Quitkin F M, McGrath P J, Stewart J W, Taylor B P, Klein D F
Department of Therapeutics, Columbia University College of Physicians and Surgeons, New York State Psychiatric Institution, New York 10032, USA.
Neuropsychopharmacology. 1996 Oct;15(4):390-4. doi: 10.1016/0893-133X(95)00272-F.
Recently the claim that there is a lag in the time of onset of antidepressant effects has been questioned. This issue rests on contrasting the time course of ultimate responders versus nonresponders on imipramine and amitriptyline. It is concluded that in 1 week on antidepressants "clinicians were capable of detecting changes in general states between the groups and the specific effects of depressed mood and anxiety and the physical expression of distress" (Katz et al. 1987). To examine this issue, we first used the design in which ultimate responders and ultimate nonresponders to antidepressants were compared at 1 and 2 weeks. Clearly there were statistically significant differences between ultimate responders and nonresponders on drug. However, the same was true on placebo. When the ultimate responders on placebo were contrasted to the ultimate responders on drug at 1 and 2 weeks using the Clinical global Impression (CGI) scale and the Hamilton Depression Rating Scale (HDRS), there was no difference between drug and placebo. This was also true for a subgroup of patients who met the criteria for melancholia. We conclude that, if the effects of nonspecific improvement are partialed out, there is no evidence of a medication effect at 1 and 2 weeks.
