Despite the intrinsic appeal of an interaction between SLEs and 5HTTLPR, and despite support from animal models demonstrating the biological plausibility of such an interaction, results from human studies fail to replicate convincingly the original finding of Caspi et al. [2003]. The meta-analyses by Munafo et al. [2008] and Risch et al. [2009] combines results from 5 and 14 studies, respectively that rigorously adhered to the conditions of the initial study, and conclude there is no evidence for an interaction. Our results strengthen these conclusions. Uher and McGuffin [2008] reviewed evidence from a range of quantitative population genetic studies, neurophysiological investigations, and experimental studies in animals and concluded that there is evidence for an interaction. All three studies are critical of many of the human studies that set out to replicate the finding and they make recommendations about design and analysis of future studies. We note however, that all fail to acknowledge potential problems with the genotyping assay of 5HTTLPR.
