Despite the fact that the use of antidepressants and other psychologic strategies may seem intuitive in the setting of moderate to severe functional bowel disease, conclusive data from well-designed trials have been lacking.

In this setting, Drossman and colleagues have conducted the largest randomized trial to evaluate the effectiveness of desipramine, a tricyclic antidepressant, as well as cognitive-behavioral therapy vs their control conditions (placebo and education, respectively) in women with moderate to severe functional bowel disorders. Additionally, these investigators assessed the clinical benefits of the active treatment regimens (desipramine or cognitive-behavioral therapy) in clinically meaningful patient subgroups (ie, with/without history of depression; with/without history of sexual or physical abuse; predominant diarrhea; predominant constipation; moderate vs severe disease).

This randomized, comparator-controlled, multicenter trial involved 431 adult women from the University of North Carolina and the University of Toronto who had moderate to severe symptoms of functional bowel disorders. Subjects received either psychologic (cognitive-behavioral therapy vs education) or antidepressant (desipramine vs placebo) therapy for a period of 12 weeks. Assessment of physiologic, clinical, and psychosocial parameters were conducted before and after end of therapy.

Overall, results suggest that the active psychologic treatment (ie, cognitive-behavioral therapy) was effective for women with functional bowel disorders, including irritable bowel syndrome, whereas the active antidepressant treatment (desipramine) appeared to only be effective in the management of those patients who were adherent (ie, able to stay on their medication). Specifically, by intention-to-treat analysis, cognitive-behavioral therapy was found to be significantly more effective than its placebo condition (education; P = .0001; responder rate, 70% vs 37%, respectively; number needed to treat [NNT], 3.1). By contrast, desipramine did not demonstrate significant efficacy vs placebo in the intention-to-treat analysis (P = .016; responder rate, 60% vs 47%, respectively; NNT, 8.1) but did show a statistically significant benefit in the per-protocol analysis (P = .01; responder rate, 73% vs 49%, respectively; NNT, 5.2). The latter was especially significant when subjects with nondetectable blood levels of desipramine were excluded.

Results of subgroup analyses demonstrated that cognitive-behavioral therapy had benefit over its control condition for all subgroups except those patients with depression. By contrast, subgroup analyses indicated that active antidepressant therapy (desipramine) was more beneficial than placebo for women with moderate vs severe symptoms, a history of abuse, without comorbid depression, and with predominant diarrhea.

The investigators highlighted several limitations associated with this study that should be considered. First, at least moderate side effects were observed in up to 25% of subjects taking desipramine, and thus some participants may have become unblinded. However, it should be noted that similar side effects occurred in up to 15% of participants receiving placebo as well. Additionally, dropouts occurred in 23% of patients allocated to treatment; therefore, data that anticipated treatment responses were "imputed" for several of these individuals. Finally, the fact that all patients fulfilling Rome I or II criteria for functional bowel disorders were included also warrants some consideration. Overall, however, as the study authors emphasize, these findings may be generalized to any patient who fulfills the selection criteria used in this trial, who is treated by their standardized psychologic protocol, or who is compliant in taking desipramine as prescribed.

Thus for women with moderate to severe functional bowel disorders, cognitive-behavioral therapy is effective and desipramine may be effective when taken adequately. It is important to note, however, that certain clinical subgroups are more or less amenable to these relative treatments.