Objective: The aim of this study was to examine whether demographic or pretreatment clinical and social characteristics influenced the response to methylphenidate (MPH) in the Preschoolers with ADHD Treatment Study (PATS).
Methods: Exploratory moderator analyses were conducted on the efficacy data from the PATS 5-week, double-blind, placebo-controlled six-site titration trial. Children (N = 165, age 3-5.5 years) were randomized to 1 week each of four MPH doses (1.25, 2.5, 5, and 7.5 mg) and placebo administered three times per day (t.i.d.). We assessed the fixed effects on the average slope in the regression outcome on moderators, weight-adjusted dose, and the moderator-by-dose interaction using SAS PROC GENMOD.
Results: A significant interaction effect was found for a number of co-morbid disorders diagnosed in the preschoolers at baseline (p = 0.005). Preschoolers with three or more co-morbid disorders did not respond to MPH (Cohen's d at 7.5 mg dose relative to placebo = -0.37) compared to a significant response in the preschoolers with 0, 1, or 2 co-morbid disorders (Cohen's d = 0.89, 1.00, and 0.56, respectively). Preschoolers with more co-morbidity were found to have more family adversity. No significant interaction effect was found with the other variables.
Conclusions: In preschoolers with ADHD, the presence of no or one co-morbid disorder (primarily oppositional defiant disorder) predicted a large treatment response at the same level as has been found in school-aged children, and two co-morbid disorders predicted moderate treatment response; whereas the presence of three or more co-morbid disorders predicted no treatment response to MPH.