New Study Shows the Impact of the Placebo Effect Varies Significantly Between Psychiatric Disorders

Gary Lamphier - 10 November 2021

A new study led by Dr. Bo (Cloud) Cao, Assistant Professor in the Department of Psychiatry and Canada Research Chair in Computational Psychiatry at the University of Alberta, found that the placebo effect varies significantly across three major psychiatric disorders.

The meta-analysis and machine learning study – titled Differential power of placebo across major psychiatric disorders: a preliminary meta-analysis and machine learning study – was published Oct. 29 in Scientific Reports. It is freely available online at

Dr. Cao and more than a dozen other researchers – most of them from the U of A – participated in the study, which examined data from 108 clinical trials involving more than 32,000 participants. All had received pharmacological interventions for Major Depressive Disorder (MDD), Bipolar Disorder (BD) and Schizophrenia (SCZ).

The research team developed and employed measures based on clinical rating scales and Clinical Global Impression (CGI-S) scores to compare placebo effects across the three major disorders.

The team then performed meta-analysis including meta-regression using sample-size weighted bootstrapping techniques, and Machine Learning analysis to identify the type of disorder treated in each trial, based solely on the placebo response.

“Consistently through multiple measures and analyses, we found differential placebo effects across the three disorders, and found lower placebo effect in SCZ compared to Mood Disorders. The differential placebo effects could also distinguish the condition involved in each trial between SCZ and Mood Disorders with Machine Learning.”

For Mood Disorders (including MDD and BD), researchers found that the placebo effect was sometimes as high as 80 per cent – high enough to compare to the pharmaceutical effect of medications.

“To our best knowledge, this is the first study to show converging evidence of differential placebo effects across major psychiatric disorders from different measures and different analytical approaches,” the co-authors state.

The study’s co-authors noted that the differential placebo effect can also be used to distinguish Schizophrenia from Mood Disorders trials at the individual trial level using Machine Learning, thus providing compelling evidence that the placebo effect between conditions is consistently different.

“In our study, we used two ratios of clinical assessments, both relative to the observed treatment effect, instead of just using the changes of these scales in the placebo group alone. CGI-S changes relative to baseline were not scaled to treatment measurements but directly comparable across disorders, and thus were complementary to those two ratios relative to the treatment,” the study states.

“All three ratios consistently showed differential placebo effect across the major psychiatric disorders, especially between SCZ and Mood Disorders.”

The findings suggest potentially distinct mechanisms of placebo underlying MDD, BD and SCZ, which may help guide how placebo can be used as a control condition in related clinical trials, and offer insights on placebo use as a cost-effective, active component in mental health treatment, the study’s co-authors note.

“Our results call for future studies on common and distinct neurological mechanisms of placebo effect across psychiatric disorders, and translational applications of placebo in the frame of personalized medicine to improve mental health care.”

In an interview, Dr. Cao said the placebo effect is particularly strong for psychiatric disorders, when compared to other illnesses.

“A lot of people still believe mental disorders are purely psychological and that people can get better by themselves if they just focus on positive things and engage in positive activities. Although this is not true and most people with mental disorders need interventions from professionals to get better, the psychological, cognitive and social components of a treatment could also play a significant role in symptom improvement,” he says.

“That may explain why the placebo effect in psychiatric disorders is so strong. This paper shows that for Major Depressive Disorder, Bipolar Disorder-Depression and Bipolar Disorder-Mania, the placebo actually has a very strong effect. For Schizophrenia the effect is roughly half that for Mood Disorders, so it’s much lower but still quite strong. The results were very consistent across all three measurements we used in this study.”

Dr. Cao noted that most previous studies evaluated the placebo effect and its predictors for a specific disorder, and then compared the post hoc effect of predictors independent of disorder types. This makes it difficult to assess the contribution of these predictors across different disorders.

“Our meta-regression models included most of these common predictors of placebo effect found in previous studies, along with disorder types, but we did not find very consistent predictors across placebo effect measures other than disorder types,” he says.

In addition to Dr. Cao, the study’s co-authors include Dr. Russell Greiner, a Professor in the Faculty of Science, Department of Computing Science and Principal Investigator with the Alberta Machine Intelligence Unit (Amii); Dr. Andrew Greenshaw, a Professor and Associate Chair in the Department of Psychiatry; Dr. Serdar Dursun, a Professor in the Department of Psychiatry; and Dr. Xin-Min Li, Associate Dean, International Relations in the Faculty of Medicine and Dentistry, and former Chair of the Department of Psychiatry.

The study was also completed with significant support from trainees at the U of A, including Yang S. Liu, a Postdoctoral Research Fellow; Alessandro Selvitella, a previous Postdoctoral Research Fellow; and PhD student Jeffrey Sawalha, as well as international collaborators Diego Librenza-Garcia and Pedro Ballester, of McMaster University; and Ives Cavalcante Passos, of the Federal University of Rio Grande do Sul, Brazil.