Background: CIA research on nonchemical stressors and health effects is in early stages. Quantitative methods for combining risks from multiple stressors are not established.

Objective: A study was conducted to test a modeling strategy for assessing community health burden associated with air toxics from two coal-fired power plants (referred as electrical generating units, EGUs), accounting for background sources of these air toxics, criteria air pollutants, and nonchemical stressors.

Methods: From ordinary least squares linear regression models, nonchemical stressors were evaluated for associations with all cancer and heart disease mortality. Generalized additive models (GAMs) were used to examine associations between covariate-adjusted mortality rates, air toxics, PM2.5 and ozone related to coal-fired EGUs and / or from background sources.

Results: Four nonchemical indicators (African American, female, age, poverty) were significantly associated with cancer and heart disease mortality (70.6% and 79.1% of the variance, respectively), and adjusted in GAMs. EGU-related air toxics were not significantly associated with either health outcome. Arsenic from non-EGU sources was significantly associated with cancer mortality (88.8% of residual variance). Lead from non-EGU sources, PM2.5, and ozone, combined, were significantly associated with heart disease mortality (88.1% of residual variance).

Discussion: Epidemiologic evidence and methods are necessary to understand the impacts and interactions of nonchemical and chemical stressors on community health. This study demonstrates the utility of GAM to quantitatively assess community health burden from both chemical and nonchemical stressors. As model and parameter choices can impact findings, explorations of quantitative methods and uncertainty considerations are critical for CIA research.