Michael E. Thase, Edward S. Friedman, Amy L. Fasiczka, Susan R. Berman, Ellen Frank, Eric A. Nofzinger, and Charles F. Reynolds III
Objective: This report compares response to
cognitive-behavioral therapy (CBT) and pharmacotherapy in
sequential cohorts of men with DSM-III-R major depression.
Method: Patients were enrolled in consecutive
standardized 16-week treatment protocols conducted in the same
research clinic. The first group (N = 52) was treated with Beck's
model of CBT, whereas the second group (N = 23) received
randomized but open-label treatment with either fluoxetine (N =
10) or bupropion (N = 13). Crossover to the alternate medication
was permitted after 8 weeks of treatment for antidepressant
nonresponders. The patient groups were well matched prior to
treatment. Outcomes included remission and nonresponse rates, as
well as both independent clinical evaluations and self-reported
measures of depressive symptoms.
Results: Despite limited statistical power to
detect differences between treatments, depressed men treated with
pharmacotherapy had significantly greater improvements on 4 of 6
continuous dependent measures and a significantly lower rate of
nonresponse (i.e., 13% vs. 46%). The difference favoring
pharmacotherapy was late-emerging and partially explained by
crossing over nonresponders to the alternate medication. The
advantage of pharmacotherapy over CBT also tended to be larger
among the subgroup of patients with chronic depression.
Conclusion: Results of prior research comparing
pharmacotherapy and CBT may have been influenced by the
composition of study groups, particularly the gender composition,
the choice of antidepressant comparators, or an interaction of
these factors. Prospective studies utilizing flexible dosing of
modern antidepressants and, if necessary, sequential trials of
dissimilar medications are needed to confirm these findings.
J Clin Psychiatry 2000;61(7):466-472
© Copyright 2000 Physicians Postgraduate Press, Inc.