Valproate for Diabetic Painful Neuropathy

Abstract & Commentary

Synopsis: Sodium valproate appears safe and effective for the management of DPSN.

Source: Kochar DK, et al. Acta Neurol Scand. 2002;106: 248-252.

Diabetic painful sensory neuropathy (DPSN) remains difficult to treat—witness the persistence of ongoing and upcoming clinical trials searching for the magic bullet and the large number of medication options available (none work superbly in all cases). In this randomized, double-blind, placebo-controlled trial, sodium valproate, 400 mg p.o. t.i.d., was administered to 30 patients with efficacy and safety compared to 30 placebo-controlled patients. All 60 patients were demographically matched with documented DPSN, and they were excluded if they demonstrated hepatic disease, TB, or other causes of neuropathy including uremia, vitamin deficiency, paraneoplastic or hereditary neuropathy, or alcoholism. Detailed clinical and neurological examinations were performed and the short-form McGill pain questionnaire was used to quantify pain severity at study entry, week 1, and after 4 weeks of therapy. Motor and sensory nerve conduction studies were performed at study onset and after 4 weeks. Student’s t-test provided statistical analysis.

Among 28 patients who completed 4 weeks of the active drug arm, 20 reported pain relief compared to 5 of 24 in the placebo arm. Pain severity was significantly decreased (P < 0.05) after 4 weeks on active therapy, though no significant change was appreciated after 1 week. Electrodiagnostic studies did not improve during this short study. Only 1 patient was withdrawn due to abnormal liver function tests while on valproate. Among the other patients who did not complete the study, 3 were for noncompliance (1 active arm, 2 placebo arm) and 2 due to lack of effect (placebo arm). Sodium valproate appears safe and effective for the management of DPSN.

Comment by Michael Rubin, MD

Why do some diabetics develop painful sensory neuropathy while in others the neuropathy is painless? Ten painful sensory neuropathy diabetics were matched with 10 diabetics with nonpainful sensory neuropathy for age, diabetes duration, insulin regimen, duration of neuropathy, and HbA1c. All were fitted with a continuous glucose-monitoring system (MiniMed Inc., Sylmar, Calif, US) for 3 days to determine if a relationship existed between glucose fluctuations and painful neuropathy. Patients with other causes for foot pain, including arterial disease, skin ulcers, or arthritis were excluded. Daily pain scores were recorded by the patients on a horizontal 10-cm scale. Analgesics were allowed as needed. Glucose excursions were measured as a mean amplitude over 24 hours, and M-values, a quantitative measure of blood glucose deviations over a specified time period, were calculated.1 Spearman’s rank correlation coefficient and the Mann-Whitney U-test provided statistical analysis.

Frequency of glucose excursions, mean glucose value, and mean M-value were significantly greater in the painful neuropathy group compared to the painless group. However, the mean amplitude of glucose excursion did not differ between groups and no correlation was appreciated between episodes of pain and number or amplitude of glucose excursions. Greater glucose flux is associated with painful, rather than painless, neuropathy in diabetics.

Dr. Rubin is Professor of Clinical Neurology, New York Presbyterian Hospital—Cornell Campus.


1. Service FJ, et al. Diabetes. 1970;19:644-655.