Background/Aims:To examine the short- and long-term effect of particulate matter 2.5 (PM2.5) air pollution on risks of premature death and mortality from cardiovascular disease and cancers across the counties of the United States (U.S.)

Methods: Data from 103 counties across the U.S. from 50 states on PM2.5 concentrations measured in 2013 to 2010, and premature death (assessed by person-years of life lost, PYLL for those who died before age 75) in 2010, and age-adjusted mortality from coronary heart disease (CHD), stroke, heart failure (HF), lung cancer and all-cancersfrom 2007 to 2015 in residents aged 35 and older were analyzed using an ecological study design. Annual average county-level PM2.5 concentrations were adjusted for individual counties’ total population. Univariate and multilevel spatial regression analysis technologies were applied to estimate the short- (a 5-yrs lag time) and long-term (10-yrs) effects of exposure to PM2.5 on a subsequence of the risk for the study outcomes of the interest. Confounders (% of African Americans, poverty rate (%) and # of primary care physicians per 100,000 population) were adjusted and potential mediators (obesity and diabetes) were tested in multivariate analysis models.

Results:Of 206 counties in the study, the annual mean (interquartile range, IQR) PM2.5 concentrations were 13.8 (13.2-15.4), 13.7 (12.8-15.1), 14.7 (13.6-16.5), 14.2 (13.0-15.9), 13.8 (12.4-15.8), 12.4 (11.9-13.1), 11.3 (11.0-11.6), and 12.4 (12.2-12.8) ug/m3, from 2003 to 2010.County-level mean PYLL (IQR)were 7824 (6297-9155) per 100,000 person-yearsin 2010. The average life expectancy was 74.6 (72.9-76.5) in men and 79.6 (78.4-81.0) in women. Multilevel spatial regression analysis indicates that every 10-unit increase in the ratio of PM2.5 to total population was associated with a decrease of 2.3 years in life expectancy in men (p<.0001), and 0.87 years in women (p=0.02). The same, increased PM2.5 concentrations were significantly associated with an increased risk of premature death and mortality from HF, stroke, CHD, lung cancer and all-cancers in men and women (p<0.001). However, these associations were greatly attenuated after adjustment for race/ethnicity (% of African Americans). Significant mediation effects of obesity and diabetes on a possible pathway of PM2.5 →mediators→the study outcomes of interest were observed.   

In conclusion, the study, using population health surveillance data from 2003-2015indicates that counties with elevated PM2.5 concentrations are associated with an decreased life expectancy and increased risks of premature death and mortality from CVD and cancers.The effect of PM2.5 on the outcomes of interest may go through a critical pathway of increasing the risk of obesity and diabetes.