Is migraine a spectrum disorder? Questioning the concept of “episodic” and “chronic” migraine using population-based data from 10,226 adults with migraine from 14 countries

BACKGROUND: The first edition of the International Classification of Headache Disorders (ICHD-I) did not recognize chronic migraine (CM). ICHD-II introduced CM as a complication of migraine, with criteria that were widely criticized and later revised to require ≥ 15 monthly headache days (MHDs) with ≥ 8 monthly migraine days (MMDs). ICHD-3 reclassified CM as a type distinct from episodic migraine, retaining these frequency thresholds but without empirical justification. We investigated whether population-based evidence on headache characteristics and burden supported distinction between episodic and chronic migraine types, or indicated that migraine was better conceptualized as a single disorder expressed along a frequency spectrum. METHODS: We performed a meta-analysis of individual participant data from 15 cross-sectional population-based surveys from the Global Campaign against Headache, all randomly selecting adults (18-65 years) and using the HARDSHIP questionnaire for data collection. Diagnoses made algorithmically observed the hierarchical order of ICHD. We identified four migraine groups: (1) migraine (definite+probable) including, and (2) migraine (definite+probable) excluding those with concomitant probable medication-overuse headache (pMOH); (3) migraine (definite only) including, and (4) migraine (definite only) excluding those with concomitant pMOH. MMDs and MHDs were regressed against headache characteristics, quality of life (QoL) and impaired participation. RESULTS: Among 36,407 participants (mean age 37.5 years; 53.4% females), 10,266 (28.2%) met criteria for migraine (14.6% definite; 13.6% probable), including 896 (2.5%) with concomitant pMOH. Distributions of MMDs and MHDs were heavily right-skewed, peaking at 3 days/month, but showed no inflections at the thresholds for CM. Age was positively associated with headache frequency, while most headache characteristics varied minimally across frequencies. QoL declined slightly but linearly with increasing frequency, whereas lost workdays, household days and social days increased more-or-less linearly (albeit that household days plateaued after 15 days/3 months). Regressions were strikingly similar across genders and the four migraine groups. CONCLUSION: Population-based evidence indicates that migraine, considered in relation to headache frequency, characteristics and attributed burden, is better understood as a spectrum disorder rather than as two (or more) entities distinguished categorically by frequency. The current thresholds for CM do not align with detectable step-increments in headache characteristics or burden. These findings fill a major knowledge gap, usefully informing the development of ICHD-4.