Socio-geographic disparity in cardiorespiratory mortality burden attributable to ambient temperature in the United States

• Published in: Environmental science and pollution research international 2019-01, Vol.26 (1), p.694-705
• Main Authors: Zhang, Yunquan, Xiang, Qianqian, Yu, Yong, Zhan, Zhiying, Hu, Kejia, Ding, Zan
• Format: Article
• Language: English
• Subjects: Aging; Air pollution; Air Pollution - analysis; Ambient temperature; Aquatic Pollution; Atmospheric Protection/Air Quality Control/Air Pollution; Cardiovascular Diseases - mortality; Climatology; Cold Temperature; Communities; Confidence intervals; Data analysis; Data processing; Demographics; Earth and Environmental Science; Ecotoxicology; Empirical analysis; Environment; Environmental Chemistry; Environmental Exposure - analysis; Environmental Exposure - statistics & numerical data; Environmental Health; Environmental science; Fatalities; Heterogeneity; High temperature; Hot Temperature; Humans; Humidity; Low temperature; Meteorological data; Meteorology; Morbidity; Mortality; Multivariate Analysis; Nonlinear Dynamics; Poisson density functions; Pollution studies; Regression analysis; Research Article; Respiratory Tract Diseases - mortality; Social Class; Socioeconomic factors; Socioeconomic status; Socioeconomics; Spatial heterogeneity; Statistical analysis; Subgroups; Temperature; Temperature effects; United States - epidemiology; Urban areas; Urbanization; Urbanization - trends; Waste Water Technology; Water Management; Water Pollution Control; Weather; Weathering
• ISSN: 0944-1344; 1614-7499
• DOI: 10.1007/s11356-018-3653-z

Compared with relative risk, attributable fraction (AF) is more informative when assessing the mortality burden due to some environmental exposures (e.g., ambient temperature). Up to date, however, available AF-based evidence linking temperature with mortality has been very sparse regionally and nationally, even for the leading mortality types such as cardiorespiratory deaths. This study aimed to quantify national and regional burden of cardiorespiratory mortality (CRM) attributable to ambient temperature in the USA, and to explore potential socioeconomic and demographic sources of spatial heterogeneity between communities. Daily CRM and weather data during 1987–2000 for 106 urban communities across the mainland of USA were acquired from the publicly available National Morbidity, Mortality and Air Pollution Study (NMMAPS). We did the data analysis using a three-stage analytic approach. We first applied quasi-Poisson regression incorporated with distributed lag nonlinear model to estimate community-specific temperature-CRM associations, then pooled these associations at the regional and national level through a multivariate meta-analysis, and finally estimated the temperature-AF of CRM and performed subgroup analyses stratified by community-level characteristics. Both low and high temperatures increased short-term CRM risk, while temperature-CRM associations varied by regions. Nationally, the fraction of cardiorespiratory deaths caused by the total non-optimum, low, and high temperatures was 7.58% (95% empirical confidence interval, 6.68–8.31%), 7.15% (6.31–7.85%), and 0.43% (0.37–0.46%), respectively. Greater temperature-AF was identified in two northern regions (i.e., Industrial Midwest and North East) and communities with lower temperature and longitude, higher latitude, and moderate humidity. Additionally, higher vulnerability appeared in locations with higher urbanization level, more aging population, less White race, and lower socioeconomic status. Ambient temperature may be responsible for a large fraction of cardiorespiratory deaths. Also, temperature-AF of CRM varied considerably by geographical and climatological factors, as well as community-level disparity in socioeconomic status.