RESUMO
The most recent meta-analysis of gas cooking and respiratory outcomes in children was conducted by Lin et al. [93] in 2013. Since then, a number of epidemiology studies have been published on this topic. We conducted the first systematic review of this epidemiology literature that includes an in-depth evaluation of study heterogeneity and study quality, neither of which was systematically evaluated in earlier reviews. We reviewed a total of 66 relevant studies, including those in the Lin et al. [93] meta-analysis. Most of the studies are cross-sectional by design, precluding causal inference. Only a few are cohort studies that could establish temporality and they have largely reported null results. There is large variability across studies in terms of study region, age of children, gas cooking exposure definition, and asthma or wheeze outcome definition, precluding clear interpretations of meta-analysis estimates such as those reported in Lin et al. [93]. Further, our systematic study quality evaluation reveals that a large proportion of the studies to date are subject to multiple sources of bias and inaccuracy, primarily due to self-reported gas cooking exposure or respiratory outcomes, insufficient adjustment for key confounders (e.g., environmental tobacco smoke, family history of asthma or allergies, socioeconomic status or home environment), and unestablished temporality. We conclude that the epidemiology literature is limited by high heterogeneity and low study quality and, therefore, it does not provide sufficient evidence regarding causal relationships between gas cooking or indoor NO2 and asthma or wheeze. We caution against over-interpreting the quantitative evidence synthesis estimates from meta-analyses of these studies.