Our study does have limitations. The findings are based on retrospective, cross-sectional analysis of childhood trauma exposure, which introduces the potential for recall bias and possible differences in recall bias between alcoholics and controls, limiting inferences regarding direct causality. Furthermore, our model is not to be taken as an exhaustive analysis of the link between childhood trauma and alcohol dependence severity. The “direct” effect of emotional abuse on alcohol dependence severity could be accounted for by other mediators not included in our analysis, such as variation in brain structure and function, and in the neuroendocrine stress response. In addition, we did not have data on recent or current life stressors or trauma that may have potentiated any relationships with early life experiences. Future studies will assess the potential moderating effects of these factors, as well as gender and genetic variation.
