Differential Placebo Responses for Pharmacotherapy and Neurostimulation in Late-Life Depression

Background: The magnitude of the placebo response depends on both the modality used as the “placebo” and the intervention with which it is compared, both of which can complicate the interpretation of randomized controlled trials (RCTs) for depression in late life. Given that neurostimulation and pharmacotherapy are among the most common interventions studied for late-life depression, comparing the relative placebo responses in studies of these interventions can aid interpretation of relative effect sizes. Materials and Methods: We analyzed data from two RCTs of adults aged ≥60 years in an episode of treatment-resistant major depression, one comparing aripiprazole and matching placebo pills and the other comparing deep repetitive transcranial magnetic stimulation (rTMS) and sham rTMS. In both RCTs, depression was assessed using the 17-item Hamilton Depression Rating Scale (HDRS-17). The primary comparison occurred after four weeks using analysis of covariance (ANCOVA) of HDRS-17 scores in participants who received placebo pills or sham rTMS. Relevant covariates included years of education, duration of depressive episode, and baseline HDRS-17 score. Results: Accounting for covariates, there was a larger reduction of HDRS-17 after four weeks in the sham rTMS group (estimated marginal mean ± SE: −5.90 ± 1.45; 95% CI: [−8.82, 2.98]) than in the placebo pills group (−1.07 ± 1.45; [−3.98, 1.85]). There were no significant differences between these groups in the binary outcome analysis of response and remission rates at four weeks or any outcome at trial end point comparison. Conclusions: Sham rTMS may have a larger placebo response than placebo pills early in the treatment of older adults with treatment-resistant depression. Differential placebo responses should be considered in both the interpretation and design of RCTs.