Medication-overuse headache¶
TL;DR — Medication-overuse headache (MOH) is diagnosed when a person with a pre-existing headache disorder has headache on ≥15 days/month and regularly overuses acute medication for >3 months at class-specific thresholds; migraine and MOH are coded together (IHS 2018, PMID 29368949). The diagnosis identifies a reversible treatment target but does not prove a one-way mechanism: rising migraine frequency drives medication exposure while repeated exposure can perpetuate headache (Ashina 2023, PMID 36732518). Education plus withdrawal and preventive therapy has the best direct strategy evidence; withdrawal plus early prevention produced the highest six-month cure rate, but at one year all three tested strategies remained effective and between-group effects were not significant (Carlsen 2020, PMID 32453406; Carlsen 2021, PMID 34325483). Withdrawal may be outpatient and advice-led for uncomplicated cases, while opioid/barbiturate use, severe psychiatric/medical comorbidity or repeated relapse may require structured care (Rossi 2006, PMID 16919060; Rossi 2013, PMID 23565591). CGRP-targeted therapies can reduce both headache and overuse without formal detoxification in trials, but this does not erase dependence, toxicity or withdrawal risks of the overused drug (Sirilertmekasakul 2024, PMID 38564060).
Diagnostic thresholds¶
| Acute class | Overuse threshold for >3 months | Important boundary |
|---|---|---|
| Ergot derivatives | ≥10 days/month | Count days, not number of doses |
| Triptans | ≥10 days/month | Multiple triptans share the class-day total |
| Opioids | ≥10 days/month | Dependence/withdrawal risk adds complexity |
| Combination analgesics | ≥10 days/month | Includes combinations with caffeine/opioid/barbiturate |
| Simple analgesics/NSAIDs/acetaminophen | ≥15 days/month | Aggregate non-opioid analgesic days per ICHD rules |
| Multiple classes without individual overuse | Combined regular use ≥10 days/month | Switching classes does not evade exposure |
Headache must occur on ≥15 days/month. “Medication overuse” below that headache threshold is a risk state, not automatically MOH. The overused medicine was taken for acute/symptomatic headache treatment; exposure for another indication requires careful causal classification (IHS 2018, PMID 29368949).
Why causality remains contested¶
The ICHD-3 diagnosis no longer requires improvement after withdrawal. That makes diagnosis prospective and treatment timely, but weakens within-person causal confirmation. The conceptual loop is:
more migraine days → more acute-treatment days → altered nociceptive/behavioural state → still more headache days
Severe baseline disease, poor acute efficacy, limited preventive access and psychiatric comorbidity can drive both sides. Population surveys show overuse associated with higher monthly headache days and disability, but cross-sectional direction is unresolved (Schwedt 2018, PMID 29797100).
| Evidence | Supports a medication contribution | Residual ambiguity |
|---|---|---|
| Headache often improves after withdrawal | Reversibility after exposure reduction | Education, prevention and regression occur simultaneously |
| Drug classes have different thresholds/relapse | Pharmacological contribution | Thresholds are consensus-informed, not individual biological cutoffs |
| Animal repeated opioid/triptan exposure alters trigeminal processing | Mechanistic plausibility | Dosing and endpoints differ from clinical MOH |
| Prevention can reverse overuse without forced withdrawal | High frequency is treatable upstream | Does not show exposure was harmless |
Repeated morphine exposure in animals produces sustained sensitization and loss of diffuse inhibitory controls in dura-sensitive medullary neurons, a plausible opioid mechanism but not a quantitative human threshold model (Okada-Ogawa 2009, PMID 20016098).
A coordinate-based neuroimaging meta-analysis reported convergent gray-matter alterations across small MOH studies, but divergent source studies and absence of diagnostic performance make these group findings mechanistic candidates rather than a clinical test (Chen 2022, PMID 35143020).
Epidemiology and risk profile¶
Reviews place general-population prevalence around 1–2%, with enrichment in women aged 30–50 and in depression, anxiety and other chronic pain conditions (Cheung 2015, PMID 25398377; González-Oria 2021, PMID 32917437). Estimates depend heavily on acute-drug availability, diagnostic criteria and survey ability to count treatment days.
| Associated factor | Evidence interpretation |
|---|---|
| High baseline headache frequency | Creates exposure opportunity and predicts progression |
| Depression/anxiety | May increase disability, self-management difficulty and relapse |
| Other chronic pain | Adds non-headache analgesic exposure |
| Opioid/barbiturate-containing products | Higher dependence and complicated withdrawal potential |
| Previous MOH/withdrawal | Marks recurrence risk |
| Limited preventive/specialist access | Sustains reliance on repeated rescue |
In a comparative study, mood disorders (OR 4.5, 95% CI 1.5–13.5), anxiety (OR 5.0, 1.2–10.7) and other substance-use disorders (OR 7.6, 2.2–26.0) were more frequent in migraine-derived MOH than migraine alone; selection and temporal direction limit causal interpretation (Radat 2005, PMID 15955038).
Treatment components¶
Treatment is not synonymous with abrupt cessation. It contains distinct components:
- explain the exposure–headache loop without blame;
- choose abrupt versus tapered withdrawal by drug class;
- plan short-term withdrawal symptoms and rescue boundaries;
- start or optimize prevention;
- replace the acute plan with an effective, limited strategy;
- monitor days and relapse;
- address dependence, sleep and psychiatric comorbidity.
Systematic review found withdrawal widely supported but limited high-quality evidence comparing setting, bridge therapy and timing of prevention (Chiang 2016, PMID 26122645). Updated French recommendations likewise organize education, withdrawal and prevention as a strategy rather than treating detoxification as a single universal procedure (Corand 2021, PMID 34332779).
Randomized strategy evidence¶
The Danish three-arm open-label RCT randomized 120 adults to withdrawal plus immediate prevention, prevention alone with optional later withdrawal, or withdrawal with optional later prevention. At six months, all strategies improved, but combined withdrawal plus prevention produced the highest MOH cure and episodic-headache conversion (Carlsen 2020, PMID 32453406).
At one year, 96 participants were evaluated. Monthly headache days fell by 10.3 with withdrawal plus prevention, 10.8 with prevention plus optional later withdrawal and 7.9 with withdrawal plus optional later prevention; the between-group comparison was not significant (p=0.377), and relapse occurred in 11% overall (Carlsen 2021, PMID 34325483). The combined strategy produced the fastest effect, while open-label treatment and a tertiary Danish setting limit transportability.
| Strategy | Advantage | Constraint |
|---|---|---|
| Withdrawal + immediate prevention | Attacks both exposure and baseline disease | More simultaneous changes and attribution difficulty |
| Prevention first | May make spontaneous reduction feasible | Toxic/dependence-prone exposure continues initially |
| Withdrawal first | Demonstrates exposure effect and simplifies regimen | Withdrawal symptoms and untreated migraine burden |
An earlier 56-person randomized open-label study found prophylaxis without detoxification, outpatient detoxification and no specific intervention difficult to distinguish robustly, illustrating underpowering and heterogeneity (Hagen 2009, PMID 18823363). Four-year follow-up shows relapse and long-term fluctuation remain material (Hagen 2011, PMID 21711255).
Advice, brief intervention and setting¶
In low-medical-need transformed migraine, strong advice to stop overused medication performed similarly to structured detoxification programs, supporting simple outpatient approaches for selected patients (Rossi 2006, PMID 16919060). In complicated MOH, structured programs were tested because psychiatric/medical comorbidity and prior detoxification change needs (Rossi 2013, PMID 23565591).
Primary-care and neurologist follow-up after withdrawal produced similar mean headache outcomes in a randomized study, suggesting that protocols need not be specialist-owned for every case (Bøe 2009, PMID 19228151). A brief intervention in general practice also identified predictors of successful detoxification (Kristoffersen 2017, PMID 28369734).
| Outpatient-favouring | Structured/inpatient-favouring |
|---|---|
| Non-opioid/simple regimen | Opioid or barbiturate dependence/withdrawal risk |
| Stable psychiatric/medical state | Unstable psychiatric or serious medical comorbidity |
| Reliable follow-up/support | Repeated failed outpatient withdrawal |
| Clear replacement acute plan | Severe withdrawal vomiting/dehydration or unsafe rescue behaviour |
Abrupt versus tapered withdrawal¶
Triptans and simple analgesics are commonly stopped abruptly. Opioids, barbiturates and some sedative combinations may require tapering and supervised management because physiological withdrawal can be hazardous. The migraine diagnosis never authorizes abrupt discontinuation of a dependence-producing drug without class-specific planning.
Withdrawal headache, nausea, sleep disruption, anxiety and autonomic symptoms commonly peak over days. Evidence for corticosteroid bridging is inconsistent; prednisone remains used in some settings without a decisive modern benefit case (Kaltseis 2022, PMID 36437611).
Preventive therapies in MOH¶
A 2023 randomized-evidence synthesis supported onabotulinumtoxinA and CGRP antibodies in chronic migraine with medication overuse but judged evidence insufficient to determine topiramate's effect in that specific population (Giri 2023, PMID 36856015). Acute withdrawal plus add-on botulinum toxin A did not show a clear additional effect in a double-blind trial, suggesting withdrawal itself can be powerful and that visible cosmetic effects complicate blinding (Pijpers 2019, PMID 30982843).
Anti-CGRP trial meta-analysis found movement from overuse to non-overuse across acute categories (Sirilertmekasakul 2024, PMID 38564060). This supports a prevention-first option for some patients, especially when withdrawal has failed, but does not answer long-term relapse after stopping preventive therapy.
Dependence-like behaviour and stigma¶
Some patients show impaired control, salience and difficulty reducing medication, while others simply follow frequent symptoms with rational repeated dosing. Treating all MOH as addiction stigmatizes; treating all use as passive exposure misses dependence.
The Severity of Dependence Scale changed across randomized MOH strategies, with withdrawal plus prevention producing larger improvements in dependence-like behaviour (Rouw 2021, PMID 33326656). Screening should guide support, not moral classification.
Relapse prevention¶
| Control | Operational form |
|---|---|
| Diary | Monthly headache, migraine and class-specific medication days |
| Effective rescue | A plan that achieves function without serial redosing |
| Preventive review | Early response, tolerability and adherence checks |
| Refill signal | Detect early excess rather than wait for ≥3 months |
| Comorbidity care | Depression, anxiety, sleep and other pain management |
| Education | State exact day limits and what to do when approaching them |
Mobile-app prevention trials aim to detect rising medication days and reinforce education, but digital engagement and clinical outcome effectiveness require confirmation (Diener 2022, PMID 35546412).
In a 215-patient inpatient-withdrawal cohort, 172 had one-year data and 22% had relapsed; psychiatric comorbidity and overused-drug features were candidate prognostic factors, but the uncontrolled specialist sample cannot define an individual relapse rule (Sances 2010, PMID 19614697).
Open questions¶
- For which drug classes and patients can prevention-first therapy replace formal withdrawal without higher relapse? (Carlsen 2020, PMID 32453406; Sirilertmekasakul 2024, PMID 38564060)
- Are ICHD class thresholds biologically distinct or pragmatic population cutoffs? (IHS 2018, PMID 29368949)
- Which bridge, if any, improves withdrawal completion without becoming a new overused treatment? (Kaltseis 2022, PMID 36437611)
- Can refill/app monitoring prevent MOH prospectively in high-frequency episodic migraine? (Diener 2022, PMID 35546412)
- What separates dependence-like MOH from rational high use caused by undertreated chronic migraine? (Rouw 2021, PMID 33326656)
Related pages¶
- Chronic migraine and chronification — the high-frequency substrate.
- Acute treatment — replacement rescue plans and class hazards.
- Preventive treatment — options started during or before withdrawal.
- Migraine in children and adolescents — sparse age-specific MOH evidence.
- Red flags and safety concerns — dangerous withdrawal and medication toxicity.
References¶
- Headache Classification Committee of the IHS. ICHD-3. Cephalalgia. 2018. PMID 29368949
- Ashina S, et al. Medication overuse headache. Nat Rev Dis Primers. 2023. PMID 36732518
- Cheung V, Amoozegar F. Medication overuse headache. Curr Neurol Neurosci Rep. 2015. PMID 25398377
- González-Oria C, et al. Revision and updating of medication overuse headache. Neurologia. 2021. PMID 32917437
- Chiang CC, et al. Treatment of medication-overuse headache: systematic review. Cephalalgia. 2016. PMID 26122645
- Carlsen LN, et al. Comparison of three treatment strategies for MOH: randomized trial. JAMA Neurol. 2020. PMID 32453406
- Carlsen LN, et al. Treatment of MOH: effect and predictors after one year. Headache. 2021. PMID 34325483
- Hagen K, et al. Management of MOH: one-year randomized multicentre trial. Cephalalgia. 2009. PMID 18823363
- Hagen K, et al. Randomized trial on MOH: outcome after one and four years. Acta Neurol Scand Suppl. 2011. PMID 21711255
- Rossi P, et al. Advice versus structured detoxification in low-medical-need transformed migraine. Cephalalgia. 2006. PMID 16919060
- Rossi P, et al. Advice versus structured detoxification for complicated MOH. J Headache Pain. 2013. PMID 23565591
- Bøe MG, et al. Chronic daily headache with medication overuse: neurologist versus primary-care follow-up. Cephalalgia. 2009. PMID 19228151
- Kristoffersen ES, et al. Predictors of successful primary-care detoxification. Acta Neurol Scand. 2017. PMID 28369734
- Corand V, et al. French recommendations for MOH treatment strategies. Rev Neurol (Paris). 2021. PMID 34332779
- Kaltseis K, et al. Prednisone for withdrawal headache in the CGRP-antibody era. Headache. 2022. PMID 36437611
- Pijpers JA, et al. Acute withdrawal and botulinum toxin A in chronic migraine with medication overuse. Brain. 2019. PMID 30982843
- Giri S, et al. Topiramate, botulinum toxin A and CGRP antibodies in chronic migraine with MOH. Cephalalgia. 2023. PMID 36856015
- Sirilertmekasakul P, et al. Transition from medication overuse after anti-CGRP antibodies. Neurol Sci. 2024. PMID 38564060
- Rouw C, et al. Dependence-like behaviour in patients treated for MOH. Eur J Pain. 2021. PMID 33326656
- Diener HC, et al. Mobile-software prevention of medication overuse and MOH: trial protocol. Trials. 2022. PMID 35546412
- Radat F, et al. Psychiatric comorbidity in evolution from migraine to MOH. Cephalalgia. 2005. PMID 15955038
- Sances G, et al. Risk factors in MOH: one-year follow-up. Cephalalgia. 2010. PMID 19614697
- Schwedt TJ, et al. Factors associated with acute medication overuse: MAST study. J Headache Pain. 2018. PMID 29797100
- Okada-Ogawa A, et al. Morphine-induced sensitization in dura-sensitive medullary neurons. J Neurosci. 2009. PMID 20016098
- Chen XY, et al. Gray-matter alteration in MOH: ALE meta-analysis. Brain Imaging Behav. 2022. PMID 35143020