Overall, the results of the present study show that the long-term outcomes for participants with ADHD when left untreated were poor compared with non-ADHD controls, and that treatment of ADHD improved long-term outcomes, but usually not to the point of normalization. The outcomes that were studied (with ADHD symptoms deliberately excluded as an outcome) most often included drug use/addictive behavior, academic, and antisocial behavior. This was followed by social function, self-esteem, occupation, driving, services use, and obesity outcomes. These trends may reflect what is of most immediate interest to society in a given time period. For example, obesity, the least-studied outcome, has come into interest only recently, likely due to the increasing obesity epidemic in developed countries. Increasing interest in the epidemiology of obesity, led to the report of an association between obesity and ADHD in 2002 . Our data also indicate that there are specific geographical trends, with academic outcomes being of greater interest for study in the US and Canada and antisocial behaviors of greater interest in Europe. This difference of interest may be a function of only more severe cases, likely to have oppositional-defiant or conduct disorder comorbidity, being diagnosed outside Northern America. These trends have been described in more detail in a separate publication .
The number of studies of long-term outcomes of ADHD has risen noticeably over the last 30 years, especially since 2000. This corresponds to a trend in awareness of the consequences of ADHD by clinicians, which appears to be on the rise .
Treatment resulted in beneficial effects for many of the outcomes reported (72% of outcome results). These beneficial effects were observed as either significant improvement over pretreatment baseline, in comparison to untreated ADHD participants, or stabilization of the outcomes (that is, prevention of the deterioration over time from baseline reported with untreated ADHD [43–46]). Driving and obesity outcomes were the most often reported to be responsive to treatment. Of course, a decrease in obesity may be due to an appetite suppressant effect of stimulants and atomoxetine. The relatively small number of studies of these two outcomes (two studies each) comparing treated with untreated ADHD and the consistently positive response to treatment support further investigation in these areas. Three other outcomes that were often reported to be responsive to treatment were self-esteem, social function, and academic outcomes. These results are supported by a relatively large number of studies (10, 12, and 21 studies, respectively) comparing participants with treated ADHD with participants with untreated ADHD. These outcomes may be more closely related to symptom relief. The outcomes reported are not independent of one another and changes in one may reflect changes in others. The wider effects of response to treatment in these two areas may warrant further investigation.
The four remaining outcomes that appeared to be least responsive to treatment were drug use/addictive behavior, antisocial behavior, services use, and occupation, with 67%, 50%, 50%, and 33% of reported outcome results demonstrating a benefit of treatment, respectively. Persistence in these cases may have to do with the existence of comorbidities, such as conduct disorder, which has for example, been associated with increased substance use disorders . Oppositional defiant disorder or conduct disorder, may contribute to long-term outcomes in people with ADHD, specifically crime and substance use [54–56], and thus may affect the response observed for these outcomes with treatment for ADHD. Other comorbidities, such as depression, obsessive-compulsive disorder, or autism may have similarly influenced the results we observed.
Services use may persist due to the incomplete amelioration of ADHD symptoms and impairments, possibly because although ADHD symptoms (like outcomes) respond to treatment, they are not completely normalized. Finally, continued impairment in occupation despite treatment may reflect the cumulative effects of ADHD symptoms and dysfunctioning over the lifespan. For example, low academic grades may later restrict employment or opportunities, impaired social function may precipitate extra friction with employers. The differential responsiveness of different outcomes to treatment is an intriguing area for future study.
It should be mentioned that if we had categorized the study outcomes by age, such as 5 to 17 and over 18, we may have observed different areas of improvement depending on the age group. In combining the groups it is possible that this distinction is lost. One also needs to consider however, that certain categories such as occupation would not be as relevant to the 5 to 17 age group as opposed to academic achievement, which would apply to all groups.
Even with treatment, worse outcomes were often reported for the ADHD group than for people without ADHD. This is not surprising, because although behavioral and drug treatment have been demonstrated to improve ADHD symptoms, these treatments do not necessarily normalize behavior to control levels [35, 57–61].
For example, in one study of the effect of methylphenidate treatment on classroom measures, a 20 mg dose produced normalization in 30% to 60% of participants, (depending on the measurement used) although 53% to 94% showed improvement . Many studies (42) in our analysis evaluated treatment effects only against non-ADHD controls, as opposed to pretreatment baseline or any untreated state. In these studies, only 24% of outcomes were reported to be similar for treated ADHD and non-ADHD controls. For all the other outcomes reported (76%), the outcomes remained worse than non-ADHD controls, and there was no mechanism with this study design by which to measure improvement with treatment that did not completely 'normalize' the outcome. In studies with other study designs (comparing participants with treated ADHD and participants with untreated ADHD or pretreatment baseline), benefit with treatment was reported for 72% of the outcomes. These study designs allowed the improvement with treatment to be demonstrated, even though the outcome may not have 'normalized'. When considering the effects of treatment reported in any one study, the comparator group used to evaluate the effectiveness of ADHD treatment is particularly important.
The results of four studies included in this analysis that used both types of comparators within the same study ((1) non-ADHD controls and (2) untreated ADHD participants or ADHD participant's pretreatment baseline) were consistent with the present overall observations that there was clear improvement or stabilization with treatment of ADHD for social function, antisocial behavior, and academic outcomes, but not to the extent that non-ADHD control outcomes were matched [13, 39, 48, 62]. This general pattern was noticed in the earliest of the four studies, as the authors conclude in their 12-year follow-up study that 'The most striking finding of the study is the repetitive pattern of finding significant differences between the stimulant-treated hyperactives and their control group (with the control group almost invariably doing better). However, there are several areas in which the stimulant-treated hyperactives seem to do better than their untreated counterparts' (referring to academic, driving, self-esteem, and social function outcomes) . This pattern was also found in an 8-year follow-up study of a different sample 'despite overall maintenance of improvement in functioning relative to baseline (pretreatment), the MTA group as a whole was functioning significantly less well than the non-ADHD classmate sample' . In the present study, this pattern was broadly replicated across the outcome groups when analyzed individually. Comparisons against non-ADHD controls only, may mask improvements with treatment.
It should be noted that ADHD Rating Scale IV total scores decline (improve) between ages 5 to 7 and ages 14 to 18 in both Caucasians and African-Americans (although in Latinos they actually increase during these time periods). Hyperactivity-impulsivity scores, especially, decline from ages 4 to 7 to age 14 and older in both boys and girls . Therefore, we cannot discount that natural decline in symptoms (and possibly other dysfunctions) occurring over time also contributes to the improvement in outcomes observed with treatment, although this idea contradicts the reported worsening of functional domains in untreated ADHD. According to one paper, symptoms and functioning are related. With full symptom remission, illicit drug use and antisocial behaviors become similar to non-ADHD controls, but while social function improves, it does not reach non-ADHD levels .
Finally, we observed that treatment outcomes for the subgroup of domains that exhibit lower percentages of outcomes that benefit from treatment (drug use/addictive behavior, antisocial behavior, services use, and occupation) were differentially improved when studies from Northern America were compared with those from the rest of the world (in this case, all 'Rest of World' studies were from Europe). Based on this analysis there appears to be a geographical bias with regard to how responsive these four outcome group results are reported to be with treatment, a result that may be accounted for by regional differences in study design along with the resulting study population age, or diagnostic practices. In this very specific comparison, the numbers of studies from 'Rest of World' countries are low (four studies, seven outcome results), thus as further investigation of these outcomes around the world are published, the results of this comparison may be clarified.
Limitations and possible sources of bias
Several possible risks of bias and limitations need to be considered regarding the included studies. First, a publication and cultural bias could have resulted from including only studies that were published in English. In addition, the analysis excluded unpublished studies that might have been presented at conferences, for example. Also, our search relied on search engines for 'peer reviewed' status. Moreover, by strictly adhering to Cochrane systematic review guidelines and only including studies that were identified in our original electronic search, it is possible that some relevant studies may have been missed, introducing a search engine and literature database bias. This bias was reduced by extensive searching of 12 databases. Nonetheless, we are aware of four studies that would have met inclusion criteria, but were not identified by the search engines due to a technical limitation or inadvertent search string exclusion. Examination of these studies shows that the reported results are consistent with the overall results of the present analysis. The results of these four studies are summarized briefly here. One study reported poorer outcomes for participants with ADHD versus non-ADHD controls with regard to academic achievement, occupational adjustment, antisocial behavior, relationships, and substance use . A second study found a high incidence of ADHD (65%) in 23 adolescents who attempted suicide . A third study reported that by young adulthood, participants with ADHD were similar to non-ADHD controls in minor aspects of social and occupational outcomes (e.g., time socializing with friends and with hobbies), but had poorer outcomes in major aspects of these outcomes (e.g., had many more offspring and most were not living with them) . A fourth study found that stimulant treatment in children with ADHD significantly improved reading scores and decreased grade retention . The results of these studies are consistent with our overall finding that untreated ADHD is associated with poor long-term outcomes and that these outcomes improve with treatment.
Researcher bias could also be a possible source of bias in this analysis, however, this was reduced by having two researchers independently agree on the articles included and strict, simple inclusion criteria were established prior to searching.
Other sources of bias could include biases of individual research groups, which was eliminated by including only electronically identified studies (as mentioned above) and not selectively including the studies of specific groups and omitting others. As observed, different study designs may also lead to different conclusions and taking comparators into consideration is critical. We included studies of various designs, which may minimize such bias.
A further possible bias could arise from changes over time in diagnostic criteria or discrepancy between classification systems, specifically differences in the definition of hyperkinetic disorder (ICD-9 or ICD-10) versus ADHD (DSM-III/DSM-III-R/DSM-IV). Differences are less likely within classification systems . One study found that 93% of children diagnosed with ADHD using DSM-III-R diagnosis also received a DSM-IV ADHD diagnosis, indicating good correspondence between classification systems . Rediagnosis of the MTA sample by ICD criteria, however, resulted in only 25% of the DSM-IV-diagnosed MTA sample of combined-type ADHD qualifying as having hyperkinetic disorder or hyperkinetic conduct disorder by ICD-10 criteria .