Predictors of stability of attention-deficit/hyperactivity disorder subtypes from childhood to young adulthood.

OBJECTIVE To determine the 5-year prospective stability of population-based and DSM-IV subtypes of attention-deficit/hyperactivity disorder (ADHD) as well as to explore predictors of stability. METHOD A total of 708 twins ages 7 to 19 years who were identified from birth records of the state of Missouri and had participated in a study of ADHD were reassessed 5 years later in a blinded fashion. Stabilities of DSM-IV and population-based ADHD subtypes were compared using percentage of agreement with significance tested by the kappa statistic. Predictors of stability of subtype diagnosis were determined using multivariate logistic regression. RESULTS In general, 5-year ADHD subtype stability was poor to modest and ranged from 11.1% to 24.0% for DSM-IV for subtypes and from 14.3% to 35.3% for clinically significant population-derived subtypes. There were no predictors of diagnostic stability that applied across subtypes. There were subtype-specific predictors including a diagnosis of oppositional defiant disorder for DSM-IV primarily inattentive ADHD; lower verbal IQ for DSM-IV combined type ADHD; and younger age, oppositional defiant disorder, and medication use for population-defined severe combined ADHD. CONCLUSIONS Population-defined ADHD subtype criteria demonstrated modestly improved diagnostic stability over 5 years compared to DSM-IV subtypes. Few correlates or predictors of stability were identified.