Placebo response in pharmacological and dietary supplement trials of autism spectrum disorder (ASD): systematic review and meta-regression analysis

86 randomized, placebo-controlled trials (pharmacological and dietary supplements) comparing with placebo; most were double-blind and parallel design; cross-over data used only from the first phase to avoid carryover effects. Participants: individuals with autism spectrum disorder diagnosed by standardized criteria; total N = 5,365; 2,360 on placebo; 75 studies in children/adolescents; eight in adults; three mixed-age. Median placebo participant age was 8 years; treatment durations typically 8–12 weeks (median ~10 weeks); minimum duration 7 days; no dosing information provided.

Placebo produced substantial improvements across core ASD symptoms: social-communication difficulties (SMC −0.32; 95% CI −0.39 to −0.25), repetitive behaviors (SMC −0.23; 95% CI −0.32 to −0.15), and overall core symptoms (SMC −0.36; 95% CI −0.46 to −0.26). About 19% of participants were at least much improved with placebo (CGI-I). Larger placebo responses were linked to caregiver ratings, lower risk of bias, flexible dosing, larger sample sizes and more sites, earlier publication year, and higher baseline irritability, with some effects varying by domain. No clear difference by age group. Conclusion: placebo responses are substantial in ASD trials and may obscure true drug effects; to improve detection of efficacy, future trials should be well-powered, minimize site variability, train raters to reduce placebo-by-proxy effects, and use developmentally appropriate, less caregiver-dependent outcome measures; standardizing reporting (including CGI anchoring) and developing robust core-symptom measures are needed.

Focus on placebo effects in core symptoms only; wide heterogeneity of scales and CGI anchoring; moderators of drug–placebo differences not analyzed; reliance on aggregate data with potential ecological bias; many eligible studies lacked data; cross-over data limited to first phase; majority of trials pediatric; potential sponsorship/reporting biases; results should be interpreted as exploratory due to data limitations.