In the general context of risk factor epidemiology, risk prediction models, such as the Framingham risk score for coronary heart disease [23] and genetic risk scores (a.k.a Genetic Risk Prediction Studies (GRIPS)) [24]–[29], have been widely used. Following from these ideas, it would be interesting to assess the predictive ability of an “Environmental Risk Score” as a follow-up to an EWAS study after identifying environmental pollutants significantly associated with health outcomes. A risk score may also facilitate targeting of preventive interventions [27].
