Effects of Weldingon Health IV

Effects of Weldingon

Health IV

An up-dated (December 1980-June 1982) literature surveyand evaluation of the data recorded since the publication ofthe first report, to understand and improve the occupational

health of welding personnel.

ByWinifred G. Palmer

Submitted by:Tracor Jitco

1776 E. Jefferson StreetRockville, MD 20852

Submitted to:SAFETY AND HEALTH COMMITTEE

AMERICAN WELDING SOCIETY550 N.W. LeJeune Road

Miami, FL 33126

International Standard Book Number: 0-87171-230-X

American Welding Society, 550 LeJeune Road, Miami, FL 33126

©1983 by American Welding Society.All rights reserved.

This report is published as a service and convenience to the welding industry and is the product of an independentcontractor (Tracor Jitco) which is solely responsible for its contents. The materials in this report have not beenindependently reviewed or verified and are only offered as information. AWS assumes no responsibility for anyclaims that may arise from the use of this information. Users must make independent investigations to determinethe applicability of this information for their purposes.

Printed in the United States of America

Contents

Personnel vAcknowledgements viiPreface ixIntroduction xiExecutive Summary xiiiTechnical Summary xv

/. The Exposure 11.1 Fumes 11.2 Gases 51.3 Protective Coatings 71.4 Electromagnetic Radiation 91.5 Noise 10

2. Effects of Welding on Human Health 112.1 Respiratory Tract 112.2 Cancer 152.3 Effects on the Ear and Hearing 192.4 Effects on the Eye and Vision 192.5 Effects on the Skin 202.6 Effects on the Cardiovascular System 212.7 Effects on the Nervous System 212.8 Effects on the Liver 222.9 Effects on the Musculoskeletal System 222.10 Effects on the Urogenital Tract 232.11 Effects on the Teeth and Oral Cavity 232.12 Metal Fume Fever 232.13 Sensitivity to Fume Components 232.14 Biochemical Changes 242.15 Human Fatality 252.16 Occupational Medicine 25

3. Toxicologic Investigations in Animals and in Cell Cultures 253.1 Animal Studies 253.2 In Vitro Studies 29

References 31

Bibliography 37

III

Personnel

The author of the report by Tracor Jitco was:

Winifred G. Palmer

AWS Research Committee

A.N. Ward, Chairman Caterpillar Tractor CompanyK.L. Brown, Vice Chairman Lincoln Electric Company

M.E. Kennebeck, Jr., Secretary American Welding SocietyJ.S. Gorski Kemper Insurance Companies

E. Mastromatteo INCO Limited

Acknowledgements

The American Welding Society gratefully acknowledges the time and effort expended by the membersof the Research Committee and the financial support of the program by industry contributions.

Supporting Organizations

Air Products and Chemicals, Inc.Airco Welding ProductsAllis-ChalmersAlloy Rods Division, The Chemetron CorporationAWS Detroit SectionAWS New Orleans SectionArcos CorporationThe Binkley CompanyCaterpillar Tractor CompanyChicago Bridge and Iron CompanyGrove Manufacturing Company, Division of Kidde, Inc.General Electric CompanyThe Heil CompanyHobart Brothers CompanyHuntington Alloys, Inc.Lincoln Electric CompanyMiller Electric Manufacturing CompanyNational-Standard CompanyA.O. Smith CorporationTeledyne-McKay, Inc.Trinity Industries, Inc.Truck Trailer Manufacturers AssociationWalker Stainless Equipment CompanyWeld Tooling Corporation

Many other organizations have made contributions to support the ongoing program from May 1979 tothe present.

VII

Preface

This literature review has been prepared for the Safety and Health Committee of the AmericanWelding Society to provide an assessment of current knowledge of the effects of welding on health, aswell as to aid in the formulation of research projects in this area, as part of an ongoing program sponsoredby the Society. Previous work has included studies of the fumes, gases, radiation, and noise generatedduring various forms of arc welding (see Bibliography). Conclusions based on this review and rec-comendations for further research are presented in the introductory portions of the report. Section 1summarizes recent studies of the occupational exposures. Section 2 contains information related to thehuman health effects of exposure to byproducts of welding operations. Section 3 discusses studies of theeffects of welding emissions on laboratory animals and in vitro cell systems. Referenced materials areavailable from Tracor Jitco, Inc.

IX

Introduction

The health of workers in the welding environment isa major concern of the American Welding Society. Tostay abreast of this subject, the health literature isperiodically reviewed and published in the reportEffects of Welding on Health. Three volumes havebeen published to date (Refs. 135, 143, and 144); thefirst covered data published prior to 1978, while thelatter two covered the periods 1978 to May, 1979 andJune, 1979 to December 1980, respectively. Thecurrent report included information that was pub-lished between December, 1980 and June, 1982, andshould be read in conjunction with the previousvolumes for a comprehensive treatment of the litera-ture on the Effects of Welding on Health.

Included in this volume are studies of the charac-teristics of welding emissions that may have impacton the control technologies necessary to protect thewelding worker (Section 1). Considerable discussionis devoted to ozone which may become a greaterproblem to welders as improved ventilation and

decreased fume exposures reduce the rate of degenera-tion of this hazardous gas. Much recent research hasfocused on chromium and nickel, since exposure tocertain chemical forms of these metals may causeserious chronic health problems.

In keeping with previous volumes, the health studiesare organized according to the organ system affected.The respiratory tract, the primary route of entry ofwelding fumes and gases into the body, also is a majortarget organ of a number of components of theseemissions. Acute (e.g., metal fume fever, cadmiumpoisoning) as well as potential chronic respiratoryeffects (e.g., emphysema, cancer) of welding emissionsare of concern. The latter are far less well understoodand whether or not there is an excess risk of cancerfrom these exposures has not been established. Con-tinued research in the form of epidemiologic studies,investigations with laboratory animals, and in vitrogenotoxicity studies will help to resolve this question.

XI

Executive Summary

A problem inherent in research concerning thehealth effects of welding is that there is a great deal ofvariability in both welding processes and in workingconditions which makes it difficult to perform studieson homogeneous populations of sufficient size topermit statistical analysis. These variations may be lesscritical when examining causes of acute physiologicalresponses to welding exposures (e.g., metal fume fever,burns, photokeratitis). However, the associationbetween chronic exposures to welding emissions anddisease conditions whose causes are less well under-stood, or which occur at a low incidence, remainsambiguous in many cases.

The Respiratory TractOf the five metals (Cr, Ni, As, Be, and Cd) that have

been shown by epidemiologic studies to be related toan increased cancer incidence in workers in certainindustries, chromium and nickel are present in signifi-cant quantities in fumes released during weldingcertain metals (e.g., stainless steel and nickel alloys).However, only some chemical compounds containingthese metals are carcinogenic, and it is not known ifthese compounds are present in welding fumes. Theavailable epidemiologic evidence is insufficient todetermine whether or not welders are at an increasedrisk of developing lung cancer from exposure tochromium and nickel in welding fumes.

Most of the reports of the effects of welding onhealth which appeared during the period of this reviewdealt with effects of welding on the respiratory tract.Metal deposits have frequently been found in welders'lungs by chest X-rays and at autopsy. The use of

magnetopneumography enables rough estimates to bemade of the quantities of metals retained in the lungs.However, the association between this retention ofmetals and impaired lung function is not clear. Forthis reason, it is important to develop more accuratemethods for the determination of lung function. Of thesix studies using pulmonary function tests of welderswhich appeared during this report period, threeindicated impaired lung function among welders, andthree reported no differences between welders andcontrols.

Four epidemiologic studies of the lung cancerincidence among welders have appeared in the recentliterature. Two of these indicated an elevated lungcancer risk. One of these two studies focused onwelders of stainless steel, but the size of the studycohort was too small for the results to be consideredconclusive. The second was a death record study ofwelders selected from the rosters of a trade union inSeattle, Washington. A significant increase(SMR*=174) in the lung cancer rate was found whendeaths that occurred more than 20 years from theinitial date of exposure were considered. A study ofwelders exposed to nickel-rich welding fumes did notindicate an elevated lung cancer risk; however, theinvestigator emphasized the need for a follow-up studyof this cohort to allow more time for the appearanceof tumors. There is a great need for further well-planned epidemiologic studies of welders. Because ofthe high cost of these studies and the limited resourcesavailable to the welding community, Stern suggested

•Standardized mortality ratio

XIII

that such studies focus on exposure to weldingprocesses in which it is suspected that genotoxic orcarcinogenic materials, such as nickel and chromiumcompounds, are released (Ref. 122).

Other Organ SystemsReports of recent health studies that examined the

effects of welding on organ systems other than therespiratory tract are highlighted below.

— With the exception of cutaneous burns anderythema of the neck which were more common inwelders, no significant differences between the fre-quency of skin and eye disorders were found betweenwelders and nonwelders from the same fabricationfacility. Good safety and health measures were prac-ticed in this plant.

— An excess of inflammation of the oral mucousmembranes, including periodontal disease, wasobserved in two separate studies of welders.

— No differences were observed in the morphology,number, and motility of sperm produced by weldersand by workers in other occupational groups.

Biological Monitoring of Exposureto Welding Emissions

The development of methods for the accuratedetermination of personal exposures are needed toestablish and measure the effectiveness of control

technologies. In conjunction with devices that collectpersonal air samples for determining the welder'sactual exposure, biological monitoring affords ameans for estimating the actual dose of contaminantstaken up by the body. Such tests are useful for thesupport of population and epidemiologic studies andfor the evaluation of the effectiveness of industrialhygiene measures introduced to reduce workerexposure. Finally, biological monitoring duringroutine medical examinations may play an importantpart in alerting the occupational physician to indivi-duals who may be exposed to unacceptable levels ofwelding emissions.

The importance of this area of research is witnessedby the number of studies that have been published onthis subject. Studies using magnetopneumographyhave demonstrated that the amount of materialsretained in the lung correlates with the number ofyears spent welding, the extent of siderosis, and therelative quantity of fumes generated by the weldingprocess. In other work, the relationship between urineconcentrations and breathing zone levels was linearfor fluorine, suggesting that this element may be usefulfor monitoring the exposure of welders who use basic-coated electrodes. Although urine chromium levelstended to increase during the work week, this metalnevertheless may be useful for estimating exposures tofumes generated by welding of stainless steel. Urinenickel levels tend to fluctuate widely and are apparentlynot a useful measure of nickel exposures.

XIV

Technical Summary

The ExposureFumes

The major objectives of research on fume emissionsare to evaluate health hazards, to develop weldingmethods which produce less toxic fumes, and to enableestimation of air exchange rates required to bringfume concentrations to acceptable levels. Total fumeemissions are greatest with shielded metal arc weldingand flux cored arc welding and vary with the electrodeused and welding parameters.

To determine the airflow requirements of ventilationsystems, Alekseeva et al. examined the emissionsreleased during gas tungsten arc welding of manganese-copper alloys (Ref. 4), copper alloys containing nickelor zinc (Ref. 3), and eight different welding wirescontaining from 2 to 70% nickel (Ref. 5). Pokhodnyaet al. (Ref. 102) found that the rate of manganesevaporization from rutile or basic electrodes increasedwith the alkalinity of the slag. Oleinchenko et al. (Ref.98) determined that fluoride emissions from electrodecoatings increased with the moisture of the coatingduring automatic aluminum welding.

A large portion of the solid phase of weldingaerosols consists of respirable particles. Particles mayexist as single entities or as chains or agglomerates(Refs. 12, 23, 51, 70 and 81). When present, chromatesappear to be condensed on the surface of metal oxideparticles (Refs. 70 and 81).

Two major interlaboratory studies are currentlyexamining methods for the determination of solubleand insoluble hexavalent chromium in weldingemissions (Refs. 11 and 51). Variations in the methods

of sample collection, storage, and extraction pro-cedures may affect the analytical results (Refs. 11,49,and 128).

GasesThe sources of toxic gases, e.g., nitrogen oxides,

ozone, and carbon monoxide, during welding arediscussed. Johansson (Ref. 71) reported that theaddition of helium or hydrogen to shield gases reducedthe ozone levels released during GMAW or GTAW ofstainless steel, and Smars (Ref. 117) reported that theaddition of nitric oxide, but not helium, significantlyreduced ozone levels during GTAW of aluminum.However, Farwer (Ref. 39) argued that because of itstoxicity and stability, the NO2 generated from nitricoxide in shield gases may be a greater hazard than theozone whose formation it prevents. The presence ofmagnesium, but not silicon, in aluminum alloyssignificantly reduces the production of ozone duringGMAW of aluminum (Ref. 117).

Production CoatingsPrimers or paints on metal surfaces may contribute

significant quantities of formaldehyde, carbonmonoxide, hydrogen cyanide, and other organicvapors to welding emissions (Ref. 21). These varyprimarily with the characteristics of the binders usedin the coatings.

Electromagnetic RadiationSince ultraviolet radiation in the vicinity of 270 to

280 nm is most likely to be absorbed by and damage

xv

the genetic material (DNA), Bartley et al. (Ref. 14)investigated the intensity of radiation with these wave-lengths generated by GTAW and GMAW of alumi-num. They found that, unlike its effect on ozone levels,magnesium in aluminum alloys effects an increase inultraviolet radiation between 270 and 280 nm.

Pabley and Keeney (Ref. 99) showed that reflectivegold polycarbonate filter plates in welders helmetsmore effectively reduced exposure to infrared radiationthan did green glass filter plates. SWnzyetal. (Ref. 116)developed design criteria for a semitransparent curtainwhich would enable safe bystander viewing of thewelding area and increase visibility within the weldingbooth.

NoiseHermanns (Ref. 65) compared the noise levels

produced by different welding processes and foundthat, while most produced at least occasional noiseabove 85 dB, plasma cutting was the noisiest.

Effects of Welding onHuman Health

Respiratory SystemUsing magnetopneumography, the quantity of

particles retained in the lungs of welders in the asbes-tos (Ref. 31) and shipyard (Ref. 42) industries wasfound to be significantly higher than that in workersnot employed in welding. Of workers using a varietyof common welding methods, stainless steel weldersemployed in SMAW had the highest quantities andthose using GTAW had the lowest quantities ofretained metals in the lung(Refs. 73 and 80). Weldersperforming SMAW of mild steel had intermediatelevels of retained metals.

Two cases of pulmonary fibrosis were reported inaluminum arc welders (Refs. 62 and 134). Cigarettesmoke and other exposures may also have contributedto the disease. Gobatto et al. (Ref. 45) examined lungtissue from 17 autopsied welders. Fibrosis was presentin more than half, and siderosis was present in all ofthe lungs examined. Areas of inflammation or fibrosisconsistently had associated iron deposits which led theinvestigators to speculate that the metal or associ-ated toxins, or both, were involved in the disease process.Zober (Ref. 147) reviewed and analyzed reports of 47cases of pulmonary fibrosis in welders publishedbetween 1955 and 1979. He concluded that, at most,only 20 cases could have been related to the weldingexperience.

Hayden et al. (Ref. 60) found that welders in theengineering industry had slightly more absences due toupper respiratory tract infections than did controls, butabsenteeism resulting from other diseases did not

differ from control groups. McMillan and Molyneux(Refs. 83 and 84) found no differences in the absenteepatterns from respiratory tract diseases in nonsmokingwelders and controls. However, they did note thatsmoking welders tended to have slightly more absencesfrom lower respiratory tract disease than did the non-welding controls who smoked tobacco products.

The results of lung function tests of welders areinconsistent. No significant differences between theresults of pulmonary function tests of welders andcontrols were found by Schneider and Rebohle amonga population of shipyard workers (Ref. I l l ) or byHayden et al. (Ref. 60) among workers in the engineer-ing industry. Zober examined 40 welders and 40 agematched controls and, with the exception of siderosisdetected by chest X-rays, no abnormalities in pul-monary function were identified that could be relatedto welding (Ref. 145). However, Akbarkhanzadeh(Ref. 2) found that welders in the shipyard industryhad impaired lung function, and Cavatorta (Ref. 25)reported changes in respiratory function which wererelated to the duration of the welding experience.Ohmori et al. (Ref. 97) found a slight decrease in peakflow rate and maximum expiratory flow rate at 75%vital capacity which was related to ozone exposuresduring GMAW. According to Stern (Ref. 122), someof these inconsistencies may result from populationdynamics, whereby workers with pulmonary diffi-culties may change occupations, and to bystandereffects whereby some welders may be exposed to toxicsubstances from nonwelding sources at the place ofemployment.

Lung CancerPolednak (Ref. 103) compared the causes of

mortality among welders of nickel alloys with aseparate cohort of workers who welded mild steel,stainless steel, and aluminum. The number of deathsfrom all causes as well as diseases of the respiratorytract, including cancer, were the same in both groupsand did not differ significantly from the general U.S.population. Because the period of time following theinitial welding exposures may have been too short torule out an increased respiratory cancer risk in weldersexposed to nickel-containing fume, Polednak (Ref.103) and Gibson (Ref. 44) stressed that follow-upstudies should be performed with this importantcohort as the number of years from the initial exposureto welding fumes increases.

Beaumont and Weiss (Ref. 16) examined thenumber of deaths from lung cancer among 3247welders selected from the rosters of a trade union inSeattle, Washington. A significant increase (SMR =174) in the lung cancer rate was found when deathsthat occurred more than 20 years from the initial dateof exposure were considered. Sjogren (Ref. 115) notedan increased risk of lung cancer (3 deaths observed

XVI

versus 0.68 expected) among a population of 234stainless steel welders in Northern Sweden. The resultsof a death certificate study by Ahonen (Ref. 1) indi-cated that the risk of lung cancer was the same forwelders and for workers in other occupations.

Ear and HearingA survey of the effectiveness of earmuffs and ear-

plugs indicated that welders and platers who useearplugs may suffer less hearing damage than thoseusing earmuffs (Ref. 38). The investigators speculatedthat their results may reflect the greater ease withwhich earmuffs can be removed and replaced by thewearer.

Eye and VisionA Bureau of Labor Statistics Survey showed that

welding and cutting injuries represented 0.5% of theWorkmen's Compensation cases in the United Statesin 1978. Of these injuries 67% involved the eye, 1 / 3 ofwhich were welder's flash. Emmett et ah (Ref. 36)found no difference between visual acuity or frequencyof ocular abnormalities between welders, nonweldersexposed to welding fumes, and workers who were notexposed to welding fumes. All workers were employedfor a comparable time period at the same facility.

A case of a pterygium (ultraviolet-induced growthon the outer surface of the eye) was reported in a 56year-old former welder (Ref. 68). It is unlikely that thisresulted from work-related exposure to ultravioletradiation (Refs. 96 and 126).

SkinScars from skin burns were found on 45%, and ultra-

violet-induced dermatitis in 8% of welders during asurvey of Soviet shipyard workers (Ref. 131). Theresults of a survey performed in a Danish hospitalindicated that welders suffered from burns morefrequently than workers in other occupations (Ref.112). A study in a U.S. facility revealed more cutaneousscars resulting from thermal and mechanical injuriesin welders and machinists than in other workers (Ref.36).

In the latter study, facial erythema, actinic elastosis,and premalignant and malignant skin lesions werefound with the same frequencies in welders and non-welders. Erythema of the neck was more frequent inwelders. The investigators noted that this was a rela-tively young population and the absence of excesspremalignant and malignant skin lesions may not beconclusive due to the relatively short period of timefollowing initial welding exposures.

Cardiovascular SystemAn experimental apparatus that simulated the

welding experience was designed and used in a labora-tory setting to determine changes in heart rate whilecarrying out typical welding procedures (Ref. 130).The variation in heart rate was greatest when the"welder" altered his posture to change electrodes. Thisvariation was greater with a stooped (downhandwelding) position than with a standing (overheadwelding) or a crouched position.

Nervous SystemChandra et ah found that urine manganese and

serum calcium levels were elevated in welders whoworked with manganese-containing electrodes andhad positive neurological symptoms (tremors anddeep brisk tendon reflexes in the extremities) (Ref. 27).The investigators suggested that serum calcium levelsmay be of value in diagnosing early stages of manga-nese poisoning.

Musculoskeletal SystemAn incidence of about 15 to 20% was found for

supraspinatus tendinitis in shipyard workers (Ref. 63).The number of years of welding experience and therated level of shoulder muscle load could not be corre-lated with the development of supraspinatus tendinitis,but these factors could not be excluded by the resultsof this study. Kadefors et ah (Ref. 72) measured themyoelectric activity in individual muscles while eightbasic welding postures were assumed by humansubjects in an experimental setting. All positionscaused localized muscle fatigue, but the overheadwelding position was potentially the most damagingto shoulder muscles. This study indicated that theassumption of certain positions may reduce the totalmuscle load in different welding situations. Tumakovand Grigorev (Ref. 132) designed an exercise programaimed at reducing tension in muscles subjected tostatic stress to be followed by welders suffering fromlumbosacral radiculitis.

Urogenital TractNo differences were observed in the morphology,

number, and motility of sperm produced by weldersand workers in other occupational groups (Ref. 76).This and an earlier study (Ref. 58), indicate thatwelding has no effect on fertility.

A study by Alsbirk et ah (Ref. 6) detected no changesin kidney function in welders of stainless steel.

Teeth and Oral CavityAn excess of inflammation of the oral mucous mem-

branes, including periodontal disease, was observedby Wulf and Seefeld (Ref. 140) and Melekhin andAgarkov (Ref. 89). To investigate the cause of com-plaints by underwater electric arc welders of a metallic

XVII

taste in their mouth and damage to their dental fillings,Rockert, Christensson, and Orthendal (Ref. 28 and 108)studied the underwater effects of electric currents ondental amalgams. They found that the current throughthe amalgam was dependent on its surface area anddistance from the electrode.

Metal Fume Fever and AllergicReactions

Nine cases of metal fume fever in welders werereported in Mexico (Ref. 18). Five of the affectedwelders had pulmonary abnormalities detected byX-rays.

Zugarman (Ref. 148) described a case of a welderwith an unusually severe allergic response to chromi-um. An in vitro study of the response of isolated whiteblood cells to manganese and fluorine indicated that50% of the welders tested were sensitive to eithermanganese or fluorine and 25% were sensitive to bothelements. None of the nonwelding controls wasresponsive (Ref. 89). Since the physiological responsewas not examined by standard allergy tests, theimportance of the in vitro results cannot be evaluated.

Biochemical ChangesElevated blood lead levels were observed in welders

of steel coated with anti-corrosive paints (Ref. 46).Elevated blood lead levels, decreased blood hemo-globin levels, and increased urine levels of delta-aminolevulinic acid were found in 22 tank workers inEgypt who were exposed to lead fumes for over 22years, but had no overt clinical signs of lead poisoning(Ref. 90). These biochemical alterations are typicalfor lead poisoning. Evidence of abnormal liverfunction was also indirectly indicated by the increasesin the serum enzymes SGOT, SGPT, and LDH.

Elevated plasma nickel levels were found in shipyardworkers (Ref. 47). Urine nickel levels, which fluctuatedthroughout the day, were less accurate markers fornickel exposure than plasma nickel levels. Sjogran(Ref. 114) examined the relationship between breath-ing zone levels and urine concentrations of fluorine,chromium, and nickel. The relationship between urineand breathing zone levels was linear for fluorine. Urinechromium content increased throughout the workweek even though airborne levels were relativelyconstant. The relationship between airborne andurinary nickel levels was nonlinear.

Human FatalityA human fatality was reported to be due to inhala-

tion of cadmium fumes from the use of an oxyacetylenetorch with silver solder (Ref. 82). The solder, whichcontained 20% cadmium, was not labeled as such bythe manufacturer.

Toxicological Investigationsin Animals and Cell Cultures

Animal StudiesThe extent of the inflammatory response to foreign

particles in the lung can be monitored by measuringthe number of migratory leukocytes (free cells), thelevels of surfactant, or the hydrolytic enzyme activityin the respiratory tract. Using these parameters,White et al. (Ref. 137) found that particles generatedby SMAW of stainless steel using a rutile electrodeelicited a slightly greater inflammatory pulmonaryreaction than did particles from SMAW of mild steelusing basic or rutile electrodes. Speculating that theresponse to the stainless steel particles may have beenlargely due to the chromates, they performed a follow-up study in which the inflammatory response elicitedby the water-soluble and insoluble fractions of particlesfrom the stainless steel welding system was comparedwith that elicited by potassium dichromate. Althoughthe chromate salt enhanced all of the parametersstudied, the effects observed are general responses topulmonary irritants, and it cannot be concluded fromthese studies that chromate is responsible for theinflammatory response elicited in the lung by fumesfrom welding stainless steel.

Repeated administration of fumes from coatedelectrodes led to the appearance of fibrotic materialsaround aggregates of welding particles in the lungs ofguinea pigs and rats (Ref. 7). The fibrotic response wasalso evidenced by the elevation of hydroxyproline in thelungs. Insufficient data were presented for a deter-mination of the factors responsible for the fibroticresponse.

The importance of examining the toxic effects ofcomplete mixtures rather than isolated componentsof industrial pollutants was demonstrated byIFnitskaya and Kalina (Ref. 69). These investigatorsfound that repeated inhalation by rats of a mixture ofaluminum oxide and welding gases (1 mg/m3 ozoneand 3.3 mg/m3 NOX ) caused less pulmonary damagethan did inhalation of the welding gases alone. Theauthors suggested that the toxicity of the gases wasreduced by adsorption to the aluminum oxide parti-cles.

English et al. (Ref. 37) compared the distribution ofNiOand NiCl2 throughout the body following admini-stration of the salts by intratracheal instillation to rats.The water-insoluble NiO salt was distributed moreslowly from the lung to other organs than was thewater-soluble NiCl2 but was eventually found in thesame organs as was NiCl2. The investigators con-cluded that NiO gradually becomes converted in thelung and the adjacent lymph nodes to a more water-soluble form which allows the nickel to becomedistributed throughout the body.

XVIII

Kawata et al. (Ref. 77) examined the acute effectson laboratory animals of inhalation of high concen-trations of fumes from basic-coated electrodes. Fumesfrom electrodes containing lithium were less toxic thanfumes from comparable lithium-free basic electrodes.

In Vitro StudiesUsing the sister chromatid exchange assay,

Niehbuhr et al. found that nickel-rich welding fumescause DNA alterations (Refs. 94 and 95). The geno-toxicity of the fumes was comparable to that of nickelsulfate. The authors concluded that the water-solu-bility of nickel compounds in welding fumes is aninadequate indicator of its potential toxicity.

White et al. (Ref. 138) compared the cytotoxicity offumes generated by SM AW of mild steel using eitherbasic or rutile-coated electrodes and by SMAW ofstainless steel using a rutile-coated electrode. Stainlesssteel fumes inhibited the growth of a rapidly prolifer-ating human cancer cell line. Fumes from basicelectrodes caused the most hemolysis of red bloodcells, and all three welding methods caused about a30% decrease in the viability of cultured macrophages.

ConclusionsA number of unanswered questions persist con-

cerning the effects of welding on health. Most of theseproblems have been reviewed in past volumes of thisseries. (Effects of Welding on Health, V o l u m e s I, II ,and III) and many are also problems encountered inother industries where dusts and fumes are prevalent.Some of the major health-related questions andrecommendations for addressing these questions arepresented.

Pulmonary Function TestsPulmonary function tests may be inadequate for

detecting lesions in the smaller airways. The reliabilityof these tests for general medical screening of weldersand prospective welders requires further study.

Cancer IncidenceNo general agreement is apparent in the findings of

the many epidemiologic studies as to whether or notthere is an elevated cancer incidence among welders.This may be, in part, because the work conditions areexceedingly variable from workplace to workplace,and studies of welders working under different con-ditions may not be comparable. Another difficulty isthat, unless the cancer risk is markedly elevated, a verylarge exposed population is needed to unequivocallydetermine a disease incidence. A clear definition of thework conditions, exposure levels, welding materials,and general health (including medical and smoking

histories) may allow the integration of results fromdifferent investigations, and thereby increase the sizeof the population studied. Coordination of studiessponsored by different associations (as suggested byStern, Ref. 123) and adoption of standard protocoldesigns (Refs. 122 and 146) would help to achieve thisgoal.

Chromium and NickelCarcinogenicity

Of the solid components released during welding ofclean metals, only nickel and hexavalent chromiumhave been implicated as human carcinogens. It is notknown whether the forms of nickel and chromiumpresent in welding fumes are carcinogenic. Only oneepidemiologic study has indicated an elevated respira-tory cancer risk in welders of stainless steel, and inthis work, the size of the study cohort was quite small(Ref. 115). As Stern suggested (Ref. 122), epidemio-logic studies should focus on welding shops in which"hot spots" for cancer may exist. In other words, thosewelding conditions (e.g., welding of stainless steel orhigh nickel alloys) which are suspected of releasinggenotoxic or carcinogenic materials should be highpriority targets for investigation.

Mutagenicity TestsIt is a widely held belief that at least some occupa-

tional and environmental exposures which causecancer do so by causing damage to DNA. As discussedearlier, many in vitro tests assess the ability of testsubstances to damage DNA. A major problem within vitro tests is that they ignore the immune defensemechanisms present in live animals as well as theability of metabolic systems to detoxify foreignchemicals. Furthermore, when complex emissions arebeing examined by in vitro tests, they do not reflect thefact that, in animal systems, there may be a selectiveuptake (especially when airborne mixtures enter thebody through the respiratory tract), absorption, andorgan distribution of the various components of themixture. What is apparently toxic in an in vitro system,therefore, may be relatively innocuous to the wholeanimal.

A technique, which has been used with workers inother industries (Refs. 10 and 141), that partiallyovercomes these difficulties involves examiningcultured white blood cells collected from exposedworkers and appropriate controls for evidence of anyDNA damage incurred in vivo following naturalexposures. This technique, combined with the analysisof personal air samples, discriminates between compo-nents that may or may not enter or remain in the bodyand allows a direct examination of effects on exposedworkers. A disadvantage of this method is that it onlyexamines one organ system and may not pick up the

XIX

chromosomal damage produced by toxins that onlyaffect specific target organs. Nonetheless, because itmore accurately reflects the actual workplace expo-sures, the examination of the effects of weldingexposures on the genetic material of white blood cellscollected from welders is a promising method fordealing with a difficult problem.

Interactions Between Componentsof Welding Fumes, Gases, andVapors

With few exceptions, threshold limit values andother exposure levels are based on the known toxico-logic effects of single elements or compounds. Thus,synergistic toxicologic interactions between chemicalsare not taken into account and for certain combina-tions of exposures or components within mixtures, thecombined effects may be greater than the additiveeffects of the individual components. In this case,permissible levels may not be sufficient to protect theworker. On the other hand, various components ofexposures may act to reduce the toxic effects of othercomponents of the exposure, and an unnecessaryfinancial burden may be placed on industry in con-forming to overly stringent regulations.

Although in vitro assays may be useful for examin-ing certain types of multiple effects, animal tests areprobably most useful for detecting synergists orantagonists because these reactants may exert theireffects through entirely different organ systems thanthe toxin whose action they are affecting (e.g., onecomponent of a mixture might suppress the immunesystem which could, by reducing the body's defensemechanisms, increase the effectiveness of a carcinogenin a different organ such as the liver). Little is knownabout interactions between components of industrialemissions, and this is a most important area for futureresearch.

Another type of interaction between components of

mixtures is that suggested in the study by Il'nitskayaand Kalina (Ref. 69). In that case, pulmonary lesionsresulting from exposure to welding gases (ozone andNOX) were greater in the absence of alumina than inits presence. This was thought to be due to the adsorp-tion of the gases onto the alumina particles. Healthprofessionals should consider the effect that the designof new control technologies to eliminate or reducesome components of an exposure could have on theremainder of the components. Another example ofthis is that reduction of fumes during welding may leadto an increase in ozone levels. Thus, during the intro-duction of any new technologies, or changes inexisting technologies, the effects on the whole systemmust be considered.

ErgonomicsWelding places a strain on the pectoral girdle, back,

and shoulder muscles. Although, as discussed earlier,the use of specific body positions while welding maydecrease some of the static load on affected muscles(Ref. 72), these strains cannot be eliminated. Asemphasized by Spelbrink (Ref. 119), there is a realneed for the development of mechanical aids whichcan perform much of the static holding functions inwelding and thereby reduce the muscle strain ex-perienced in this type of work.

Health and SafetyOnce identified, factors responsible for adverse

health effects are likely to be controllable. In only oneof the reports of workplace studies cited above (Ref.36), it was stated that data were collected in a workenvironment in which there was a strong emphasis onworker safety and health. In this case, skin and eyelesions among welders were minimal. As discussed byHinrichs (Ref. 66), studies of the health effects ofwelding in workplaces where recommended safepractices are used would be most important indetermining the effectiveness of these practices.

xx

Effects of Welding on Health IV

1. The Exposure

Welders may be exposed to the fumes and gases ofwelding emissions, radiation, noise, and organicvapors released during the welding of primed ordegreased metals. The emissions produced by weldingof clean, umprimed metals consist of fumes containingmetal oxide particles, and gases which are primarilyoxides of nitrogen (NOX), ozone (O3) and carbonmonoxide. The concentrations of these componentsvary greatly with the welding method. Although allmay be present in differing amounts during welding bythe various methods, in general, the greatest quantitiesof fumes are produced during shielded metal arcwelding (SM AW) and flux cored arc welding (FCAW);ozone is produced in the highest concentrations duringgas metal arc welding (GMAW), especially of alumi-num; and the NOX are most prevalent during gasburning welding processes. Carbon monoxide levelsmay be significantly elevated during welding of metalcoated with primers which contain organic bindersand during gas metal arc welding with a CO2 shield.

1.1 FumesThe fumes, which consist primarily of minute metal

oxide particles, constitute the solid phase of weldingemissions and originate mainly from the weldingconsumables. The quantity and constituents of fumesdepend on the welding process, the electrode type, thewelding current and voltage, and other parameters.The objectives of much of the research performed onwelding fume generation are to develop low fumeelectrodes and also to enable estimation of airexchange rates required to bring the fume concentra-tions below established permissible levels.

Fume generation is generally measured either on thebasis of the total quantity of fume emission per unittime, the fume generation rate or FGR (gram/ min),or the relative fume formation rate, RFFR, which isthe quantity of emissions produced when a certainmass of consumable electrode (or filler metal) is melt-ed (weight fume/weight electrode) (Ref. 8). Similar

2/EFFECTS OF WELDING ON HEALTH IV

measures are used in the Scandinavian countrieswhere the fume generation rate is symbolized as E(mg/sec) and the relative fume formation rate issymbolized as R and is measured in terms of mgfume/kg electrode.

1.1.1 Effects of Voltage and Current

Of the commonly used welding methods, with theexception of GMAW of aluminum, SM AW generatesthe highest fume levels and gas tungsten arc welding(GTAW) the lowest (Refs. 70 and 133). For SMAW,total fume emission, FGR, increases with weldingcurrent. Since the increased fume emissions reflect anincreased melting rate, the welding current has littleinfluence on the ratio of fume emissions to electrodeconsumption (RFFR). However, both FGR andRFFR increase with increasing voltage. Johanssonet al. explained that this may be due to the longer arcwhich, by evoking greater air turbulence, facilitatesthe penetration of oxygen through the protective gasshield with resulting increases in metal oxide forma-tion (Ref. 70).

Eichhorn et al. (Ref. 35) reported that raising thewelding current has a greater effect on fume emissionfrom acid or basic-coated electrodes than on that fromacid rutile-coated electrodes. Other factors, such aspolarity, may alter fume generation rates. The extentof the alteration, and the polarity which produces thehighest fumes, are dependent on the electrode type.Alternating current tends to produce less fumes thandirect current (Ref. 35).

In GMAW with an argon shield, RFFR is indepen-dent of voltage, once the voltage is high enough toproduce a stable metal spray. However, using amixture of argon and CO2, RFFR increases withvoltage once a stable spray arc is established. With aCO2 shield, the total fume emission, FGR, increasesgreatly as the voltage is increased (Ref. 70).

Castagna and Spagnoli (Ref. 24) reported that smalldifferences in the arc voltage and current resulted inmajor differences in the concentration of fumes andozone in the breathing zones of welders using argonshielded GMAW with continuous feed electrodescomposed of copper-coated mild steel. By increasingthe power from 280A and 28V to 300A and 30V, theconcentration of ferric oxide was reduced five-fold,the oxides of manganese, zinc, copper, and nitrogenwere reduced approximately ten-fold, and the ozonelevel increased by a factor of eight.

1.1.2 Effects of Electrode Composition

Several investigators compared the quantity offumes generated with different types of electrodes ofelectrode-base metal combinations (Refs. 3, 4, 5, 35,88 and 102). Eichhorn (Ref. 35) reported that acidrutile-coated electrodes produce less fumes than doacid or basic-coated electrodes. Of the electrodes

tested, the cellulose-coated electrodes used in open airpipeline construction generate the highest quantitiesof fumes. In general, the total quantity of fumes(FGR) generated increases with the diameter of theelectrode. However, in relation to the total amount ofmetal deposited, fume emission decreases withincreasing electrode diameter.

To determine the airflow requirements of ventila-tion systems, Alekseeva et al. examined the emissionsreleased during GTAW of manganese-copper alloyswith an argon shield (Ref. 4) and of copper with anitrogen shield (Ref. 3). In the first study, the quantityof emissions during welding of three different basemetals composed of alloys containing copper, smallamounts of aluminum, nickel, and iron and approxi-mately 43, 53, and 63% manganese, respectively, werestudied. The two alloys containing the higher percent-age of manganese also contained small amounts ofzinc. The filler wire was composed of a copper alloycontaining 43% manganese. The manganese contentof the fumes was directly related to the manganesecontent of the base metal and it varied from 2.4 to 26.2mg/ M3 in front of the shield and 0.8 to 2.2 mg/ M3

behind the shield. Manganese, copper and zinc oxide,but no aluminum or nickel oxides, were present in theweld emissions. For every alloy examined, manganeseoxide, but not copper or zinc oxide, reached levelsbehind the face shields that were higher than thepermissible exposure levels in the USSR. The alloyscontaining zinc gave off more total solid materialsthan did that alloy which lacked zinc. In the formercase, zinc represented close to 50% of the total particu-late material in the welding emissions.

In the second study (Ref. 3), emissions from fillerand base metal combinations with varying quantitiesof nickel and zinc were compared. As in the previoussystem, the total emissions were greatest when zincwas present in the copper alloy. In systems with littleor no zinc, copper oxide was the primary componentof the emissions that had to be considered during thedesign of ventilation systems. When zinc alloys wereused, zinc oxide levels were sufficiently high that theyhad to be considered as well.

A third study by Alekseeva et al. (Ref. 5) examinedthe concentrations of nickel in front of the welder'sshield during GTAW with an argon shield of eightsamples of welding wire which contained 2.2 to 70%nickel. The welding conditions were kept constantduring this test. Nickel was not found in the weldingfumes (lower limit of detection - 0.2 //g/ml sample)when the nickel content of the welding wire was lessthan 5%. When the welding materials contained nickelin concentrations equal to or greater than 5%, theamount of nickel in the fumes increased with the nickelcontent of the welding wire.

Pokhodnya et al. (Ref. 102) compared the releaseof iron and manganese from slags from rutile or basicelectrodes of differing alkalinities. The rate of man-

The Exposure 15

ganese vaporization, but not that of iron, increasedwith the alkalinity of the slag.

Oleinchenko et al. (Ref. 98) found that the quantitiesof soluble (NaFand KF)and insoluble (A 1F3) fluoridecompounds, as well as the quantity of gaseous fluoride(HF), increased with the moisture content of theelectrode coating (44% cryolite, 38% KC1, 15% NaC 1,3% silicon) during automatic aluminum welding.Precalcination of the welding electrodes at 380 to420° C significantly reduced emissions of the fluoridecompounds.

Buki and Feldman (Ref. 22) developed equationsfor estimating the concentrations of metal compoundsin the emissions generated by gas-shielded welding ofsteels and copper and aluminum alloys. Tests wereperformed which demonstrated that the calculatedconcentrations of a number of metal oxides (nickel,aluminum, manganese, copper, silicon, chromium,and iron) generated during welding of steel with stripelectrodes and during argon-shielded arc welding andplasma arc welding were in the same range as thosedetermined experimentally. The authors concludedthat their equations can be put to practical use, forexample, to obtain the initial data for calculating therequirements of ventilation systems.

1.1.3 Particles

The size and shape of airborne particles govern theiraerodynamic behavior within the respiratory tractwhich, in turn, determines whether or not they aredeposited in the lung. Spherical particles greater than5 /xm in aerodynamic diameter will generally betrapped within the upper respiratory tract (e.g. nasalpassages, sinuses, trachea) and expelled, whereasparticles between O.I and 5 /xm in diameter areconsidered to be respirable; that is, they can be inhaledand retained within the deep lung. Particles smallerthan 0.1 jim may not be slowed down sufficiently bymechanisms such as impaction and may remain withinthe airstream to be removed during the expirationcycle. The aerdynamic behavior of fibers such asasbestos differ from that of spherical particles, andfibers up to 20 /xm in length have been found withinthe lung.

In general, particles present in welding fumes arespherical. Irregular shaped particles may also bepresent in the air in welding shops, but these usuallyarise from grinding and other mechanical operations(Ref. 13). Most reports indicate that the greatestnumber of particles in welding fumes tend to havediameters which range from less than 0.1 Mm up to0.5 /urn (Refs. 23, 35, 51, 70 and 88). Johansson et al.(Ref. 70) reported that the mass median diameters(the total mass of the particles with diameters largerthan the mass median diameter is equal to the totalmass of the particles with diameters smaller than themass median diameter) of particles produced by

different welding processes ranged from less than0.25 /xm for fumes produced by GTAW to 0.3 to 0.6/xm for GMAW and SMAW of stainless steel.

The particles may exist in welding fumes as unique,single entities or they may be present as agglomeratesor chains (Refs. 12, 23, 51, 70 and 88). Carsey (Ref. 23)reported that particles in the solid phase of fumes fromSMAW of mild steel consist of two fractions; the firstis represented by single, nonagglomerated particles0.05 to 0.1 /im in diameter, and in the second, theparticles are agglomerated into aggregates 0.25 to0.5 /im in diameter. Particles from SMAW of stainlesssteel tended to be less heavily agglomerated and had amore even size distribution, which was in the range of0.05 to 0.4 /xm.

Other investigators (Refs. 51 and 70) reported thatmany of the airborne particles from SMAW of stain-less steel are present as chains or agglomerates ofsmaller particles. The count median diameter (halfthe total number of particles are larger in diameter andhalf smaller than the count median diameter) wasreported to be 0.15 /im by Gray et al. (Ref. 51) and tobe less than 0.1 /xm by Johansson et al. (Ref. 70). Grayet al. noted the presence of some large, "glossy coatedspherical particles" in the SMAW fumes that were upto 10 /xm in diameter. They also found that particlesfrom GMAW tend to be spherical and are present infumes both as independent entities as well as in chains.They postulated that the chains and free particles formby different mechanisms. Differences in the observa-tions of these investigators are probably due tovariations in the welding processes rather than theanalytical techniques employed.

A number of investigators have applied relativelynew or experimental techniques to examine theelemental composition of particles in welding fumes.These techniques included X-ray fluorescence, inwhich particle are irradiated with X-rays and theresultant fluorescing X-ray spectra are analyzed (Ref.23); Particle-Induced X-ray Emission (PIXE) in whichthe X-ray spectra emitted by particle bombarded withprotons are analyzed (Refs. 70 and 86); and scanningelectron microscopy coupled with Energy DispersiveX-ray Analysis (EDXA) in which X-ray spectra fromparticles irradiated with an electron beam areexamined (Refs. 81 and 133). In one study, ElectronSpectrocopy for Chemical Analysis (ESCA), whichalso provides information on the oxidation state ofelements, was used (Ref. 70). With this method, onlythe elements in a very thin layer on the particle surfacecan be analyzed.

Malmqvist et al. (Ref. 86) and Johansson et al.(Ref. 70) used Particle-Induced X-ray Emission(PIXE) to analyze the elements present in particles ofwelding aerosols. Thirteen different combinations ofelectrodes and base metals with SMAW, GMAW, orGTAW were used. With low alloy steels, fluorine,potassium, calcium (these three were with SMAW

4/EFFECTS OF WELDING ON HEALTH IV

only), iron and manganese were most abundant. Withstainless steel electrodes, chromium, nickel, and lowlevels of molybdenum were found. Aluminum oxidewas the main component of aluminum weldingaerosols and only very low quantities of a few otherelements were detected. The fumes generated bySMAW, GMAW, and GTAW were separated by acascade impactor and the elements in each particlesize range were analyzed. No marked differences wereobserved in the distribution of elements.

With SMAW using stainless steel electrodes,chromium, manganese and iron increased and potas-sium and fluoride decreased with increasing voltage.The difference may arise from the fact that potassiumand fluorine originate from the electrode coating whilethe other elements generally originate from the meltingmetal core (Ref. 86). The elemental composition offumes from GMAW of stainless steel did not vary withvoltage (Refs. 35 and 70).

Koponen et al. (Ref. 81), using scanning electronmicroscopy with EDXA to examine particles fromSMAW of stainless steel, found that nickel may existas discrete particles. Chromates appeared to bepresent as condensates on the surface of metal oxideparticles. Using ESCA, Johansson et al. (Ref. 70)found that 60 to 100% of the chromium on the particlesurfaces is in the hexavalent state with SMAW ofstainless steel, while less than 15% of that on theparticles generated by GMAW is hexavalent. Washingthe SMAW particles with a pH 7.4 buffer significantlyreduced the chromium on the outer surface.

Barfoot et al. (Ref. 13) studied the variation withtime of day of the chemical and physical character-istics of particulates and pollutants in a welding shopusing a streak sampling method. Attempts were madeto correlate analyses with activity in the shop. Theanalyses were done in a welding shop ventilatedthrough ceiling ducts. Occasional melting and grind-ing operations were also performed on the premises aswell as argon arc, oxyacetylene welding, and brazing ofa variety of metals. Samples were collected at adistance of about 2 m from the welding table and 1 mabove the ground. PIXE and scanning electronmicroscopy with EDXA were used for the determina-tions.

Amounts of elements such as aluminum, silicon,zinc, sodium, potassium, and calcium varied dramati-cally (up to three orders of magnitude) with time, andpollution levels were, on average, ten times higherduring the day than at night. Most of the particles wereless than 3 /zm, and a large fraction were betwen 0.5and 1.5 im in diameter. Iron and aluminum wereoccasionally found in discrete spherical particlescomposed of a single element, and with a meandiameter of about 1 /im. Silicon and aluminum weresometimes found in needle-like shapes of up to 2 fim

in length. The other particles contained combinationsof two or more elements. Many irregular shapedparticles were found, some of which could be associatedwith the grinding operations.

1.1.4 Chromium

Much emphasis has been placed on the developmentof reliable methods for determining the concentrationof hexavalent chromium (Cr-VI) in welding fumes,primarily because of reports that it may be carcino-genic for humans (Ref. 91). Although an increasedrisk has only been demonstrated in workers in thechromium industry and not in welders, new stringentthreshold limit values (TLV) have been proposed byNIOSH (Ref. 91) which would have industry-wideeffects. Totally reliable methods for the extraction ofCr-VI from all types of welding fumes have not yetbeen established (Ref. 52). The potential need forcompliance with strict standards and concern forworker safety make it necessary to be able to evaluateconfidently the levels of hexavalent chromium inwelding fumes.

An "interlaboratory round robin" study, organizedby the American Welding Society and described byAndrews and Hanlon (Ref. 11), evaluated techniquedeveloped by Blakeley and Zatka (Ref. 19) for thedetermination of Cr-VI in welding fumes. Similar tothe technique described by Thomsen and Stern (Ref.129), the Blakeley-Zatka method relies on the ex-traction of chromates with sodium carbonate at highpH to avoid the reduction of Cr-VI to Cr-III by Fe-II,which can occur under the acid conditions describedin the method recommended by NIOSH (Ref. 91).

In this multiple laboratory study, bulk fumesamples collected from GMAW and SMAW methodswith stainless steel consumables were analyzed using aslight modification of the Blakeley-Zatka method; thefinal chromium determination was made by atomicabsorption spectrophotometry. The bulk of thechromium in GMAW fumes was not hexavalent; thesefumes contained very low quantities of water solubleand insoluble Cr-VI. On the other hand, most of thechromium in SMAW fumes was soluble Cr-VI, withsmaller quantities of non-Cr-VI and very low quanti-ties of insoluble Cr-VI. A second round of testing wasdone with a fraction of SMAW fumes collected bypolyvinyl chloride membrane filters. The averagevalue for water-insoluble Cr-VI was slightly higher,and that of non-Cr-VI slightly lower than thatobtained by the participating laboratories for the bulksample. The results of the tests of SMAW fumesobtained by six of the laboratories are shown inTable 1.

The Exposure!5

Table 1Chromium content of SMAW fumes collected

on polyvinyl chloride membrane filters

MEAN(RelativeStandard

Deviation)

Soluble Cr-VI

Insoluble Cr-VI

Non-Cr-VI

Total Cr(Sum of above)

1

4.35 4.72 4.86 4.48 5.02 4.26

0.81 0.12 0.36 0.37 0.19 0.34

0.43 0.37 0.94 1.38 0.71 0.48

5.59 5.21 6.16 6.23 5.92 5.08

4.62(6.5%)0.365(66%)0.718(54%)

5.70(8.6%)

Andrews and Hanlon Ref. 11

Andrews and Hanlon concluded that the potentialfor reduction of Cr-VI during sampling can be mini-mized by collecting samples on inert filters, storingsamples under dessication, and performing analyses assoon as possible after collection. Alterations that canbe produced in the Cr-VI content of fume samples bydifferent collection and storage methods are furtherdiscussed by Gray et al. (Ref. 49) and by Thomsenand Stern (Ref. 128).

Preliminary results of a similar interlaboratorystudy, organized by the International Institute ofWelding, were reported by Gray et al. (Ref. 51).Cr-VI was extracted from the fumes with a solution ofNa2CO3 and NaOH. Some variation was observedbetween the results obtained by the four participatinglaboratories, as in the preceding study. Fumes fromSMAW with low hydrogen or rutile covered electrodeshad considerably more Cr-VI but less total chromiumthan did fumes from GMAW of stainless steel.

That there is substantially more Cr-VI in emissionsfrom SMAW than from GMAW of stainless steel wasalso reported by other investigators (Refs. 114 and133). In the latter case, the Cr-VI content of the fumemay vary with the welding parameters; fumes producedby GMAW with short circuiting transfer arcs werereported to contain significantly more water solubleCr-VI than fumes generated by spray arc conditions(Ref. 128). As described by Koponen et al. (Ref. 81),with SMAW of stainless steel using basic electrodes,the chromium in the slag becomes oxidized to Cr-VIwith the oxidation of the slag forming elements andthen reacts with alkali oxides from the electrode coat-ing to form soluble chromates (e.g., CaCrO4, K2Cr207)which condense on the surface of tne slag and metal

oxide particles. The absence of both oxygen and alkalioxides in welding with inert gas shields contribute tothe absence of soluble Cr-VI in emissions generatedby the GMAW method.

Because the carcinogenicity as well as other formsof biological activity may vary with the solubility ofCr-VI compounds in aqueous media (Ref. 91), thedetermination of water soluble Cr-VI is becomingroutine in the analysis of stainless steel fume samples.Gray et al. (Ref. 52) discussed problems inherent in theavailable methods for the analysis of total chromiumand the determination of soluble Cr-VI. These authorsdescribed a method which can more completely dis-solve fume samples in preparation for analysis byatomic absorption spectrometry than can the standardmethods which may leave a solid residue with somefume samples. They also demonstrated that, for thedetermination of the soluble Cr-VI compounds,different methods will lead to varying results, and theystressed the need for standardizing both the definitionof Cr-VI solubility and methods for its determination.

1.2 GasesThe major gases that are generated during welding

of umprimed metals include ozone, carbon monoxide,carbon dioxide, and the nitrogen oxides. These gasesarise from interactions between ultraviolet radiation,intense heat or electric currents with normal atmos-pheric constituents, and also with some shieldinggases (e.g., carbon dioxide). As described below,recent research has been performed on the generationof ozone, the nitrogen oxides, and carbon monoxideduring welding.

6/EFFECTS OF WELDING ON HEALTH IV

1.2.1 Nitrogen Oxides

At temperatures greater than l000°C, molecularnitrogen in the air is coverted into nitric oxide (NO).NO is readily oxidized in air at ambient temperaturesto form nitrogen dioxide (NO2). Although nitrogenoxides (NOJ may be produced during most weldingprocesses, the levels of NOX are greatest during gaswelding and plasma cutting (Ref. 107).

Press (Refs. 105 and 106) examined levels of nitrogenoxides produced by gas welding processes. As pre-viously reported, (see Effects of Welding on Health,Volume III), his studies demonstrated that the quantityof nitrogen oxides produced is proportional to thelength of the flame. A family of curves for differentflame lengths was developed, showing the variationin the nitrogen oxides produced per unit time withdifferent size nozzles. The quantity of nitrogen oxidesincreased with the size of the nozzle and the length ofthe flame. With small variations, the curves weresimilar for welding with oxyacetylene, propane, andnatural gas. A practical comparison between thequantities of nitrogen oxides produced with differentfuel gases could not be made because nozzles of thesame size do not necessarily give comparable per-formances with different gases. In an experimentalsetting, where the same heat transfer to the workpieceand equal temperatures at the underside of the platewere achieved, the least amount of nitrogen oxides wasproduced with propane gas.

1.2.2 Ozone

Ozone is produced by the interaction of ultravioletradiation of wavelengths lower than 200 nm withoxygen. Ozone is readily decomposed in air to molecu-lar oxygen. The rate of decomposition is acceleratedby metal oxide fume which also prevent its formationby blocking the transmission of ultraviolet light sothat ozone presents little problem to welders usingFCAW or SMAW.

Farwer (Ref. 39) reported that ozone does notpresent a hazard, even without local fume extraction,with GT A W of Cr-Ni steel or aluminum. With GT A W,a helium shield resulted in higher levels of ozone thandid an argon shield. Significant quantities of ozone aregenerally formed during GMAW. The highest levelsof ozone are produced by argon-shielded welding ofaluminum (Ref. 133). With this welding process,ozone is more likely to be a problem, relative topermissible levels, than other emissions.

A factor which may reduce ozone levels is thepresence of nitric oxide (NO) which reacts readily withozone to produce molecular oxygen and nitrogendioxide (NO2).

NO + O3 -> O2 + NO2

Because of this, the addition of small quantities ofnitric oxide in shield gases has been recommended to

reduce ozone levels during SMAW (Refs. 43 and 117).However, Farwer does not agree with this method andhe argues that the reaction may not be rapid orcomplete enough to effectively reduce ozone levels.In support of this, he cites evidence that significantquantities of NO and O3 may be found together inwelding aerosols (Refs. 15 and 40). He suggests thatanother reason these two gases may coexist in the weldarea is that NO2 may be photochemically convertedback to NO by the action of ultraviolet light. Since NOis readily converted to NO2 which, because of itsstability, represents a greater hazard than ozone,Farwer cautions strongly against the use of NO as anadditive in shield gases to control ozone levels.

A difference between the generation of ozone andother welding emissions is that ozone forms not onlyin the immediate vicinity of the arc, but also furtheraway from it, albeit in decreased quantities. Accordingto Farwer, the addition of additives to the argon shieldmay actually change the spatial distribution of ozoneproduction and although it may decrease the ozone inthe vicinity of the arc, it may increase the ozoneconcentration in the breathing zone of the welder(Ref. 39).

In contrast to the report by Farwer, Johansson(Ref. 71) reported that the levels of ozone duringGMAW and GTAW welding could be significantlyreduced by inclusion of helium or hydrogen in theshield gases. For GTAW welding of stainless steel withan argon shield, the ozone level decreased as the con-centration of hydrogen in the shield gas was increasedfrom 0 to 5%. For GMAW welding of stainless steel,the ozone concentration was reduced by 15% when theshield gas was changed from 98% argon + 2% CO2 to63.8% argon + 3.2% CO2 + 32% helium + 1% hydrogen.In this work (Ref. 71), measurements were made atvarious distances from the arc. The concentration ofozone in front of the welder's helmet, resulting fromGMAW of aluminum, decreased markedly as thehelium concentration of the argon shield was increasedfrom 0 to 30%. Further reductions in ozone levels werefar less dramatic as the helium concentration wasincreased to 100%. Differences between Farwer's data(Ref. 39) and that of Johansson (Ref. 71) cannot beresolved with the information available, but Johasson'swork is in accord with other reports that shield gasadditives can effectively reduce ozone levels (Refs. 41,43 and 117).

Smars (Ref. 117) found that varying the amount ofhelium from 0 to 100% in the argon shield with GTAWof aluminum had only a minimal effect on the ozoneproduced. A slight increase in ozone levels in theregion of the weld occured as helium increased from0 to 30% of the shield gas, and a small decrease in theozone level occurred as the helium concentration wasincreased from 70 to 100%. The nitrogen dioxide levelsvaried in the opposite direction. On the other hand, hereported that inclusion of nitric oxide in the shield gas

The Exposure 17

significantly reduced ozone levels.Smars (Ref. 117) also found that ultraviolet radia-

tion and ozone, but not nitrogen dioxide increase withthe arc length, and that as the arc current is increasedfrom 150 to 300A, ozone levels reach a maximum be-tween 225 and 250 A. Welding rods of AlSi5 generated themost ozone followed closely by nonalloyed aluminumrods. The ozone levels generated by welding rods con-taining magnesium (AlMg3 and AlMg5) wereapproximately half that of rods containing silicon.Similarly, Frei et al. (Ref. 43) reported that ozonelevels were much lower when gas metal arc weldingbase metals composed of AlMg5 than when weldingAlSi5.

1.2.3 Carbon Monoxide

Carbon monoxide forms during the incompletecombustion of carbonaceous material and is generallynot formed in significant quantities during SMAW ofuncoated metals with the majority of electrodes.Carbon monoxide can be produced during GMAWwith a CO2 shield (Refs. 39 and 119), and occasionallyelevated levels with peak readings of 150 ppm wererecorded with this method (Ref. 133). Significantcarbon monoxide levels can also form by combustionof the welding gases during oxyfuel gas welding.

1.3 Protective CoatingsSubstances released during welding of primed or

painted metals and from degreasing agents have, inrecent years, been recognized as a potential source oftoxic materials and have received increasing attentionin health and research studies of the welding profes-sion. The gases and organic vapors emitted frommetals with protective coatings were examined byBohme and Heuser (Ref. 21) using GTAW without afiller metal to minimize airborne contaminants frommaterials other than the coatings. Of the componentsmeasured in the welding emissions, carbon monoxide,the nitrogen oxides hydrogen cyanide (HCN), formal-dehyde, and toluene diisocyanate (released frompolyurethane binders) were found to be the mostimportant when considered in relation to MAK (maxi-mum safe workplace concentrations) values establishedin the Federal Republic of Germany (see Table 2 forTLV's and MAK values for some substances found inwelding fumes). HCN and carbon monoxide, whicharise from binders in the production coatings, wereproduced in the least amounts by ethyl silicate and thehighest amounts by epoxy resins. The quantity ofHCN, which was proportional to the amount of binderin the coating, was also low with alkyd resins and highwith polyurethane resins. Although levels of nitrogenoxides were significant, they were not influenced bythe coatings.

Levels of phenol released from phenol resins werelow in relation to MAK values. Of the aldehydes

detected, butyraldehyde and acetaldehyde were presentin concentrations too low to be of concern in theworkplace. On the other hand, formaldehyde, whichhas a lower MAK value (1 ml/m3) than the otheraldehydes, was present in significant amounts.Formaldehyde was not released in detectable quanti-ties from binders containing zinc dust, which wasattributed to the low percent of binders in zinc-containing coatings but, as the authors suggested, mayalso have been due to an inhibition of formaldehydeformation by the zinc.

The levels of gases were measured in front of andbehind a face shield placed at a normal workingdistance from the arc. In front of the face shield, theMAK levels were exceeded for carbon monoxide,formaldehyde (this was high for only one of thecoatings tested) and toluene disocyanate (this wasthree times the MAK level with the one polyurethaneresin tested). Although none of the gases exceededMAK levels behind the face shield, the investigatorswarned that care should still be exercised when weldingover protective coatings since the thickness of thepaint films often exceeds that suggested by the manu-facturer, especially in corners and with manualpainting. The resultant increase in emissions couldeasily exceed MAK values behind the face shield.

The fumes emitted from steel plates coated with 29shop primers or components of primers (pigments -ferric oxide, zinc, aluminum, zinc chromate; ex-tenders - barium sulfate, magnesium silicate, calciumcarbonate; and six different binder systems) duringGTAW welding were examined by Koerber andFissan (Ref. 79). With the exception of primerscontaining zinc pigments, the total mass of the emis-sions was relatively low compared with those releasedfrom SMAW or GMAW with a CO2 shield of un-painted steels. Zinc oxide represented the bulk of theemissions from zinc primers and, according to theinvestigators, the total quantity of emissions fromthese primers was similar to those typically releasedfrom SMAW or CO2 welding of uncoated steels.

Low levels of lead and cadmium were released fromall primers; the primers which produced the highestlead levels contained high contents of zinc or ferricoxide and the zinc-containing primers produced thehighest cadmium levels. More of the particles gene-rated from the primers had diameters greater than0.3 Mm than did those from uncoated steel. Most of theparticles were smaller than 1 urn and their massmedian diameter was approximately 0.1 ,.m.

During a field study of fumes released from SMAWand GMAW of non-alloy steel either unprimed orprimed with a thin film of red shop primer containingthe pigments zinc tetraoxychromate and iron oxide,Ulfvarson reported that iron, zinc, and chromiumwere released in greater amounts from primed steel, aswould be expected (Ref. 133). However, manganeseand copper also increased, even though these elements

8/EFFECTS OF WELDING ON HEALTH IV

O)

I

M

X CO

| E•a CO

(0CO

CO

a>~. 73CM • *

abl<

ited

•" cD

in t

he/e

ls

.2

tanc

CO

3CO

O

25

man

y

a>O

"o

3 2W

|

a> c•o —

31caoo

8V

'5. uO c«

BO

uVrv

ISsi

guBOI

T3O

M J=

•§

8.Ss

B•C

C8

a.

EuO

.5 B

SIa*

IIo oa. a.E Eo oCJ c j60 60B .E'?>'>>'O O

< u a,M i> i> -5 eo •= -a <e E

| | |B E •£ 5 2 58 8-5-5 I . «

B >>

CJ CJ

u, iiO cu

.b o. :

S

s gflss.il1313 a a s.

IIIca ? 0

I 2>. x Sx o -5n. x ¥&> a> >

E B Bcu u cac E gOH O-4-tE E °o o *j

60 60 3E E .«

•>/>* toO O E

S S3 Oca ca o

E g E

•at fM O O

E E n

o ir> m

"60 M E £ - ^ M.P £ I r 'c cjfi E bo M

- * E : E - E E

-S .EBE S60 — - — - - - - -C 60 60 60 00

.„ E E E E

23

a,

#B

T3

o'S*360E

•312!

E'3

B3

BCO

O

Eocj60B

' > ,O

BO

aoCJ00B

'>.o

£ 1qd

E

"h

0 C E M M E

i ! 2 : E E -

OXI

1 J3

3D-E

x;n.

a.ME'S'3Eou

E 3

si11

E geu O00 'C" u u60 " S

S 2 Ss^3o

8 g2 <j

oo n. E•S M §

2.5'S

- ^ ca

ill

•g.s -* 8£E & ^

<2 O

E 5 ^ E ^ ir

»O w O V> O O

u

oa.IIn3a.

2Len

ca

S « §§ 1 1

E Eo oK to « M M M 3M 3 g E.E B.-S

S1

o S_ ca .5 .a .s

•a .5 S o oB « as a(9 ? A Ig

"oj~ao E ^6E E - E

11 g- E o.O. Q, D.

— D.—

a.

— d o o o d. . . - oooooo<Ncs — wio —

•3 Iged ^ oj

E3

U i

E3

Ecau

ide

X0

•5c0

-DU i

caCJ

cu

"x0BO

Ec0

Xuca0

Io cau. -D

X OCJ cj

cu

o- 2Sc

1-12 S>I 2

5

cu

XoE T3o cau. cu

•ac3oa.Boo•acCB '

•a to

11.E «

CJ

ca ^ ^ ^

oj3 3

3 —

i> caB 0060 E

11

3 °O ^ ,

E E

cuT3

_>>"0E

cuT3

cu

ce

J3

'B EO u

'S 'B

uB

— euCU O &Qc c KO cu ON Xi X\O CX D ,

Q-cnOXC

e l l2'•ate

'£ '£ > H

oO

Bu

1oUBC3

CU

<

. "8E Oo -a

Eca

Ui

eu en

o ao c2.a -a•= E

C83a.cu

2cu

2

3•a

oo

S E2 &

- ? -S

2 2•S 2

2 S

E « wcu 3 *-•

• - E

1

u3ca

_o'caVIE

OO

cuX

O

Hcu

0JCJCO

cu cu

op g,5

•asS 6

O o

a. a.S 2 2.2? a a

E -g S

111

u 5

3 a:

E o5 5a ou u< <

The Exposure 19

were not present in the primer. An explanation offeredby the investigator for this observation was that largerdiameter electrodes with a higher current are used forwelding primed metals, releasing more metal oxidesfrom the base metal. However, it was also noted thatthe thickness of the welded materials were notrecorded, and a higher current would have been usedhad the thickness of the primed metal pieces beenlarger than that of the unprimed metals, which wouldhave caused the same effect (Ref. 133).

Kireev et al. (Ref. 78) analyzed the emissionsproduced by GM AW with a CO2 shield on steel platescoated with four different protective primers. Man-ganese oxides, styrene, and unsaturated hydrocarbonswere the components of the emissions of greatestconcern generated by welding of the Russian primercoats, EF-0121, VL-023, MS-067, and EP-057. (Thecomposition of the primers was not stated.)

1.4 Electromagnetic RadiationElectromagnetic radiation in the ultraviolet, visible,

and infrared region of the spectrum is produced duringmost welding processes. Ultraviolet radiation shorterthan 175 nm is readily absorbed by oxygen, and istherefore, dissipated close to the weld. Ultravioletradiation of longer wavelengths is less readily dissi-pated and can be a problem with some weldingprocesses. Ultraviolet radiation tends to be insignifi-cant during SM AW because its transmission is readilyblocked by the heavy fumes. Of the commonly usedwelding processes GMAW, especially of aluminum,produces the most intense ultraviolet radiation.

Bartley et al. (Ref. 14) measured the intensity of theultraviolet radiation with wavelengths greater than250 nm produced by helium shielded GTAW andargon-shielded GMAW of aluminum. In this work,they were most interested in radiation in the vicinityof 270 to 280 nm since ultraviolet radiation with thesewavelengths is most likely to react with genetic material(DNA) and, by implication, may be the spectral regionassociated with skin cancer. Their results showed thatwith GMAW the use of consumable electrode wireswith 5% magnesium produced markedly more radia-tion between 275 and 300 nm than did electrode wireswithout magnesium. With helium-shielded GTAWwithout filler metal, about ten times more ultravioletradiation in this wavelength range was produced whenwelding aluminum plates with 2.5% magnesium thanwhen magnesium was not present. The latter findingsare in contrast with the work of other investigators(Refs. 39, 71 and 117) who found that ozone, whicharises from the interaction of ultraviolet light withmolecular oxygen, is less intense with GMAW ofaluminum containing magnesium than when weldingunalloyed aluminum. Since Bartley et al. (Ref. 14)measured only the intensity of ultraviolet radiationwith wavelengths between 250 and 350 rm, the

apparent conflict between this work and that of theother investigators cannot be resolved.

Exposure to the electromagnetic radiation producedby welding in the ultraviolet (greater than 175 rm),visible (especially in the blue-violet range), andinfrared regions can cause severe skin and eye injuries.Glass or polycarbonate filter plates in welder's helmetscan effectively prevent injury from most electromag-netic radiation (as well as metal spatter). However,according to Pabley and Keeney (Ref. 99), whenmaterials (e.g., ferric oxide) which absorb infrared areincorporated into the filter plates, the absorbedinfrared may be re-radiated from the plates into thearea of the welder's eyes. Reflective metal coats (usuallygold, silver, or aluminum) on the outer surface of thepolycarbonate filter plate inhibit the absorption ofinfrared radiation and reflect much of it away fromthe plate. Pabley and Keeney compared the heatbuildup on green glass filter plates and reflective goldcoated polycarbonate plates during welding of safetyplates; the helmet filter plate temperature was 216%and 54% higher than ambient with green glass andreflective polycarbonate filters, respectively. With thesafety plates in place, the difference in the heat buildupon the back safety plate was far less dramatic, andthere was little difference between the two types offilters. Thus, reflective polycarbonate filters mayintroduce a real advantage, in terms of infrared radia-tion exposure to the welder's eyes, only when the backsafety plate is not in use.

To enable observation of welders for supervisory orsafety purposes, to reduce the welder's sensation ofconfinement, and to increase visibility within thewelding booth, Sliney et al. (Ref. 116) developeddesign criteria for a semitransparent curtain whichwould shield the bystander from hazardous radiationfrom the arc. The absorption of infrared radiation wasnot an important consideration since at the bystander'sdistance of 1 to 2 m from the arc, the intensity is toolow to be damaging. Thus, the primary concern in thedesign of the curtain was to filter out ultravioletradiation and the intense blue visible light from the arcthat is not necessary for visibility. Incorporated intothe curatin design are fluorescent dyes (to absorbultraviolet radiation and reemit visible light) andsubmicron zinc oxide particles (to absorb short waveultraviolet radiation, to serve as light scattering centersfor the diffusion of transmitted light, to reduce thebrightness of the arc spot, and to scatter the visiblelight returned to the curtain from the arc spot toincrease general illumination of the work area.)

Unstable arc temperatures, variations in arc length,drafts, and shifts in air currents cause fluctuationsin the electromagnetic radiation produced duringwelding. Gvozdenko et al. (Ref. 55) discussed theinadequacy of available instrumentation to measureaccurately the total quantity of electromagneticradiation produced with time by welding operations,

10/EFFECTS OF WELDING ON HEALTH IV

and presented a report of his study with a portableinstrument that had been developed in the USSRwhich integrates the total light output in the range of280 to 400 nm. He found that the strength of theradiation varied with the angle from the plane of thewelding arc.

1.5 NoiseIn a study of noise produced by various welding

processes, Hermanns (Ref. 65) reported that, with theexception of gas welding with a size 3 nozzle andGTAW, all processes studied (including gas welding,

SMAW, GMAW shielded with inert gases, CO2 ormixed gases, GTAW, and acetylene, propane, orplasma cutting) produced at least occasional noisesabove 85 dB (A) at a distance of 400 mm from thesource (Table 3). With GMAW, the noise generatedvaried with the electrode used and more noise wasproduced than with SMAW. Of the welding processesstudied, plasma cutting was the noisiest, and the noiselevel varied slightly with the metal being processed(plasma cutting of copper or aluminum was slightlynoisier than that of steel). Hermanns recommendedthat reduction of the size of the flame be used to reducenoise levels during oxyacetylene welding.

Table 3Noise levels of various welding processes*

ActivityPulse sound level

dB(A)

Gas welding, size 3 nozzle, 5 mm thick steel plateGas welding, size 4 nozzle, 5 mm thick steel plateGas welding, size 5 nozzle, 5 mm thick steel plateGas welding, size 6 nozzle, 5 mm thick steel plateSMAW, 150A, AC, Ti VIII s rod electrode,

core diameter 3.25 mmSMAW, 180A, AC, Ti VIII s rod electrode,

core diameter 4 mmSMAW, 180A, AC, Kb IX s rod electrode,

core diameter 4 mmSlag chipping on plate (200 mm x 150 mm x 10 mm)SMAW, 110A, core diameter 3.25 mm, pipe,

including slag chippingSMAW, 200A, core diameter 4 mm, fillet

weld, including slag chippingGMAW - mixed gas shield, 300A, wire diameter 1.2 mm,

spray arc, steelGMAW - CO2shield, 100A, wire diameter 0.8 mm, short arcGMAW using a pulsed arc (100 Hz), 200/300A,

wire diameter 1.6 mm, aluminumGMAW, 200A, wire diameter 1.6 mm, aluminumGTAW, 100A, rod diameter 2.4 mm, DCGTAW, 60A, rod diameter 2.4 mm, ACFlame cutting, acetylene, nozzle 10 to 25 mm, pipe,

6mm wall thicknessFlame cutting, propane, 10 mm platePlasma cutting, 100A, 10 mm steel platePlasma cutting, copper sheetPlasma cutting, aluminum sheetPlasma cutting, thick workpiecesArc-air gougingGrinding, manual grinderDressing, needle hammer

849297

103

86

84

86105

92

96

9791...95

951026574

9598

100100

>110>103

105103

•Measured at a distance of approximately 400 mm from source

Hermanns, Ref. 65

Effects of Welding on Human Health 111

2. Effects of Welding onHuman Health

Potential health hazards are associated with allaspects of the welding exposures described in Section 1,including gases, vapors, fumes, noise, and radiation.Other hazards associated with welding are burns fromhot flying metal pieces and the risk of electrocution(particularly for underwater welders). Welding ex-posures may directly affect the eyes (radiation,irritating gases, burns from flying metal particles), ears(noise, burns from flying metal particles), skin(radiation and burns), and musculature (strainedworking positions). With the exception of the latter,protective clothing and gear, when used properly, caneffectively protect the welder. On the other hand, thepotential health effects of welding emissions (gases,vapors and fumes) on the internal organs are moredifficult to assess and control. The chronic effects ofsome of the individual components of weldingemissions on humans are poorly understood and safeexposure levels cannot be determined with the avail-able information. Also, little is known about thecombined effects of many components of weldingemissions. Some of these may interact synergistically,making their combined effects greater than, or differentfrom, the additive effects of the individual components.Antagonistic effects are also possible, and the com-bined effects of two or more toxins could be less thantheir individual effects. These poorly understoodproperties of complex industrial emissions make itdifficult to determine realistic permissible exposurelevels for workers. For this reason, it is important forthe industry to maintain an active surveillance of thehealth research literature and to remain well informedregarding any reports of adverse health effects ofwelding. Described below are health reports whichappeared in the literature from December, 1980 (thetime of writing the Effects of Welding on Health,Volume III) to June 1982.

2.1 Respiratory TractThe respiratory tract represents the primary route

of entry of welding fumes and gases into the body.Toxic components of the emissions may have eithera direct effect on the lung or may be transported fromthe lung via the lymphatic and circulatory system toother parts of the body where their toxic effects maybe manifested. Examples of gases that directly affectthe lung are ozone and nitrogen dioxide which arehighly irritating and may cause pulmonary edemaafter acute exposures. Emphysema and fibrosis mayresult from chronic exposures to these gases. Carbonmonoxide has a systemic effect by combining withhemoglobin and interfering with the transport andutilization of oxygen. This will cause anoxia which canbe fatal with sufficiently high exposures.

As described earlier, a large proportion of the parti-cles in welding fume are respirable and may be de-posited in the lung. Most of the particles which enterthe lungs become ingested by macrophases within aday or two and will either be eliminated in this formby normal coughing mechanisms, or, for particles suchas ferric oxide and alumina, may remain quiescent inthe lung for long periods of time in pockets surroundedby these cells. Some particles may be transported fromthe deeper part of the lung to the lymphatic system.Relatively insoluble compounds, such as nickel oxide,may gradually become solubilized, leave the lung andlymphatic system, and become distributed to otherparts of the body.

2.1.1 PneumoconiosisPneumoconiosis is a general term for dust deposi-

tion in the lungs. Siderosis, or deposits of iron particles,is the most commonly occurring form in welders.Aluminum welders may have large deposits of alumi-num particles in the lung. Both iron (Ref. 32) andaluminum particles (Ref. 110), in the absence of otherforeign materials, are generally considered to bebenign. Thus, they may remain in the lung for longperiods without evidence of pathological changes intheir vicinity. Fibrosis, the growth and spreading ofconnective tissues with resultant formation of scartissue, results from the deposition of particles andfibers such as quartz (silica) and asbestos. This is apathologic condition which, when extensive, can causesevere impairment of lung function.

The pathogenicity of ordinarily benign particles,such as those composed of aluminum, may changewhen they are trapped in the lung in conjunction withother materials. Bauxite pneumoconiosis, or Shaver'sdisease, a condition found in some aluminum smelters,is associated with inhaled particles of alumina andcrystalline silica. The characteristic pathological lesionof Shaver's disease is a diffuse interstitial fibrosis asopposed to the focal or nodular type of fibrosis inducedby silica alone.

It should be noted that most of the fibrotic particlesreferred to above (e.g., asbestos and silica) are notproduced by welding and in those cases where theymay be present in the welding environment, they arisefrom other processes.

2.1.2 Estimation of Retained Particles in the LungA magnetic method for the measurement of

ferromagnetic particles in the lungs was developed byCohen a decade ago (Refs. 29 and 30) and has beenused by several investigative groups to estimate theretention of particles in the lungs of welders andpersons in other dusty occupations. With this method,the subject's lungs are magnetized with an externalmagnetic field. Then the field is removed and theremanent field produced by the magnetized particleswithin the subject's upper torso is measured. In a recent

12/EFFECTS OF WELDING ON HEALTH IV

study of asbestos miners and millers, Cohen andCrowther (Ref. 31) noted that those workers withwelding experience had, on the average, five timesmore ferromagnetic particles in their lungs than didnonwelders.

Using this method, Freedman et al. (Ref. 42) foundthat the amount of ferromagnetic minerals retainedin the lungs of eleven electric arc welders from theshipyard industry was several orders of magnitudegreater than that present in three machinists from thesame shipyard, 16 former asbestos insulators, and 24rural male controls. The amount of materials retainedin the lungs of welders correlated well with the numberof years spent in the welding trade as well as the extentof siderosis indicated by chest X-rays, but not withsmoking histories. The latter finding was unexpectedby the investigators since an earlier report had shownthat fewer particles are deposited deeply in the alveolibecause of constriction of the smaller airways in thelungs of smokers (Ref. 110).

Kalliomaki et al. (Refs. 73 and 80) used magneto-pneumography for estimating the amount of metalcontaminants retained in the lungs of workers per-forming SMAW of mild steel, stainless steel weldersengaged primarily in either GTAW or SMAW, iron

factory and foundry workers, and stainless steelgrinders. They found that stainless steel shielded metalarc welders had the highest mean quantity (approxi-mately 4 grams of dust per person), and mild steel arcwelders had the second highest mean quantity (ap-proximately 1 gram of dust per person) of retainedmetals in the lung (Table 4). Gas tungsten arc weldershad considerably less retained metals (0.2 gram perperson) than the other welders. The magnetic momentand size of particles generated by GMAW, GTAW,and SMAW of stainless steel and SMAW of mild steelwere examined, and it was found that the remanentmagnetic field was indirectly related to the particlesize. GMAW and GTAW generated the smallestparticles which produced the greatest remanentfield (Ref. 74).

Later, Kalliomaki et al. attempted to correlate levelsof metals retained in the lungs with chromium andnickel levels in the urine of 83 SMAW and GTAWstainless steel welders (Ref. 75). Only the SMAWwelders had elevated urinary chromium levels. Theirurinary chromium levels were significantly related tothe quantities of the metals in the lung. Urinary nickellevels were only slightly elevated in the shielded metalarc welders and were normal in the gas tungsten arc

Table 4Groups, numbers, exposure time and the approximate amount of metalcontaminants in the lungs of groups exposed to different metal aerosols

1. Mild steel arcwelders fromtwo shipyards

2. Stainless steelwelders:GTAW

(10% SMAW)SMAW

(90% SMAW)

3. Workers in aniron factory

4. Foundry workers

5. Stainless steelgrinders

Kalliomaki et al, Ref. 73

N

44

21

29

27

10

21

Exposure

Mean

20

13

19

13

33

11

time (years)

SD

7

9

10

5

7

4

Amount

Mean

1.0

0.2

4.0

0.2

0.2

0.1

of dust (G)

Range

0.2...8.0

-

0.6...10

0.02...1.0

0.06...4.0

0.02...1.8

Effects of Welding on Human Health/13

welders. The investigators attributed the absence ofelevated urinary nickel levels in shielded metal arcwelders to the solubility characteristics of the nickelfumes. However, as discussed below, other investi-gators observed elevated urine nickel levels in stainlesssteel welders (Refs. 47 and 114). Urinary nickel valuesmay fluctuate markedly throughout the day, and 24hour, or at least 8 hour, samples should be collected toobtain accurate values. In the study performed byKalliomaki et ah, urine samples were collected onlyonce for each person at the end of the work week.

2.1.3 Pulmonary FibrosfsTwo cases of pulmonary fibrosis in aluminum

welders were recently reported. The first case involveda man, who had been an aluminum arc welder workingprimarily in confined spaces in the shipbuildingindustry for 17 years (Ref. 134). This man was report-edly exposed exclusively to aluminum during hiswelding career and was also a tobacco smoker. He hadreduced pulmonary function, and X-rays showed anarea of the lung with a dense peripheral infiltratesuggestive of obstructive pulmonary disease. This partof the lung was surgically removed, and histologicalexamination revealed diffuse and focal fibrosis withdense accumulations of macrophages containingmetallic material, shown by scanning electron micro-scopy using EDXA to be composed primarily ofaluminum. Pulmonary fibrosis is not generally associ-ated with exposure to aluminum vapors, and the in-vestigators do not exclude the possibility that otherexposures, such as cigarette smoke, may have con-tributed to the pathogenic processes observed in thelung.

The second case was a 35 year-old aluminum welderwhose clinical symptoms included moderate dyspnea(breathing difficulties) on exertion and decreasedpulmonary function indicative of fibrosis (Ref. 62).Hazy basal infiltrates and a lung opacity were evidenton chest X-rays. A biopsy showed consolidation inthe opaque area and diffuse chronic interstitialpneumonitis. EDXA indicated that there were largeamounts of aluminum in the lung. Small quantitiesof carbon dust, iron, and asbestos were also present.The authors concluded that the pulmonary fibrosiswas largely due to the aluminum deposits since carbonand iron are known to be relatively inert in the lungand the small amounts of asbestos appeared to be"insufficient to account for his disease". However, asnoted earlier, pulmonary aluminum deposits aregenerally benign unless they are associated with othertoxic materials.

Gobbato et al. (Ref. 45) histopathologically ex-amined the autopsied lungs from 17 individuals whohad been electric arc welders in shipyards. Six diedfrom nonpulmonary diseases, two from respiratoryfailure, eight from lung cancer, and one from meso-thelioma. More than half the lungs had fibrotic lesions.

Siderosis was observed in all of the lungs, sometimesin conjunction with infiltrates of inflammatory cellsand sometimes in areas of fibrosis. The areas ofinflammation and fibrosis always had associated irondeposits. These observations, as well as the largeproportion of respiratory tract neoplasms, led theauthors to suggest cautiously that there is a casualrelationship between the pulmonary deposition ofwelding fumes and the development of advancedrespiratory tract disease. These data should be inter-preted with restraint since (1) the manner in which thecases were chosen for study was not indicated andany selectivity in this process could bias the results;(2) no history of tobacco use was obtained by theinvestigators; and (3) asbestos bodies were observedin three of the lungs.

Zober (Ref. 147) reviewed 29 reports publishedbetween 1955 and 1979 in which a total of 47 casesof histologically verified fibrotic changes in the lungswere documented. The mean age of the subjects was51 years, with a range of 21 to 69 years. In 18 of thecases, pulmonary fibrosis was identified histologicallyduring autopsy and in 29 cases, fibrosis was identifiedin biopsies or specimens that were collected surgically.

According to Zober, each of the authors of the 29articles related the lung fibrosis to the welding ex-posure, usually because of the close proximity offibrotic lesions and iron deposits. However, there wasinsufficient information concerning the history oftobacco use, previous medical history, extra-occupa-tional and vocational exposures, and exposures topollutants other than welding fumes at the workplaceto categorically come to this conclusion. Among the47 cases studied, 13 individuals had previous pul-monary diseases (e.g., tuberculosis or pneumonia),and in 14 cases there were indications of asbestosis orsilicosis. Zober eliminated these and other cases inwhich there were no physiological symptoms ofreduced lung function and concluded that, at the most,20 cases remained in which welding fumes may havemade a significant contribution to the development oflung fibrosis. Zober stressed the importance of per-forming studies which examine sufficient parametersusing the appropriate methodology to enable com-parisons between observations of different investiga-tors at different occupational locations.

2.1.4 Acute Respiratory InfectionsBecause the inhalation of gases such as NO2

reportedly reduces the resistance to respiratory tractinfections, there has been some interest in whetherwelders have an increased frequency of acute respira-tory tract diseases. Hayden et al. (Ref. 60) found thatamong 258 welders in the engineering industry, workabsences due to upper respiratory tract infectionswere slightly, but significantly, greater than those ofcontrol workers. There was no difference in thefrequency of absence resulting from other diseases

14/EFFECTS OF WELDING ON HEALTH IV

between the two groups.These results were only partially corroborated in a

study by McMillan and Molyneux of workers in threeshipyards in Devonport, Gibraltar (Ref. 84). Absencepatterns of 533 welders were compared with twogroups of controls. Group I was composed of menwith intermittent exposure to welding fumes andconsisted of 517 boilermakers and 835 shipwrights;Group II consisted of men who were not exposed towelding fumes and included 999 electrical fitters,403 painters, and 765 joiners, all of whom were em-ployed in the dockyards for at least 6 consecutivemonths during a recent 5 year period. Even though themean age of the welders (44.8 years) was moderatelyhigher than that of controls in Group I (37.5 years) andGroup II (40.1 years), the absence rates were com-parable among all three groups. After the data wereadjusted for age, no substantial differences in theabsence patterns from upper respiratory tract infec-tions were seen. The number of episodes of lowerrespiratory tract disease was small in all groups andthe proportion of workers in each group who wereabsent due to lower respiratory tract disease wascomparable. However, welders and workers in ControlGroup I appeared to take longer to return to workafter illnesses of this nature than those who worked inthe cleaner environments (Control Group II).

The same workers were later evaluated in terms ofabsences due to respiratory tract illnesses and smokinghabits (Ref. 83). It was found that a higher proportionof welders smoked than did controls. Among the non-smokers in all groups, welders had the lowest frequencyand severity of absences due to upper respiratorytract diseases. There was no difference between theabsentee rates of controls and welders due to lowerrespiratory tract disease. Among the smokers, theseverity of absence due to all respiratory tract diseaseswas similar in all three groups. In the case of lowerrespiratory tract disease, smoking welders appeared tohave slightly more and slightly longer spells of absencethan did controls. McMillan attributed this to theirsmoking rather than to exposures to welding fumes,stating that smoking welders are less likely to returnto work before their symptoms are completely re-solved since "they are more susceptible to the irritatingand thus aggravating effects of welding pollutantsduring acute respiratory disease".

2.1.5 Lung Function TestsPulmonary function tests are used to detect disease

processes which restrict lung expansion or reducepulmonary elasticity. (The reader is referred toAppendix C of the Effects of Welding on Health,Volume I for more thorough explanation of the termsused in this section.) In brief, these tests measure, byspirometry, the air volume which can be inhaled orexpelled either forcefully or under normal breathingconditions. An obstructive or restrictive ventilatory

defect is a change in the ability to move air into andout of the lungs.

The forced vital capacity (FVC) is a measure of thevolume of air that can be expelled forcefully followinga maximal inspiration. This measure reflects resistanceto airflow in the lungs. The forced expiratory volumeduring the first second of exhalation (FEV,), amaximal mid-expiratory flow volume (MMFV), ormaximal expiratory flow rate (MEFR) measure thepercent of the total volume or the rate of air expelledduring a poriton of the expiratory cycle. Thesemeasures all serve as indicators of airway resistance.Lung function tests are frequently used to monitorindustrial workers for damage to the respiratory tract,such as changes in pulmonary compliance, and todetect early signs of pneumoconiosis or airway ob-structions.

A factor that must be considered in interpretingthese tests is the natural variation that occurs betweenindividuals and with age. Another problem is that thesmaller airways have little influence on the total air-ways resistance and, thus, lung function tests tend todetect only changes in the larger airways duringdynamic compression of the lungs. The value ofpulmonary function tests as a means of detectingpulmonary insufficiencies in welders has been con-troversial, and the results of tests by different investi-gators at different industrial locations are variable. Atthe very least, caution must be exercised in inter-preting the results of lung function tests since they maynot deviate from normal, even in the presence of smallpulmonary lesions apparent on X-rays.

In a study of workers in four shipyards, Schneiderand Rebohle examined pulmonary function in 432electric arc welders and 420 control workers who werenot exposed to welding fumes (Ref. 111). The weldersand controls had similar work loads, were matched onthe basis of age, and had no significant differences insmoking habits. The vital capacity, FEV, and mid-expiratory flow volumes and rates were the same forwelders and nonwelders. The investigators concludedthat, since welders are known to have a higher incidenceof chronic bronchitis and pulmonary siderosis, thevalue of pulmonary function tests for medical screen-ing of welders is questionable. However, data on thefrequency of bronchitis and siderosis in the examinedworkers was not given.

Cavatorta et al. (Ref. 25) examined pulmonaryfunction in welders who were exposed to relativelyhigh levels of chromium through the use of chromiumalloyed electrodes in an Italian machine shop for up to2 years (average exposure to chromium fumes was0.8 years). A total of 28 workers were examined whohad been welders for a total of 2 to 40 years, with anaverage total work experience of 19 years. Althoughchanges in respiratory function were related to theduration of time spent welding, a correlation betweenreduction of lung function and chromium exposure

Effects of Welding on Human Health 115

could not be established. As the investigators stated,this work may not have been conclusive because theexposure to fumes containing high levels of chromiumwas too short to have manifested toxic effects andalso possibly because the total number of workersstudied was insufficient for statistical analyses.

A comparison of the effects of welding and tobaccosmoke on lung function was carried out with shipyardworkers by Akbarkhanzadeh (Ref. 2). Pulmonaryfunction tests were performed with 209 welders and109 nonwelders. Approximately half of each groupwere cigarette smokers and the data were analyzedwith respect to age, the extent of tobacco use, and thenumber of years spent welding. Tobacco smokers hada significant reduction of FEV1, FVC, and T, (thetransfer factor in the lung for CO). Welding fumes alsoimpaired lung function; this effect was independent oftobacco use. The effects of smoking and weldingexposures on pulmonary function were approximatelythe same and no synergism was observed between thetwo exposures.

In a study performed in Japan (Ref. 97), exposure toozone was found to have a definite effect upon pul-monary function. Flow volume curves were obtainedfrom 68 healthy welders who had been exposed torelatively high levels of ozone while performingGMAW. The welding population was 20 to 29 years ofage and was categorized according to the ozone con-centrations in the areas in which they worked. Dustexposures and smoking habits were considered duringthe analysis of the data. Peak flow rate and maximumexpiratory flow rate at 75% of the vital capacity (V75)were decreased slightly in the welders. However, nocorrelation between these values and the ozone con-centration to which the welders had been exposedcould be made. On the other hand, the maximumexpiratory flow rate at 25% vital capacity (V25) ofwelders who had been exposed to ozone concentra-tions of 1.05 ppm or more was significantly lower thanthat of welders exposed to less than 1.05 ppm. Thedecrease in the V25 as well as the FVC was related tothe length of exposure to ozone.

Hayden et al. (Ref. 60) found that welders in theengineering industry had no significant abnormali-ties in lung function. The values of FEV,, PVC, peakexpiratory flow, and maximum expiratory flow at50% vital capacity (MEFV50) were the same among258 welders and 258 controls. The only differenceswere that the MEFV25 was lower in all welders and theMEFV50 was lower in welders who smoked tobaccoproducts than in other groups. No correlation couldbe made between the results of lung function tests andthe number of years spent welding, the work condi-tions (e.g., confined spaces, use of fume extraction),or the type of welding employed. These investigatorsalso found that, during an experiment with ninewelders, pulmonary function values were the samewhen measured before and after 15 minutes of con-

tinuous electric arc welding (Ref. 61).Zober examined 40 welders and 40 age matched

controls who were not exposed to welding fumes(Ref. 145). In each group, 18 of the subjects weresmokers. The welders excreted four times as muchchromium and nickel in the urine as did controls,confirming actual exposures to these substances.Acute, but not chronic, bronchitis was a more frequentcomplaint of welders than nonwelders. Among thenonsmokers, welders exhibited no increased abnor-malities in pulmonary function. More smokers hadabnormal forced expiration, FVC and MEFV25.27

than did nonsmokers, and these abnormalities couldnot be related to the welding exposures. Of thesmokers, six welders and two nonwelders had bron-chitic murmurs whereas signs of bronchitis were seenin only two controls and in none of the non-smokingwelders. The only significant pulmonary finding whichwas clearly related to welding exposures and not totobacco use was siderosis, indicated by small roundishshadows on chest X-rays. Zober concluded that the"greater incidence of significant findings was clearlyapparent in welders who smoked" and suggested thatthere was "probably a combined effect by tobaccosmoke and welding fumes". The groups studied in thiswork were relatively small, however, and furtherstudies using larger cohorts seem necessary beforesuch conclusions can be substantiated.

It is apparent from these and from earlier studies(see Effects of Welding on Health, Volumes 1, II andIII) that inconsistent results may be obtained withpulmonary function tests. Stern (Ref. 122) discussedthree factors which may confound the results ofpulmonary function tests. The first is the effect ofpopulation dynamics, whereby self-selection amongthe welding population may encourage workers whoexperience ventilatory difficulties to change occupa-tions. The second is the effect of smoking on pulmon-ary function. However, many investigators today areaware of this problem and attempt to control for theeffect of cigarette smoke in the analysis of their data.The third factor is a bystander affect. Stern cites theexample that many welders are employed in shipyardswhere there is known to be a higher incidence ofchronic respiratory disease, among all workers than inother populations. Thus, he states that some of thepulmonary function test results may be related toexposures other than welding at the place of employ-ment.

2.2 Cancer2.2.1 Lung Cancer

A positive association between welding exposuresand respiratory tract cancer has not been establishedalthough at least two components of some weldingfumes, nickel and chromium, have been associatedwith respiratory tract cancer in workers from otherindustries. Certain nickel compounds have been impli-

16/EFFECTS OF WELDING ON HEALTH IV

cated as causative factors in the development of nasalcarcinomas and lung cancer in nickel refinery workers(Refs. 33 and 101). However, not all the specific nickelcompounds involved in the development of thisdisease are known, and to date, evidence for an associ-ation between respiratory tract cancer and exposuresto nickel-containing welding fumes has not beendemonstrated. As with nickel, an association betweenrespiratory tract cancer and exposure to chromiumhas been recognized for many years (Ref. 124). Onlychromium in the hexavalant state (Cr-VI) has beenimplicated as an etiologic factor in lung cancer (Ref.93). As with nickel, a clear cut association betweenexposure to chromium in welding fumes and thedevelopment of lung cancer has not been established.The identity of the Cr-VI compounds that may beresponsible for the development of respiratory tractcancer is unknown, but the fact that the water-solu-bility of the Cr-VI may influence its bioavailability andhence its ability to cause cancer and other healthproblems has been the basis for regulatory decisionsinvolving acceptable exposure limits (Ref. 91).

To explore the effects of exposure to nickel fumes,Polednak (Ref. 103) examined the causes of mortalityof 1059 white male welders employed between 1943and 1973 at three plants in Oak Ridge, Tennessee.Approximately half of these welders worked at theGaseous Diffusion Plant where they welded nickel-alloy pipes and received high exposures to nickelfumes relative to the other cohorts who welded mild

steel, stainless steel, and some aluminum. Standardmortality ratios (SMR) were calculated by comparingthe number of actual deaths from different causeswith those expected on the basis of death rates in sex-and age-matched controls in the U.S. population.Histories of tobacco use could be obtained only forabout 30% of the study population.

The SMR's of all welders for deaths from all causesand for deaths from diseases of the respiratory tractand cancer did not significantly differ from those ofthe general U.S. population nor did they differ signifi-cantly when the two groups of welders were comparedwith each other. A nonstatistically significant excessof deaths from emphysema (SMR = 221) and lungcancer (SMR = 150) as well as a slightly elevatednumber of deaths from diseases of the nervous system(three deaths from brain cancer as compared with1.55 expected and three from neurological diseasesas compared with 1.89 expected) was observed whenthe entire cohort of welders was compared with thecontrol population.

No correlation could be made between nickelexposure and respiratory tract disease when the twogroups of welders were compared with each other. Theeffect of length of employment on the development ofrespiratory tract cancer was examined by comparingthe SMR for welders with more than 50 weeks exper-ience to those with less than 50 weeks experience(Table 5). For this comparison, the welders wereplaced into two categories; the first included all

Table 5Mortality from selected causes by length-of-employment

Selected Causes of Death

Length ofemployment as

a welder

All welders<50 wk as

welder (N=491)

>50 wk aswelder (N=568)

Nickel alloy welders<50 wk as

welder (N=255)

All causesNo. SMR

100 97(79;118)*

73 76(60;96)

44 84(61;113)

All cancerNo. SMR

21 110(68;168)

11 63(31;113)

9 94(43; 179)

Respiratorycancer

No. SMR

10 157(75;289)

7 121(49;249)

2 62

Respiratorydiseases

No. SMR

9 176(81;334)

4 86

3 116

>50wkas 40 80 8 90 5 175 2 83welder (N=281) (57;109) (39; 177) (57;408)

•Confidence limits (95%) for SMR, if the observed number of deaths is five or more.

Polednak Ref. 103.

Effects of Welding on Human Health /17

welders in the study, and the second consisted of thewelders exposed to nickel fume. The only SMR thatcorrelated with the duration of welding experience wasthat for respiratory tract cancer in the group of weldersexposed to nickel fumes, but the confidence limitswere high. Polednak concluded that the sample sizeand the length of follow-up were insufficient to assessthoroughly the risk of respiratory cancer in welderswith nickel exposure. Because a small but insignificantexcess of diseases of the respiratory tract and nervoussystem was observed among welders, Polednak con-cluded that follow-up studies of this and relatedgroups of welders are warranted. This was supportedby Gibson (Ref. 44) who noted that in Polednak'sstudy, the "cut-off date for the ascertainment of vitalstatus was January 1, 1974, and at least six additionalyears of follow-up are already available". In thewelding population studied, 53% were hired before1950 and had a 23 year follow-up period while another40% were hired before 1960 and had a 13 year follow-up period. Cancer of the respiratory tract is generallythought to have a latency period of 20 to 30 yearsbeyond the initial exposure to carcinogens. Therefore,a more conclusive analysis could be obtained byre-evaluating the causes of death among this popula-tion at a later time. The groups examined in this studyare of particular value since welders from the samegeographical location, with and without exposure torelatively high concentrations of nickel fumes, can becompared.

A cohort of 234 men who welded stainless steel forat least 5 years between 1950 and 1965 was selectedfrom eight companies in Northern Sweden (Ref. 115).Their death rates from all causes did not differ fromthose of the general Swedish population. However,three welders died from lung cancer as opposed to0.68 expected deaths (p = 0.03). Smoking historieswere not obtained in this study, but the investigatorscited other studies which indicated no importantdifferences between smoking habits of welders and thegeneral population. Sjogren suggested that the excesslung tumors may be related to exposure to hexavalentchromium and stressed the need for follow-up studieswith larger populations of stainless steel workers.

Working on the hypothesis that the greater the ex-posure to pulmonary carcinogens, the earlier the age atwhich a person will succumb to lung cancer, Ahonen,using death certificates, determined the mean age ofdeath of 1820 Finnish workers in different occupations(Ref. 1). The mean age of all cases of death from lungcancer was 65 years. Plumbers and insulators who,according to Ahonen have a greater than normal riskof lung cancer, had the most significantly depressedage of death from this disease. Welders and Smiths didnot deviate from the average age of death from lungcancer, suggesting that their risk of dying from thisdisease is no greater or less than that of workers inother occupations.

Beaumont and Weiss (Ref. 16) examined thenumber of deaths from lung cancer among 3247welders who were selected from the InternationalBrotherhood of Boilermakers, Iron Ship Builders,Blacksmiths, Forgers, and Helpers in Seattle, Wash-ington. These welders were part of a larger study forwhich the number of deaths from diverse causes wasanalyzed earlier (Ref. 17 and Effects of Welding onHealth, Volume III). The welders worked in thisoccupation for at least 3 years between 1950 and1976 and, for each, the initial day of exposure wasbefore January, 1950. Of these welders, 529 haddied by 1976. Their lung cancer rates were determinedfrom death certificates and compared with age, sex,and race adjusted statistics for the total U.S. popu-lation and with the lung cancer death rates of 5432nonwelders from the same union. Fifty of the weldersdied from lung cancer as opposed to 38 expected(SMR = 132, p = 0.06). The number of lung cancerdeaths became statistically significant only whendeaths occurring 20 or more years from the first dateof employment were evaluated (observed: 39, expected:22.4; SMR - 174, p < 0.001). No information con-cerning the smoking history of the study cohort wasobtained.

A basic problem with health studies in the weldingindustry is that the occupational conditions are ex-tremely heterogeneous; work conditions vary widelyat different industrial locations, and the day-to-dayvariations in the working conditions at any one placemake the occupational history of individuals difficultto document. Thus, it is important that the workingconditions of study populations be carefully described.Zober (Ref. 146) examined 25 epidemiological studiesof the chronic effects of welding on the respiratorytract which had been performed in 14 differentcountries between the years 1965 and 1980. He statedthat variations in the design of the studies and insuf-ficient background information made it difficult tocompare them or to come to meaningful conclusions.

Some of the areas which Zober indicated wereinadequately investigated or documented in many ofthe studies examined were: the descriptions of thewelding processes used, dust levels in the breathingzones of the workers, duration of exposure, history oftobacco use, and the age of the welding and controlpopulations. In many studies, the number of weldersor the number of controls examined was too small toyield statistically meaningful results. Different para-meters were examined in the various studies (e.g., chestX-rays were evaluated in eleven studies, clinical exam-inations were performed in three studies, examinationsfor bronchitis in one study, and pulmonary functiontests in six studies). He concluded that the parametersstudied and the analytical methods used should bestandardized in order to make meaningful comparisonsbetween welding populations possible.

In another critique of published investigations of

18/EFFECTS OF WELDING ON HEALTH IV

the respiratory tract disease incidence among welders,Anderson (Ref. 9) concluded that pulmonary diseases,including chronic bronchitis and cancer, are far morelikely to be due to cigarette smoking and asbestosexposure than to welding fumes. In studies wherewelders in shipyards are compared with nonwelders inthe general population, the data must be suspectbecause of the asbestos exposures that were frequentlyencountered in shipyards. Comparisons betweenage-matched workers in occupations other thanwelding within the same shipyard are more meaningful,because occupational exposures to contaminantsother than welding fumes should be similar for boththe welding and control populations. He concludedthat "epidemiological studies must be controlled forsmoking, for the use of drugs, for asbestos exposure,and for each industry in which the welder has worked,even if no welding was done in that industry". Pros-pective epidemiological studies should includequantitative measurements of the exposure to weldingfumes, and care should be taken to control the habitsof individual welders which may introduce variabilityinto their total exposures.

Although these criticisms are valid, only a few ofthe requirements of these reviewers can be met in deathcertificate studies. Accurate data concerning smokinghabits, exposure levels, extra-occupational andbystander exposures, physiological status, etc., canonly be obtained with prospective studies; however,such investigations are enormously expensive. Deathcertificate studies, with their known limitations (e.g.,a secondary cause of death may be listed on the deathcertificate; personal histories and habits are difficult toobtain), can provide an overall picture of the healthstatus of a specific population and are useful fordesignating populations with environmental oroccupational exposures in which health problems may

exist and for which more intensive study is indicated.Stern recently reviewed 19 published epidemiologic

studies of welders. From the combined results of thesestudies, he calculated a small increase in the lung cancerrisk (1.3 : 1) for welders in general (Ref. 120). Afterexcluding the effects of tobacco smoking and shipyardemployment where bystander exposures to asbestosand possibly other carcinogens frequently existed,Stern found that there was still a small increase in thecancer risk for welders. Arguing that chromium andnickel from welding fumes have the same toxicologicproperties in in vitro and in vivo systems as do nickeland chromium salts, he discounted arguments thatthese elements in welding fumes present fewer risksthan those in equivalent exposures in nonweldingindustries. Carrying this argument further, Sternspeculated that a large part of the excess lung cancerrisk of welders could be attributed to the lung cancerrate in welders working with stainless steel and alloyscontaining nickel and chromium. Stern stronglyemphasized the need to explore the possibility of"high-risk hot spots" within the industry. This wouldenable the limited resources of the welding communityto be concentrated in areas where they are vitallyneeded (Refs. 120 and 122).

Stern's calculations are speculative and the elevationof the cancer risk which was derived from the 19 epi-demiologic studies surveyed is quite low. However, theidea that "hot spots" may exist within the industry isa valuable concept, and the concentration of financesand research efforts in areas where there are exposuresto suspect carcinogens or mutagens would seem to beindicated. One population that would serve well forthis type of study is the nickel-alloy pipe welders inOak Ridge, Tennessee, investigated by Polednak (Ref.103). Another relevant study, currently being conduct-ed by the German Cancer Research Center, will

Table 6Mesothelioma rates in dockyard occupations

OccupationNumber of

tumorsMan-years

at riskRate per 100,000

man year

Laggers and sprayersBoilermakersPaintersWelders and burnersShipwrightsEngine and electrical fittersJoinersOther intermittently exposed occupationsUnexposed workers

3646

1114162

1,4533,7203,3667,510

17,89035,990

5,14051,29012,250

206161119806139191216

G. Sheers and R.M. Coles, Ref. 113

Effects of Welding on Human Health/19

compare the cancer incidence in welders who workedwith CrNi-alloyed filler metals with that in nonweldersin the same factory. The welding cohort of more than1000 persons and a control group of similar size willcome from over 20 different companies (Ref. 54).

2.2.2 Asbestos and MesotheliomaThe development of pleural mesotheliomas is closely

associated with asbestos exposure. In a recent study,Sheers and Coles evaluated the incidence of meso-thelioma that occurred between the years 1964 and1978 among shipyard workers in Plymouth, England(Ref. 113). The mesothelioma rate of welders wasfound to be 80/100,000 man years as compared with206 for laggers and sprayers, 161 for boilermakers,119 for painters, and 16/100,000 man years for unex-posed workers (Table 6). The time between the initialexposure and death from mesothelioma was 30 to 50years. Thus, people developing mesotheliomas duringthe period of this study may have received the bulk oftheir exposure between 1945 and 1965, a time when theuse of asbestos as an insulating material becamewidely accelerated, but the dangers of exposure to thismaterial were not known.

Asbestos fibers were found to be in all locations inthe shipyard used in this investigation. The concentra-tions varied widely in different parts of the shipyardat different times during the workshift. The fibercounts, which reached peaks in confined spaces suchas engine and boiler rooms, were highest during theremoval of lagging. The welders moved freely betweenwork areas, and much of their exposure to asbestosapparently resulted from the general distribution ofthe fibers throughout the shipyard. Other exposuresmay have resulted from the use of asbestos cloth,gloves, and other materials. Since the results of thisstudy reflect exposures that occurred some 40 yearsago, at a time when there was little understanding ofasbestos toxicity and the necessity for stringent pre-cautions against asbestos exposure, the frequency ofmesothelioma observed in welders and other workersmay not be predictive of the future incidence of thisdisease.

2.3 Effects on the Ear and HearingWelders may suffer hearing loss from noises gener-

ated during welding as well as by other processes in thework area. For example, at a distance of 400 mm fromthe source, the mean pulse sound levels from manualgrinding, dressing with a needle hammer, and slagchipping are approximately 105 dB. Noises generatedby GMAW, SMAW, and gas welding are in the rangeof 85 to 95 dB (A) (Table 2) (Ref. 65). The OSH A noisestandard is 90 dB (A), averaged over an 8 hour day.

Erlandsson et al. (Ref. 38) compared the effective-ness of earmuffs and earplugs in reducing hearingdamage to welders and platers in an assembly and aboiler shop. No difference was seen between wearers

of the different protective devices in the boiler shop,whereas assembly workers who used earmuffs sufferedmore severe hearing impairment and had a higher rateof progress of the hearing impairment than did users ofearplugs. The investigators concluded that, althoughearmuffs reputedly afford somewhat better protectionthan earplugs, users of the latter may have less hearingdamage because earplugs are less convenient to put onor remove and thus may be worn more constantly thanearmuffs.

2.4 Effects on the Eye and VisionIn the absence of adequate protection, welders may

suffer eye damage from exposure to nonionizingelectromagentic radiation from the ultraviolet to theinfrared regions.

Ultraviolet radiation is primarily absorbed by theouter structures of the eye (cornea, conjunctiva, iris).Exposure to ultraviolet radiation from the welding arccan result in acute keratoconjunctivitis, also knownas arc eye, welder's flash, and actinic ray photokeratitis.The symptoms of this inflammatory disorder appearwithin 4 to 12 hours after exposure, last for up to 2days, and include blurred vision, lacrimation, burningpain, and headache. A chronic photoophthalmia mayalso occur which is manifested by functional visualdisturbances, increased light sensitivity, and symptomsof chronic inflammation of the conjunctiva and eyelids.

Visible radiation penetrates the eye and is absorbedby the retina and choroid. Thermal retinal damage canresult from short, intense exposure to visible and nearinfrared radiation between 400 and 1400 nm. Bluelight between 400 and 500 nm can cause a photochemi-cal retinal injury (e.g., solar eclipse burn) referred to asthe "blue light hazard". It is caused by an exposurelasting several seconds or longer to sources that are notsufficiently intense to cause retinal burns (Refs. 56 and87).

Infrared radiation with wavelengths longer than1400 nm is absorbed primarily by the cornea andaqueous humor where it may cause thermal damage.Infrared radiation of less than 1400 nm can be trans-mitted to the deeper structures of the eye and has beenassociated with the formation of lenticular cataracts.

In 1978, welding and cutting injuries represented0.5% of the Workman's Compensation cases in theUnited States (Ref. 136). The results of a Bureau ofLabor Statistic's survey showed that, of these injuries,67% involved the eye and 1/3 of these were welder'sflash. According to the welders surveyed, about 2/3 offlash burns were due to exposure to other welder'sequipment. Most of the remaining ocular injuries weredue to hot flying metal or slag particles. About half ofthe workers were wearing goggles or helmets with filterlenses at the time of their accidents. However, pro-tective curtains or shields were used only at 25% of thework stations.

Emmette/a/. (Ref. 36) found virtually no difference

20/EFFECTS OF WELDING ON HEALTH IV

in the frequency of eye abnormalities among 77persons who worked as welders for at least 10 years,75 nonwelders who worked in the welding area for atleast 8 years, and 58 persons who worked for 10 yearsin the same fabrication facility but were never in thewelding area. Visual acuity was the same in all threegroups. The only statistically significant differencebetween the groups revealed by ophthalmologic ex-aminations was the prevalence of dust in the eye lids.All persons in the welding shop area in this plant wererequired to wear safety glasses so these results may notbe representative of shops where hygienic practicesare less adequate. Despite the good practices, many ofthe welders reported having experienced flash burns.The frequency of flash burns that resulted fromaccidentally striking their own arcs was about thesame as that from inadvertant exposure to otherwelder's arcs.

The case of a 56 year-old man who was formerly awelder and had a recurrent pterygium (an abnormaltriangular fold of membrane, extending from the con-juctiva to the cornea) in one eye was recently reported(Ref. 68). The development of pterygia is apparentlyrelated to exposure to ultraviolet radiation (Refs. 126and 127). However, welders are not known to have anincreased incidence of this disorder (Refs. 36 and 96), afact which has been attributed to the use of protectiveshields (Refs. 36 and 126). Work-related exposure toultraviolet radiation was considered to be an unlikelycause of the pterygium case cited above (Refs. 96 and126).

2.5 Effects on the SkinWelders are subject to a variety of skin conditions,

the majority of which can be prevented by protectiveclothing. Among these are burns from contact withhot metals, ultraviolet-induced erythema, and embed-ded airborne metal particles. In some individuals,allergic dermatoses can develop from repeated skincontact with components of welding fumes such aschromium, nickel, zinc, cobalt, cadmium, moly-bdenum, and tungsten compounds. As evidencedbelow, burns represent the most frequent skin lesionsin these workers.

In a recent study in the Soviet Union (Ref. 131), 117welders in a shipyard were examined for skin lesions.Work-related lesions were found in 45% of the welders.Of these, superficial and deep burns were the mostcommon and were present in 41% of the workers.Ultraviolet radiation-induced dermatitis was detectedon the face, hands, and forearms of 8.3% of the workers.

Burns were found to be more prevalent amongwelders than workers in other occupations in theOdense district of Denmark. Of 918 occupational skinburns and 948 ocular burns treated in the OdenseHospital, nearly half occurred in the iron industry,particularly during welding, cutting, and flameplaning (Ref. 112).

Emmett et al. (Ref. 36) compared the frequency ofskin abnormalities among 77 welders (Group 1), 75nonwelders who worked in areas where they wereexposed to welding fumes (Group 2), and 58 personswho worked in the same plant under conditions inwhich they were not exposed to welding fumes (Group3). This was the same population that was used in theeye study cited above. The principle welding processesused were GMAW shielded predominantly with CO2

and FCAW of mild steel. Occupational and smokinghistories and details of sun exposures were obtainedfor each participant in the study. To control for varia-tions in susceptibility to ultraviolet radiation, onlyCaucasian males were included in the study.

No differences were observed between the incidenceof skin tumors or premalignant lesions among thegroups. Erythema, when present, was generallylocalized and confined to unprotected areas of thebody. Erythema of the neck was considerably morefrequent in welders. The incidences of facial erythemaand actinic elastosis (degeneration of elastic tissue ofthe dermis resulting from exposure to ultravioletradiation) did not differ significantly between occupa-tional groups.

Small cutaneous scars, resulting from thermal andmechanical injuries, were most prevalent in weldersand machinists (Table 7). No significant differences

Table 7Percentage of workers with

small cutaneous scars

Number

Emmett

of Small Scars0

1 -5>5

et al., Ref. 36

Group 1

7%29%64%

Group 2

76%15%9%

Group 3

65%16%19%

in the frequency of various other dermatoses wereobserved. The investigators noted that, although pre-malignant and malignant skin tumors could not berelated to welding exposures, good hygienic practiceswere employed in this plant, and a relatively youngwork population was examined. Further studies areneeded in which ultraviolet-induced diseases andpremalignant and malignant skin conditions areexamined in an older population of welders in whichthe duration of exposure and period of time after theinitiation of exposure are sufficiently long to enablethe expression of skin diseases that may result fromchronic exposure to ultraviolet radiation.

Effects of Welding on Human Health/21

2.6 Effects on the CardiovascularSystem

Heart rates during simulated welding in a laboratorysetting were measured while subjects were standingand performing overhead welding, stooping in adownhand welding position, or crouching with thewelding table tilted 30° from the horizontal (Ref. 130).The apparatus which was used to simulate weldingconsisted of a table with a V-shaped slit across thesurface. The weld site was represented by a reflectingphotoelectric cell disc. A light bulb at the end of thetip of a metal rod (the "electrode") represented the arc.Welding was simulated by keeping the bulb in contactwith the photoelectric cell disc as it was automaticallydrawn down the length of the slit without resting theelectrode on the table edge. The experimental situationwas designed so that procedures normally practicedduring welding, such as changing the electrode andstriking the arc, were performed. The heart rate varia-tion, as measured by electrocardiography, was greatestwhen the welder altered his posture to change elec-trodes. The heart rate was lower when welding thanduring the change of electrodes. The variation in theheart rate occurring between welding and electrodechanges was greatest when the downhand positionwas used.

Hayden et al. (Ref. 61) found that there was nosignificant change in carboxyhemoglobin levels innine subjects before and after 15 minutes of continuouselectric arc welding. Breathing zone levels of carbonmonoxide were measured during the welding periodand were found to be low relative to the thresholdlimit value.

2.7 Effects on the Nervous System2.7.1 Manganese Poisoning

Chronic manganese poisoning is a progressivedisease with neurological and psychological mani-festations. The susceptibility to manganese poisoningvaries widely; some highly resistant persons may showno signs of disease after years of exposure while othersmay become ill after as little as 2 months of exposure(Ref. 92).

Early symptoms of manganese poisoning includeapathy, anorexia, headache, spasm, weakness, andirritability. Signs of psychoses may develop includingeuphoria, impulsive behavior, absent-mindedness,confusion, and aggression. The disease can be arrested,and may be reversible at this stage, if exposure isstopped. Otherwise, the disease will continue to pro-gress; disturbances in speech, gait and balance mayappear; and tremors may develop with eventualserious neurological involvement. Advanced stages ofchronic manganese poisoning are readily confusedwith Parkinson's disease.

Chandra et al. (Ref. 27) measured the manganesecontent of the urine, signs of neurological injury, and

serum calcium in 60 welders, 20 each from threedifferent plants. In Plant A, the work area was wellventilated and the welding electrodes contained 2.1%manganese. In Plant B, the work areas were partiallyventilated and the welding electrodes contained 0.55%manganese. In Plant C, the welders worked in con-fined spaces, and the welding electrodes contained0.45% manganese.

Positive neurological signs (tremors and deep brisktendon reflexes in the extremities) were seen in 5, 10,and 9 of the welders in Plants A, Band C, respectively.The urine manganese and serum calcium levels werehigher than those of the controls in all workers withpositive neurological signs. As expected, the presenceof neurological signs was not related to the durationof exposure to welding fumes. In both this and aprevious study (Ref. 26), Chandra et al. reported thatincreased serum calcium may frequently accompanymild and moderate (but not advanced) stages ofmanganese poisoning. The investigators suggestedthat serum calcium levels may be of value in diagnos-ing early stages of manganese poisoning.

2.7.2 Vibration DisordersThe use of vibrating equipment such as glazing

hammers and grinding machines may produce "whitefinger syndrome" (Raynaud's disease) or intermittantblanching and numbness of the fingers. Edling (Ref.34) observed indications of this condition in 10 of 24welders, all of whom had used vibrating equipment forat least 6 years.

A study of the prevalence of vibration damage in thefingers of 526 shipyard workers in the Soviet Union(including 202 ship assemblers, 185 electric arcwelders, 46 gas cutters, and 93 painters) was performedby Yatsenko et al. (Ref. 142). The controls were 60technicians and engineers from the same industriallocation. The extent of damage was tested bymeasuring the threshold sensitivity of the middlefingers of both hands to frequencies of 63, 125, and250 Hz.

A moderate (threshold sensitivity +15 to +25 dB) orlarge (threshold sensitivity equal to or greater than25 dB) decrease in the sensitivity to vibration wasobserved in 18% of the ship assemblers, 10.8% of theelectric arc welders, 4.3% of the gas cutters, 21.7% ofthe painters, and none of the controls. The extent ofthe impairment was related to the duration of ex-posure. In this shipyard, the welders frequentlyassisted the ship assemblers with their work whichincluded the use of vibrating instruments. That theexposure to vibrating instruments was solely responsi-ble for the decreased sensitivity to vibration could notbe substantiated in this study since the painters, whodid not use vibrating equipment, had a high incidenceof decreased vibration sensitivity. On the other hand,the incidence of decreased vibration sensitivity cor-related with the degree of exposure to dusts and fumes

22/EFFECTS OF WELDING ON HEALTH IV

and with the prevalence of pulmonary disorders. Theinvestigators speculated that a decreased sensitivityof the hands to vibrational stimuli is an early sign ofwhite finger syndrome or the more advanced occupa-tional sensory polyneuritis and suggested that vibra-tion sensitivity measurements should be a part ofroutine medical examinations of workers in the ship-building industry.

2.8 Effects on the LiverAlsbirk et al. (Ref. 6) evaluated liver function in

stainless steel welders by measuring the levels of serumalanine transaminase and aspartate amino transferase.A small increase in these enzymes was observed whichwas most marked in welders who consumed signifi-cant daily quantities of alcohol.

2.9 Effects on the MusculoskeletalSystem

2.9.1 Shoulder PainShoulder pain is the second most frequent cause of

visits to orthopedic clinics (Ref. 57) and is a frequentcomplaint of industrial workers. Because there is someuncertainty about the correlation between the strenu-ousness of the work and the occurrence of shoulderpain, Herberts et al. (Ref. 63) conducted an epidemio-logic study of shipyard workers in which the occurrenceof shoulder pain in 131 welders was compared withthat of 57 office clerks.

The prevalence ratio of supraspinatus tendinitis inwelders was 18%, which was significantly higher thanthat of the controls. The number of years of weldingexperience and the rated level of shoulder muscle loadcould not be correlated with the development of supra-spinatus tendinitis, but these factors could not beruled out by this study.

Kadefors et al. previously reported that musclefatigue and pain is greatest when welding in overheadpositions (Ref. 72). In that study, muscle fatigue wasmeasured by electromyography of the back andshoulder muscles. These same investigators havefurthered their study of the fatigue of the musclescontrolling shoulder movement with a new instru-ment, the spectral moment analyzer. This instrumentmeasures the myoelectric activity in individual musclesand yields hard data upon which to base recommenda-tions concerning work postures which would be leastlikely to be harmful.

Using this equipment, the myoelectric activity inshoulder muscles was measured in men who wereholding 2 kg weights while simulating work positionsoften maintained by welders (Ref. 64). Eight basicpostures were studied. With each, the elbow was bentin a 90° angle and the position maintained for oneminute. Three working levels were investigated: thehand at the waist, the hand at shoulder level, and the

hand in an overhead position. Each position causedlocalized fatigue in all muscle studies, but the over-head position was the most potentially damaging toshoulder muscles. The investigators concluded that,when possible, work positions should be changed sothat welding is done with the hand at shoulder or waistlevel. However, if welding must be done in the over-head position, the results indicated that the staticloading on the supraspinatus muscle is significantlyaffected by elbow positioning, and potential damagecan be reduced by assuming certain positions duringoverhead welding.

One approach to muscle strain and related disordersin welders (and other workers) is to provide them withperiods of rest and exercise designed to overcomestrain. The symptoms of a group of welders whosuffered lumbosacral radiculitis (inflammation of theroot of a spinal nerve) were treated by a program ofexercises to be followed routinely during the workbreak which were aimed at reducing tension in thosemuscles most subject to static stress. The beneficialresults were most evident during inclement weather,which normally aggravates this condition (Ref. 132).

2.9.2 PostureWelders differ considerably in the postures they

assume relative to the fumes. In one study, it wasfound that there was no statistical difference in dustlevels in the breathing zones of different welders,regardless of whether the welder was standing erect,bent over his work, sitting, or varying his work posi-tions (Ref. 133). However, several investigators haveindicated that posture has a significant influence onthe exposure to fumes and gases (Refs. 39,53, and 104).This is related to the characteristics of the plume inwhich welding emissions rise from the arc. As thedistance above the arc increases, upward velocity dueto thermal uplift decreases and the diameter of theplume increases. According to Grosse-Wordemannand Stracke (Ref. 53), to minimize exposure to fumes,the horizontal distance of the welder's nose from thearc should be as great as possible and the vertical dis-tance above the arc should be minimized. The weldingpostures affording the minimum exposure are, inorder: the horizontal-vertical position; vertical,upward welding while seated; vertical, upward whilestanding; flat position-seated; flat position-standing.According to the authors, the latter two positions,while being the least advantageous due to fumeexposure, are the most economical because they yieldthe largest weld pool (Ref. 53), and probably are theleast stressful to the musculoskeletal system (seeprevious section). Thus, a compromise must be struckbetween economy and muscle strain on the one handand fume exposure on the other, and in this case, fumeextraction and use of an air curtain may be the bestsolution.

Effects of Welding on Human Health/23

2.10 Effects on the Urogenital Tract2.10.1 Fertility

Kandracova (Ref. 76) evaluated fertility disordersin a total of 4200 male workers that were referred tohim by medical facilities in various factories in plantsin Eastern Slovak. Among these men, there were 69welders, 192 pipe fitters, and 57 car mechanics. Theremainder, who served as controls, worked in otherprofessions and occupations in which there was neverany exposure to welding fumes.

The quantity, viscosity, color, and pH of semen andthe number, quality, morphology, and motility ofsperm were determined. The frequency of abnormali-ties was the same among all occupational groups, andit was concluded that the frequency of abnormalspermiogenesis among welders was the same as thatamong workers in other professions. These data arein accord with those previously reported by Haneke(Ref. 58).

2.10.2 Effects on Kidney FunctionAlsbirk et al. (Ref. 6) detected no alterations in

kidney function of stainless steel welders, as evaluatedby the measurement of albumin and beta-2-micro-globulin in the urine.

2.11 Effects on the Teeth and OralCavity

During a study of the condition of teeth and gums,Wulfand Seefeld(Ref. 140) observed that the incidenceof periodontal disease was significantly greater in 100electric arc welders than in 100 nonwelders (lock-smiths, smiths, and cutters). The investigatorsspeculated that particles in welding emissions couldadhere to the teeth and gums, influencing the quantityand quality of dental plaque. The nitrogen oxides,following conversion to acids in the moist environ-ment of the upper respiratory tract, could inhibitsaliva production. The resultant dryness of the mouthcould lead to inflammatory changes in the gingiva.Finally, the authors suggested that carbon monoxidemay cause changes in the blood vessels in the mouth,producing gingival edema.

The observations of Wulf and Seefeld (Ref. 140)find support in the investigation of Melekhin andAgarkov (Ref. 89) who, during their study of fluorineand manganese sensitization in 247 electric arc welders(described below), also examined the condition of theoral cavity. They found that welders had 1.7 timesmore film deposit on their teeth and 3.1 times greaterinflammation of the mucous membranes of the mouth(including periodontitis) than did the 100 controlworkers.

Another effect of welding on the oral cavity is thepossible degeneration of dental amalgams in diverswho perform underwater electric arc welding. Tostudy the cause of complaints by divers of a metallictaste in their mouth and damage to their dental fillings,

Rockert, Christensson, and Orthendal (Refs. 28 and108) performed laboratory experiments in which theelectric current through different types of amalgamwas measured underwater at various distances from anumber of types of welding electrode. They reportedthat the current through the amalgam was dependenton the surface area and the distance between the toothand the welding electrode.

2.12 Metal Fume FeverMetal fume fever may be caused by exposure to

zinc, copper, magnesium, or other metal fumes, butit is most frequently associated with zinc fumes. Thesymptoms appear 4 to 12 hours after exposure and areheralded by a sweet or metallic taste in the mouth anddryness or irritation of the throat followed by cough-ing, shortness of breath, general malaise, nausea,muscle and joint pains, fever, and chills. Thesymptoms last for 24 to 48 hours and have no sequelae.A short-lived tolerance to metal fumes develops. Thistolerance may be lost during a short absence fromfume exposure, and the symptoms may recur on thereturn to work after a weekend or holiday; hence, thealternate name "Monday fever".

Nine cases of metal fume fever were recentlyreported in Mexican electric arc welders (Ref. 18). Thesymptoms, which appeared approximately 10 hoursafter exposure, included fever, dry cough, vomiting,and a metallic taste in the mouth. Abnormalities, fourof which were suggestive of chronic bronchitis, wereapparent in chest X-rays of five of the welders.

2.13 Sensitivity to Fume ComponentsBecause the incidence of contact sensitivity to

chromium is very high in Poland, a study was per-formed to attempt to determine the primary sources ofexposure to this allergen (Ref. 109). Two hundred andfifty persons with dermatitis and positive patch tests topotassium chromate were questioned to identify themost probable sources of the exposures that initiatedor exacerbated their sensitivity to chromate. Of the250 subjects, 132 had contact with chromates in anoccupational setting; 16 of these reported exposure towelding fumes. The study determined that tannedleather and matches are the two most frequent sourcesof chromate exposure for the general population. Itis not known if sensitized welders developed theirinitial response to chromium through non-occupa-tional sources such as matches or shoe leather, orwhether they became sensitized to chromium throughwelding exposures.

A case of a welder with an unusually severe allergicresponse to chromium was recently reported byZugarman (Ref. 148). The worker had a recurrenteczematous eruption involving the cheeks, neck, andear which developed 3 1/2 years after he began weldingchrome-plated steel desks and chairs. He had a strongpositive reaction to a potassium chromate patch test,

24/EFFECTS OF WELDING ON HEALTH IV

but did not react significantly to other allergens.The eczema cleared up within 2 weeks after stopping

work but recurred within 3 days after returning to theworkplace. Neither sunscreens, barrier creams, topicalcorticosteroids, nor improved ventilation at his workstation prevented the eruption. He became so sensi-tized to chromium that he could not work in any partof the plant without developing a cutaneous eruption.

The allergic sensitization of shielded metal arcwelders to manganese and fluorine was studied in apipe plant in the USSR (Ref. 89). The sensitivity wastested with an in vitro assay which measured theresponse of isolated white blood cells to manganeseand fluorine. The sensitivity of 176 welders wascompared with that of 100 controls who were notexposed to welding fumes. Not one of the controls wasfound to be sensitive to manganese or fluoride by thistest whereas about half of the welders were sensitiveto either manganese or fluorine, and 20 to 25% of themwere responsive to both elements. The relationship ofthe results of this in vitro assay to the actual physio-logical response of the welders to manganese orfluorine exposures was apparently not investigated;hence, the significance of the results cannot beevaluated.

2.14 Biochemical ChangesMikhail et al. (Ref. 90) evaluated serum protein and

lipid levels, as well as alterations in blood hemoglobinand heme synthesis in 16 tank welders in Egypt whohad worked at this trade for up to 22 years. Theaverage blood lead levels were 42 figj 100 g in weldersand 27.5 fig/100 g in 10 control workers who had nothad occupational lead exposure. None of the weldershad clinical signs of lead poisoning even though therewere some changes in the diagnostic parametersexamined. There was a significant decrease in thequantities of blood hemoglobin and an increase in theurine levels of delta-aminolevulinic acid, both ofwhich are indicative of the interference with hemesynthesis that typifies lead toxicity. Increased levels ofthe serum enzymes glutamic oxaloacetic transaminase,glutamic pyruvic transaminase, and lactic dehydro-genase were observed which suggested altered liverfunction. A significant increase in serum triglyceridesand beta-lipoprotein concurrent with a decrease in thephospholipid/cholesterol ratio was also observed. Theinvestigators suggested that this may indicate a pre-mature development of atherosclerosis; however, thisinterpretation is highly speculative.

Grandjean and Kon determined blood lead levels in59 welders, 67 workers in other departments, and 42retirees from a repair in refitting shipyard (Ref. 46)where lead exposures resulted primarily from repairwelding of steel parts coated with anti-corrosive paints.The average blood lead levels of the welders (39lug/100 ml blood) was significantly higher than that ofworkers in other departments (26 ^g/100 ml) and

retirees 23 /Ug/100 ml). Among the welders, 53 of 59had levels above 30 ugj 100 ml (the authors considered20 figj 100 ml to be the norm in the average workingpopulation) and 18 of these were above 40 figj 100 ml.In contrast, 19 or 67 workers in other departments hadelevated blood lead levels and only two of these wereabove 40 /ig/ (00 ml. Blood lead levels in all retireeswere less than 30 fig/100 ml. They gradually returnedto normal (20 jug/100 ml or less) within 4 years ofretirement in the nonwelders and 7 to 8 years inwelders.

Grandjean et al. also measured nickel concentra-tions in body fluids of welders and other workers in thesame repair shipyard as well as in a constructionshipyard (Ref. 47). The control population consistedof 15 hospital personnel and 28 retired workers fromthe repair shipyard. Fitters and painters had higherplasma nickel levels than welders and riggers at bothshipyards. Only the welders at the repair shipyard hadelevated plasma nickel levels. In this group, 9 of 38 hadsignificantly elevated levels as compared with 1 of 43controls, and 4 of 24 welders at the constructionshipyard. Of the welders, 3 had plasma nickel levelshigher than 10 fig/ liter which is considered to be acritical concentration (Ref. 67).

Urine nickel levels were examined only in workersfrom the construction shipyard. In this case, 9 of 24welders, but only 3 of 13 painters, had nickel levelsthat exceeded control values. This contrasts with theobservation that plasma nickel levels were higher inpainters. As the authors pointed out, urine nickelvalues fluctuate markedly during the day. Therefore,the use of random samples in this study, rather thanurine samples collected from a full 8 or 24 hour period,may have rendered the urine nickel values less accuratethan the plasma values.

Suzuki et al. (Ref. 125) studied the relationshipbetween urine levels of cadmium, copper, and zinc inwelders using cadmium-containing silver solder. Theirresults suggested that cadmium accumulation in thebody affects the excretion of copper and zinc.

Sjogren (Ref. 114) compared urinary fluorine,chromium, and nickel levels in welders and nonweldingindustrial workers. Welders using basic electrodes hadhigher fluorine levels, and those welding stainless steelhad higher chromium and nickel levels in the urine atthe end of the day than did controls. There was a linearrelationship between the breathing zone and urinefluorine concentrations in the welders. Stainless steelwelders had higher urinary chromium and nickellevels at the end of the work day than did nonweldersin the metal industry. Unlike fluorine, urinarychromium levels in stainless steel workers increasedthroughout the work week even though airborne levelsdid not so vary. The relationship between thechromium concentration in the air and urine samplescollected over the whole week was linear. However,the deviation was great and Sjogren concluded that

Toxicologic Investigations 125

urinary chromium levels can only be used to approxi-mate airborne exposures. The relationship betweenairborne and urinary nickel level was not linear.

Sjogren's results are in general agreement withearlier reports (see Effects of Welding on Health,Volume I) and they suggest that of the three elementsexamined, fluorine, and to a lesser extent chromium,but not nickel, hold promise for quantitative biologicalmonitoring of the exposure of welders to fumes.

Work in the area of monitoring personal exposureby measuring accumulation of metals in biologicaltissues is in progress at the Lund Institute of Tech-nology in Sweden. The relationship between personalexposure and levels of different metals in parts of thehuman body, e.g., body fluids, hair, nails, etc., is beingstudied (Ref. 85). Personal exposure levels will bemonitored by lightweight samplers, also developed atthat institute, which allow detection of elements atlevels well below existing TLV's in samples auto-matically collected during short, sequential timeintervals (Ref. 20).

2.15 Human FatalityThe death of a 34 year-old welder, apparently due to

inhalation of cadmium fumes, was reported in 1981(Ref. 82). This man had been working for about 30minutes with an oxyacetylene torch and silver solder.The solder, which contained over 20% cadmium, wasnot labeled as such by the manufacturer. Althoughthe building was "large and airy with a high roof", theimmediate work area was not well ventilated. Withinhours after exposure, the welder developed a persistentcough and had difficulty breathing. His urine cadmiumlevels were 20-fold higher than normally acceptablelevels. He died 5 days later, and acute sterile pneu-monitis, typical of cadmium poisoning, was found atautopsy.

2.16 Occupational MedicineIn the past, workers in occupational medicine have

stressed the need for pre-employment examinationsto screen from work in this industry those who mayhave health conditions that would render them parti-cularly susceptible to the potential health effects ofwelding. Equally important are periodic checkups ofcurrently employed welders to ensure that their healthstatus remains unaltered by their employment, andthat they have not developed any health conditionsthat would predispose them to sensitivity to weldingexposures. The assessment of the health status ofprospective and active welders was recently re-emphasized by Spelbrink(Refs. 118 and 119). Becausethe respiratory tract is the major route of entry forwelding emissions into the body and is also a majortarget organ for the potential toxic effects of weldingemissions, health examinations focus on the respira-tory tract but also examine other organ systems.

Spelbrink recommended that pre-employment exam-inations include sputum analyses, pulmonary functiontests, chest X-rays, and hearing tests, and that infor-mation concerning smoking and medical histories beobtained. Employment in the welding trade should bediscouraged for those with a history of bronchialasthma, allergic conditions of the upper respiratorytract, chronic bronchitis, tuberculosis, or high bloodpressure which does not respond to treatment. Theauthor further recommended that pulmonary functiontests and chest X-rays be obtained periodically afteremployment has begun and after any extended sick-ness or period of disability.

3. Toxicologic Investigationsin Animals and in Cell Cultures3.1 Animal Studies3.1.1 Pulmonary Effects

When particles or other foreign materials enter thelungs, they frequently elicit an inflammatory responsewhich is marked by increased surfactant levels andthe migration of leukocytes (macrophages andpolymorphonuclear leukocytes) into the lung. As partof the natural line of defense against foreign bodies,levels of certain hydrolytic enzymes become elevatedwithin leukocytes and in their immediate surround-ings. Although this response is usually short-lived,certain foreign substances, such as some bacteria orquartz particles, may elicit a chronic inflammatoryreaction. Some investigators believe that leukocytes,actively responding to inflammatory substances,release biological mediators which stimulate thedivision of fibroblasts (connective tissue cells) and thatthe extensive stimulation of fibroblasts resulting froma continuous assault with highly inflammatory foreignparticles is responsible for pulmonary fibrosis.

White et al. (Ref. 138) examined the acutepulmonary inflammatory response in rats to threechemically distinct welding fume particulates. Twofumes were generated by SMAW of mild steel witheither a basic or rutile-coated electrode; the thirdfume was generated by SMAW of stainless steel with arutile-coated electrode (Arosta 316L - 18% Cr, 11%Ni). The results were compared with those obtainedwith 99% pure titanium dioxide. One week after theparticles were introduced into the lung by intratrachealinstillation (a technique whereby materials are in-jected directly into the lung through a cannula whichis passed into the pharynx and extends through mostof the length of the trachea), the relative lung weight,the number of leukocytes (free cells) in the lung, andthe quantity of surfactant and hydrolytic enzymespresent in the lung were elevated. Each of these para-meters was elevated by all three types of weldingparticles; the particles from the stainless steel systemhad a slightly greater effect than the other particles

26/EFFECTS OF WELDING ON HEALTH IV

Table 8Body weight/lung weight ratio, free cell numbers and

levels of pulmonary surfactant in rats 1 week afterinstillation of TiO2 or SMAW particulates

Parameter studied

Lung/body weight ratio x 104

No. free cells x !0-6/g lung

Weight pulmonary surfactant(mg/g lung)

Particulate

TiO2

S. steel

Basic

Rutile

TiO2

S. steel

Basic

Rutile

TiO2

S. steel

Basic

Rutile

(F)(M)(F)(M)(F)(M)

(F)(M)(F)(M)(F)(M)

(F)(M)(F)(M)(F)(M)

Dose of compound

0

685070507050

9.2ll.O8.0

ll.O8.0

ll.O

0.400.620.290.620.290.62

0.5

6364*72597757

11.99.0

11.89.6

19.5*12.4

0.371.170.81*0.920.85*—

1.0

7271*89*568266*

12.312.414.2*8.2

14.1*9.3

0.310.811.08*1.120.70*0.97

used

2.5

„

61*89*68*7668*

„

18.7*11.87.2

16.2*9.4

„

1.220.930.740.840.80

(mg)

5.0

8172*93*68*90*67*

19.0*10.614.1*15.215.2*14.6

0.591.24*

* 0.98*1.33

* 0.93*1.17

M = male; F = female

* = significant change from equivalent control group of rats

White et al. Ref. 138

(Table 8). The investigators speculated that themaximum response elicited by stainless steel particlesresults from the Cr-VI present in the samples.

In a follow-up study, White et al. (Ref. 137)attempted to determine the extent to which theinflammatory changes could be attributed to the water-soluble fraction of the particles which contains most ofthe Cr-VI and whether any of the observed pulmonarychanges were persistent. The water-soluble andwater-insoluble fractions of the stainless steel particleswere instilled into the lung and the inflammatoryeffects compared with those resulting from theinstillation of comparable quantities of potassiumdichromate. The effects were studied over a periodof 1, 4, and for the potassium dichromate only, 13

weeks after instillation of the test material. Theinflammatory response to the instilled substancestended to decline after one week. The insolubleparticles induced a greater increase in levels ofhydrolytic enzymes whereas the number of leukocytesand the quantity of pulmonary surfactant were greaterwith the soluble fraction. Potassium dichromateenhanced most of the parameters examined. Althoughthe welding fumes mimicked the effects of the potas-sium dichromate, the effects studied by these authorsare the same as the general responses to many typesof inflammatory stimuli, and hence it is not possibleto conclude from this study that the inflammatoryeffects of welding particles in the lung are related to theCr-VI content of the particles.

Toxicologic Investigations 127

In a similar study (Ref. 7), rats and guinea pigs wereexposed by inhalation and intratracheal instillationto flux-coated electrode fumes (the welding processfrom which the fumes were derived was not described).A normal inflammatory response was observedimmediately after a single exposure to the fumes byinhalation. Continuous administration of weldingfumes by intratracheal instillation eventually led tothe appearance of fibrous tissue around aggregates ofwelding particles. Hydroxyproline was elevated in thelungs of animals exposed for a prolonged period. Thisamino acid is unique to collagen, the backbonematerial of connective tissue, and its elevation isindicative of an active fibrotic process. The investiga-tors speculated that the fibrogenic response was due tosilicon, which constitutes 7.7% (w/w) of the fumesstudied. However, this is unlikely since, when it ispresent in welding fumes, silicon is generallyamorphous and not in the quartz form (silica)associated with pulmonary fibrosis (Ref. 35).

Il'nitskaya and Kalina (Ref. 69) examined theeffects in rats of inhaled mixtures of aluminum oxide(gamma-alumina) and electric arc welding gases. Theconcentrations administered were similar to those towhich welders may be exposed. The first group of ratswas exposed to aluminum oxide dust at a concentra-tion of 48.6 mg/m3 (particle size 8-11 urn); the secondgroup was exposed to electric arc gases in which theozone concentration was 1 mg/m3, and the nitrogenoxides were present at 3.3 mg/m3/- The third groupexposed to a mixture of alumina dust and weldinggases at the same concentrations administered to thefirst two groups. Animals in the fourth group were notexposed and served as controls.

By the end of the 4-month exposure period, minorchanges in the hematologic profile and serum enzymeswere seen in all of the experimental groups. Pulmonarychanges occurred after 2- to 3-month exposures in allgroups. The animals exposed to alumina dust (groups1 and 3) developed moderate catarrhal bronchitis andswelling of alveolar septa with a small amount ofperivascular edema. (The edema declined in group 3towards the end of the observatin period.) Thepulmonary effects observed in animals exposed to thegases alone were somewhat more severe than those inthe other two groups. Bronchitis, "spreading"emphysema, swelling of the intra-alveolar septa, andperivascular edema were found in these animals. Theinvestigators suggested that the adsorption of theelectric arc gases by the alumina particles may havedecreased the toxicity of the gases. Experiments suchas these demonstrate the importance of evaluating thetoxic effects of the complete mixture, rather than

isolated components, of industrial or environmentalsubstances to which man is exposed.

Havrankova and Skoda used the wet weight andDNA content of the lung as indicators of pathologicchanges in the respiratory tract (Ref. 59) resultingfrom intratracheal instillation of fume samples fromwelding with three different basic electrodes and onerutile elecrode. All four samples caused an increase inthe lung weight, and the fumes from the rutileelectrode and one of the basic electrodes caused astatistically significant decrease in the DNA contentof the lung.

The American Welding Society is currently plan-ning a study which will compare the toxicity andfibrogenicity of standardized welding emissionsgenerated by six processes. The emissions from threeSMAW methods, two GMAW methods using eitherargon or CO2 shielding gases, and one FCAW processusing a mixed shielding gas of argon/CO2 will beadministered to rats by inhalation. The exposurelevels will simulate those experienced during thewelding in the workplace. The fibrogenicity of theseemissions and toxic effects on alveolar macrophagefunction, in particular viability, will be compared.Studies of the mutagenicity of urine from erposedanimals, cytogenetic effects in bone marrow, andsystemic effects of the welding emissions are also beingconsidered.

3.1.2 Whole Body EffectsSome question exists about the relationship

between the water-solubility and the biologicalactivity of the nickel compounds present in weldingfumes. To investigate the influence of its water-solubility, English et al. (Ref. 37) studied the wholebody distribution and excretion of soluble (NiCl2) andinsoluble nickel (NiO) in rats. The radiolabeledcompounds were administered by intratrachealinstillation and nickel levels in 14 organs, and fecesand urine were determined at various times during the90 days following treatment. NiCl2 was readilydistributed throughout the body and was rapidlycleared. About 70% of the administered dose wasremoved by the third day, and the main route ofexcretion was via the kidneys. (Ten times as muchradioactivity was measured in the urine as in the fecesover a 90 day period.)

NiO was more slowly distributed from the lungs tothe rest of the body. Both the lungs and associatedlymph nodes retained significant quantities of the NiOover the 90 day period whereas only negligiblequantities of NiCl2 were still present in those organs bythe end of that time (Fig. 1). Whereas the NiCl2 rapidly

28/EFFECTS OF WELDING ON HEALTH IV

4000

1000

a 100

Lung - NiO

Days

Fig. 1 — Change in nickel concentration in ratlung and associated lymph nodes following an

intratracheal injection of NiCl2 or NiOEnglish et al., Ref. 37

entered and disappeared from the blood stream, theblood nickel levels remained almost constant duringthe first 2 months after instillation of NiO, suggestingthat NiO was slowly and continuously cleared fromthe lung. Although the distribution rates were slower,significant levels of nickel were eventually found inother organs. Unlike NiCl2 about half of the excretedNiO was present in the feces. Because the decline ofnickel from all organs, with the exception of the lung,lymph nodes, and blood, was similar with the twoforms of nickel, the investigators suggested that NiOmay be removed from the lung by the slow conversionto a more water-soluble form whose behavior in thebody is similar to that of NiCl2. This concept is inaccord with observations that the solubility of nickelin welding fumes is markedly increased by the additionof certain biological fluids (Ref. 128).

In a recent study, Kawata et al. examined thetoxicity of basic-coated electrodes containing lithiumsalts (Ref. 77). (The use of lithium in electrode coatingsis currently being explored as a means of reducingsome of the health hazards in welding fumes; e.g., re-duction of Cr-VI content of fumes - Ref. 128.) Animals

exposed by inhalation to high levels of fumes fromelectrodes in which some of the potassium and sodiumsalts were replaced by lithium survived for 24 hourswhile those exposed to equivalent concentrations offumes from electrodes without lithium died. The totalquantity of sodium, potassium, and lithium oxideswere lower in the fumes from electrodes containinglithium than in the other electrode tested. Therefore,it is possible that differences in the toxicity of theelectrodes may have been related to differences in thealkalinity of the fumes. Of the tissues studied (lung,liver, kidney, and blood), the lung had the highestconcentration of lithium after exposure to lithium-containing electrodes. The kidney, with the nexthighest level, had approximately one tenth the con-centration found in lung. Fifty eight percent of thelithium in the kidney was excreted within 24 hours, and93% was excreted by the end of one week. The investi-gators stated that further experiments are needed todetermine the effects of lithium on kidney function.

Toxicologic studies of welding fumes are currentlyin progress at the University of Bradford in England(Ref. 48). In this work, fumes are being administered toanimals by inhalation, intratracheal instillation, intra-muscular injection, and skin painting.

3.1.3 Generation of Test SamplesThe method by which samples are generated and

prepared for biological studies is very important.When collecting and preparing samples, care shouldbe taken that they are not physically or chemicallydenatured. Particles should not be allowed to aggre-gate or agglomerate if they are to be administered toanimals by inhalation or intratracheal instillation. Themethod by which welding emissions are generatedshould be reproducible so that experimental data canbe compared from laboratory to laboratory. Forinhalation experiments, the time between generationof samples and delivery to animals should be the sameas that between generation of fumes in the worksituation and actual human exposures because ofchanges that occur naturally in air after the emissionsform.

Recently, Gray et al. (Ref. 50) developed anapparatus for the reproducible generation of fumesfrom semiautomatic welding equipment which can beused for toxicological investigations. Fumes can becollected on filter pads and used later for in vitro in-vestigation or for delivery to animals by methods suchas intratracheal instillation. Alternatively, for animalexposures, the fumes can be delivered directly toinhalation chambers as they are generated. Palmeret al. (Ref. 100) developed a system for the collectionof complex industrial emissions which are to bedelivered to animals by intratracheal instillation. Withthis system, samples are collected directly into anaqueous medium and are ready for delivery to animalsas they are removed from the sample collectiondevice.

Toxicologic Investigations/29

3.2 In Vitro StudiesIn vitro assays of the toxicity of test substances are

frequently performed in isolated cell or tissue culturesbecause they are quicker, easier to control, andgenerally less expensive than studies in live animals.Such tests can be used for routine screening ofchemicals and complex mixtures such as weldingemissions of unknown toxicity. Test substances thatare positive in these assays can then be subjected to themore extensive and costly animal toxicity tests.

Assays of genotoxicity examine whether a substancecauses mutations or alterations in the genetic material(DNA). This is important because some diseases inman, e.g., cancer, may be caused by such alterationsin DNA. One such test of the genotoxicity of weldingfumes was recently performed by Niebuhr et al. (Refs.94 and 95). The method employed was the sisterchromatid exchange assay which tests the ability of achemical to induce certain types of chromosomaldamage (e.g., the exchange of pieces of DNA strandsbetween sister chromatids). The welding fumes testedin this assay were generated by argon-shieldedGMAW of mild steel and cast iron using nickel-rich(95% nickel, 4% manganese) electrodes. The solubilityof the fumes in water, which was extremely low (onthe order of 0.1%), was increased (to about 1%) by theaddition of fetal calf serum, which apparently convertsthe nickel salts to organic nickel complexes. Thesolubility of the fumes was markedly enhanced by theaddition of whole blood, which, by its bufferingcapacity, maintains the pH at 7.4 and prevents theformation of the insoluble salt Ni(OH)2 (Ref. 128).

The mutagenicity of the water-soluble and serum-soluble fractions of the fumes was tested independentlyand compared with that of water-soluble NiSO4. Thenickel fumes were found to be mutagenic in this test.The serum-soluble fraction of the fumes caused morechromosomal damage (per mol nickel) than did the

water-soluble fractions derived either from thewelding fumes or from nickel sulfate. The investiga-tors concluded that an important influence on thegenotoxicity of nickel in industrial fumes is thebioavailability of the nickel which may vary with thebody fluids in which it is bathed. Because of this, theyconcluded that it would be inappropriate to use eitherthe water-solubility or the total nickel content offumes for the regulation of permissible exposures toindustrial fumes (Ref. 95).

Other in vitro tests can examine the effects ofchemicals on cell function or viability. Recently,White et al. (Ref. 139) compared the cytotoxicity ofparticles collected with 0.2 pm filters from SMAWusing rutile-covered stainless steel electrodes and basicor rutile mild steel electrodes. Hexavalent chromiumrepresented about 3.5% of the particles collected fromthe stainless steel system whereas only trace quantitiesof hexavalent chromium were detected in the samplescollected from the other two electrodes. In the firsttest, saline-suspended particles were added to culturesof rapidly proliferating human cancer cells. Only theparticles from the stainless steel electrodes inhibitedproliferation. Since the dose-response curve obtainedwhen hexavalent chromium was added to the cultureswas almost superimposable on that obtained with theparticles from the stainless steel system, the investi-gators conjectured that the toxic effects were due tohexavalent chromium.

In a second test, hemolysis (disruption of the redblood cell membrane with subsequent liberation ofhemoglobin) of red blood cells by each of the particlesamples was examined. In this case, dusts from thebasic electrode were the most toxic. Finally, all threedusts caused about a 30% decrease in the viability ofcultured macrophages. The investigators did notspeculate about the factors responsible for thecytotoxicity to red blood cells and macrophages.

References

1. Ahonen, A. "Age at Death from Lung Cancer inDifferent Occupations." In Abstracts Second WorldConference on Lung Cancer, June9-13, 1980. Hansen,H.H. and Dombernowsky, P. (eds.) New York:Elsevier North-Holland, 1980.

2. Akbarkhanzadeh, F. "Combined Effects of Cig-arette Smoking and Occupatio lal Air Pollution inWelding Environment." Med. Lav., 72(l):61-6 (1981).

3. Alekseeva, I.S., Norkin, Y.I., and Andronik, V.A."Hygienic Assessment of Electric Arc Welding ofCopper with a Nonmelting Electrode in Nitrogen."Svar. Proizvod., (10):37-8 (1981).

4. Alekseeva, I.S., Norkin, Y.I., and Chumakova, I.V."Hygienic Characteristics of the Air in the ManualArgon Arc Welding of Copper Alloys." Svar.Proizvod., (2):40-l (1981).

5. Alekseeva, I.S., Ryzhikov, I.E., and Norkin, Y.I."Hygienic Evaluation of Argon Arc-Welding with aNonmelting Electrode Using Nickel-ContainingMaterials." Gig. Tr. Prof. Zabol., (5):56 (1981).

6. Alsbirk, K.E., Mogensen, C.E., Husted, S.E., andGeday, E. "Liver and Kidney Function in StainlessSteel Welders." Ugeskr. Laeg., 143(3): 112-6 (1981).

7. Al-Shamma, K.J., Hewitt, P.J., and Hicks, R."Effects of Flux Coated Electrode Welding Fume onthe Lungs of Exposed Animals." 2nd Int. Congress onToxicology, Brussels; Belgium July 6-11, 1980. Toxi-col. Lett. (Amst.), (Special Issue 1):211 (1980).

8. American Welding Society. Laboratory Methodfor Measuring Fume Generation Rates and TotalFume Emission of Welding and Allied Processes.Miami, FL: American Welding Society, F1.2-79(1979).

9. Anderson, R.S. "The Effects of Welding on Health:Evidence, Problems, Future Research Suggestions."Welding J., 61(4):40-4 (1982).

10. Andersson, H.C., Tranberg, E.A., Uggla, A.H.and Zetterberg, G. "Chromosomal Aberrations andSister-Chromatid Exchanges in Lymphocytes of Men

Occupationally Exposed to Styrene in a Plastic-BoatFactory." Mut. Res., 73:387-401 (1980).

11. Andrews, L.R. and Hanlon, R.G. "Theoreticaland Practical Considerations in the Analysis ofChromium in Welding and Other MetallurgicalFumes". Proc. 2nd Chromates Symp., pp. 244-58.Pittsburgh, PA: Industrial Health Foundation, 1981.

12. Arthaud-Richter, Y., Bresson, J.R., Chambon, P.,and Prost, G. "Qualitative Study of Atmospheric Dustin a Welding Shop." Arch. Mai. Prof. Med. Trav.Secur. Soc, 41(l):33-4 (1980).

13. Barfoot, K.M., Mitchell, I.V., Verheyen, F., andBabeliowsky, T. "Combined PIXE and X-ray SEMStudies on Time-Resolved Deposits of Welding ShopAerosols." Nucl. Instrum. Methods, 181(l-3):449-57(1981).

14. Bartley, D.L., McKinnery, W.N., Jr., andWiegand, K.R. "Ultraviolet Emissions from the ArcWelding of Aluminum-Magnesium Alloys." Am. Ind.Hyg. Assoc. J., 42(1):23-31 (1981).

15. Baum, K. "Determination of the Ozone Contentsof Tungsten - Inert Gas Welding with AnalyticalEquipment and Test Tubes - Comparison of Methods".Schweissen Schneiden, 30(11):439-41 (1978).

16. Beaumont, J.J., and Weiss, N.S. "Lung CancerAmong Welders." J. Occup. Med., 23(12):839-44(1981).

17. Beaumont, J.J., and Weiss, N.S. "Mortality ofWelders, Shipfitters and Other Metal Trades Workersin Boilermakers Local No. 104, AFL-CIO, USA."Am. J. Epidemiol., 112(6):775-86 (1980).

18. Bernal-Tapia, J.A., Sanchez Bernal, M., andFranco Trujillo, J. "Fever of Some Solderers. Informa-tion on New Cases." Rev. Med. Inst. Mex. SeguroSoc, 19(3):335-6 (1981).

19. Blakeley, St. J.H. and Zatka, V.J."Determinationof Hexavalent Chromium, Total Chromium, andTotal Nickel in Welding Fume." INCO Metals Co.,

31

32/EFFECTS OF WELDING ON HEALTH IV

internal report (1978) as cited in Andrews and Hanlon(1981).

20. Bohgard, M., Malmqvist, K.G., Johnson, G.I.and Akelsson, K.R. "Personal Aerosol Sampler withSize Fractionation and Time Resolution." In Aerosolsin the Mining and Industrial Work Environment, Vol.3. Woburn, MA: Ann Arbor Science Publishers, 1982.

21. Bohme, D., and Heuser, H. "The Production ofAir-Polluting Substances When Welding OverProduction Coatings." Schweissen Schneiden,34(2):72-77 (1982).

22. Buki, A.A., and Feldman, A.M. "Prediction ofthe Composition of Aerosol Formed in Welding inShielding Gases." Weld. Pr. (Engl. Transl)., 27(2):8-12(1980).

23. Carsey, T.P. "Feasibility Study of X-RayFluorescence for Analysis of Welding and BrazingFumes." U.S. NTIS, PB82-116807.

24. Castagna, R., and Spagnoli, G. "ExperimentalStudy on Hazards Related to Toxic Gas EmissionsDuring Continuous Arc Welding in Argon Atmos-phere." Securitas, 65(4):227-32 (1981).

25. Cavatorta, A., Falzoi, M., Mutti, A., Pedroni, C ,Frigeri, G., and Franchini, I. "Spirometric Alterationsin Chromium-Exposed Welders." Ateneo Parmense(Acta Bio-Med.), 51(4):299-304 (1980).

26. Chandra, S.V., Seth, P.K.,and Mankeshwar, J.K."Manganese Poisoning: Clinical and BiochemicalObservations." Environ. Res., 7:374-80 (1974).

27. Chandra, S.V., Shukla, G.S., Srivastava, R.S.,Singh, H., and Gupta, V.P. "An Exploratory Studyof Manganese Exposure to Welders." Clin. Toxicol.,18(4):407-16 (1981).

28. Christensson, G., and Rockert, H. "Effects ofElectrical Welding Under Water on Dental Amalgamsand Enamel In-Vitro." IRCS Med. Sci.: Libr.Compend, 8(10):757-8 (1980).

29. Cohen, D. "Measurements of the Magnetic FieldsProduced by the Human Heart, Brain and Lungs."IEEE Trans. Magn., MAG-11, No. 2, 694-700(Mar. 1975).

30. Cohen, D. "Ferromagnetic Contamination inthe Lungs and Other Organs of the Human Body."Science, 180:745-8 (1973).

31. Cohen, D., Crowther, T.S., Gibbs, G.W., andBecklake, M.R. "Magnetic Lung Measurements inRelation to Occupational Exposure in AsbestosMiners and Millers of Quebec Canada." Environ.Res., 26(2):535-50(1981).

32. Creasia, D.A. and Nettesheim, P. "RespiratoryCocarcinogenesis Studies with Ferric Oxide: A TestCase of Current Experimental Methods." In Karbe, E.

and Park, J.F. (eds.) Experimental Lung Cancer:Carcinogenesis and Bioassays, New York: Springer-Verlag, 1974.

33. Doll, R., Mathews, J.D., and Morgan, L.G."Cancers of the Lung and Nasal Sinuses in NickelWorkers: A Reassessment of the Period of Risk."Brit. J. Ind. Med., 34:102-5 (1977).

34. Edling, C. "Welders' Health Examinations(letter)." J. Soc. Occup. Med., 30(2):86 (1980).

35. Eichhorn, F., Trosken, F., and Oldenburg, T."The Production of Air Polluting Substances duringManual Arc Welding." Schweissen Schneiden, 34(2):63-7 (1982).

36. Emmett, E.A., Buncher, R., Suskind, R.B., andRowe, K.W., Jr. "Skin and Eye Diseases amongArc Welders and Those Exposed to Welding Opera-tions." J. Occup. Med., 23(2):85-90 (1981).

37. English, J.C., Parker, R.D., Sharma, R.P., andOberg, S.G. "Toxicokinetics of Nickel in Rats AfterIntratracheal Administration of Soluble and InsolubleForm." Am. Ind. Hyg. Assoc. J.,42(7):486-492(1981).

38. Erlandsson, B., Hakanson, H., Ivarsson, A., andNilsson, P. "The Difference in Protection Efficiencybetween Earplugs and Earmuffs." Scand. Audiol.,9(4):215-21 (1980).

39. Farwer, A. "Air Pollutants in Shielded ArcWelding; Relationship between Their Formation andthe Concentration in the Breathing Zone -The Resultsof Laboratory and In-Plant Measurements."Schweissen Schneiden, 34(2):97-101 (1982).

40. Farwer, A. "Measurements of Ozone Concen-trations during Inert Gas Tungsten Arc Welding ofHigh Alloy Steel and Aluminum in a FabricationShop." Schweissen Schneiden, 32(11):437-442 (1980).

41. Frant, R. "Formation of Ozone in Gas ShieldedWelding." Ann. Occup. Hyg., 6:113-25 (1963).

42. Freedman, A.P., Robinson, S.E., O'Leary, K..,Goodman, L. and Stellman, J.M. "Non-invasiveMagnetopneumographic Determination of LungDust Loads in Steel Arc Welders." Br. J. Ind. Med.,38(4):384-8 (1981).

43. Frei, A., Pirner, M., and Walser, H. "Measuresagainst Ozone Formation in Arc Welding ofAluminum under Gas Protection." Schweisstechnik(Zurich), 70(10):215-20 (1980).

44. Gibson, N.E. "Mortality in Welders (letter)."Arch. Environ. Health, 37(l):60-l (1982).

45. Gobbato, F., Melato, M., and Bucconi, S."Welder's Lung of Arc Welders: HistopathologicAspects and Possible Neoplastic Complications."Med. Lav., 71(2): 132-40 (1980).

References 133

46. Grandjean, P., and Kon, S.H. "Lead Exposureof Welders and Bystanders in a Ship Repair Yard."Am. J. Ind. Med., 2(l):65-70 (1981).

47. Grandjean, P., Selikoff, I.J., Shen, S.K., andSunderman, F.W., Jr. "Nickel Concentrations inPlasma and Urine of Shipyard Workers." Am. J. Ind.Med., 1(2)181-9 (1980).

48. Gray, C.N. University of Bradford, BradfordWest Yorkshire, U.K. Letter to the author. Oct. 25,1982.

49. Gray, C.N., and Dare, P.R.M. "Chromium inMetallic Aerosols." Abstract of Royal Society ofChemistry Papers from Conference on the Detectionand Measurement of Hazardous Substances in theAtmosphere. City University, London. Dec. 1982.

50. Gray. C.N., Dare, P.R.M., and Hewitt, P.J."Apparatus for the Controlled Generation of Fumefrom Semiautomatic Welding Processes." Interna-tional Institute of Welding, IIW Document VIII1019-82, 9pp. (Nov. 1981).

51. Gray, C.N., Hewitt, P.J., and Dare, P.R.M. "NewApproach Would Help Control Weld Fumes atSource, Part One: Biomedical Background." WeldMetal Fabric, pp. 318-24, (Sept. 1982).

52. Gray, C.N., Hewitt, P.J., Hicks, R.,and Blunsum,M. "Problems Involved in the Analysis of Chromiumin Welding Fumes." International EnvironmentalSafety Exhibition and Conference held at WembleyConference Centre, London, Sept. 2-4, 1981, 8 pp.

53. Grosse-Wordemann, J., and Stracke, E. "Re-lationships between Exposure to Fumes WhenWelding and the Welding Position Dictated by DesignConsiderations." Schweissen Schneiden, 34(2): 109-12(1982).

54. Grothe, I. "Steps Taken by the West GermanEmployers' Liability Insurance Associations to Pro-tect the Health and Safety of Welders at Work."Schweissen Schneiden, 34(2):87-90 (1982).

55. Gvozdenko, L.A., Sokolov, M.V., Primak, V.N.,Kuzina, A.S., and Shishkina, N.S. "Methods ofHygiene Research on Optical Radiation." Gig. Sanit,(2):51-22 (1980).

56. Ham, W.T., Mueller, H.A., and Sliney, D.H."Retinal Sensitivity to Damage from Short Wave-length Light." Nature, 260(5447): 153-5 (1976).

57. Hammond, G., Torgerson, W., Dotter, W., andLeach, R. "The Painful Shoulder InstructionalCourse Lectures. The Am. Acad. OrthopaedicSurgery 20:83-90. St. Louis, MO: C.V. Mosby Co.,1971.

58. Haneke, E. "Findings in the Ejaculate of ElectroWelders." Dermatol. Monatsschr., 159(11): 1036-40(1973).

59. Havrankova, J., and Skoda, V. "DNA Content inthe Lungs of Rats After Dust Application." Bratisl.Lek. Listy, 76(5):552-62 (1981).

60. Hayden, S.P., Hayden, J., Pincock, A.C., Tyler,L.E., Cross, K..W. and Bishop, J.M. "Health ofWelders in the Engineering Industry." 77th AnnualMeet. Am. Lung Assoc. and 76th Annual Meet.Thoracic Soc, Detroit, MI, May 9-13, 1981.Am. Rev. Respir. Dis., 123(4 Part 2): 125 (1981).

61. Hayden, S.P., Pincock, A.C., and Bishop, J.M."Pulmonary Function and Fume Concentration inRelation to a Brief Intensive Period of Electric ArcWelding." 77th Annual Meet. Am. Lung Assoc. and76th Annual Meet. Thoracic Soc, Detroit, MI, May9-13, 1981. Am. Rev. Respir. Dis., 123(4 Part2):125(1981).

62. Herbert, A., Sterling, G., Abraham, J., andCorrin, B. "Desquamative Interstatial Pneumonia inan Aluminum Welder." Human Pathol., 13(8):694-9(1982).

63. Herberts, P., Kadefors, R., Andersson, G., andPetersen, I. "Shoulder Pain in Industry: An Epidemio-logical Study on Welders." Acta Orthop. Scand.,52(3):299-306 (1981).

64. Herberts, P., Kadefors, R., and Broman, H."Arm Positioning in Manual Tasks: An Electro-myographic Study of Localized Muscle Fatigue."Ergonomics, 23(7):655-65 (1980).

65. Hermanns, I. "Noise Problems When Welding-Causes, Effects and Prevention." SchweissenSchneiden 34(2): 107-8 (1982).

66. Hinrichs, J.F. "The Welding Environment andResearch." Weld. J. (Miami), 59(1):64,66 (1980).

67. Hogetveit, A.C., Barton, R.T., and Kostol, C O ."Plasma Nickel as a Primary Index of Exposure inNickel Refining." Ann. Occup. Hyg., 21:113-20 (1978).

68. Iliff, W.J. "Recurrent Pterygium in a Welder(letter)." J. Am. Med. Assoc, 246(9):1019 (1981).

69. Il'nitskaya, A.V., and Kalina, O.V. "HealthEffects of Dust-Gas Mixtures in Plasma Technology."Gig. Tr. Prof. Zabol., (10):5-9 (1980).

70. Johansson, G.E., Malmqvist, K.G., Bohgard, M.,and Akselsson, K.R. "Characteristics of WeldingFumes." Institutionen fur Karnfysik, Lunds TekniskaHogskola, Lund, Sweden. Report LUTFD2/TFKF-3030, 32 pp. Aug. 21, 1981.

71. Johansson, M.I. "Gas Shielded Arc Welding andWorking Environments." Svetsen, 39(6):5-8 (1980).

72. Kadefors, R., Petersen, I., and Herberts, P."Muscular Reaction to Welding Work: An Electro-myographic Investigation." Ergonomics, 19:543-58(1976).

34/EFFECTS OF WELDING ON HEALTH IV

73. Kalliomaki, K., Aittoniemi, K., Kalliomaki, P.L.,and Moilanen, M. "Measurement of Lung RetainedContaminants In-Vivo among Workers Exposed toMetal Aerosols." Am. Ind. Hyg. Assoc. J., 42(3):234-8(1981).

74. Kalliomaki, P.L., Kalliomaki, K., Kelha, V.,Sortti, V., and Korhonen, O. "Measurement of LungContamination among Mild Steel and Stainless SteelWelders." Soudage Tech. Connexes, 35(l-2):50-2(1981).

75. Kalliomaki, P.L., Rahkonen, E., Vaaranen, V.,Kalliomaki, K., and Aittoniemi, K. "Lung-RetainedContaminants, Urinary Chromium and Nickel amongStainless Steel Welders." Int. Arch. Occup. Environ.Health, 49(l):67-76 (1981).

76. Kandracova, E. "Fertility Disorders in Welders."Cesk. Dermatol., 56(5):342-5 (1981).

77. Kawata, K., Tanabe, Y., Adachi, T., Nagasaki, Y.,Kobayashi, M., Maki, S.,and Matsumoto, H. "Exper-imental Investigation of Hazards of Lithium inWelding Fume."Tekko Rodo Eisei, 29(3):22-4 (1980).

78. Kireev, V.I., Moslov, N.I., Roshchupkin, N.P.,Schevchenko, S.A. and Matyazh, P.Y., "HygienicAppraisal of an Aerosol Formed in the Welding of aMetal Coated with Protective Primers." Svar.Proizvod., (2):41-2 (1980).

79. Koerber, D., and Fissan, H.J. "Particle Emissionfrom Welding of Painted Steel." Int. J. Environ. Anal.Chem., 10(l):13-21 (1981).

80. Koponen, M., Gustafsson, T., Kalliomaki, K.,Kalliomaki, P.L., Moilanen, M., and Pyy, L. "Dustsin a Steel Making Plant: Lung Contamination amongIron Workers." Int. Arch. Occup. Environ. Health,47(l):35-46 (1980).

81. Koponen, M., Gustafsson, T., Kalliomaki, P.L.,and Pyy L. "Chromium and Nickel Aerosols inStainless Steel Manufacturing, Grinding and Weld-ing." Am. Ind. Hyg. Assoc. J., 42(8):596-601 (1981).

82. Lucas, P.A., Jariwalla, A.G., Jones, J.H., andGough, J. "Fatal Cadmium Fume Inhalation."Lancet, 2(8187):205 (1980).

83. McMillan, G.H. "The Health of Welders in NavalDockyards: Welding, Tobacco Smoking and AbsenceAttributed to Respiratory Diseases." J. Soc. Occup.Med., 31(3):112-8 (1981).

84. McMillan, G.H., and Molyneux, M.K. "TheHealth and Welders in Naval Dockyards: The WorkSituation and Sickness Absence Patterns." J. Soc.Occup. Med., 31(2):43-60 (1981).

85. Malmqvist, K. Lund Institute of Technology,Lund, Sweden. Letter to the author. Dec. 1, 1982.

86. Malmqvist, K.G., Johansson, G.I., Bohgard, M.,and Akselsson, K.R. "Elemental Concentrations inAirborne Particles from Welding and Metal SprayingOperations." Nucl. Instrum. Methods, 181(l-3):465-71(1981).

87. Marshall, W.J., Sliney, D.H., Lyon, T.L., et al."Evaluation of the Potential Retinal Hazards fromOptical Radiation Generated by Electric Welding andCutting Arcs." Nonionizing Radiation ProtectionSpecial Study 42-0312-77 (U.S.A. EnvironmentalHygiene Agency, Aberdeen Proving Ground, Md.,1977) U.S. NTIS, AD-A043023.

88. Mayer, A. and Salsi, S. "Chemical Pollution inWorkplaces for Arc-Welding." Cah. Notes Doc,101:485-97 (1980).

89. Melekhin, A.V. and Agarkov, V.I. "Sensitizationof Fluorine and Manganese Compounds and the Stateof the Oral Cavity in Workers in the Electric PipeWelding Industry." Gig. Tr. Prof. Zabol., (8): 12-4(1981).

90. Mikhail, T.H., El-Sawaf, H.A., Ibrahim, K.M.,Awadallah, R., and El-Dessoukey, E.A., "Evaluationof the Effect of Lead Exposure on the Liver inEgyptian Lead Tank Welders." Z. Ernaehrungswiss.,19(l):50-6 (1980).

91. National Institute for Occupational Safety andHealth. "Criteria for a Recommended Standard, Oc-cupational Exposure to Chromium (VI)" U.S. Dept.of Health, Education and Welfare, Public HealthService, Center for Disease Control. DHEW Publ.No. (NIOSH) 76-129. U.S. NTIS, PB-248595 (1975).

92. National Institute for Occupational Safety andHealth. "Occupational Diseases, A Guide to TheirRecognition." U.S. Dept. of Health, Education andWelfare, Public Health Service, Center for DiseaseControl DHEW (NIOSH) Publ. No. 77-181:608pp.(June 1977).

93. National Research Council. Committee onBiologic Effects of Atmospheric Pollutants.Chromium. Washington, D.C.: National Academy ofSciences, 1974.

94. Niebuhr, E., Stern, R.M., Thomsen, E., andWulf, H.C. "Genotoxicity of Nickel Welding FumesAssayed With Sister Chromatid Exchange in HumanBlood Cells in Culture." Ann. Clin. Lab. Sci., 11(1):85-6(1981).

95. Niebuhr, E., Stern, R.M., Thomsen, E., and Wulf,H.C. "Relative Solubility of Nickel Welding FumeFractions and Their Genotoxicity in Sister ChromatidExchange In Vitro." U.S. NTIS, PB82-158411, 13pp.(cl980).

96. Novak, J.F. "Recurrent Pterygium in a Welder(letter)." J. Am. Med. Assoc, 246(9):1019 (1981).

References I'35

97. Ohmori, K., Tozawa, T., Ikemi, Y., Kobayashi, Y.,and Kitazume, M. "Pulmonary Functions Observedby Means of Flow-Volume Curve in Welders ExposedTo Ozone." Sangyo Igaku, 23(4):394-400 (1981).

98. Oleinchenko, K.A., Korneev, A.D., Zusin, V.Y.,and Kabanets, A.N. "Effect of Ceramic Flux ZhA-64Moisture on the Amount of Harmful Emissionsduring Automatic Aluminum Welding." Svar.Proizvod., (2):42-3 (1980).

99. Pabley, A.S., and Keeney, A.H. "WeldingProcesses and Ocular Hazards and Protection." Am.J. Ophthalmol., 92(l):77-84 (1981).

100. Palmer, W.G., Scholz, R.C., and Moorman, W.J."An Innovative Approach to Sampling ComplexIndustrial Emissions for Use in Animal Toxicity Tests:Applications to Iron Casting Operations." Am. Ind.Hyg. Assoc. J., 44(3):184-189 (1983).

101. Pedersen, E., Hogetveit, A.C. and Anderson, A."Cancer of Respiratory Organs Among Workers at aNickel Refinery in Norway." Int. J. Cancer, 12:32-41(1973).

102. Pokhodnya, I.K., Yavdoshchin, I.R., Bulat,A.V., and Shvachko, V.I. "Sources of Manganeseand Iron Admission into a Welding Aerosol." Avtom.Svarka, (3):37-9 (1981).

103. Polednak, A.P. "Mortality Among Welders,Including a Group Exposed to Nickel Oxides." Arch.Environ. Health, 36(5):235-42 (1981).

104. Pomaska, H.U. "The Effects of the WeldingConditions and of the Workplace Conditions on theAir-Polluting Substances Produced During ShieldedArc Welding." Schweissen Schneiden, 34(2): 102-7(1982).

105. Press, H. "The Production of Air-PollutingSubstances When Working with Fuel-Gases andOxygen." Schweissen Schneiden, 34(2):68-72 (1982).

106. Press, H. "Formation of Nitric Oxides in GasWelding. Measures for the Prevention of Injuries toHealth." Soudage Tech. Connexes, 35(5-6):207-ll(1981).

107. Riediger, G., "Respiratory Protection WhenWelding." Schweissen Schneiden, 34(2):112-4 (1982).

108. Rockert, H.O.E., and Orthendal, T. "Effects ofElectrical Welding Under Water on Dental Amalgamsand Enamel." Undersea Biomed. Res., 8(1):64 (1981).

109. Rudzki, E., and Kozlowska, A. "Causes ofChromate Dermatitis in Poland."Contact Dermatitis,6(3): 191-6 (1980).

110. Sanchis, J., Dolovich, M., Chalmers, R., andNewhouse, M.T. "Regional Distribution and LungClearance Mechanisms in Smokers and Non-Smokers." In Proc. of Symposium: Inhaled Particles

and Vapours III, Walton, W.H. (ed.) 1:183-8. Surrey,England: Unwin, 1971.

111. Schneider, W.D., and Rebohle, E. "EarlyDiagnosis by Means of Flow Volume Curves inWorkers with Respiratory Exposure." Z. Erkr.Atmungsorgane, 157(3):291-6 (1981).

112. Seerup, K.K., and Juhl, M. "OccupationalBurns. Frequency, Causes and Distribution in Occu-pational Groups." Ugeskr. Laeg., 143(21): 1356-60(1981).

113. Sheers, G., and Coles, R.M. "MesotheliomaRisks in a Naval Dockyard." Arch. Env. Health,35(5):276-82 (1980).

114. Sjogren, B. "Work Environment Problems inWelding. No. 13. Relationship between Air and UrineConcentrations of Fluorides, Chromium and Nickel inWelding." Arbete och Halsa, (9): 1-25 (1981).

115. Sjogron, B. "A Retrospective Cohort Study ofMortality among Stainless Steel Welders." Scand. J.Work Environ. Health, 6(3): 197-200 (1980).

116. Sliney, D.H., Moss, C.E., Miller, C.G., andStephens, J.B. "Semintransparent Curtains forControl of Optical Radiation Hazards." AppliedOptics, 20(14):2352-66 (1981).

117. Smars, E. "The Problem of Ozone during MIGWelding of Aluminum." Schweiz. Alum. Rundsch.,31(6):255-61 (1981).

118. Spelbrink, H. "Preventative Health CareExaminations for Welders - Some Industrial HealthCare and Administrative Aspects." SchweissenSchneiden, 34(2):84-6 (1982).

119. Spelbrink, H. "Measures in the Field of MedicineAimed to Improve the Working Conditions ofWelders." Annual Welding Conference, Sept. 16-23,1981, Essen, West Germany, pp. 16-20. Dusseldorf,West Germany: Deutscher Verband fur Schweiss-technik, 1981.

120. Stern, R.M. "Assessment of Risk of LungCancer for Welders." Arch. Environ. Health, 38(3)(1983) in press.

121. Stern, R.M. "Occupational Health RiskAssessment: A Preparatory Study of the Exposureof Welders to Toxic Substances and the ResultingHealth Effects: The Working Environment ResearchGroup." U.S. NTIS, PB82-119041. 555pp. (cl981).

122. Stern, R.M., "Process-Dependent Risk ofDelayed Health Effects for Welders." Environ.Health Perspect., 41:235-53 (1981).

123. Stern, R.M., "A Protocol for the Study ofHealth Effects of Welding: The Working Environ-ment Research Group." U.S. PB82-119421. 33pp.(cl981).

36/EFFECTS OF WELDING ON HEALTH IV

124. Sunderman, F.W., Jr. "Carcinogenic Effectsof Metals." Fed. Proc, 37(l):40-6 (1978).

125. Suzuki, Y.,Toda, K., Koike, S., and Yoshikawa,H."Cadmium, Copper and Zinc in the Urine ofWelders Using Cadmium Containing Silver Solder."Ind. Health, 19(4):223-30 (1981).

126. Taylor, H.R. "Ultraviolet Radiation andPterygium."J. Am. Med. Assoc, 247(12): 1698 (1982).

127. Taylor, H.R., "Aetiology of Climatic DropletKeratopathy and Pterygium (letter)." Br. J. Optha-mol., 64:154-63 (1980).

128. Thomsen, E., and Stern, R.M. "Collection,Analysis and Composition of Welding Fumes." U.S.NTIS, PB82-155862. 34pp. (cl981).

129. Thomsen, E., and Stern, R.M. "A SimpleAnalytical Technique for the Determination ofHexavalent Chromium in Welding Fumes and OtherComplex Matrices." Scand. J. Work Environ. Health,5(4):386-403 (1979).

130. Tredre, B.E., and Collins, J.C. "A TrackingTask Used for the Simulation of Manual Metal ArcWelding." Ergonomics, 23(4):401-4 (1980).

131. Tsyrkunov, L.P. "Skin Diseases in ElectricWelders." Vestn. Dermatol. Venerol., (10):40-3 (1981).

132. Tumakov, A.Z., and Grigorev, I.I. "Preventionof Meteorotropic Reactions in Workers in VariousOccupations Undergoing Treatment in a SanitariumPreventorium." Gig. Tr. Prof. Zabol., (4):22-6 (1981).

133. Ulfvarson, U. "Survey of Air Contaminantsfrom Welding." Scand. J. Work Environ. Health,7(Suppl. 2): 1-28 (1981).

134. Vallyathan, V., Bergeron, W.N., Robichaux,P.A., and Craighead, J.E. "Pulmonary Fibrosis in anAluminum Arc Welder." Chest, 81(3):372-4 (1982).

135. Villaume, J.E., Wasti, K., Liss-Suter, D. andHsiao, S. Effects of Welding on Health, Volume I.Miami, FL: American Welding Society, 1979.

136. Weymueller, C.R. "Uncle Sam Surveys WeldingSafety." Weld. Des. Fabr., 53:138-9 (1980).

137. White, L.R., Hunt, J., Richards, R.J, andEik-Nes, K.B. "Biochemical Studies of Rat LungFollowing Exposure to Potassium Dichromate orChromium-Rich Welding Fume Particles." Toxicol.Lett., 11(1-2): 159-63 (1982).

138. White, L.R., Hunt, J., Tetley, T.D., andRichards, R.J. "Biochemical and Cellular Effects ofWelding Fume Particles in the Rat Lung." Ann.Occup. Hyg., 24(l):93-101 (1981).

139. White, L.R., Richards, R. J., Jakobsen, K., andOestgaard, K. "Biological Effects of Different Typesof Welding Fume Particulates." Proc. Meet., 1979.The In Vitro Effects of Mineral Dusts, Brown, R.C.(ed.) pp. 211-8 London: New York: Academic Press,1980.

140. Wulf, B., and Seefeld, G. "Comparative Studiesof the Periodontal State in Welders." Stomatol DDR,30(7):519-26 (1980).

141. Yager, J.W. and Benz, R.D. "Exposure toEthylene Oxide at Work Increases Sister ChromatidExchanges in Human Peripheral Lymphocytes."Science, 219:1221-3 (1983).

142. Yatsenko, K.S., Trubnikov, G.A., Afanasev,Y.A., and Molokanov, V.M. "Vibration Sensitivityin Workers Associated with Shipbuilding." Sov. Med.,(5):78-80 (1981).

143. Zakhari, S. and Anderson, R.S. Effects ofWelding on Health, Volume II. Miami, FL: AmericanWelding Society, 1981.

144. Zakhari, S., and Strange, J. Effects of Weldingon Health, Volume III. Miami, FL: American Weld-ing Society, (In press).

145. Zober, A. "Possible Dangers to the RespiratoryTract from Welding Fumes—Methods of Approach inan Industrial Health Care Context and Results."Schweissen Schneiden, 34(2):77-81 (1982).

146. Zober, A. "Symptoms and Findings at theBronchopulmonary System of Electric Arc Welders.1. Epidemiology." Zentralbl. Bakteriol., 173(l-2):92-119 (1981).

147. Zober, A. "Symptoms and Findings at theBronchopulmonary System of Electric Arc Welders.2. Pulmonary Fibroses." Zentralbl. Bakteriol.,173(1-2): 120-48 (1981).

148. Zugerman, C. "Chromium in Welding Fumes."Contact Dermatitis, 8(l):69-70 (1982).

Bibliography

S & H Research Reports

WEFGWAWNEWH1EWH2EWH3ULRWFCWFDPLVOS

Welding EnvironmentFumes and Gases in the Welding EnvironmentArc Welding and Cutting NoiseEffects of Welding on Health, Vol. 1Effects of Welding on Health, Vol. 2Effects of Welding on Health, Vol. 3Ultraviolet Reflectance of PaintWelding Fume ControlWelding Fume Control - A Demonstration ProjectLaboratory Validation of Ozone Sampling with Spill Proof Impingers

S & H Standards

Z49.1-73A6. 2-73Fl.1-76

F 1.2-79

Fl.3-83

F2.1-78F4.1-80

F6.1-78

AWS

Safety in Welding and CuttingLens Shade Selector (wall chart 11 x 17)Method of Sampling Airborne Particulates Generated by Weldingand Allied ProcessesLaboratory Method for Measuring Fume Generation Rates and Total FumeEmission of Welding and Allied ProcessesEvaluating Contaminants in the Welding Environment - A SamplingStrategy GuideRecommended Safe Practices for Electron Beam Welding and CuttingRecommended Safe Practices for the Preparation for Welding andCutting Containers and Piping That Have Held Hazardous SubstancesMethod for Sound Level Measurement of Manual Arc Welding andCutting ProcessesArc Welding Safety (information booklet)

37