Dissecting Antidepressant Placebo Expectancy-Mood Dynamics

About This Grant

Abstract Despite the high prevalence of major depressive disorder (MDD) and its projected rise as the leading cause of global disease burden by 2030, treatment efficacy remains suboptimal. First-line antidepressants have modest efficacy (~50%), and high placebo response rates (~40%) contribute to the failure of antidepressant trials and hinder new drug development. While research underscores the role of antidepressant expectancies in modulating mood across various brain regions, there is a critical need to elucidate how expectancy-driven neural dynamics interact with downstream mood regulation processes to induce sustained mood improvement. Our recent work provides the first computational account of antidepressant placebo effects, where reinforcement learning (RL) model-predicted expectancies—encoded in the salience network (SN)—trigger mood changes perceived as reward signals, which reinforce antidepressant expectancies through an expectancy-mood loop. Furthermore, we and others have demonstrated that enhanced functional connectivity (FC) between the SN and default mode network (DMN) during expectancy processing and at rest predicts long-term antidepressant placebo effects. This evidence suggests that antidepressant expectancies, originating from contextual treatment cues, are represented in the SN and influence mood regulation through top-down connections with the DMN. To test this hypothesis, this study will investigate the causal roles of the SN, DMN, and SN-DMN FC in antidepressant placebo effects using Theta Burst Stimulation (TBS). In a 2x3 factorial design, 200 patients with MDD will be randomized to three counter-balanced TBS conditions (intermittent, continuous, and sham, within-subject) targeting either the SN or DMN (between-subject). These acute experimental manipulations will modulate trial-by-trial expectancy and mood ratings and the neural encoding of model-based expectancies and mood reward signals during the “antidepressant placebo fMRI task”, which manipulates placebo-associated expectancies using visually cued fast-acting antidepressant infusions and sham visual neurofeedback. Led by experts in placebo effects, reinforcement learning, depression, and neuromodulation, this study combines a robust theoretical framework, state-of-the-art neuroimaging, precision functional mapping for personalized TBS targeting, and accelerated TBS, ensuring scientific rigor. The insights gained from this study will deepen our understanding of the neural mechanisms behind placebo effects, enhancing clinical trial design, advancing neuroimaging predictors of treatment response, and accelerating the development of expectancy-based interventions for MDD.