Abstract
The purpose of this study was to identify variables useful for predicting a positive response to the tricyclic antidepressant desipramine, amongst bulimic subjects. Using a randomized double-blind cross-over design, 24 normal weight bulimics completed a 15-week protocol in which they received either desipramine (150 mg/day)for six weeks, no drug for three weeks, followed by placebo for six weeks, or the reverse sequence. At weeks 0, 2, 4, 6, 9, 11, 13, and 15, each subject was assessed using the EDI, SCL-90, POMS and binge records. The DST, Diagnostic Interview Schedule (DIS), and a personal and family medical-psychiatric history questionnaire were administered at initial assessment, while plasma desipramine levels were obtained at weeks 4 and 13. Responders were defined in terms of both binge frequency reduction, and decrease in depressive symptoms.
In the sample of 24 subjects, desipramine was significantly more effective than placebo in reducing the frequency of weekly hinging and vomiting, as well as causing a reduction in the fatigue scale of the POMS. No significant effect of the drug was obtained on the EDI or the SCL-90. In terms of reduction in binge frequency, seven responders were identified; another seven were found to be borderline responders, while 10 were labeled as non-responders. The three groups did not differ in terms of their initial scores on the SCL-90, POMS, DST, DIS results, or psychological subscales of the EDI. However, responders were found to have lower EDI bulimia subscale scores, but a higher frequency of purging episodes than were non-responders. Eight patients were identified as borderline responders with respect to depressive symptoms. They differed from depression non-responders in that they had abnormal DST results, and were more likely to report a family history of depression. These results indicate that the predictors of the antibulimic effect of desipramine in bulimia are independent of its antidepressant properties.