Estimating risk using the novel Particulate Matters years index (ePM-years index). A report from the nationwide EP-PARTICLES study

Abstract Background Air pollution is the largest environmental risk factor, especially for cardiovascular disease (CVD). In this study, we present an original concept of estimating the risk associated with exposure to air pollution: the Particulate Matters years index (ePM-years index). Objective To assess the predictive performance of the ePM-years index. Methods The ePM-years index is an indicator of hypothetical exposure to air pollution. It is estimated by multiplying the number of years spent in a given environment by the difference between the average annual PM2.5 concentration and the allowed level of pollution given by World Health Organization (WHO) guidelines (Figure 1). We collected data from 2011 to 2020 on mortality and hospitalizations from National Health Fund and the Central Statistical Office in Poland, respectively. For further analysis, we used an age-standardized rate (SR), established on the European Standard Population structure. The source of air pollution data was GEM-AQ air quality modeling system. All data were connected with TERYT codes, all analyses were conducted at the county level (LAU-1) and then aggregated to the voivodeship level (NUTS-2). Poisson regression models were used to estimate associations between CVD mortality, ePM-years, and population level social and health status. We also estimated the attributable fraction of CVD mortality due to air pollution using AirQ+ methodology from the WHO. Results We recorded 4,010,531 deaths including 1,669,464 (41.6%) deaths due to CVD with a median age of 78.16 (SD=12.54), and 53.4% of females (N=914,932). Current smokers amounted to 29.9% (23.4%-33.2%). The main causes of death were coronary artery disease [405,268 (24.3%)] and heart failure [376,592 (22.6%)]. The mean Gross Domestic Product (GDP) per capita was 9644 EUR (7517-12,048), and the mean yearly PM2.5 concentration was 23.3 (SD=6.3) μg/m3, corresponding with the mean ePM-index 35.8 (21.5-66.6) in analysed time. The SR for CVD mortality was 500.3 (min. 461.9 – max. 604.5), the SR hospitalization rates were as follows: hypertension 312 (299-342), diabetes 90 (78.9-95.9), and HF 557.9 (310-728.4). There was an association of the ePM-years index with CVD mortality (shown as increase per 10 ePM-years): In Model 1 adjusted for GDP, risk ratio (RR) 1.023 (95% confidence interval (CI): 1.002-1.044); in Model 2 adjusted for comorbidities, RR=1.025 (95%CI: 1.004-1.046); in Model 3 adjusted for GDP, comorbidities and weather conditions RR=1.027 (95%CI: 1.005-1.049). We estimated that 16.2% (N=324,460) [95%CI: 12.54% -17.9%] of CVD deaths were attributable to PM2.5 exposure, using the WHO's air quality guideline as the reference. Conclusions The ePM-years index may be used as a tool for risk stratification for patients and incorporated into CVD outcomes prediction models. The index can help with planning of appropriate primary and secondary prevention schemes related to long-term pollution exposure.