Commentary: Gas cooking and child respiratory health—time to identify the culprits?

In this issue of the IJE, Lin and colleagues¹ report the results of a meta-analysis of the effect of indoor nitrogen dioxide (NO₂) and gas cooking on asthma and wheeze in children. Effect estimates summarizing 19 studies show that the risk of asthma increases by 32% when a gas cooker is present in the home, and 7 studies combined show that the risk of wheeze increases by 15% for a 15 ppb increase in NO₂. The presence of gas cookers inside the home is common in developed countries (around 50–70%) and has long been established as a main source of indoor air pollution, in particular NO_2.² Young children are among the population subgroups with greater exposure to indoor air pollution, as they spend a larger part of their day at home; they can indeed experience the largest part of their air pollution exposures from indoor rather than outdoor sources. Concerns about the respiratory effects of indoor air pollution from gas cooking are not new; since the 1980s there has been a plethora of studies. These were first combined into summary effect estimates by Hasselblad in 1992.³ The Hasselblad meta-analysis was particularly influential as it formed the basis for the World Health Organization (WHO) outdoor NO₂ guideline⁴ of 40 μg/m³, recently recommended also for indoor exposures.⁵ The quantitative effect estimates produced by Lin et al. are timely because many new studies have been published since 1992. Notably, even though the outcome definitions are not directly comparable, the estimates provided by Lin et al. are in line with the Hasselblad estimates from 20 years ago, and support the occurrence of respiratory health effects at the level of the WHO guideline: an important finding that should have implications for future reviews of this guideline.

Inconsistencies remain in some of the results published by Lin et al.: asthma was more clearly associated with gas cooking than with NO₂ concentrations, whereas for wheeze the reverse was found. As discussed by the authors, if gas cooking increases the risk of asthma more than NO₂ exposure does, or vice versa for wheeze, this raises important questions about the role of co-pollutants and co-risk factors. Is NO₂ the sole culprit? In developed countries, gas cooking is a main predictor of indoor NO₂ concentrations together with other sources: gas heaters and boilers, cigarette smoking and outdoor traffic-related NO₂. The presence of a gas cooker in the home is therefore often used as a proxy for indoor NO₂ exposure—as in the Lin et al. meta-analysis. However, gas cooking produces a complex mixture including volatile organic compounds, fine particulate matter (PM_2.5) and ultrafine particles (UFPs) of nanometre size, polycyclic aromatic hydrocarbons, carbon monoxide and nitrous acid. Possible synergistic effects of exposure to these pollutants are not known at this point. It is also not known whether NO₂ acts as surrogate for the other pollutants, such as PM_2.5 and UFPs, for which respiratory effects are also suspected.⁶

Outdoor air pollution studies often use NO₂ as a surrogate for exposure to the complex mixture of traffic exhaust-related pollutants. The challenges involved in identifying the pollutant(s) responsible for particular effects in a complex mixture are large. In outdoor air pollution research, multi-pollutant approaches are increasingly being considered, for example by applying more enhanced statistical techniques (dimension reduction and Bayesian hierarchical methods) to existing multi-pollutant exposure data.⁷ Indoor air pollution studies have not yet attempted to tackle the mixture challenge and the measurement of multiple indoor air pollutants should be a priority for future studies. Taking a step further, personal monitoring of exposure to indoor air pollutants would help to provide more accurate estimates of combined exposure to air pollutants from multiple indoor and outdoor sources. Increasingly available technologies for personal exposure measurement,⁶ combined with prediction modelling and information on time-activity patterns,⁸ may open new perspectives for combining indoor and outdoor air pollution exposure estimates. Of course, and again using the outdoor air pollution analogy, measures taken to control air pollution frequently lead to the reduction in concentrations of more than one pollutant by controlling sources of pollutants.⁵ In the indoor environment control measures for NO₂, such as improved ventilation, will be beneficial in reducing co-exposures also.

Of further concern is the role of other common indoor factors and housing quality indicators that may be closely related to the use of gas cookers and indoor NO₂. Several studies have shown that gas cookers are more often present in older houses, in more crowded houses and in houses with other gas appliances, and that their presence is strongly associated with other asthma and wheeze risk factors such as dampness, mould, pets and environmental tobacco smoke.^9,10 The risk estimates studies included in the meta-analysis by Lin et al. were those adjusted for asthma risk factors, but it is not described which these were, apart from smoking in the household. In any case, it is hard to disentangle effects of strongly associated indoor factors and housing quality through traditional confounder adjustment. Stratification of the meta-analysis by common indoor factors would have been valuable, but the published risk estimates were probably not available for this. Some of the heterogeneity in the effects between studies reported by Lin et al. is likely to be due to the above co-risk factors.

To summarize, the quantitative effect estimates published by Lin et al. are a timely and important check on the current NO₂ guidelines. They call for effective control measures of gas cooking-related emissions. At the same time, questions about co-exposures and co-risk factors in the indoor environment remain unanswered. In assessing the totality of indoor exposures we may move towards more reliable assessment of the causal mechanisms in the association between gas cooking and children’s respiratory health.

Conflict of interest: None declared.

References