Nutritional supplements for migraine prevention: evaluating evidence sufficiency and effect modifiers through network meta-analysis, trial sequential analysis, and meta-regression

Abstract

Background This network meta-analysis (NMA) integrated trial sequential analysis (TSA) limited to eligible direct pairwise comparisons and network meta-regression to compare efficacy and safety of nutritional supplements for migraine prevention, assess evidence maturity and identify potential effect modifiers. Methods We retrieved randomized controlled trials (RCTs) of nutritional supplements for adult migraineurs from four databases up to January 2025. Primary outcomes were migraine frequency, duration and severity; adverse events were the secondary outcome. Frequentist random-effects NMA calculated mean differences (MD) for continuous indicators and odds ratios (OR) for adverse events with 95% confidence intervals (CIs). Treatment rankings were assessed using rank probabilities and the surface under the cumulative ranking curve but were interpreted as exploratory. Inconsistency was tested where network structure allowed valid evaluation. Network meta-regression detected between-study heterogeneity, TSA was restricted to eligible direct pairwise comparisons and the CINeMA framework evaluated overall evidence certainty. Results Fourteen RCTs (791 participants) formed a six-node network: vitamins, magnesium, coenzyme Q10, probiotics, vitamin-probiotic combinations and placebo. For migraine frequency, probiotics produced the largest frequency reduction vs. placebo (MD = −2.60 attacks/month, 95% CI − 3.96 to −1.20; 72% top-rank probability), followed by vitamin-probiotics (MD = −1.92, 95% CI: −3.12 to −0.73), coenzyme Q10 (MD = −1.91, 95% CI: −2.51 to −1.00) and vitamins (MD = −1.85, 95% CI: −2.65 to −1.05). Magnesium showed non-significant benefits (MD = −0.82, 95% CI − 2.03 to 0.13), and no active supplements differed statistically from each other. No intervention significantly shortened migraine duration; vitamin-probiotics ranked first (MD = −4.68 h/attack, 95% CI − 9.76 to 0.15). Only magnesium significantly lowered severity (MD = −3.56 points on the 0–10 VAS, 95% CI − 6.50 to −0.52; 85% top-rank probability). Treatment duration correlated with frequency improvements (coefficient = −0.09, 95% CI − 0.17 to −0.02, p = 0.01). Only one probiotic-placebo trial precluded definitive TSA evidence conclusions for probiotics, and CINeMA graded most comparisons as moderate-certainty evidence. Conclusion Probiotics best reduced migraine frequency yet lacked statistical superiority over other supplements via indirect placebo comparisons. Only magnesium alleviated migraine severity, while no supplement significantly shortened attack duration. Treatment rankings remain exploratory. Large standardized head-to-head RCTs are required for validation. Systematic review registration https://www.crd.york.ac.uk/PROSPERO/view/CRD420251182820, CRD420251182820.