To the best of our knowledge, this is the first study to provide quantitative assessments and important insights regarding the interactions between hot and cold extremes and air pollution on public health. We focused on the effects of interactions at the regional level because of the similarities in climate, demographical, and socioeconomic status among cities in the specific region, so that the results can be applied to a larger spatial. We detected that exposure to PM2.5 pollution and extreme temperatures were associated with excess risk of total and cause-specific mortalities due to interaction. The interactions between PM2.5 pollution and cold extremes were mainly positive but less significant. However, significant synergistic effects were detected between PM2.5 pollution and hot extremes. Interactions were more robust in respiratory than cardiovascular mortalities across Jiangsu province.
We revealed similar adverse health effects of extreme temperatures on mortality as existing studies 29–32. We focused on the acute effects in this study, where the detected independent effect was stronger and more robust in hot seasons. Fast urbanization, large populations, and the dramatically changing environment have made China one of the most vulnerable regions experiencing frequent hot extremes 33–35. It is also plausible that high temperatures tend to trigger more acute health effects while cold events are associated with more chronic influence 13, 36. In contrast with previous studies 37, 38, at the regional level, PM2.5 pollution might not always increase the respiratory mortality risk in Jiangsu. However, PM2.5 pollution was associated with higher city-level respiratory mortality risks in Jiangsu. This difference may be due to local characteristics at a smaller scale spatial scale, which needs further exploration.
Our results go beyond previous studies by disclosing the notable difference in interactions between hot and cold extremes with PM2.5 pollution 18, 25, 39. The synergistic effect was stronger and more robust in hot seasons. Similarly, the excess risk of preterm birth due to interactions between PM2.5 and extreme temperatures ranged from 0.10 to 2.45 at high temperatures, higher than that detected at cold temperatures 40, 41. Even though the PM2.5 level tends to be more hazardous during cold seasons than hot seasons 42, people are also less likely to take outdoor physical activity and more inclined to take actions to reduce indoor PM2.5 levels when the PM2.5 concentration becomes more hazardous 43–45. Thus, personal PM2.5 exposures in hot seasons might be much lower than ambient levels in cold seasons. The uptake of PM2.5 may further increase through raised skin blood flow, minute ventilation, and sweating in hot seasons 42. The toxicity of PM2.5 also increases as the temperature elevates 42. Together, the relatively increased uptake and elevated toxicity of PM2.5 aggregate the interactions between PM2.5 and hot extremes. With global climate change, the co-occurrence of heatwaves and extreme PM2.5 pollution events are expected to increase by 175% in frequency 46. Causes and solutions to climate change and air pollution are closely linked 47. Internationally coordinated strategies regarding environmental challenges are warranted for global health benefits.
The synergistic effects on respiratory mortality in hot and cold seasons were consistent across regions; spatial heterogeneity exists in the interactive effects on cardiovascular mortality. The three regions in Jiangsu had higher interactions on respiratory mortality in hot seasons than in cold seasons. Similarly, a stronger synergistic effect of PM2.5 on respiratory disease was reported in hot seasons in Beijing and New England 48, 49. However, the studies in Chengdu and Hong Kong suggested an opposite trend 39, 50, which might be related to the special local geological and meteorological conditions. Regional differences exist in the interactions on respiratory mortality. For cardiovascular mortality, the disparities in interactions between regions may be explained by the environmental inequalities that could lead to disproportionate exposure to air pollutants and extreme temperatures for people in higher-middle income regions in China 51, which differs from the U.S. 52. The high-ranked average daily temperatures, PM2.5 concentrations, vehicle ownership, population density, and gross domestic product in middle Jiangsu may explain the highest synergistic effects on cardiovascular mortality in hot seasons. Regarding the spatial heterogeneity in the interactions, regional-level early warning systems for the occurrence of PM2.5 pollution and extreme temperatures are suggested to protect populations from the potential effects in identified areas.
Our estimates bear out the suggestions that the less educated and females may be more vulnerable to joint exposure to extreme temperatures and PM2.5 pollution. The association between cause-specific mortalities and high temperatures was more profound in the less educated 53, 54. The Air pollutants-related respiratory mortality risk was also higher in women than in men under both hot and cold extremes 55–57. Lower education levels are associated with lower socioeconomic levels 58, which may lead to poor housing environments, little preventive knowledge, and limited access to health care 59. Behavioral differences because of socioeconomic status may explain the differences between education levels. Compared with males, females have slightly greater airway reactivity, smaller airways than males, and different deposition of particles 60. Different physiological responses to ambient temperatures and air pollutants by sex might be one reason for disparities in females and males 56. The higher illiterate rate in females and the less unfavorable socioeconomic status of females also make them more vulnerable than males 55, 56.
Limitations also exist. First, the exposure was estimated at the city level, which may lead to an estimation bias due to the difference from the personal exposure. Further investigations with individual-level exposure information are suggested to explore the possible interactive effects of extreme temperatures and air pollution on cause-specific mortality. Second, this study focused on the acute effects and joint effects of PM2.5 and extreme temperatures. The lag time and lag effects of extreme temperatures and PM2.5 pollution were not considered. Last, joint and independent effects had spatial variations, so the results of this study cannot be generalized to other areas.