American Headache Society Migraine Guidelines

What the Guidelines Actually Say About Treating Migraine

The American Headache Society Migraine Guidelines are evidence-based recommendations for managing migraine, published in the journal Headache. The most recent major update came out in 2021, covering both acute and preventive treatment strategies. They're not a clinical decision support tool that auto-generates treatment plans. They're a framework that clinicians use to decide between options when they're sitting across from a patient who just told you their current medication stopped working. I ran into a situation last year where the guidelines pointed in two different directions for the same patient. She had migraine with frequent aura, cardiovascular risk factors, and a history of inadequate response to topiramate. The guideline section on CGRP monoclonal antibodies had a strong recommendation, but the preventive treatment hierarchy still listed beta-blockers and topiramate as first-line. The recommendation for erenumab or galcanezumab was solid, but insurance prior authorization was going to be a nightmare given her insurance plan's formulary tiering. I ended up going with the CGRP mAb after a detailed conversation with the patient about the trade-offs, documenting the medical necessity argument thoroughly in the chart. The guidelines support this path, but they don't walk you through the insurance reality that follows.

Where to Find the American Headache Society Migraine Guidelines

The full guideline document is available through the journal Headache, which is the official publication of the American Headache Society. It's also referenced and summarized on the AHS website. You can access it through standard medical journal subscriptions or institutional access if you have it. Many clinicians find the summary tables more useful than reading the full document end-to-end, since the evidence ratings are broken out clearly by condition and intervention. The acute treatment recommendations are organized around medication classes. Triptans remain the first-line specific therapy for moderate to severe attacks. The guidelines rate them as Level A evidence. NSAIDs like naproxen and ibuprofen are also Level A for mild to moderate attacks. Gepants, specifically rimegepant and ubrogepant, received new recommendations with B-level evidence after the clinical trial data came out. They're positioned as alternatives for patients who can't tolerate triptans or have contraindications like cardiovascular disease. One thing the guidelines don't emphasize enough is the timing issue. A patient who takes a triptan three hours into an attack gets a noticeably worse outcome than someone who takes it at onset. The evidence here is clear but the guidelines bury it somewhat in the discussion rather than making it a headline recommendation. I tell every patient the same thing: take the medication when the pain starts, not when it becomes unbearable. This single habit change is probably the biggest factor in whether acute treatment succeeds or fails for most people.

Overuse medication headache is another area where the guidelines are precise but the clinical application is messy. The threshold is fifteen days per month for triptans and combination analgesics, or ten days for gepants and ergots. But patients routinely underestimate their usage. I had a patient who insisted she was only taking her medication four days a week until I asked her to log it day by day. She was at twenty-two days. The guideline recommendation to stop the overused medication and transition to a preventive is straightforward in the text, but the withdrawal period can be brutal. Patients often need a bridge with a short course of corticosteroids or a non-specific agent during that transition. The guidelines mention this but don't give it the prominence it deserves in the prevention algorithms.

Preventive Treatment: What Beginners Get Wrong

The preventive section is where most clinicians I work with struggle with implementation. The guidelines give clear recommendations for first-line agents: topiramate, propranolol, metoprolol, valproate, amitriptyline, and candesartan. Each has a Level A or B rating depending on the indication. But the guidelines don't account for comorbidities the way real practice does. A patient with migraine and depression will respond differently to amitriptyline than one with migraine and hypertension where candesartan makes more sense. The guidelines list these options side by side without much nuance about selection. The bigger pitfall I see is the dosing timeline. Patients and clinicians both want to evaluate effectiveness too early. The guideline recommendation is to try an adequate dose for at least eight to twelve weeks before declaring failure. But many primary care providers reassess at four weeks, see no meaningful reduction, and switch. That's too soon. Topiramate especially needs gradual upward titration and a longer observation period. I've seen too many patients cycled through three or four preventives in six months because nobody gave them adequate trials. Each switch resets the clock. The math works out badly for the patient. CGRP monoclonal antibodies changed the landscape significantly. The 2021 guidelines gave them a Level A recommendation for prevention, which was a substantial shift. Erenumab, fremanezumab, galcanezumab, and eptinezumab all have robust trial data behind them. But the cost barrier is real and immediate. These medications can run several thousand dollars per month before insurance kicks in, and even with coverage, prior authorization delays of two to four weeks are common. During that delay, patients often go untreated and worsen. The workaround I've found useful is starting an oral preventive that the patient can get filled immediately while the prior authorization runs in parallel. It's not ideal, but it keeps the patient from falling through the cracks.

Botox and Neuromodulation Devices

OnabotulinumtoxinA for chronic migraine is a Level A recommendation. That part is well established. The guidelines specify the injection protocol clearly: thirty-one sites across seven muscle regions, repeated every twelve weeks. The evidence shows roughly a fifty percent responder rate, meaning about half of chronic migraine patients see their headache days drop by half or more. The other half get modest benefit or none at all. There's no reliable predictor upfront of who will respond. I've learned to manage expectations honestly at the first consultation rather than letting the patient assume Botox is a guaranteed solution. Neuromodulation devices like gammaCore, Cefaly, and Nerivio have Level B recommendations. They're positioned as non-pharmacologic options, which makes them attractive for patients who want to minimize medication burden. The clinical data is decent but not as strong as the pharmacologic options. Response rates hover around thirty to forty percent in real-world studies, and device adherence drops off significantly after the first few months. Most patients I see who get prescribed these end up using them sporadically. The guidelines don't address the behavioral component of device-based treatment, which is a genuine gap.

Special Populations and Where the Guidelines Fall Short

The pediatric and adolescent population is covered but the recommendations are less granular. The guideline supports topiramate and propranolol for prevention in younger patients, but the dosing adjustments and monitoring requirements aren't spelled out the way they would be in a dedicated pediatric neurology resource. For pregnant patients, the guidelines are essentially silent on many of the newer agents. Valproate is contraindicated, which is well known, but what about the CGRP antibodies? The pregnancy registry data is limited. Clinicians are left to extrapolate, and the guidelines acknowledge this gap without offering a clear alternative framework. Menstrual migraine is another area where the guidelines provide reasonable recommendations but clinical practice requires more granularity. The short-course preventive approach using naproxen or frovatriptan around the menstrual window is well supported. But I've encountered patients whose attacks don't align cleanly with their cycle and who end up on daily preventives when a targeted approach might have sufficed, or vice versa. Tracking headache diaries alongside cycle tracking is essential, but the guidelines don't emphasize the diary method as much as it deserves.

Limitations You Should Know About

The guidelines are a snapshot of the evidence at the time of publication. They don't incorporate data from trials that come out after the cutoff date. The CGRP oral preventives like atogepant and rimegepant were included in the 2021 update, but real-world effectiveness data is still accumulating. There's a difference between how these drugs performed in randomized controlled trials and how they perform in a busy clinic with patients who have comorbid conditions and polypharmacy. The guidelines can't capture that gap. Another limitation is the evidence hierarchy itself. Level A recommendations come from high-quality randomized trials, but those trials often exclude the exact patients you see in practice. Patients with multiple comorbidities, those on six or more medications, and those who've failed multiple preventives are systematically underrepresented in the evidence base. The guideline recommendations may not apply cleanly to your most complex cases. That's true for nearly all clinical practice guidelines, but it's worth keeping in mind when you're trying to decide what to do with a patient who doesn't fit the trial profiles. The guidelines also don't address health equity concerns directly. Access to CGRP therapies, neuromodulation devices, and even specialist referral varies enormously by geography, insurance type, and socioeconomic status. A Level A recommendation means nothing to a patient whose insurance won't cover it. The guidelines are clinically sound. The implementation reality is where the friction sits.