Temporal trends in associations between ozone and circulatory mortality in age and sex in Canada during 1984–2012

Abstract

Background Considerable research has been conducted on the association between ground-level ozone (ozone) and various causes of mortality, but the relationships by age and sex (biological) have been inconsistent, and temporal trends remain unexplored. Objectives The study goals are to investigate the adverse health effects of short-term exposure to ozone on circulatory mortality by age and sex, and to examine trends in annual health effects. Methods Daily ozone, temperature, and circulatory mortality counts (ICD I00-I99) were collected for 24 urban cities for 29 years (1984–2012). Associations between ozone and circulatory mortality were estimated using generalized additive Poisson models for season (warm vs. cold), age [base (≥1) vs. seniors (>65)], and sex, accounting for confounders (calendar-time, temperature, day of the week). City-specific estimates were pooled to represent national associations through Bayesian hierarchical models. Results While the cold season returned insignificant estimates, the warm season showed statistically significant associations: a 10 ppb increase in ozone was associated with 0.7% increase in circulatory mortality with a 95% posterior interval of 0.2%, 1.1%. One-day lagged ozone in the warm season showed little age differences [0.7% (0.23%, 1.12%) vs. 0.8% (0.22%, 1.27%)], but visible sex differences: females were at a higher circulatory mortality risk than males [1.1% (0.31%, 1.71%) vs. 0.3% (−0.46%, 0.98%)]. Annual estimates suggest overall up-down temporal changes; a slightly increasing trend until 2002–2004, and a generally decreasing trend thereafter. Conclusion This study found noticeable sex-related differences in circulatory mortality attributable to short-term exposure to ozone. Further research is warranted to understand whether sex alone, or unknown interactions with other factors derived the differences, and to clarify the specific biological mechanisms underlying differences in risk estimates between females and males.