The efficacy of interventions might be underestimated or even go undetected as a main effect when it is hidden in gene-by-environment (G×E) interactions. This review moves beyond the problems thwarting correlational G×E research to propose genetic differential susceptibility experiments. G×E experiments can test the bright side as well as the dark side of the moderating role of genotypes traditionally considered to represent vulnerability to negative conditions. The differential susceptibility model predicts that carriers of these risk genotypes profit most from interventions changing the environment for the better. The evolutionary background of G×E and differential susceptibility is discussed, and statistical methods for the analysis of differential susceptibility (versus diathesis stress) are reviewed. Then, based on results from 22 randomized G×E experiments, meta-analytic evidence for the differential susceptibility model is presented. Intervention effects are much stronger in the susceptible genotypes than in the nonsusceptible genotypes. The final sections suggest possibilities to broaden the G component in the G×E equation by including genetic pathways, and to broaden the E component by including methylation level and gene expression as promising ways to probe the concept of the environment more deeply and address the perennial issue of what works for whom.
Keywords: G×E interaction; diathesis stress; differential susceptibility; dopamine; microtrial; randomized controlled trial; serotonin.