A national case-crossover study on ambient ozone pollution and first-ever stroke among Chinese adults: Interpreting a weak association via differential susceptibility

摘要

Evidence suggesting an association between ozone exposure and stroke risk remains inconsistent; variations in the distributions of susceptibilities of the study populations may explain some of it. We examined the hypothesis in a general Chinese population.During 2013-2015, 1356 first-ever stroke events were selected from a large representative sample, the China National Stroke Screening Survey (CNSSS) database; daily maximal 8-hour ozone concentrations were obtained from spatiotemporally interpolated estimates of in-situ observations over China. We conducted a time-stratified case-crossover design to assess associations between stroke risk and ambient ozone exposure. Next, potential effect modifiers were identified using interaction analyses. Final, a well-established approach was applied to estimate individual-level susceptibility (i.e., the individual-specific effect given a certain combination of multiple effect-modifiers) and its probability distribution among all the CNSSS participants (n = 1,292,010).With adjustments for temperature, relative humidity and ambient fine particulate matter exposure, a 10-mu g/m(3) increment in mean ozone levels 2-3 day prior to symptom onset was associated with a 3.0% change in stroke risk (95% confidence interval: -12%, 73%). This association was statistically significantly enhanced by male gender, rural residence and low vegetable and fruit consumption. The subgroup results suggested that a fraction of the population might be considerably affected by ozone, regardless of the insignificant association in average level. The analysis of susceptibility distribution further indicated that the ozone-stroke association was statistically significantly positive in 14% of the general population.Susceptibility to ozone-related stroke significantly varied among Chinese adults. Characterizing individual-level susceptibility reveals the complexity underlying the weak average effect of ozone, and supports to plan subpopulation-specific interventions to mitigate the stroke risk. (C) 2018 Elsevier B.V. All rights reserved.