Recently, an increasing number of adolescents and young adults have presented to clinics with ADHD symptoms, requesting stimulant medication. It is not certain whether this is due to an increase in individuals searching for cognitive-enhancing medication or whether this represents the development of late-onset ADHD that requires treatment. Birth cohort studies report a 2.5–10.7% prevalence of late-onset ADHD that first emerges in adolescence or adulthood, with the majority of adults with ADHD (67.5–90.0%) not experiencing symptoms in childhood. These data contrast with the concept of ADHD being a chronic neurodevelopmental disorder that appears before the age of 12 years. Therefore, critics speculate that the high prevalence of late-onset ADHD may be due to other factors such as methodological artefacts, reliance on ADHD screening instruments, inability to recognise symptoms between assessments, false positives, failure to consider other mental health disorders and undetected childhood symptoms of ADHD. Using data from the Multimodal Treatment of ADHD study,* this article retrospectively assessed the proportion and clinical profile of individuals with ADHD symptoms with the aim to determine the true prevalence of late-onset ADHD.
Participants† (n=239; mean age 9.89 years at baseline; mean age 24.4 years at final adult assessment) who did not meet the criteria for ADHD during childhood were enrolled in the study. Study assessments included measures of ADHD symptoms (using the Swanson, Nolan and Pelham Rating Scale in childhood and the Conners’ Adult ADHD Rating Scale in adulthood‡), impairments (using the Columbia Impairment Scale§), substance use (using the Diagnostic Interview Schedule for Children [DISC¥] and Substance Use Questionnaire, with self-report of more than twice a week classified as heavy substance use) and mental disorders (using the DISC).
Results showed that 96/239 (40.2%) participants met the Diagnostic and Statistical Manual of Mental Disorders – 5th Edition (DSM-5TM) symptom threshold for ADHD based on combined parent-, teacher- and self-reports. Of these participants, 32/96 (33.0%) met the criteria for ADHD onset in adolescence with clinically significant impairment; however, as 11 of these cases first met DSM-5TM symptom criteria at ≤12 years of age, these cases were considered childhood-onset ADHD. Of the remaining 21 participants who qualified for adolescent-onset ADHD, 3 had symptoms attributed to heavy substance use and 9 had a history of pre-existing or concurrent mental disorders. Therefore, a total of 13 participants were identified as having onset of ADHD symptoms and impairment in adolescence, with 6 of these participants (2.5%) having their symptoms reported by a teacher. Among these 6 participants, the mean age of adolescent-onset ADHD was 14.22 years, with 4 participants exhibiting symptoms during teenage years and 2 exhibiting symptoms until their 20s.
Of the 239 participants enrolled in the study, based on self- and parent-reporting, 47 (19.7%) met the criteria for DSM-5TM symptom threshold during ≥1 adult assessment. The majority of these participants (40/47 [85.1%]) experienced clinically significant impairment, with 12 demonstrating ADHD symptoms in childhood, 18 during adolescence (4 identified as adolescent-onset ADHD) and 10 during adulthood. Therefore, 24 participants met the criteria for adult-onset ADHD, as 14 experienced symptoms in adolescence, with the majority (14/24) of these cases due to heavy substance use. Presence of another mental health disorder affected 5 of the remaining 10 participants, and 2 participants were excluded as they did not have a DISC score for adulthood. Of the 3 adult-onset ADHD cases, symptoms for 1 participant were only detected in one setting, therefore only 2 out of 239 (0.8%) participants (aged 21.05 and 27.45 years, respectively) showed evidence of adult-onset ADHD, although both possessed a variety of past or current mental health symptoms.
Limitations of this study included: lack of detailed and frequent data collection to carefully assess psychiatric functioning over time; the fact that one study did not perform full childhood assessments (which may have led to missed symptoms in some cases); lack of symptom detail up until the age of 10 years; and the fact that participants were from the same school and therefore some characteristics (e.g. gender, family income) may have been over-represented. In addition, impairments were only available from parents during adolescence, therefore some participants may have met the criteria for adolescent-onset ADHD if teacher- and self-ratings had been available. Participants were only assessed until their mid-to-late 20s, and some cases of late-onset ADHD may appear at a later age. Finally, physical health, personality disorders and impulsive features were not collected, which may have better explained some late-onset cases.
By using a stepped diagnostic procedure that considers multi-informant data, longitudinal symptom patterns, impairment, and co-occurring mental disorders and substance use, these data indicate that the prevalence of late-onset ADHD (2.5% for adolescent-onset and 0.8% for adult-onset) is lower than reported in previous studies (2.5–10.7%) and is not without a complex psychiatric history. The authors concluded that although some adolescents and adults may initially present with ADHD-like symptoms, clinicians should carefully assess substance use and comorbid mental health disorders before diagnosing and treating individuals. Further research is required to understand how neurocognitive changes during adolescence may mimic or facilitate ADHD symptoms.
Read more about late-onset ADHD here
*The Multimodal Treatment of ADHD (MTA) study compared the effects of pharmacological and non-pharmacological treatment in children (aged 7.0–9.9 years) with ADHD over a 2-year period. Two years following baseline, 289 children without ADHD were enrolled in the study to serve as the local comparison group. This study was followed up for 16 years post-baseline
†Participants must have had at least one assessment for ADHD in adolescence (12–17 years) and during adulthood (≥18 years). Of the original 289 normative comparison participants enrolled in the MTA study; 31 participants with a baseline Diagnostic Interview Schedule for Children diagnosis of ADHD and 19 participants with insufficient follow-up data were excluded
‡Diagnostic and Statistical Manual of Mental Disorders – 4th Edition – text revision (DSM-IV-TR) ADHD symptoms were measured using the Swanson, Nolan and Pelham (SNAP) Rating Scale, which was completed by parents, teachers and adolescents, and the Conners’ Adult ADHD Rating Scale (CAARS), which was completed by participants and parents. Study respondents were asked to assess the presence of ADHD symptoms on a scale of 0 (“not at all”) to 3 (“very much”). Scores of 2/3 indicated presence of symptoms
§The Columbia Impairment Scale assesses impairments across multiple domains, and in this study, parent- and self-versions of the scale were used to determine impairments related to “getting along with kids own age”, “schoolwork”, “behaviour at home” and “behaviour at school”. Response options ranged from 0 (“no problem”) to 6 (“extreme problem”), with a score ≥3 in ≥1 of the 4 domains sufficient to meet the threshold for impairment
¥The Diagnostic Interview Schedule for Children (DISC) is a parent- or self-report interview that indicates the presence of a mental disorder that may overlap with late-onset ADHD-related symptoms. In this study, eight clinicians reviewed the onset and chronicity of all mental symptoms and unanimously concluded whether ADHD symptoms or impairment could be attributable to another disorder
Sibley MH, Rohde LA, Swanson JM, et al. Late-onset ADHD reconsidered with comprehensive repeated assessments between ages 10 and 25. Am J Psychiatry 2017; Epub ahead of print.