Objective: A retrospective study was undertaken to examine predictors of weight loss in children with attention deficit hyperactivity disorder (ADHD) treated with stimulant medication.
Design: Children diagnosed with ADHD and treated with methylphenidate hydrochloride (MPH) or dextroamphetamine sulfate (DEX) for at least 5 months were identified by retrospective chart review. Analyses were performed on 32 children (29 boys) treated with MPH and 28 children (25 boys) treated with DEX. Variables examined included pretreatment weight, pretreatment body mass index (BMI), duration of treatment, total daily dose, and weight-adjusted dose.
Results: The MPH-treated group was initially seen at a mean age (+/- standard deviation) of 7.5 years (+/- 2.1). Mean duration of follow-up was 11.2 months (+/- 3.8). Mean total daily dose was 25.5 mg (+/- 8.2) and mean weight-adjusted dose was 1.0 mg/kg/day (+/- .5). Mean change in weight as expressed in z scores was -0.4 SD (+/- .7). The DEX-treated group was initially seen at a mean age of 9.3 years (+/- 2.7). Mean duration of follow-up was 10.8 months (+/- 4.1). Mean total daily dose was 14.9 mg (+/- 8.2) and mean weight-adjusted dose was .5 mg/kg/day (+/- .3). Mean change in weight was -0.6 SD (+/- .6). Although the MPH group was younger than the DEX group, there were no between-group differences in weight or height either before treatment or at follow-up when adjusted for age and gender. Using linear regression analysis, significant correlations were found between pretreatment weight and change in weight for both the MPH-treated group (r = .63) and the DEX-treated group (r = .47). Stepwise multiple-regression analysis indicated that pretreatment age, weight-adjusted dose, and duration of follow-up did not significantly contribute to the variance in change of weight for either the MPH- or DEX-treated groups. To adjust for height, changes in BMI were also analyzed. For both medication groups, no differences were found between the heavier (BMI > or = 50%) and thinner (BMI < 50%) children in duration of follow-up, total daily dose, or weight-adjusted dose. When the medication groups were combined, a greater proportion of heavier children experienced a decrease relative to their predicted BMI compared with thinner children (80% vs 52%). Comparisons of mean BMI slopes revealed that the heavier group experienced an absolute decrease in their BMI (-.139) in contrast to the thinner group (+.014) (t = 2.70).
Conclusions: Pretreatment weight, adjusted for age, gender, and height, is a significant predictor of weight loss in children with ADHD treated with either MPH or DEX. In contrast, pretreatment age, duration of treatment, and weight-adjusted dose were not found to be significant predictors.