Abstract
Purpose of Review Trials of non-pharmacological treatment in chronic migraine report inconsistent effects on headache frequency. This review groups recent trials and syntheses by mechanism and asks whether mechanism explains the inconsistency and what it implies for treatment order. Recent Findings A primary-care education and self-management programme (CHESS) improved pain self-efficacy, while headache impact, headache days and medication overuse were unchanged at 12 months. Mindfulness added to withdrawal and prophylaxis (MIND-CM) increased the proportion with at least 50% fewer headache days (78% versus 48%). In chronic migraine with medication overuse, eptinezumab added to brief education reduced monthly migraine days in weeks 1–4 by 3.2 more than placebo (RESOLUTION). Syntheses rate the frequency effects of cognitive-behavioural, relaxation and mindfulness treatments as low-certainty; acceptance-based treatment improved disability without a significant frequency effect, and mindfulness reduced attack frequency slightly. During withdrawal, a nurse-led behavioural programme reduced acute-medication days at week 24 without changing migraine days. Summary Grouped by mechanism, treatments that lower ascending nociceptive drive, including calcitonin gene-related peptide (CGRP) pathway blockade and, on weaker evidence, behavioural treatment of insomnia, reduce headache frequency. Treatments aimed at the appraisal of pain improve self-efficacy and, in several trials, disability; no trial identified isolated their effect on frequency after a documented fall in drive, most tested packages mix both mechanisms, and some trials do not fit. The proposed rule is to start drive-lowering treatment first, behavioural forms included, and to offer interpretation-directed treatment alongside it when waiting is inappropriate. Five falsifiable predictions follow.