Hyperactive Boys Almost Grown Up

Abstract

\s=b\We report a prospective longitudinal study of 101 male adolescents (ages 16 to 23 years) who had been diagnosed hyperactive in childhood (ages 6 to 12 years), compared with 100 normal controls. The DSM-III diagnoses were made blind to group membership. Information was obtained for 98% of the original cohort. The full attention deficit disorder with hyper-activity (ADDH) syndrome persisted in 31% of the probands vs in 3% of the controls. The only other two conditions that distinguished the groups significantly were conduct and substance use disorders. These disorders aggregated significantly among the probands with continued ADDH. The results indicate that the greatest risk factor for the development of antisocial behavior and drug abuse is the maintenance of ADDH symptoms. Substance use disorders followed the onset of conduct disorder in the overwhelming majority of the cases. (Arch Gen Psychiatry 1985;42:937-947) The follow-up of any psychopathologic entity of childhood is of importance because knowledge of later status fosters understanding of the significance of the early condi¬ tion. In turn, this knowledge may provide refinements in the nosologie classification that ultimately may facilitate the search for discrete causes. Among child psychiatric conditions, the outcome of hyperactive children has drawn considerable attention because, on the one hand, they are common among clinic attendees, and on the other, they are believed to contribute to the pool of antisocial adults.1 Antisocial, criminal behavior is of great public concern, and the identification of at-risk groups is essential for its prevention. Therefore, the study of the eventual fate of hyperactive children has important scientific and possibly public health goals. (The term hyperactive used throughout this article refers to children diagnosed in DSM-III as having attention deficit disorder [ADD] with hyperactivity [ADDH].) A number of investigators have reported on the later status of hyperactive children.216 Table 1 summarizes the studies and their designs. Very few have met the de¬ sideratum of being prospective in nature. Reliance on retrospective childhood diagnoses is problematic and ren¬ ders replication difficult, as the information available from clinical chart summaries varies across patients, as well as across treatment centers. Unless one is identifying clear-cut behaviors, such as court referral for antisocial acts, or low IQ, it is difficult to have confidence in the exact nature of the original condition. Therefore, the childhood psychiatric status of cases studied retrospectively is controversial. Furthermore, with one exception,6 cases in all studies have been evaluated with full knowledge of their early status. The degree to which this possible bias may have colored judgments during later assessment cannot be deter¬ mined. Another limiting factor in the published literature is that many studies had a large proportion of dropouts, so that the residual group studied often represents an insufficient proportion of the original cohort to engender confidence about the accuracy of the findings for the group as a whole. In several instances2·3'9·11·1316 the age at follow-up is early, 13 or 14 years. As the vast majority of hyperactive children are referred during elementary school years, the modal age being 9 years or so, status at 13 years does not provide information relevant to outcome much beyond the peak of the age-risk period for diagnosis. An early, unsatisfactory study12 provided a grim picture of the outcome of hyperactive children, with many being institutionalized. The most striking finding of another early study8 was the need for special education in 76% of the cases. In addition, 30% had been in trouble with the police; however, none had been in jail, and none reported any habitual drug use. Mendelson et al11 reported that close to 20% of a young adolescent sample had appeared in juvenile court, and over 70% of the mothers reported that their child was more active, restless, and impulsive than normal peers. Similar results were reported by Borland and Heckman,4 who compared cases, diagnosed retrospectively, with their siblings. Loney et al9 have related a number of early characteristics