several machine-smoking regimens, TobReg recom-mends that the ISO select the Canadian ‘intense’ regimen.

Significance for public health policies

Continued misuse of the ‘per cigarette yields’ generated with the current ISOregimen is harmful to public health and results in inadequate characterizationof the smoke generated by different products. In the Canadian ‘intense’ reg-imen, more intense smoking conditions are tested, resulting in better charac-terization of cigarette smoke for use in public health. The yields derived fromthis regimen can be used, for example, to set product performance standards.

Machine smoking testing is useful for characterizing cigarette emissions fordesign and regulatory purposes; however, it is not intended to be and is nota valid measure of human exposure or risk. Care should be taken that themeasures are not misinterpreted by consumers as differences in exposure orrisk.

Implications for WHO programmes

Accurate characterization of tobacco products and disclosure to regulatoryagencies are essential for tobacco product control, as outlined in Articles 9and 10 of the Framework Convention. Machine smoking regimens that allowbetter characterization of the smoke generated by different products are es-sential for improving public health and might result in reductions in the levelsof known toxicants in emissions. WHO must continue to support TobReg’srecommendation that a new machine smoking regimen be standardized.

277