Association between indoor air pollution measurements and respiratory health in women and children in Lao PDR

  • Published on

  • View

  • Download

Embed Size (px)


<ul><li><p>Association between indoor air pollution measurements and</p><p>respiratory health in women and children in Lao PDR</p><p>Introduction</p><p>There is a well-established association between ambientair pollution and adverse health outcomes in developedand developing countries (WHO 2001, 2002; Schwartz,1994). Owing to the substantially longer time periodsspent indoors where dispersal of pollutants may bepoor, indoor air pollution can pose many hundreds oftimes greater exposure than outdoor air pollution(Bruce et al., 2000). Indoor air pollution is of partic-ular concern for women and children in developing</p><p>countries (Bruce et al., 2000; Smith and Mehta, 2003;Ezzati, 2005; Rinne et al., 2007), where up to 90% ofrural households cook and heat using unprocessedbiomass such as dung, crop residues, wood andcharcoal (Bruce et al., 2000). A number of studieshave identied indoor use of biomass fuel as a majorcontributor to indoor air pollution (Zhang and Smith,1999; Wornat et al., 2001; Oanh et al., 1999; Daiseyet al., 1989; Fischer and Koshland, 2007), which isclaimed to account for 2.22.8 million deaths annuallyin developing countries (Tuckett et al., 1998). Indeed,</p><p>Abstract This article presents the results of a study on the association betweenmeasured air pollutants and the respiratory health of resident women and chil-dren in Lao PDR, one of the least developed countries in Southeast Asia.The study, commissioned by the World Health Organisation, included PM10, COand NO2 measurements made inside 181 dwellings in nine districts within twoprovinces in Lao PDR over a 5- month period (12/0504/06), and respiratoryhealth information (via questionnaires and peak expiratory ow rate (PEFR)measurements) for all residents in the same dwellings. Adjusted odds ratios werecalculated separately for each health outcome using binary logistic regression.There was a strong and consistent positive association between NO2 and CO foralmost all questionnaire-based health outcomes for both women and children.Women in dwellings with higher measured NO2 had more than triple of the oddsof almost all of the health outcomes, and higher concentrations of NO2 and COwere signicantly associated with lower PEFR. This study supports a growingliterature conrming the role of indoor air pollution in the burden of respiratorydisease in developing countries. The results will directly support changes inhealth and housing policy in Lao PDR.</p><p>K. Mengersen1, L. Morawska2,H. Wang2, N. Murphy1,F. Tayphasavanh3, K. Darasavong3,N. S. Holmes2</p><p>1School of Mathematical Sciences, QueenslandUniversity of Technology, Brisbane, Qld, Australia,2International Laboratory for Air Quality and Health,Queensland University of Technology, Brisbane, Qld,Australia, 3Ministry of Health, Vientiane, Laos</p><p>Key words: Indoor air pollution; Developing countries;Respiratory health outcome; Pulmonary disease;Exposure; Statistical association.</p><p>L. MorawskaInternational Laboratory for Air Quality and HealthQueensland University of Technology2 George Street, BrisbaneQld 4001, AustraliaTel.: +61 7 3138 2616Fax: +61 7 3138 9079e-mail:</p><p>Received for review 28 October 2009. Accepted forpublication 2 July 2010.</p><p>Practical ImplicationsThis is the rst study that investigated indoor air quality and its impact within residential dwellings in Lao PDR,which is one of the poorest and least developed countries in south-east Asia, with a life-expectancy of 56 years in 2008.While there have been other studies published on indoor air quality in other developing countries, the situation inLaos is dierent because the majority of houses in Laos used wood stoves, and therefore, emissions from woodburning are the dominant sources of indoor air pollution. In other countries, and studies, while emission from woodburning was investigated, wood was rarely the main or the only fuel used, as the houses used in addition (or solely)dung, kerosene or coal. The study quantied, for the rst time, concentrations in houses two provinces in Laos PDRand shed light on the impact of human activities and urban design on pollutant concentrations and respiratory health.This study contributes to the accumulation of evidence to provide more reliable estimates of risk and a more informedbasis for decision-making by concerned governments and communities.</p><p>Indoor Air 2011; 21: in Singapore. All rights reserved</p><p> 2010 John Wiley &amp; Sons A/S</p><p>INDOOR AIRdoi:10.1111/j.1600-0668.2010.00679.x</p><p>25</p></li><li><p>the household use of solid fuels has been claimed as thelargest single environmental cause of ill health (Smithet al., 2005).Poor indoor air quality in general and the use of</p><p>biomass fuels for cooking and heating in particular havebeen shown to be associated with reduced lung functionwhich, in turn, has been linked to short- and long-termrespiratory and cardiac illnesses, particularly in womenand children (Smith et al., 2005; Desai et al., 2004; Karaet al., 2003; Ozbay et al., 2001; Rudan et al., 2008;Dherani et al., 2008; Liu et al., 2007; Diaz et al., 2007).Carbon monoxide (CO), nitrogen dioxide (NO2), par-ticulate matter (PM) and other chemicals have beenconrmed in relationships between biomass fuel andadverse respiratory health (Rinne et al., 2007; Dionisioet al., 2008; Franklin, 2007; Rumchev et al., 2007; Liuet al., 2008; Koistinen et al., 2008; Kumar et al., 2007).Biological mechanisms for the eect of biomass smokeon lower respiratory infections in children, and math-ematical modelling of intake of pollutants from episodicemissions have also been considered (Grigg, 2007;Nazaro, 2008). These associations are potentiallymodied by biological contaminants such as dust, pets,damp and mould, as well as other chemical contami-nants such as tobacco smoke, o-gassing emissions,other carcinogens, gender and age (Wigle et al., 2007;Dales et al., 2008). The instruments used for determin-ing respiratory health, which include clinical measuressuch as hospital admission, questionnaire-based assess-ment, or direct measurements such as forced expiratoryvolume, may also impact on the observed results (Eptonet al., 2008).This article reports on a study of the association</p><p>between measured air pollutants in residences in LaoPDR, and the reported respiratory health and mea-sured lung function of women and children who livedin these residences. The study also evaluates thecontribution of relevant indoor air activities, includingtime spent near a cooking place and near a re, dustinside and outside, smoking and drying clothes inside.This is part of a larger epidemiological and measure-ment study, detailed in Methods. The Lao populationis distinguished by a number of factors that impactsubstantially on this association. The population is themost rural in south-east Asia, and wood is almostuniversally used for cooking. However, the climate ismore temperate than many other countries in whichsimilar studies have been conducted, and the majorityof houses are well ventilated because of their construc-tion from wood or bamboo (see Figure 1).</p><p>Methods</p><p>Study design</p><p>The study was comprised of a casecontrol studyembedded in a cross-sectional study and was designed</p><p>in collaboration with thirteen representatives of theLao Ministry of Health, the National Statistics Centreand the Science, Technology and EnvironmentalAgency. Participants in the study were chosen accord-ing to a cluster design comprising two provinces, ninedistricts within these provinces and 20 hospitals orhealth centres within the districts. Six districtsfrom Vientiane province (Phonhong, Mad, Feuang,Thoulakhom, Kasy and Vangvieng) and three fromBorikhamxay province (Bolikhanh, Khamkeut andPakkading) were chosen to meet the prerequisites ofwide representation of ethnic groups, wide range ofhousing characteristics, high prevalence of respiratoryillness, accessibility and adequate sta resources. Foreach province, a list of hospitals, health centres,villages and village population within each districtwas obtained from the National Statistics Centre. Atotal of 20 health centres were selected as follows: 12from Vientiane province (Thoulakhom (3), Phonhong(1), Kasy (1), Vangvieng (3), Feuang (3), Mad (1)) and8 from Bolikhamxay province (Bolikhanh (2), Pakk-ading (2), Khamkeut (4)).For each health centre (or group of centres), four</p><p>children admitted for acute respiratory illness, aged 14 years sequentially from 1st October 2005, wereenrolled in the study. No more than two such childrenwere enrolled from any one village. One to twochildren were randomly selected from the same village(in a dierent house), matched by age, ethnic groupand location of kitchen. The houses in which thesechildren resided were included in the study. For eachhouse in the study, experimental measurements ofindoor air quality were taken. For each householdmember, lung function measurements and question-naire information related to indoor air factors, expo-sure and respiratory and cardiovascular health weregathered; details are provided in Sections Question-naire and Indoor air pollutants measurement below.</p><p>Fig. 1 An example of a typical Laos house</p><p>Mengersen et al.</p><p>26</p></li><li><p>The study was designed to be conducted in the coolseason during which indoor exposure and respiratoryillness are potentially highest.</p><p>Questionnaire</p><p>A questionnaire was delivered to a representative ofeach household enrolled in the study. The survey wasconducted in 356 houses, including the 199 houses, inwhich the air-quality measurements were performed.The questionnaire was based on the American Tho-racic Society Questionnaire, the World Health Survey,World Bank approach, ISAAC questionnaire andWestern Australia Heart Survey, and critically evalu-ated by two international experts. Separate questionswere developed for children aged 14 years, youth aged514 years and adults. They were then tailored to localconditions and interests by the 13-member local team,many of whom have medical degrees and are profes-sionals in public health and environmental practice.The questionnaire was translated into Lao and pilotedin Pong Song.A small team of local health experts was selected to</p><p>administer the questionnaire to all households in thestudy. The team participated in the questionnairedevelopment, a prestudy training workshop and thepilot exercise.The questionnaire comprised six parts: interviewer</p><p>information, completed by the chief interviewer; a childhealth questionnaire for each child aged 14 years inthe household, completed by a parent or primary carer;a child health questionnaire for each child aged 514 years in the household, completed by a parent orprimary carer; an adult health questionnaire for eachadult in the household, completed by the adultsthemselves where possible or by a close relative orhead of the household; a household characteristicsquestionnaire, completed by the head of the household;and a town/village questionnaire, completed for eachcommunity by the community head, community healthworker or the head of the household.Questions on household characteristics focused on</p><p>indoor combustion sources, other exposures such assmoking and dust, and ventilation. Health questionsfocused on symptoms, medical diagnoses and hospi-talization associated with acute respiratory infectionsin the past 2 weeks, the past month and in the pastyear. Direct and indirect indoor, outdoor and occupa-tional exposures to pollutants were also solicited.</p><p>Indoor air pollutants measurement</p><p>Measurements of pollutants were conducted between23rd December 2005 and 14th April 2006. Measure-ments were made within the living area in eachdwelling for a period of 12 h, from 06:00 to 18:00from December to April (corresponding to the dry</p><p>season) during which cooler outdoor temperatureswere expected to result in higher indoor exposure andincidence of respiratory illness.Particle mass (PM10), nitrogen dioxide (NO2) and</p><p>carbonmonoxide (CO)weremeasuredwithin the housesusing a portable Escort ELF sampling pump (MSA),connected to a cyclone (PM10), long duration colourdetectionCO tubes (Draeger) andNO2 adsorption tubes(SKC 224-40-02). The cyclone contained preweighed0.8- lm pore PVC lters (MSA) through which the owrate was adjusted to 1.7 l/min. After equilibrated in adesiccator (20C, 65%RH), the lters were weighed on amicrobalance (Mettler Toledo), with an upper limit of3 g and 0.002 mg readability. The CO detection tubeswere designed for measurement of time-weighted aver-ageCOconcentration.The chemicals in the tubes changein colour in response to the presence of CO gas in thesample. By knowing the volume of sampled air, theamount of colour change can be translated into a veryaccurate measurement of the level of CO, described as apercentage of the total air or in parts per million (ppm).NO2 was sampled and analysed following the Occupa-tional Safety and Health Administration TechnicalManual (6014). In brief, NO2 was sampled by adsorp-tion onto two consecutive molecular sieve tubes treatedwith triethanolamine, which reacted with NO2 to formnitrite and nitrate ions. NO2 concentration was detectedusing aUV-Vis spectrometer (ThemoSpectronic) at k =540 nm to measure the sum of nitrite and nitrate ions.The location of instruments was chosen based on the</p><p>following constraints: (i) representative of the mainliving area of house; (ii) away from the sources; (iii)away from the windows and the doors; and (iv) safe forthe instrumentation and people.</p><p>Lung function measurement</p><p>Peak expiratory ow rate (PEFR, measured in l/min)was chosen as the indicative measure of lung functionin this study, because of its ability to detect moderateor severe airway obstruction, the simplicity of admin-istration, the portability of the measuring instrumen-tation and its applicability for a wide cross-section ofthe population.PEFR measurements were taken for each available</p><p>member of the household in which air pollutantmeasurements were made. Measurements were con-ducted by the same trained medical and health sta whoadministered the questionnaires, using a hand-heldPEFRmeter. The comparative unreliability of measure-ments obtained for children</p></li><li><p>using two types of respiratory health measures. Therst was based on the questionnaire responses and thesecond was based on lung function measurements.Analysis of respiratory health outcomes based on</p><p>questionnaire data focused on reported events in thepast 2 weeks and the past month for children 14 years, and the past month and the past 12 months forchildren aged 514 years and for women aged15+ years. Exploratory statistical analyses comprisedsummary statistics, plots and t-tests. Adjusted oddsratios were calculated for each health outcome sepa-rately using binary logistic regression, with the loga-rithm of the pollutant measurements as explanatoryvariables. Models were t with and without a constant,and adjustment for age and gender of the children wasalso considered. Results are reported for modelswithout a constant or adjustment. Multivariate anal-yses of variance based on general linear model...</p></li></ul>