Magnitude and Determinants of Placebo Response in Acute Migraine Trials: A Systematic Review and Meta-Analysis

Neurology. 2026 Aug 11;107(3):e218213. doi: 10.1212/WNL.0000000000218213. Epub 2026 Jul 10.

ABSTRACT

BACKGROUND AND OBJECTIVES: Migraine is a highly prevalent and disabling neurologic disorder. Beyond drug-specific mechanisms, the role of contextual, individual-related, and disease-related factors remains poorly characterized. We quantified placebo response in randomized controlled trials (RCTs) of acute migraine treatments and identified patient-level, drug-level, and study-level determinants.

METHODS: This meta-analytic review searched PubMed, Embase, Cochrane Central, ClinicalTrials.gov, and the European and Japanese Trial Registers from inception to November 2025 for published/unpublished, double-blind, parallel-group, phase 2 or 3 RCTs evaluating acute pharmacologic migraine treatments. Placebo rates with 95% CIs and prediction intervals (PIs) were pooled using random-effects models. Subgroup analyses and meta-regressions explored potential drivers. The primary outcome was 2-hour pain freedom after placebo administration. Secondary outcomes included 2-hour pain relief, sustained pain freedom, rescue medication use, global impression of improvement, and most bothersome symptom (MBS) freedom.

RESULTS: We analyzed 126 RCTs published between 1991 and 2025, comprising 24,614 placebo-treated participants (mean age range, 34.8-45.2 years; female, 65.2%-96.9%). The pooled 2-hour pain freedom rate was 11% (95% CI 10%-12%; PI: 4%-27%). The highest estimates were observed for intranasal (14%; 95% CI 12%-16%) and subcutaneous (13%; 95% CI 8%-20%) administration, compared with oral (10%; 95% CI 9%-12%) (p < 0.05). A lower proportion of patients with severe baseline headache and a lower probability of receiving an active treatment were associated with greater responsiveness (β -1.38; 95% CI -2.37 to -0.38; p < 0.01 and β -1.20; 95% CI -2.02 to -0.38; p < 0.005, respectively). Placebo response increased significantly over time (β 0.03; 95% CI 0.01-0.04; p < 0.001), even after multivariable adjustment. Placebo rates were 34% for 2-hour pain relief, 9% for sustained pain freedom, 52% for rescue medication use, 24% for positive global impression, and 34% for MBS freedom. Exploratory analyses of pain relief showed high placebo responses among non-Caucasian individuals, in Asian countries, and in gepant RCTs.

DISCUSSION: Placebo response represents a substantial and increasing component of outcomes in acute migraine RCTs. Improved trial design, standardized communication, and transparent reporting may enhance effect-size interpretation and preserve assay sensitivity in future trials.

PMID:42430679 | DOI:10.1212/WNL.0000000000218213