Although adverse effects of high temperature on mortality have been studied extensively in urban areas, little is known of the heat–mortality associations outside of cities.
We investigated whether heat–mortality associations differed between urban and nonurban areas and how urbanicity affected the vulnerability to heat-related mortality.
We first analyzed heat-related mortality risk in each of 102 counties in Jiangsu Province, China, during 2009–2013 using a distributed-lag nonlinear model. The county-specific estimates were then pooled for more urban (percentage of urban population ≥ 57.11%) and less urban (percentage of urban population < 57.11%) counties using a Bayesian hierarchical model. To explain the spatial variation in associations by county, county-level characteristics affecting heat vulnerability were also examined.
We found that the overall mortality risk comparing the 99th vs. 75th percentiles of temperature was 1.43 [95% posterior intervals (PI): 1.36, 1.50] in less urban counties and 1.26 (95% PI: 1.23, 1.30) in more urban counties. The heat effects on cardiorespiratory mortality followed a similar pattern. Higher education level and prevalence of air conditioning were significantly associated with counties having lower risks, whereas percentage of elderly people was significantly associated with increased risks.