Introduction: Sorting Through the Prevention Options

If you get frequent migraine attacks, your doctor may suggest a preventive (prophylactic) medication, something you take regularly to make attacks less frequent rather than to stop one in progress. The problem is that there are a lot of options: beta-blockers, anticonvulsants, antidepressants, calcium channel blockers, the newer CGRP-targeted drugs, and more. Which one should come first?

A new systematic review and network meta-analysis published in BMJ Evidence-Based Medicine in June 2026 tackles exactly that question. It pooled 199 randomized controlled trials covering more than 47,000 adults to compare preventive drug classes head to head, not just on how well they reduce attacks but on how likely people are to stick with them. That second part matters, because a drug only helps if you can tolerate taking it.

What the Researchers Did

The team searched six major medical databases from their inception through January 2026 for randomized controlled trials of preventive drugs in adults with episodic or chronic migraine. They graded each trial for risk of bias using the Cochrane risk-of-bias tool, and they rated the overall certainty of each comparison using the GRADE system, which is the standard approach guideline panels use to decide how much to trust a finding.

The main effectiveness outcome was the reduction in monthly migraine days compared with placebo. They also looked at monthly migraine attacks and at tolerability, measured as how often people dropped out of a trial because of the treatment.

The Headline Findings

Three drug classes came out on top for cutting the number of migraine days each month: calcium channel blockers, CGRP-targeted therapies, and beta-blockers. The reductions were modest in absolute terms (roughly one and a half to two fewer migraine days per month versus placebo) but consistent.

When the researchers narrowed the lens to monthly migraine attacks (as opposed to total days), beta-blockers, calcium channel blockers, anticonvulsants, and CGRP-targeted therapies all probably reduce attack frequency, with moderate certainty.

The catch shows up in tolerability. People were significantly more likely to stop taking calcium channel blockers and anticonvulsants than placebo, meaning the side effects pushed more of them off the drug. Beta-blockers and CGRP-targeted therapies did not carry that same tolerability penalty.

The Drug Classes, Briefly Explained

If some of these names are unfamiliar, here is a quick orientation to what each class is and how it is generally used in migraine.

Effectiveness and Tolerability Side by Side

How the Classes Compared

This table summarizes the analysis for the main preventive drug classes, including the effect on monthly migraine days, the effect on attacks, tolerability, and the GRADE certainty of the evidence. Negative numbers mean fewer migraine days or attacks compared with placebo.

Why Tolerability Is Half the Story

It is tempting to just pick whichever drug shows the biggest number next to it. Calcium channel blockers had the largest reduction in migraine days here, so why not start there? Because the same analysis found people were 40% more likely to quit them due to side effects. A preventive you abandon after a month does not prevent much.

This is the practical value of looking at effectiveness and tolerability together. Beta-blockers and CGRP-targeted therapies landed in a sweet spot: they meaningfully reduced migraine frequency and people generally stayed on them. That combination is what makes the authors point to these two as the strongest options when you restrict the picture to the most trustworthy evidence.

"The best preventive on paper is not always the best preventive in real life. A medication only works if you can actually keep taking it, which is why effectiveness and tolerability have to be weighed together rather than one at a time." - Cerebral Torque

A Note on Certainty: Not All Findings Are Equal

One of the most useful things about this study is that it does not treat every result as equally solid. Using GRADE, the authors flagged how confident we can be in each comparison. The reduction in migraine days from CGRP-targeted therapies earned a high-certainty rating, the strongest tier. Most other findings landed at moderate certainty, meaning the effect is probably real but future research could shift it.

That matters because the trials themselves were uneven. Only about 15% were rated low risk of bias, fewer than half clearly described how they randomized participants, and missing outcome data was the single most common weakness, affecting more than half the trials. The network meta-analysis is only as strong as the trials feeding it, and the authors are upfront about that.

What This Means If You Have Migraine

This is one study, and it does not replace a conversation with your own clinician. But it does give you a clearer map for that conversation. A few practical points worth keeping in mind:

• Prevention is a numbers game. Even the best classes here cut migraine days by roughly one and a half to two per month on average. Some people respond far better than the average, but it is reasonable to expect a meaningful reduction rather than a cure.
• Beta-blockers remain a sensible first step for many. They are effective, well tolerated in this analysis, inexpensive, and decades of experience back them up.
• CGRP-targeted therapies earned the strongest evidence rating. They are newer and often more expensive, but the high-certainty result for reducing migraine days is notable.
• If a drug is not tolerable, that is real information. Calcium channel blockers and anticonvulsants worked but saw more people drop out. Side effects are a legitimate reason to switch, not a failure on your part.
• Give a preventive a fair trial. Preventives usually take weeks to show their full effect, so stopping too early can look like failure when the drug simply needed more time.

Limitations to Keep in Mind

The analysis compared drug classes, so it cannot tell you which specific beta-blocker or which CGRP drug is best for you. It also reports averages across thousands of people, and migraine is famously individual. Two people on the same preventive can have very different experiences. Finally, because many of the underlying trials had design weaknesses, several findings sit at moderate rather than high certainty.

Key Takeaways

The big message is reassuring: several preventive options genuinely work, and the field now has a clearer ranking that weighs benefit against tolerability. If you are weighing whether to start a preventive or switch from one that is not working, this study gives you and your clinician solid ground to stand on. Bring it up at your next appointment.