2026 AAN/AHS Migraine Prevention Guideline Summary

This is a condensed summary of the 2026 Pharmacologic Treatment for Migraine Prevention in Adults Practice Guideline (Potrebic et al.), jointly developed by the American Academy of Neurology (AAN) Guidelines Subcommittee and the American Headache Society (AHS) and endorsed by the American Academy of Family Physicians. It applies to adults (≥18 years) with migraine and is informed by a companion systematic review covering literature through June 6, 2024. The document does not formally name a superseded guideline.

Read the grades carefully — this is an AAN guideline, not an AHA/ASA one. It uses AAN levels of obligation, not Class of Recommendation / Level of Evidence. Verbatim from the paper: “Must” corresponds to Level A (very strong), “should” to Level B (strong), and “may” to Level C (weak). Separately, the systematic review grades confidence in the evidence for efficacy as high / moderate / low — a different axis from level of obligation. There are 52 recommendation statements: 13 Level A, 34 Level B, 5 Level C. Throughout this guideline, chronic migraine means headache on ≥15 days/month for >3 months which, on at least 8 days/month, has the features of migraine headache.

🔹 Bottom Line: Key Changes & High-Impact Points

  • A concrete threshold to start prevention — offer preventive treatment at ≥4 migraine days/month OR ≥4 moderate-to-severe headache days/month, or for substantial disability from migraine (Level B)
  • Drug choice is driven by what the patient prioritises, not by a single ranked ladder — the guideline compares options on four properties (efficacy, tolerability, safety, cost) and gives named, explicitly episodic-vs-chronic drug lists for the efficacy, tolerability and long-term-safety priorities, plus a list for patients not responding to higher-confidence choices. The cost recommendation carries no drug list — it defers to the patient’s formulary and insurance system (all Level B)
  • When efficacy is the priority — the high- or moderate-confidence lists (Level B): episodic migraine — atogepant, eptinezumab, erenumab, fremanezumab, galcanezumab, propranolol, topiramate, valproate; chronic migraine — the same list plus onabotulinumtoxinA, minus propranolol
  • Best tolerability = the CGRP-targeted agents (atogepant + the four mAbs; plus onabotulinumtoxinA in chronic migraine), while the longest safety track record = propranolol and topiramate (onabotulinumtoxinA and topiramate in chronic migraine)
  • Pregnancy is the Level A block — 4 of the 13 Level A statements are pregnancy-related: people with migraine of childbearing potential who require preventive therapy must be made aware of fetal risk in the event of an unplanned pregnancy, and known teratogens (divalproex sodium, topiramate) should be avoided if possible; maximise nonpharmacologic approaches; use drugs only after a full risk/alternatives discussion; and minimise total exposure
  • Named pregnancy options are all Level C (weak): nifedipine first; then metoprolol or propranolol if nifedipine is not an option or is ineffective, with their risks explicitly balanced against benefit; and onabotulinumtoxinA for chronic migraine only, again with risks balanced against benefit
  • Comorbidity-directed single-agent picks: amitriptyline for comorbid fibromyalgia (Level B), topiramate for increased BMI (Level B), and enalapril / nifedipine / telmisartan for comorbid untreated hypertension (Level C) — candesartan has insufficient evidence
  • Medication overuse is an indication to start prevention, not a barrier — offer preventives to those meeting medication-overuse criteria and to those with medication-overuse headache, preferring agents with evidence in that population (CGRP mAbs, atogepant, onabotulinumtoxinA, topiramate) (all Level B)
  • Explicit efficacy-assessment clock: wait 8–12 weeks at the recommended tolerated dose for most drugs, but 24 weeks for onabotulinumtoxinA (both Level B)
  • Adverse-effect counselling and monitoring are Level A — counsel on common and serious/life-threatening AEs before prescribing, and monitor as part of routine follow-up
  • Discontinuation carries a real relapse signal — counsel that stopping risks more headache days and worse headache-related QOL, and discuss the risks and benefits of tapering after 6 months of treatment (Level B)

1. Deciding to Start a Preventive Medication

  • Inform all patients with migraine that effective preventive treatments exist for frequent migraine attacks (Level B)
  • Offer preventive treatment to patients with migraine who experience ≥4 migraine days per month, or ≥4 moderate to severe headache days per month, to reduce headache frequency (Level B)
  • Offer preventive treatment to patients with substantial disability from migraine to reduce migraine-associated disability (Level B)

2. Shared & Informed Decision Making

  • Inform patients of appropriate medication choices, taking into account medical and psychiatric history, currently prescribed medications, and any contraindications (Level B)
  • Discuss and understand the patient’s preferences regarding potential adverse effects (Level B)
  • Discuss and understand the patient’s preferences regarding treatment modality — oral vs injectable (Level B)

3. Choosing a Preventive in Patients Without Significant Comorbidities or Contraindications

The panel found that no single medication has been shown to be clearly superior for efficacy in reducing migraine attack frequency, although some have greater confidence in evidence of efficacy than others; head-to-head trials have generally not provided evidence of sufficient quality to establish superiority of one medication over another. Options are therefore compared on four properties — strength of evidence for efficacy, tolerability, safety, and cost — and the recommendation is matched to whichever the patient prioritises. Each list below is population-specific: episodic and chronic migraine lists differ.

🔹 Matching the Drug to the Patient’s Priority (Recommendation 3)

  • Efficacy is the priority — offer preventives with high or moderate confidence in the evidence for efficacy (Level B):
    • Episodic migraine: atogepant, eptinezumab, erenumab, fremanezumab, galcanezumab, propranolol, topiramate, and valproate
    • Chronic migraine: atogepant, eptinezumab, erenumab, fremanezumab, galcanezumab, onabotulinumtoxinA, topiramate, and valproate
  • Tolerability is the concern — offer preventives with fewer side effects (Level B):
    • Episodic migraine: atogepant, eptinezumab, erenumab, fremanezumab, and galcanezumab
    • Chronic migraine: atogepant, eptinezumab, erenumab, fremanezumab, galcanezumab, and onabotulinumtoxinA
  • Long-term or unknown harms are the concern — offer preventives widely used for a long period of time, while observing specific contraindications (Level B):
    • Episodic migraine: propranolol and topiramate
    • Chronic migraine: onabotulinumtoxinA and topiramate
  • Cost is the concern — offer less expensive options depending on the patient’s specific pharmaceutical formulary plan and health care insurance system (Level B)
  • Not responding to higher-confidence choices — offer preventives with low confidence in the evidence for efficacy (Level B):
    • Episodic migraine: amitriptyline, flunarizine, metoprolol, pizotifen, rimegepant, and telmisartan

4. Pregnancy & Patients Planning Pregnancy

Randomised trials of drug safety in pregnancy are not ethically possible, so no drug can be declared completely safe in pregnancy. Some preventive medications have prolonged half-lives and therefore need to be stopped early in female patients planning a pregnancy, and many pregnancies are unplanned, so unexpected fetal drug exposure may occur. This section carries the densest concentration of Level A obligations in the guideline.

  • People with migraine of childbearing potential who require preventive therapy must be made aware of the potential risks to the fetus in the event of an unplanned pregnancy, and agents with known teratogenic effects (e.g., divalproex sodium, topiramate) should be avoided if possible (Level A)
  • In people who are pregnant or planning pregnancy, nonpharmacologic approaches must be maximised — behavioral interventions, acupuncture, exercise, trigger management (Level A)
  • If pharmacologic prevention is considered potentially necessary during pregnancy, it must be used only after a thorough discussion of risks and alternatives, with joint decision making (Level A)
  • If pharmacologic prevention is necessary during pregnancy, clinicians must develop a strategy that limits overall pharmacologic exposure while maintaining the best possible migraine control (Level A)
  • If a preventive is considered necessary during pregnancy, review the ACOG guideline medication recommendations and the potential associated risks determined from their systematic review (Level B)

🔹 Named Drug Options in Pregnancy — All Level C (Weak)

  • First named option: nifedipine (Level C). ACOG lists no potential associated risks for nifedipine; AAN confidence in efficacy is low, while ACOG considers it for first-line use in prevention.
  • If nifedipine is not an option or is ineffective: metoprolol or propranolol (Level C) — but the risks (cardiovascular abnormalities, cleft lip or palate, neural tube defects, and fetal growth retardation) must be balanced against the potential benefits.
  • For chronic migraine only: onabotulinumtoxinA (Level C) — risks must be balanced against benefits. Outcome data are extremely limited and mainly pertain to preconception and first-trimester exposures. A Class III study of 397 exposed women (95% preconception or first trimester) found an overall fetal defect prevalence of 2.6%, consistent with general-population rates.
  • Confidence-in-evidence vs ACOG position, for the agents that appear on both lists: nifedipine (AAN low / ACOG consider first line), amitriptyline (AAN low / ACOG balance risks and benefits), metoprolol (AAN low / ACOG balance), propranolol (AAN moderate / ACOG balance), onabotulinumtoxinA (chronic migraine only; AAN high / ACOG balance).

5. Lactation

  • Clinicians must counsel patients who are lactating that migraine preventives pass into breast milk in varying amounts and can affect the health of breastfed infants, including counselling on the risks of specific medications using evidence-based reference databases such as LactMed (Level A)
  • Information on passage into breast milk varies between drugs and is scarce for newer treatments

6. Treatment Over the Adult Lifespan (Older Adults)

  • When prescribing a preventive for older adults, assess — depending on the drug considered — for vascular disease, potential drug interactions, and reduced renal and/or hepatic function (Level B)
  • When prescribing a preventive with sedative potential to an older adult, discuss sedation and confusion and, if appropriate, initiate therapy at lower-than-usual doses (Level B). Preventives more likely to cause sedation: amitriptyline, valproate, topiramate, and pizotifen
  • When choosing a preventive for an older adult, the clinician must discuss the possibility of hypotension or postural hypotension and, if appropriate, avoid drugs that significantly lower blood pressure or start at lower-than-usual doses (Level A). RCTs of telmisartan, propranolol, and metoprolol in migraine prevention have shown significant decreases in blood pressure

7. Sex-Related Considerations in Drug Choice

  • Clinicians prescribing valproic acid must notify female patients that it can increase the risk of polycystic ovary syndrome (Level A). In women with epilepsy, PCOS incidence was higher with valproate (OR 3.04, 95% CI 2.09–4.43)
  • Clinicians prescribing topiramate must notify female patients of childbearing potential that it can make hormonal contraception less effective at doses above 200 mg/d (Level A). Doses ≤200 mg/d produce minimal or no reduction in estrogen and norethindrone; doses >200 mg/d cause potentially clinically important reductions in serum etonogestrel in implant users, while 100 mg/d causes only minor decreases. Low-dose topiramate as used in migraine prevention does not increase unintended pregnancy versus propranolol, amitriptyline, or other oral preventives
  • Clinicians prescribing a preventive for older adult male patients should evaluate whether there is risk of urinary retention and either avoid drugs with anticholinergic effects (e.g., amitriptyline) or discuss the possibility with the patient (Level B). Tricyclic antidepressants were significantly more likely than placebo to cause urinary retention (OR 2.52, 95% CI 1.29–4.90)

8. Comorbid Medical & Psychiatric Conditions

General Approach

  • Discuss with the patient who has a medical or psychiatric comorbidity whether it is preferable to treat both conditions with a single medication (combined approach) or to treat them independently (Level B)
  • If a combined approach is chosen, monitor efficacy and side effects closely to determine whether to switch to treating both conditions independently (Level B)

Comorbid Hypertension

  • Clinicians may inform people with migraine and comorbid untreated hypertension that monotherapy options are available to treat both conditions, including enalapril, nifedipine, and telmisartan (Level C) — each assessed as possibly effective in migraine prevention with low confidence in the evidence
  • Candesartan is widely used in migraine prevention, but there is insufficient evidence to support its use

Comorbid Fibromyalgia

  • Offer amitriptyline for migraine prevention in patients with comorbid fibromyalgia (Level B). In fibromyalgia, amitriptyline 10–50 mg/d versus placebo reduced sleep disturbance (SMD −0.97, 95% CrI −1.10 to −0.83) and fatigue (SMD −0.64, 95% CrI −0.75 to −0.53) and improved QOL (SMD −0.80, 95% CrI −0.94 to −0.65)

Increased Body Mass

  • When an oral preventive is being considered in a patient with increased body mass index, offer topiramate (Level B) — topiramate produced statistically significant weight decreases versus placebo at all dosages studied (50–200 mg)
  • Preventives that promote weight gain in some patients: amitriptyline, propranolol, pizotifen, and flunarizine. Obesity is a risk factor for chronification of migraine

9. Patients With Medication Overuse

Medication overuse is defined here as regular use of acute and/or symptomatic migraine drugs with intake of >9 days/month for prescription migraine pain-relieving medication or >14 days/month for nonspecific pain medications.

  • Offer preventive medication to those with migraine who meet criteria for medication overuse (Level B)
  • Offer preventive medication to those who have medication-overuse headache (Level B)
  • Prefer preventives with evidence in this population — CGRP monoclonal antibodies, atogepant, onabotulinumtoxinA, and topiramate — over medications lacking such evidence (Level B)
  • Available evidence suggests that simply starting a preventive is equally effective to starting a preventive plus implementing acute medication withdrawal — but those studies included only small numbers of participants who overused opioids or barbiturate-containing medications, so there is insufficient evidence to draw the same conclusion in those scenarios

10. Interaction With the Patient’s Acute Treatments

  • When prescribing a preventive, clinicians must evaluate the risk of drug interactions with the patient’s acute migraine medications (Level A)
  • When prescribing a preventive, clinicians must evaluate the risk of side effects from combining that preventive with the patient’s acute migraine medications (Level A)

11. When to Assess Treatment Efficacy

🔹 The Efficacy Clock (Recommendation 14 — all Level B)

  • Most medications: wait at least 8–12 weeks at the recommended tolerated dose before assessing efficacy
  • OnabotulinumtoxinA: wait 24 weeks at the recommended dose before assessing efficacy
  • Suboptimal response by 8 weeks → optimise the dose to the maximum tolerated dose, or to the predetermined maximum based on efficacy studies
  • No efficacy after 8–12 weeks with other contributing factors identified (e.g., inconsistent adherence, frequent trigger exposure) → shared decision making to continue an additional observation period
  • Side effects not tolerable → consider a dose reduction; and offer an alternative treatment option
  • Suboptimal response after 12–24 weeks (per the 8–12-week and 24-week thresholds above), with other non-contraindicated options available → shared decision making to initiate an alternate preventive
  • Encourage continued lifestyle and trigger management in every patient taking a preventive medication

Underlying evidence: superiority to placebo was typically demonstrated between 8 and 12 weeks for oral medications (most common timepoint 12 weeks); onabotulinumtoxinA studies evaluated response at 24 weeks; CGRP mAb studies evaluated response at both 12 and 24 weeks.

12. How to Assess Treatment Efficacy

  • Document the frequency of migraine attacks to optimise assessment of treatment efficacy (Level B)
  • Clinicians’ assessment of treatment efficacy must be patient-centric and based on individual goals (Level A)
  • Use frequency, severity, migraine-associated symptoms, QOL, and use of acute therapy when assessing efficacy (Level B)
  • Clinicians may use a measurement tool — headache diary, HIT-6, or MIDAS — consistently across patient visits to monitor efficacy and treatment response (Level C). QOL instruments sensitive to improvement in migraine trials: MIDAS, HIT-6, MSQ, and MPFID

13. Counselling & Assessment of Adverse Effects

  • Clinicians must provide counselling on common and serious or life-threatening adverse effects, based on product monographs and drug information databases, before prescribing a migraine preventive (Level A)
  • Clinicians must monitor patients for common and serious or life-threatening adverse effects as part of routine follow-up (Level A)

14. Stopping Preventive Therapy

  • Counsel patients that there is limited evidence suggesting a risk of increased headache days and decreased headache-related QOL after discontinuing a preventive medication (Level B)
  • Discuss the potential benefits and risks of tapering a preventive after 6 months of treatment (Level B)
  • Only two discontinuation studies were of sufficient design to inform recommendations: after 6 months of topiramate 50–200 mg, randomised discontinuation (by 100 mg/week) produced a slight increase in migraine attack days (1.19 days in 4 weeks, 95% CI 0.71–1.66; p < 0.0001), more days on acute medication, and worse migraine-related QOL on MIDAS. After 6 months of galcanezumab, discontinuation was associated with >50% of month-6 responders (≥50% reduction in monthly migraine headache days) losing their response 5 months after discontinuation

🔴 Cautions & What NOT To Do

This AAN/AHS guideline has no COR 3 “no benefit / harm” class. The items below are the explicit avoid-, caution- and insufficient-evidence statements found in the recommendations and their rationales.

  • Avoid known teratogens — divalproex sodium and topiramate — if possible in people of childbearing potential who require preventive therapy (Level A)
  • Do not start a drug in pregnancy without first maximising nonpharmacologic approaches and holding a full risks-and-alternatives discussion (both Level A)
  • Tricyclic antidepressants: despite amitriptyline appearing in table 2 of the ACOG guideline, the text of that guideline states TCA use is not recommended in pregnancy
  • Candesartan — insufficient evidence to support its use in migraine prevention, despite widespread use
  • Opioid or barbiturate overuse — insufficient evidence that starting a preventive alone (without acute medication withdrawal) is adequate; do not extrapolate the general medication-overuse finding to these patients
  • Older adult men: avoid anticholinergic drugs (e.g., amitriptyline) where there is risk of urinary retention, or explicitly discuss the risk (Level B)
  • Older adults generally: avoid drugs that significantly lower blood pressure where postural hypotension is a concern, or start below usual doses (Level A) — note telmisartan, propranolol, and metoprolol all significantly lower BP
  • Do not judge onabotulinumtoxinA a failure before 24 weeks, or most medications before at least 8–12 weeks at the recommended tolerated dose (Level B)
  • Topiramate >200 mg/d makes hormonal contraception less effective — female patients of childbearing potential must be notified (Level A)
  • Valproic acid — female patients must be notified of increased PCOS risk (Level A)
  • Note which preventives promote weight gain — amitriptyline, propranolol, pizotifen and flunarizine — when body mass is a consideration; where an oral preventive is being considered in a patient with increased BMI, the guideline’s recommended choice is topiramate (Level B)

Reference

Potrebic S, Tanveer S, Becker WJ, et al. Pharmacologic Treatment for Migraine Prevention in Adults Practice Guideline Recommendations: Report of the AAN Guidelines Subcommittee and the American Headache Society. Neurology. 2026;107(7):e214881. doi: 10.1212/WNL.0000000000214881

Companion systematic review: Pringsheim T, Smith DB, Tanveer S, et al. Pharmacologic treatment for migraine prevention in adults systematic review: report of the AAN Guidelines Subcommittee and the American Headache Society. Neurology. 2026;107(4):e218112. doi: 10.1212/WNL.0000000000218112