Heterogeneity and Variability of Placebo Effects within and between Randomized Controlled Trials across Major Psychiatric Conditions: A Systematic Review and Meta-analysis

Background: Research has increasingly highlighted the magnitude of placebo responses in psychiatric trials, but the variability of these responses has been largely. This systematic review and meta-analysis examines the heterogeneity and variability of placebo and medication effects across major psychiatric disorders .Methods: This PROSPERO registered systematic review and meta-analysis identified published randomized clinical trials (RCTs) examining treatment efficacy among people with one of four major psychiatric conditions (depressive, anxiety, obsessive- compulsive-OCD, and post-traumatic stress disorders-PTSD). We included clinical trials with at least one placebo and one medication arm (citalopram or escitalopram) reporting at least one quantitative clinical outcome. Natural logarithm of adjusted Coefficient of variation (ln CV), indicating changes in variance from baseline to final measurements within the placebo and medication groups and CV ratio between the placebo and medications groups (ln CVR) were calculated to assess variability of treatment response in each group and its correlates. Higgins’s I2 estimated the heterogeneity in the placebo and medication effect sizes (ES). Results: The review included 47 studies with 126 outcome measures, involving 6,887 participants. Studies participants showed a moderate to high effect (Placebo ES: 0.66, Medication ES: 1.01, P<0.01). There was a significant increased variance, compared to baseline, within both placebo and medication groups (ln CV: 0.52 and 0.65, respectively, P<0.001), with a relatively more pronounced increase in the medication group (ln CVR: -0.13, P<0.001). Variance changes were most prominent in PTSD (ln CV: 1.48 for placebo and 0.97 for medication, P<0.01). The relative variance changes in the placebo and treatment arms were significantly different between the conditions, with depressive disorders showing a greater relative increase in variance within the medication group (ln CVR: -0.21, P<0.01), whereas OCD and PTSD saw larger variance increases in the placebo group (ln CVR: 0.08 and 0.45, respectively, P<0.01). Anxiety disorders did not exhibit significant variance differences between groups (ln CVR: -0.08, P>0.05). Clinician-rated (ln CVR: -0.15) and secondary outcomes (ln CVR: -0.24) showed higher relative variability in medication groups (P<0.01).Conclusion: Both placebo and medication effects showed a substantial effect with significant variability across psychiatric conditions within and between RCTs, potentially influenced by psychopathological features, participants characteristics, and study methodologies. These findings highlight the nuanced complexity of placebo responses, emphasizing the need for careful consideration of these variances in research and clinical settings.