Psychopharmacology Bulletin, 28:309314. Methodo logical issues, such as a lack of physician verification of selfreported diabetes, variability in the time frame of depression being assessed, small sample sizes, and low participation rates could have also hindered the validity of findings 1. The design of all studies should be publicly available. On the other hand, something has to be done quickly in HIV patients with MCD: the course of disease can be extremely fulminant.