Skip navigation

Sodium valproate for migraine prevention

Clinical bottom line

Sodium valproate is effective for the prevention of migraine attacks. About half of the patients had a reduction in the number of migraine attacks or days with migraine by about 50%. Numbers-needed-to-treat for at least 50% reduction in migraine frequency for individual studies were three to four. The incidence of adverse effects was not insignificant, with nausea, dizziness and drowsiness associated with its use. Women of childbearing age should use the drug with caution due to the possibility of birth defects.

The publication of some new RCTs of the anticonvulsant sodium valproate for migraine prophylaxis prompted Bandolier to pull the evidence together for this drug. We have previously looked at the evidence for anticonvulsants and neuropathic pain which showed the anticonvulsants, carbamazepine, clonazepam and sodium valproate were effective for migraine prevention, though there was only one trial of each drug.

Systematic review

Inclusion criteria were randomised, double blind trials of sodium valproate for migraine prevention in adults.

Search methods

Comprehensive searches of the following databases were made: MEDLINE (1966 - Sept 2000), EMBASE (1980 - Sept 2000), Cochrane Library (Issue 3, 2000) and the Oxford Pain Relief Database (1950 - 1994)[1] and other reviews [2]. A series of free text searches were undertaken, using generic and trade names for sodium valproate. There was no restriction to language. Retrieved reports were searched for additional trials. Neither individual authors nor pharmaceutical companies were contacted for unpublished data.


The outcome most often quoted was the number of patients with at least 50% reduction in their migraine headache fequency over four weeks at the end of 12 weeks treatment. Adverse effect withdrawal and the number of people having particular adverse effects was also noted.


Five randomised, double blind trials were found, all were placebo controlled [3-7]. One trial was excluded as it was a duplicate of Hering and Kuritzky, 1992 [4]. Table with individual trial details is here.

Doses of valproate were varied, often were titrated, and often were titrated against plasma concentration. The lowest starting dose was 250 mg/day and the highest ending dose was 1500 mg/day. Patients at baseline had at least two migraine attacks per month, usually for at least six months. The average number of attacks per month at baseline was six. Migraine was diagnosed using International Headache Society criteria in three studies, and using very similar criteria in the other.

Trials were of higher methodological quality, with quality scores of 3 and 4 out of 5. Trials were disparate with respect to dosing regimes and outcome measures but pooling of data for quantitative analysis was still possible, as all measured at least 50% reduction in migraine attacks over about 12 weeks. Dichotomous data were used to calculate NNTs for individual studies.

There were three RCTs [5-7] with outcomes of at least 50% reduction in migraine attacks. All three trials used IHS criteria for migraine diagnosis, baseline values of migraine frequency were established during a four week run-in period. Twelve week treatment periods were followed by four week washout in the two crossover trials before the second phase of treatment. All three trials showed that between 44 to 50% patients achieved at least 50% reduction in the frequency of migraine days or attacks on sodium valproate compared with 14 to 21% on placebo (Figure 1). The NNTs for individual studies were three to four (Table 1).

Table 1: Numbers-needed-to-treat for one patient to achieve at least 50% reduction in migraine frequency (attacks/days) over four weeks on sodium valproate versus placebo

Study Sodium valproate dose Number of patients with at least 50% reduction in migraine frequency on valproate Number of patients with at least 50% reduction in migraine frequency on placebo NNT (95% CI)
Jensen et al 1994 1000 to 1500 mg/day, plasma concentration around 50 mg/ L 17/34 6/34 3.1 (1.9 to 8.9)
Mathew et al 1995 Dose titrated to give trough plasma valproate levels 70 to 120 mg/ L 33/69 5/36 3.0 (2.0 to 5.7)
Klapper et al 1997 500 to 1500 mg/day 58/132 9/44 4.3 (2.6 to 11)
Combined 108/235 20/114 3.5 (2.6 to 5.3)

Combining all three studies, 108/235 patients (46%) had more than 50% reduction in migraine frequency with valproate, compared with 20/114 (18%) with placebo. The pooled relative benefit was 2.6 (1.7 to 4.0) and the pooled NNT was 3.5 (2.6 to 5.3).

The earlier study [4] was not as methodologically rigorous in design as it lacked a washout period between crossover from one treatment to the other, and did not report baseline values of migraine frequency. Sodium valproate was more effective than placebo in reducing the frequency of migraine attacks by almost 50% (Figure 2).

Adverse effects

Information on adverse effects was extracted from each trial (Table 2). There were significantly more withdrawals due to sodium valproate than placebo with a number needed to harm (NNH) of 9.4 (6.0 to 21). Particular adverse effects were nausea, tremor, dizziness and drowsiness with NNHs of 3.1, 6.2, 6.5 and 6.3, respectively. Asthenia, weight gain and hair loss also occurred with sodium valproate but were not significantly more frequent than with placebo.

Table 2: Numbers-needed-to-harm for adverse effects on sodium valproate versus placebo

Adverse effect category Number of studies Number with AE with valproate Number with AE with placebo Relative risk (95% CI) Number-needed-to-harm (95% CI)
Any adverse effect 3 127/245 44/124 1.4 (0.69 to 3.0) nc
Withdrawals from adverse effects 3 38/245 6/124 3.1 (2.1 to 4.7) 9.4 (6.0 to 21)
Particular adverse effects
Tremor 3 33/220 0/124 47 (10 to 207) 6.2 (4.7 to 9.1)
Nausea 3 107/245 14/124 3.3 (2.0 to 5.4) 3.3 (2.5 to 4.2)
Dizziness/vertigo 2 37/175 5/87 2.6 (1.1 to 6.1) 6.5 (4.3 to 13)
Drowsiness 3 58/245 9/114 2.6 (1.4 to 5.2) 6.3 (4.3 to 12)



We know that anticonvulsants for neuropathic pain are effective [8] so it is no surprise that sodium valproate is so effective at preventing migraines. The adverse effects associated with its use may be a concern for some patients. Newer anticonvulsants coming onto the market with less adverse effects may present better treatment options. So far there are no RCTs to support the use of drugs like Gabapentin for migraine prevention, but Bandolier will keep a keen eye out for them.


  1. Jadad AR, Carroll D, Moore A, McQuay H. Developing a database of published reports of randomised clinical trials in pain research. Pain 1996; 66:239-46.
  2. Tran BN, Vivian VS, Burch KJ. Can valproate prevent migraine headaches. Journal of Pharmacy Technology. 1997; 13:163-8.
  3. Gordon CR, Zuritzky A, Doweck Iea. Vestibulo-ocular reflex in migraine patients: the effect of sodium valproate. Headache 1993; 33:129-32.
  4. Hering R, Kuritzky A. Sodium valproate in the prophylactic treatment of migraine: a double-blind study versus placebo. Cephalalgia 1992; 12:81-4.
  5. Jensen R, Brinck T, Olesen J. Sodium valproate has a prophylactic effect in migrain without aura: a triple-blind, placebo-controlled crossover study. Neurology 1994; 44:647-51.
  6. Klapper J. Divalproex sodium in migraine prophylaxis: a dose-controlled study. Cephalalgia 1997; 17:103-8.
  7. Mathew NT, Saper JR, Silberstein SDea. Migraine prophylaxis with divalproex. Arch Neurol 1995; 52:281-6.
  8. McQuay H, Carroll D, Jadad AR, Wiffen P, Moore A. Anticonvulsant drugs for management of pain: a systematic review. BMJ 1995; 311:1047-52.