The results from a Bayesian reinforcement learning model showed the antidepressant placebo trial-wise expectancies were updated by composite learning signals multiplexing sensory evidence and trial-wise mood.
Investigators have begun to parse together why there is a pronounced placebo following antidepressant treatment in patients with major depressive disorder (MDD).1
A team, led by Marta Peciña, MD, PhD, Department of Psychiatry, University of Pittsburgh, identified the neurobehavioral mechanisms underlying the evolution of antidepressant placebo effects using a reinforcement learning framework.
The mechanism resulting in the persistence of the placebo effects patients treated with antidepressants is not well understood, despite high usage rates.
“The antidepressant placebo functional magnetic resonance imaging task manipulates placebo-associated expectancies using visually cued fast-acting antidepressant infusions and controls their reinforcement with sham visual neurofeedback while assessing expected and experienced mood improvement,” the authors wrote.
In the previous placebo analgesia studies, investigators have induced placebo effects by setting expectancies using verbal instructions and/or Pavlovian conditioning, through the pairing of an inactive placebo stimulus with an unconditioned stimulus.
Recently, investigators have found that reinforcement learning can provide a computational account of this phenomena where expectancy learning starts with an a prior and depends on prediction errors or the mismatch between what is expected and what it is experiences instead of the contiguity between the conditioned and the unconditioned stimuli.
In the acute within-patient cross-sectional study of antidepressant placebos, the investigators examined 60 patients aged 18-55 years not receiving medication for MDD at the University of Pittsburgh between February 21, 2017 and March 1, 2021. The mean age of the patient population was 24.5 years and 85% (n = 51) of the participants were female.
The team compared 5 alternative reinforcement learning models predicting the expectancy ratings to characterize the evolution of participants’ placebo expectancies or learned expectancies.
The investigators examined the trial-by-trial evolution of expectancies and mood using multilevel modeling and reinforcement learning, relating model-predicted signals to spatiotemporal dynamics of blood oxygenation level-dependent (BOLD) response.
The results from a Bayesian reinforcement learning model showed the antidepressant placebo trial-wise expectancies were updated by composite learning signals multiplexing sensory evidence and trial-wise mood (Bayesian omnibus risk <0.001; exceedance probability = 97%). The placebo expectancy, neurofeedback manipulations, and composite learning signals modulated the visual cortex and dorsal attention network (threshold-free cluster enhancement [TFCE] = 1 − P >.95).
The learned placebo expectancies modulated the salience network (SN, TFCE = 1 – P >.95), positively scaling with depression severity as patients anticipated antidepressant infusions.
“Results of this cross-sectional study suggest that on a timescale of minutes, antidepressant placebo effects were maintained by positive feedback loops between expectancies and mood improvement,” the authors wrote. “During learning, representations of placebos and their perceived effects were enhanced in primary and secondary sensory cortices. Latent learned placebo expectancies were encoded in the SN.”
Peciña M, Chen J, Karp JF, Dombrovski AY. Dynamic Feedback Between Antidepressant Placebo Expectancies and Mood. JAMA Psychiatry. Published online March 01, 2023. doi:10.1001/jamapsychiatry.2023.0010