The results of this meta-analysis should not deter investigators from including environmental risk factor information in their studies, once robust marginal gene associations have been identified. Characterization of gene-environment interaction has been most successful for diseases or traits that allow the study of a single gene with a major effect in the context of a relevant environmental exposure of varying magnitude, and also when the environmental exposure has a strong effect.53 For example, the identification of the protective influence of the Δ32 mutation in the CCR5 chemokine receptor against human immunodeficiency virus infection was based on evaluation of unaffected controls with high exposure to the virus.54
