In criticizing the null finding from the earlier article from this group [Gillespie et al., 2005], Uher and McGuffin [2008] argue that our sample was recruited and selected for alcohol addiction: in fact, it is a volunteer population sample assessed on a range of indicators of physical and mental health that included addiction, depression, and SLE, and does not differ in these respects from the larger-base sample from which it was drawn [Heath et al., 1997]. However, Uher and McGuffin [2008] noted other valid limitations applicable to our study sample: that the interaction (i) may be age-dependent, occurring in younger adults, and (ii) may occur more in females. Further, the strongest effects of SLE on depression occur in the month immediately following the SLE, with the effect then decaying over several months [Kendler et al., 1998; Surtees and Wainwright, 1999]. We address these points by performing supplementary analyses that assess the specific hypotheses they raise.
