Skip to content

Chronic migraine and chronification

TL;DR — Chronic migraine is defined by ≥15 headache days/month for >3 months, with migraine features or migraine-specific response on ≥8 days; the boundary is operational, while biology and disability vary continuously (IHS 2018, PMID 29368949). It affects roughly 1–2% of the population and about 8% of people with migraine; older synthesis estimated annual episodic-to-chronic conversion near 3% and two-year remission near 26% (May 2016, PMID 27389092). High baseline headache frequency, acute-medication overuse, obesity, depression/anxiety, stressful events and sleep problems recur as predictors, but observational association and causal lever are not equivalent (Buse 2019, PMID 30589090). OnabotulinumtoxinA and CGRP-targeted therapies have randomized evidence in chronic migraine with medication overuse; a 2023 synthesis judged topiramate evidence insufficient for that specific subgroup (Dodick 2010, PMID 20487038; Giri 2023, PMID 36856015). No trial identified in the live PubMed search through 2026-08-30 proved disease modification after treatment stops.

A threshold on a continuum

State Monthly headache days Migraine-feature days Caveat
Low-frequency episodic Commonly 0–3 Below chronic threshold Not a formal ICHD subtype
Moderate-frequency episodic Commonly 4–7 Below chronic threshold Often prevention-eligible by disability
High-frequency episodic Commonly 8–14 Below chronic threshold Can resemble chronic migraine more than low-frequency disease
Chronic migraine ≥15 for >3 months ≥8 Can coexist with MOH

The 15-day boundary enables trials, reimbursement and communication, but a person moving from 14 to 15 days has not undergone a categorical biological transformation. Diaries should record total headache days, migraine days, severity/function and acute-medication days rather than only the label.

Disease course

Migraine can progress and remit. “Chronification” describes increasing frequency and persistence; “transformation” historically mixed phenotype and frequency change (Bigal 2009, PMID 19833063). Systematic review found inconsistent definitions, follow-up intervals and predictor adjustment across progression studies, limiting pooled causal inference (Buse 2019, PMID 30589090).

Transition Approximate signal Interpretation boundary
Episodic → chronic ~3% per year in older synthesis Average across cohorts/definitions (May 2016, PMID 27389092)
Chronic → episodic ~26% within 2 years “Remission” may mean crossing below 15 days, not freedom from migraine (May 2016, PMID 27389092)
Month-to-month category movement Common in longitudinal web cohorts Regression to mean and treatment both contribute (Adams 2015, PMID 25304766)

The reversible component is clinically important: chronic migraine is not inevitably progressive. It also complicates uncontrolled treatment reports because participants selected at a frequency peak may improve even without the intervention.

Risk architecture

Factor Evidence status Modifiability problem
Baseline headache frequency Most consistent predictor Also partly defines proximity to outcome
Acute-medication overuse Strong association Severe disease drives medication exposure
Depression/anxiety Recurrent association Bidirectional with disability and sleep
Obesity Population association Intervention evidence for preventing chronic migraine is limited
Sleep disturbance Recurrent association Often measured cross-sectionally
Stressful life events Association with progression Measurement and reverse causation
Low socioeconomic status/access Associated with burden and care gaps Bundles multiple exposures
Female sex Higher prevalence Not a treatment target; confounding by hormones/social exposure

The safest causal statement is that baseline frequency and overuse identify high-risk states, while evidence that deliberately changing each associated factor prevents chronification varies (May 2016, PMID 27389092; Buse 2019, PMID 30589090). A risk factor becomes a prevention target only after intervention evidence.

Mechanisms proposed

Repeated attacks may increase peripheral and central sensitization, alter descending control and stabilize a high-frequency network state. Allodynia and lower sensory thresholds can accompany frequent disease, but whether they cause progression or mark current attack load is unresolved (Bigal 2009, PMID 19833063).

Medication overuse may amplify nociceptive signaling, yet high frequency generates the opportunity and need for repeated acute use. CaMEO found acute-medication overuse more frequent in chronic than episodic migraine, but its observational design cannot determine direction (Schwedt 2021, PMID 34476122). Genetic susceptibility, psychiatric/sleep comorbidity and environmental exposures plausibly interact rather than form one linear pathway (May 2016, PMID 27389092).

OnabotulinumtoxinA

OnabotulinumtoxinA is administered using the PREEMPT fixed-site/fixed-dose protocol with optional follow-the-pain sites, generally 155–195 units across head/neck muscles every 12 weeks; the evidence and implementation history are reviewed by Escher et al. (2017, PMID 29204191). Its clinical effect is attributed to inhibition of peripheral sensory-neuropeptide release and reduced afferent input, not muscle paralysis alone (Whitcup 2014, PMID 25399521).

PREEMPT component Design/result Interpretation
PREEMPT 1 341 active vs 338 placebo; primary headache-episode endpoint not significant, several secondary endpoints positive One pivotal trial missed its original primary endpoint (Aurora 2010, PMID 20647170)
PREEMPT 2 347 active vs 358 placebo; headache-day primary endpoint favoured treatment Supported the pooled program (Diener 2010, PMID 20647171)
Pooled double-blind 1,384 adults; greater reduction in headache days and multiple secondary outcomes Pooling prespecified similar trials increased precision (Dodick 2010, PMID 20487038)
56-week program Two blinded cycles then three open-label cycles Long exposure but placebo comparison ends at week 24 (Aurora 2011, PMID 21883197)
Medication-overuse subgroup Benefit observed in baseline overuse subgroup Secondary analysis, not a withdrawal-strategy trial (Silberstein 2013, PMID 23790235)

The injection paradigm is standardized to preserve evidence fidelity. Customized dosing is common but has less randomized validation (Green 2018, PMID 29596847). Real-world series support effectiveness after oral failures but are subject to selection, missing follow-up and expectation (Pedraza 2015, PMID 25897415).

Neck pain, injection-site pain, ptosis and localized weakness are characteristic harms. Benefit may accumulate over multiple cycles, so stopping after one cycle risks false nonresponse; open-label completion analyses, however, preferentially retain tolerators/responders (Aurora 2014, PMID 24107267).

Pooled PREEMPT patient-reported outcomes also improved HIT-6 and migraine-specific quality-of-life measures through 56 weeks, but the controlled comparison ended after the 24-week double-blind phase (Lipton 2016, PMID 27288354).

CGRP-targeted prevention

Systematic review of chronic-migraine RCTs and real-world studies supports antibodies and gepants, including people with prior failures and medication overuse (Oliveira 2024, PMID 38575868). An indirect likelihood-to-help-or-harm analysis compared CGRP antibodies with older preventives but explicitly noted the absence of head-to-head trials at that time (Sacco 2021, PMID 33567891). PROMISE-2 enrolled 1,072 participants with baseline ~16.1 MMD: change over weeks 1–12 was −7.7 days with eptinezumab 100 mg, −8.2 with 300 mg and −5.6 with placebo (Lipton 2020, PMID 32209650).

The placebo change of 5.6 days is not noise to ignore; it includes expectancy, diary effects, regression to mean and natural fluctuation. A meta-analysis estimated a substantial contextual proportion of the observed MMD reduction in antibody trials, reinforcing use of active–placebo differences and responder rates (Forbes 2020, PMID 32668023).

Anti-CGRP antibodies can move participants from medication-overuse to non-overuse status without formal detoxification in trial data, but this does not prove withdrawal is unnecessary for every drug class or patient (Sirilertmekasakul 2024, PMID 38564060).

Conventional preventives

Topiramate has chronic-migraine evidence and is included with onabotulinumtoxinA and CGRP antibodies in MOH-focused randomized synthesis (Giri 2023, PMID 36856015). Beta-blocker evidence is much broader for episodic migraine; a systematic review found benefit for several agents but heterogeneity and fewer chronic-specific trials (Jackson 2019, PMID 30893319).

Selection should integrate comorbidity and harms rather than simply stacking agents. Cognitive adverse effects, reproductive risk, blood pressure, weight, constipation and mood can be more decision-relevant than small mean efficacy differences.

Difficult-to-treat and resistant migraine

“Resistant” and “refractory” definitions depend on number of failed classes, adequate dose/duration, contraindications and persistent disability. Apparent failure may reflect:

  • incorrect diagnosis or unrecognized secondary headache;
  • inadequate dose, duration or adherence;
  • absorption/timing failure for acute therapy;
  • medication overuse or opioid/barbiturate exposure;
  • untreated sleep or psychiatric comorbidity;
  • an outcome that ignores meaningful partial benefit;
  • access-driven discontinuation rather than biological nonresponse.

Trial adherence to IHS chronic-migraine guidance has been variable, making cross-study definitions and endpoints less comparable than labels imply (Deen 2019, PMID 31042062).

Combination therapy

OnabotulinumtoxinA plus a CGRP antibody is widely reported in refractory chronic migraine, but evidence is observational. A 2026 exploratory meta-analysis found a pooled 7.9-day MHD reduction (95% CI 5.7–10.2) across six extractable cohorts, with moderate-to-high heterogeneity and no controlled comparative designs (Sarvari Soltani 2026, PMID 41721358). The mechanistic rationale—reducing different afferent fiber signaling—does not substitute for randomized additive-effect and withdrawal trials. Combination decisions should record the marginal change after the second component, not only improvement from the original baseline.

Combination question Needed design
Is dual therapy additive? Randomized add-on with stable first component
Which sequence is better? Pragmatic head-to-head strategy trial
Can one component be removed? Blinded withdrawal after stable response
Does benefit justify cost? Societal cost-effectiveness with productivity
Are rare harms additive? Large registry with active comparator

Preventing progression

Risk-factor counseling is reasonable, but “disease modification” requires showing that early intervention changes later course after exposure ends. Existing preventive trials demonstrate suppression during treatment. None establishes that treating four MMD today prevents chronic migraine years later independently of ongoing therapy (May 2016, PMID 27389092).

A prevention trial would need high-risk episodic participants, early effective versus usual step care, standardized acute-use management, multi-year follow-up and post-withdrawal outcomes. Monthly-day transition alone is insufficient; disability and sustained state should co-primary.

Open questions

  • Can early high-efficacy prevention reduce multi-year chronic-migraine incidence after treatment withdrawal? (Buse 2019, PMID 30589090)
  • Which associated factors are causal levers rather than markers of rising frequency? (May 2016, PMID 27389092)
  • Does standardized onabotulinumtoxinA plus CGRP blockade provide randomized additive benefit? (Oliveira 2024, PMID 38575868)
  • What is the optimal number of onabotulinumtoxinA cycles before declaring nonresponse? (Aurora 2014, PMID 24107267)
  • Can remission be defined by sustained low disability rather than a binary crossing below 15 headache days? (Adams 2015, PMID 25304766)

References

  1. Headache Classification Committee of the IHS. ICHD-3. Cephalalgia. 2018. PMID 29368949
  2. May A, Schulte LH. Chronic migraine: risk factors, mechanisms and treatment. Nat Rev Neurol. 2016. PMID 27389092
  3. Bigal ME, Lipton RB. Migraine chronification—concept and risk factors. Discov Med. 2009. PMID 19833063
  4. Buse DC, et al. Migraine progression: a systematic review. Headache. 2019. PMID 30589090
  5. Adams AM, et al. Impact of chronic migraine: CaMEO methods and baseline results. Cephalalgia. 2015. PMID 25304766
  6. Dodick DW, et al. OnabotulinumtoxinA for chronic migraine: pooled PREEMPT results. Headache. 2010. PMID 20487038
  7. Aurora SK, et al. PREEMPT 1 randomized phase. Cephalalgia. 2010. PMID 20647170
  8. Diener HC, et al. PREEMPT 2 randomized phase. Cephalalgia. 2010. PMID 20647171
  9. Whitcup SM, et al. Development of onabotulinumtoxinA for chronic migraine. Ann N Y Acad Sci. 2014. PMID 25399521
  10. Aurora SK, et al. Pooled 56-week PREEMPT program. Headache. 2011. PMID 21883197
  11. Aurora SK, et al. Five treatment cycles in PREEMPT. Acta Neurol Scand. 2014. PMID 24107267
  12. Silberstein SD, et al. PREEMPT subgroup with acute-medication overuse. J Neurol Sci. 2013. PMID 23790235
  13. Green MW, et al. PREEMPT-derived versus customized injection paradigm. Toxicon. 2018. PMID 29596847
  14. Pedraza MI, et al. OnabotulinumtoxinA in 52 chronic-migraine patients using PREEMPT paradigm. Springerplus. 2015. PMID 25897415
  15. Lipton RB, et al. OnabotulinumtoxinA improves quality of life over one year: PREEMPT. Cephalalgia. 2016. PMID 27288354
  16. Chiang CC, et al. OnabotulinumtoxinA in chronic migraine: clinical evidence and experience. Ther Adv Neurol Disord. 2017. PMID 29204191
  17. Oliveira AB, et al. CGRP-targeted medication in chronic migraine: systematic review. J Headache Pain. 2024. PMID 38575868
  18. Lipton RB, et al. Eptinezumab in chronic migraine: PROMISE-2. Neurology. 2020. PMID 32209650
  19. Forbes RB, et al. Efficacy and contextual effects of CGRP antibodies: meta-analysis. Headache. 2020. PMID 32668023
  20. Drellia K, et al. Anti-CGRP monoclonal antibodies: likelihood to help or harm analysis. Cephalalgia. 2021. PMID 33567891
  21. Giri S, et al. Topiramate, botulinum toxin A and CGRP antibodies in chronic migraine with MOH. Cephalalgia. 2023. PMID 36856015
  22. Sirilertmekasakul P, et al. Transition from medication overuse after anti-CGRP antibodies. Neurol Sci. 2024. PMID 38564060
  23. Jackson JL, et al. Beta-blockers for prevention of headache: systematic review. PLoS One. 2019. PMID 30893319
  24. Deen M, et al. Adherence to IHS chronic-migraine preventive-trial guidelines. Cephalalgia. 2019. PMID 31042062
  25. Schwedt TJ, et al. Medication overuse and headache burden: CaMEO. Neurol Clin Pract. 2021. PMID 34476122
  26. Sarvari Soltani M, et al. Dual-mechanism onabotulinumtoxinA and anti-CGRP antibody prevention: exploratory meta-analysis. Eur J Med Res. 2026. PMID 41721358