Valproic acid-associated weight gain in older children and teens with epilepsy.

This research was undertaken to determine the incidence and predictors of weight gain in older children and teens treated with valproate for epilepsy. Subjects who were 10 to 17 years of age, who began valproate treatment between January 1, 1996, and December 31, 2000, and who had documented weight and height measurements at medication initiation and at least one follow-up visit were retrospectively identified. Exclusion criteria were follow-up <2 months, discontinuation of valproate within 2 months, concurrent therapy with medication known to affect weight, or gastrostomy feeding. Body mass index was calculated at initiation and either discontinuation of valproate or last follow-up and stratified into four categories: underweight, (<fifth percentile), appropriate (fifth to eighty-fifth percentile), potentially overweight (>/=eighty-fifth to ninety-fifth percentile) and overweight (ninety-fifth percentile or higher). Potential predictors of change in body mass index were examined. Mild-to-moderate weight gain was observed in 58% of the 43 subjects treated with valproate (median increase, 2; twenty-fifth to seventy-fifth percentile, 1-6.4). Seventy-nine percent remained in the same body mass index category, and 14% moved up to a potentially overweight or overweight category. The only predictor of an overweight category at follow-up was a potentially overweight or overweight category at initiation (P <0.0002). Two factors tended to predict an increase in body mass index: normal neurocognitive status (P = 0.06) and primary generalized seizure type (P = 0.07).