Toward a Cohort Perspective of Climate Epidemiology: The Case of Examining Intergenerational Inequalities in Susceptibility to Non-Optimal Temperatures in Japan.

Abstract

Background: Younger generations are projected to experience more severe climate exposure impacts during their lifetimes than older generations as global warming progresses. Despite the increasing evidence of the recent temporal changes in heat-related mortality risks, there remains a lack of research exploring this association from a cohort perspective.

Objective: Our objective was to quantify the variation in susceptibility to short-term effects of non-optimal temperature on mortality, across generations and over the life course of specific generations, using a novel age-period-cohort approach.

Methods: An extended two-stage analytical approach was applied to a nationwide mortality dataset covering individuals born from 1866 to 2019 in all 47 prefectures in Japan. Daily mortality counts observed between 1972 and 2019 were aggregated into 5-year birth cohorts and corresponding age groups of the decedents. For each prefecture and birth cohort, the age-dependent association between ambient temperature and daily mortality was modeled using conditional quasi-Poisson regression. Then, the prefecture-specific associations were pooled across cohorts, separately for each age group, using a repeated-measure meta-regression. To model the intergenerational changes in risks, a nonlinear, continuous term for cohort was applied in the meta-analysis.

Results: A total of 29 million all-cause deaths were analyzed. The relative risk (RR) of heat-related mortality (99th temperature percentile compared to minimum mortality temperature) decreased across generations for elder adults (65-89 years of age), from $\text{RR}=1.18$ [95% confidence interval (CI): 1.13, 1.23] for those born in 1901-1905 to $\text{RR}=1.04$ (95% CI: 1.01, 1.07) for those born in 1926-1930 ($p\text{-trend}=0.004$). Similar to heat-related risk, the cold-related mortality risk (at first percentile) also decreased across the same cohorts ($p\text{-trend}<0.001$). The predicted continuous trends in heat- and cold-related mortality risks exhibited a nonlinear decline across generations. An inconsistent pattern was observed for middle-aged people (40-64 years of age) born between 1930 and 1960, with a slight increase in risks for cold and heat over generations. For cohort-specific risk trajectories, heat- and cold-related mortality risks generally increased with age, after 60 years old.

Discussion: This nationwide, individual-level study adopted a novel cohort perspective to investigate how population susceptibility to short-term non-optimal temperature exposure varies across generations. Our findings revealed disparities in susceptibility between generations, highlighting the importance for researchers and policymakers to consider cohort differences in efforts to promote future health advancements and reduce inequalities. https://doi.org/10.1289/EHP15226.