In a sample of males diagnosed with ADHD in childhood followed up for ten years, prior treatment with stimulants was associated with a subsequently decreased risk for depressive, disruptive, and anxiety disorders, and for grade retention. These findings suggest that treatment of ADHD with stimulants has protective effects against the development of some psychiatric and functional outcomes.
Our study confirms the results of Daviss et al (12
), who also found a protective effect of pharmacotherapy against subsequent MD in ADHD youth. However, our results extend this finding by documenting the protective effect while statistically adjusting for parental lifetime history of MD, as well as baseline severity of ADHD.
Our results failed to detect an association between stimulant treatment and the risk for BD. Similarly, published chart reviews of stimulant treatment in patients with BD have also been inconclusive (20
). Given the large and bi-directional comorbidity between ADHD and BD, and the standing of stimulants as the first-line therapeutic approach for ADHD, additional research into the effect of stimulant therapy on the subsequent risk of BD is warranted.
Our results also suggest that stimulant therapy, in addition to exerting a therapeutic effect in the short-term (17
), may also reduce the risk for CD and ODD across much longer time intervals. However, our study was not consistent with a report by Molina et al (39
), which found a positive association between past year treatment and delinquency in a large sample of children with ADHD. Two factors may explain this discrepancy. First, our study examined the onset of CD, while Molina et al examined delinquency; in the same study (39
), the proportion of subjects with delinquency who also had CD was only 33%. Second, the Molina et al study examined the use of any prescription medication for ADHD, while our study focused on stimulants specifically.
Our study disagrees with prior work suggesting that stimulant treatment does not protect against grade retention (28
). Several factors may have contributed to this discrepancy. For example, the sample size of Powers et al was smaller (n=90) than that tested in this study (n=122 for the repeated grade model), providing relatively less statistical power in the former study. Secondly, the mean age of the subjects from the Powers et al study was younger at follow-up (mean age in years of 18.2). Thus, the subjects from the Powers et al study may not have completely aged through the period of risk for grade retention.
What explains this inverse association between stimulant treatment and adverse psychiatric and educational outcomes? Efficacious stimulant therapy may interrupt the pathogenic trajectory leading toward other disorders. For example, a child under efficacious ADHD treatment may experience improved self-esteem and behavioral compliance in school, attendant improvement in his standing with teachers as well as association with a normative peer group. Over time, these detours may provide an alternate developmental trajectory for this child, ultimately resulting in decreased risk for major depression and disruptive behavior disorders. However, this explanation assumes continued efficacy of stimulant therapy across time, which may not be tenable (40
). Alternatively, our findings may be due to unmeasured confounding factors that predict which children with ADHD are provided treatment, and also are predictive of psychopathology, such that children who are unlikely to receive treatment for ADHD are also more likely to develop comorbid disorders, regardless of treatment.
These results must be considered in the light of methodological limitations. Our sample was originally ascertained according to DSM-III-R criteria, and our results may not generalize to samples ascertained by DSM-IV. However, considering the high overlap between the two definitions (93% of DSM-III-R cases received a DSM-IV diagnosis (41
)), any effect should be minimal. Because our sample consisted of referred Caucasian boys, we do not know if our results will generalize to ADHD children in the general population, or to other racial groups, or to females. Although our study was prospective, we relied on retrospectively (i.e. within the intervals between assessments) reported ages of onset for treatment and comorbid disorders to establish the temporal sequence. Thus, our results may suffer from misclassification (and thus a reduction in precision) to the degree that these ages were incorrectly recalled. However, while the exact ages may not have been recalled accurately, the relative ordering is likely to be correct, so that any misclassification of our exposure and outcome variables should be minimal. Also, we cannot address in these data the reasons why subjects were treated with stimulant versus non-stimulant medications. Finally, our naturalistic study design cannot provide the more informative evidence that would be produced by a randomized, controlled study.