Laserfiche WebLink
INDOOR NITROGEN DIOXIDE, GAS COOKING AND ASTHMA 1725 <br />illness risk. This early quantitative analysis became a <br />benchmark study for the relationship between indoor <br />NO, and respiratory illness in children, and an <br />important reference for the outdoor NO, Air Quality <br />Guideline value established by the World Health <br />Organization (WHO) in 1997' and confirmed in <br />2005.7 More recently, WHO has reviewed studies on <br />indoor NO, exposure, but without doing a formal <br />meta-analysis.8 Recent journal reviews of the <br />issue""` have also been qualitative. In view of <br />the dearth of quantitative meta-analyses based on <br />recent studies, we decided to review studies on <br />asthma, wheeze, gas cooking and indoor NO, in chil- <br />dren with the purpose of obtaining quantitative effect <br />estimates. <br />Methods <br />Selection criteria <br />We searched for studies from which quantitative <br />effect estimates of the relationship between gas cook- <br />ing, indoor NOz and respiratory health effects in chil- <br />dren could be obtained. We attempted to identify all <br />population studies in relation to this topic. The litera- <br />ture was searched with PubMed and ISI Web of <br />Knowledge from 1977 up to 31 March 2013 with <br />the following search terms: (i) indoor nitrogen <br />1 1064 articles from databases I <br />dioxide and children; (ii) personal nitrogen dioxide <br />and children; (iii) gas cooking and children; (w) gas <br />appliance and children; (v) unvented and children; <br />(vi) gas heating and children; and (vii) gas heater <br />and children. The seven search results were combined <br />with the Boolean operator 'or'. All of the 34 epidemio- <br />logical studies included in Table 5.2 of the recent <br />WHO guidelines for indoor air quality and citations <br />from previous reviews and identified articles were <br />considered as well. Duplications were removed. <br />To be eligible for inclusion, studies had to: (i) be <br />published in English; (if) be primary study, not re- <br />views; (iii) examine respiratory disease in infancy or <br />in childhood (defined by a maximum age of subjects <br />18 years) as outcomes; (iv) examine exposure to <br />indoor NO, or household gas cooking or gas heating; <br />(v) be conducted within family houses, not in schools <br />or classrooms; and (vi) report an odds ratio or other <br />effect estimator12 or sufficient data to estimate them. <br />Articles fulfilling all six criteria were included for <br />further review (Figure 1). <br />All studies were reviewed according to the six inclu- <br />sion criteria. Commentaries, and studies not per- <br />formed in children or exposures not relevant or <br />without respiratory outcomes, were excluded. The <br />remaining articles were reviewed independently by <br />the three authors. Articles that did not report on the <br />association between selected exposure variables and <br />295 duplicates and 34 non-English articles discarded <br />1 735 articles selected for absa'act review I <br />571 excluded: <br />Studies did not meet the inclusion criteria (reviews, <br />commentaries, studies not performed in children, <br />exposure not relevant, without respiratory outcomes) <br />7 additional articles, from search of reference lists <br />171 articles selected Por review of whole article to <br />determine whether exposure and outcome is relevant <br />67 excluded: <br />Studies not reporting on respiratory outcomes or selected <br />exposure assessments (including indoor gas heating); not <br />reporting a quantitative effect estimate <br />63 excluded: Studies not about asthma and wheeze: <br />exposure was gas heating, coal fuel and unvented geysers; <br />gas cooking was compared to biomass burning or use of <br />fossil fuels; indoor nitrogen dioxide was dominated by <br />outdoor source; studies based on same dataset <br />41 at fides using asthma and wheeze as health outcomes selected <br />for data abstraction, quality assessment and meta-analysis <br />Figure 1 Study selection flow chart <br />