Treatment Effectiveness and Harms

• Adults: Interventions including relaxation (alone or combined with other components) may slightly lower migraine/headache attack frequency, mindfulness-based cognitive therapy may improve migraine-related disability, and an education component may also improve migraine-related disability (strength of evidence [SOE]: All low).
• Evidence regarding adverse effects among individuals of any age, effectiveness for children and adolescents, and biopsychosocial factors is inconclusive.

Comparative Effectiveness and Harms

• Adults: mindfulness-based stress reduction (MBSR) may offer greater benefit for disability and quality of life (QOL) than education; relaxation + education might offer lower migraine attack frequency but higher QOL than propranolol; relaxation may outperform cognitive behavioral therapy (CBT) + relaxation + education in QOL; MBSR + education and biofeedback may reduce attack frequency compared to stress management training + education and CBT + relaxation, respectively (SOE: All low).
• Children/adolescents: a combination of CBT, biofeedback, and relaxation may lead to lower migraine attack frequency and disability compared to education alone (SOE: Low).

Effects of Individual Behavioral Components

• Evidence is insufficient to address this Key Question.

Effects of Non-headache Focused Interventions

• In adults with chronic migraine, a behavioral sleep modification may reduce headache frequency at 6 weeks (SOE: Low); no studies were included for children.

Objectives. Behavioral interventions for migraine prevention can offer an important alternative or complement to medications. We conducted this systematic review to evaluate the effectiveness and comparative effectiveness of behavioral interventions for migraine prevention.

Data Sources. MEDLINE, Embase, PsycINFO, PubMed, the Cochrane Database of Systematic Reviews, clinicaltrials.gov, and grey literature sources for randomized trials published from 1975 to January 11, 2023.

Review Methods. We asked five Key Questions of preventative behavioral migraine interventions, including efficacy, comparative efficacy (including comparisons to medications), contributions of specific behavioral components, telehealth implementation, non-migraine-focused treatments, and associations with biopsychosocial factors. For scoping and methodologic decisions, we consulted numerous experts, including primary care physicians, adult and pediatric neurologists, clinical psychologists, researchers, funders, pediatric and adult patients with migraine, and caregivers. Six reviewers screened references in duplicate against a priori eligibility criteria, tabulated information from included studies, rated risk of bias, conducted pairwise meta-analyses as well as network meta-analyses of three primary outcomes, and rated the strength of evidence (SOE).

Results. We included 57 randomized trials (44 trials of adults and 13 of children/adolescents) published since 1978. Most preventive interventions were multicomponent, using one or more of five primary components (cognitive behavioral therapy, biofeedback, relaxation, mindfulness-based therapies, and/or education). Most studies were at high risk of bias, primarily due to concerns about measurement bias and incomplete data. Given the small amount of evidence on any given intervention/comparator/outcome, data were often insufficient to permit conclusions. We found possible effectiveness (for adults) of mindfulness-based cognitive therapy, relaxation, and education; clinically important advantages of mindfulness-based stress reduction over education; and relaxation + education over propranolol for improving migraine-related quality of life (SOE: Low). For children/adolescents, the combination of cognitive behavioral therapy, biofeedback, and relaxation may result in lower migraine attack frequency and lower disability than education alone (SOE: Low). Evidence was insufficient to draw a conclusion regarding delivery of behavioral interventions using digital and telehealth modalities.

Conclusions. Evidence consisted primarily of small trials of multicomponent interventions compared with various types of control groups. Some preventative behavioral interventions may reduce headache frequency and disability. Future research should standardize intervention components to improve reproducibility, use attention control groups to control for expectation confounds, consider digital/telehealth for care delivery, improve the completeness of data collection, evaluate biopsychosocial patient factors including various social determinants of health, and report meta-analyzable effect sizes.