Although the importance of evaluating the health effects of multi-pollutant exposures has recently been recognized [18], [42], only a few studies have been conducted, mostly focused on multiple air pollutants [10], [21], [43]–[46], probably due to methodological challenges, such as collinearity, measurement errors, potential interaction between pollutants and potential non-linear exposure-health relationships [16]. Patel et al. adopted newer techniques used in genomics and proposed an Environment-Wide Association Study (EWAS) [5], [6]. This approach provided excellent insight to identify ‘top hit’ pollutants. However, few epidemiologic studies have provided methods to estimate combined effects or to predict risks from multi-pollutant exposure [43], [47].
