Current issues of ACP Journal Club are published in Annals of Internal Medicine


Review: Anticonvulsants are better than placebo for reducing the frequency of migraine attacks


ACP J Club. 2005 Jan-Feb;142:21. doi:10.7326/ACPJC-2005-142-1-021

Clinical Impact Ratings

GIM/FP/GP: 5 stars

Neurology: 6 stars

Source Citation

Chronicle E, Mulleners W. Anticonvulsant drugs for migraine prophylaxis. Cochrane Database Syst Rev. 2004;(3):CD003226. [PubMed ID: 15266476]



In patients with migraine, are anticonvulsants more effective than placebo for preventing or reducing the intensity of migraine attacks?


Data sources: Studies were identified by searching MEDLINE (1966 to April 2003), the Cochrane Central Register of Controlled Trials (up to April 2003), reference lists of review articles, and books related to headache; contacting drug companies, authors of reports, and experts in the field; and hand-searching Headache and Cephalalgia.

Study selection and assessment: Studies were selected if they were randomized controlled trials (RCTs) that compared anticonvulsants given regularly during headache-free intervals in adults > 18 years of age with placebo, no intervention, other drug treatments, or behavioral or physical therapies. Methodological quality was assessed using the Jadad 5-point scale.

Outcomes: Headache frequency (number of migraine attacks measured at 28 d), headache index measures (frequency and intensity or duration), and adverse events. A ≥ 50% reduction in headache frequency or index was used to distinguish between treatment success and failure.

Main results

15 RCTs met the selection criteria. Anticonvulsants investigated were divalproex sodium (500 to 1500 mg daily [4 RCTs]), topiramate (50 to 200 mg daily [3 RCTs]), sodium valproate (800 to 1500 mg daily [2 RCTs]), gabapentin (1200 to 2400 mg daily [2 RCTs]), carbamazepine (dose not reported [1 RCT]), clonazepam (1 mg daily [1 RCT]), and lamotrigine (200 mg daily [1 RCT]). Mean duration of the maintenance phase of trials was 9.6 weeks (range 4 to 18 wk). The median quality score of studies was 4 (range 1 to 5). Compared with placebo, anticonvulsants as a class reduced mean migraine frequency by 1.4 attacks per 28 days (meta-analysis of 8 RCTs; weighted mean difference [WMD] −1.43, 95% CI −2.2 to −0.65) and increased the number of patients in whom migraine frequency is reduced by ≥ 50% (meta-analysis of 10 RCTs) (Table). For individual anticonvulsants compared with placebo, migraine frequency was reduced by sodium valproate (2 RCTs, WMD −4.31, CI −8.32 to −0.30), topiramate (3 RCTs, WMD −1.27, CI −1.74 to −0.79), and gabapentin (1 RCT, WMD −1.89, CI −2.35 to −1.43). The results of individual anticonvulsants for the number of patients in whom migraine frequency is reduced by ≥ 50% are in the Table. Lamotrigine or clonazepam did not differ from placebo for reduction of migraine frequency. The most common adverse events were nausea (number needed to harm [NNH] 6.6, CI 5.0 to 9.8), asthenia or fatigue (NNH 12.3, CI 7.6 to 31.8), tremor (NNH 12.4, CI 8.9 to 20.1), weight gain (NNH 16.0, CI 8.5 to 154.4), and dizziness or vertigo (NNH 16.3, CI 9.5 to 57.9).


In patients with migraine, anticonvulsants as a class are more effective than placebo for reducing the frequency of migraine attacks.

Source of funding: International Headache Society.

For correspondence: Dr. E. Chronicle, University of Hawaii at Manoa, Honolulu, HI, USA. E-mail

Table. Anticonvulsants vs placebo for number of patients who responded with ≥ 50% reduction in migraine frequency at mean 9.6 weeks*

Number of trials (n) Comparisons Weighted event rates RBI (95% CI) NNT (CI)
10 (1341) Anticonvulsants as a class vs placebo 49% vs 20% 71% (63 to 80) 4 (3 to 5)
4 (579) Divalproex sodium vs placebo 47% vs 21% 74% (57 to 91) 4 (3 to 12)
1 (68) Sodium valproate vs placebo 29% vs 18% 68% (47 to 89) 4 (2 to 10)
3 (498) Topiramate vs placebo 48% vs 22% 74% (66 to 82) 4 (3 to 6)

*Abbreviations defined in Glossary; weighted event rates, RBI, NNT, and CI calculated from data in article using a random-effects model.


The review by Chronicle and colleagues supports the use of anticonvulsant medications to prevent migraine attacks. However, it is unclear how to incorporate antiepileptic medication into the care of patients with frequent migraine attacks. Few studies have compared medications directly, and none have combined medications for prophylaxis.

Any meta-analysis risks combining studies that are too heterogeneous to be “lumped” together. The underlying assumption of this Cochrane review is that all antiepileptic medications can logically be considered together as a class. However, the mechanisms of action of antiepileptic drugs are either typically unclear or vary if known. In addition, the studies described different rules for administering rescue medication.

Many patients in studies come from such specialized trial settings as headache or neurology clinics. Patients presenting to these centers might have headaches that are more difficult to treat. However, this bias should strengthen any findings that support antiepileptic treatment.

Although the benefit of anticonvulsant drugs appears worth the risk, cost may be prohibitive (e.g., topiramate is much more expensive than propranolol), and the rates of side effects and discontinuation of medication were high.

Antiepileptic medications have shown similar results to other classes of drugs commonly used for prophylaxis in reducing the frequency and intensity of migraines by about 40% (1). Only 2 studies compared antiepileptic medications with other medications used for such migraine prophylaxis as propranolol or amitryptiline (2). It would be interesting for future research to investigate the synergistic effect of combining 2 medications (e.g., β-blockers and antiepileptic drugs) to improve outcomes.

Alexander W. Chessman, MD
Medical University of South Carolina
Charleston, South Carolina, USA


1. Brandes JL, Saper JR, Diamond M, et al. Topiramate for migraine prevention: a randomized controlled trial. JAMA. 2004;291:965-73. [PubMed ID: 14982912]

2. Linde K, Rossnagel K. Propranolol for migraine prophylaxis. Cochrane Database Syst Rev. 2004(2):CD003225. [PubMed ID: 15106196]