Within-Year PM2.5 Exposure Structure Provides Mortality-Predictive Information Beyond the Annual Mean

The study examined whether the distribution of PM2.5 concentrations throughout the year contains mortality-related information not captured by the annual mean. The research analyzed data from 27,289 county-years across 3,064 U.S. counties during 2003-2011. Researchers used the PMRI (PM2.5 magnitude-rank index), which measures the structure of daily concentrations during the year. Results showed that adding PMRI to the annual mean improved mortality prediction accuracy - the R-squared coefficient increased from 0.0802 to 0.1391. The improvement was consistent across different geographic regions, time periods, and all six studied causes of death. The study demonstrates that the way PM2.5 is distributed throughout the year contains important mortality risk information that the annual mean does not capture.