Background: Although active maintenance treatments appear superior to placebo in preventing depression recurrence in older adults, few data are available to guide maintenance modality selection to maximize the probability of continued wellness for a given patient. Patients' temporal patterns of acute treatment response may predict who requires which maintenance therapy to remain well.
Methods: Depression levels were observed over 16 weeks of combined nortriptyline (NT) and interpersonal psychotherapy (IPT) in 140 persons aged >or=60 years with recurrent major depression. Subjects were empirically classified into four groups: rapid, sustained responders; delayed, sustained responders; mixed responders without sustained improvement; prolonged nonresponders. Groups were compared on subsequent recovery rates and on time to depression recurrence after randomization to 3 years of combined maintenance therapy (monthly IPT with NT), monotherapy (either IPT or NT alone), or medication clinic with placebo. Pretreatment psychosocial and clinical variables were controlled.
Results: Initial response profile predicted ultimate recovery rates, as well as who remained well, given the maintenance treatment received. Rapid initial responders showed lower recurrence risk with either combined or monotherapy, relative to placebo. Specific types of monotherapy appeared equally effective in rapid responders. In initially mixed responders, only combined therapy was superior to placebo. It was marginally superior to monotherapy. For delayed responders, combined therapy was superior to placebo; monotherapy did not differ from the other maintenance conditions. Prolonged nonresponders did not benefit from maintenance treatment.
Limitations: Subjects had only recurrent, unipolar depression. Initial response profile groups were established empirically and require replication. Sample sizes in initial response profile by maintenance treatment cells were small.
Conclusion: The ability to match patients to maintenance treatments more likely to prevent recurrence may be enhanced by considering the temporal profile of initial response to acute treatment.