Abstract / Summary
Abstract Background ADHD is conceptualized as neurodevelopmental disorder with childhood onset of symptoms, but prospective studies have documented adolescents and adults meeting criteria for ADHD who had no childhood diagnosis or non-clinical level of symptoms in childhood. What these cases might represent (i.e., residual childhood ADHD, neurodevelopmental delay , or something else) is an important diagnostic question. Objective To compare five prospective longitudinal cohorts and address the effects of sample ascertainment of childhood symptoms, assessment of subthreshold symptoms, impairment, comorbidity, substance use, and repeated assessment on detection of apparent late-onset ADHD. Methods A targeted comparative narrative synthesis was performed using published reports of five prospective longitudinal cohorts (Dunedin Multidisciplinary Health and Development Study, E-Risk Longitudinal Twin Study, 1993 Pelotas Birth Cohort, Avon Longitudinal Study of Parents and Children (ALSPAC), Multimodal Treatment Study of ADHD Local Normative Comparison Group (MTA LNCG). The narrative synthesis examined three core areas: (a) cohort features, (b) childhood and (c) later assessments of ADHD, subthreshold symptoms, comorbid disorders, and substance use, and (d) operational definitions of apparent late-onset ADHD. Substantial heterogeneity existed among cohorts in population, assessment procedures, instruments, developmental age at assessment, and operationalizations of apparent late-onset ADHD, precluding quantitative meta-analyses. Results All five cohorts gave different estimates of adult or apparent late‑onset ADHD. In the Dunedin (Moffitt et al., 2015) 3.0 percent of 38‑year‑olds met criteria for adult ADHD and 90 percent of adult ADHD cases had no documented childhood history in the health records. In the E‑Risk study (Agnew-Blais et al., 2016) 7.9 percent of 18‑year‑olds met criteria for ADHD and 5.4 percent fit our definition for a late‑onset case. In the Pelotas study (Caye et al., 2016) 12.2 percent of 18‑19‑year‑olds met symptom criteria for ADHD, which dropped to 6.3 percent after excluding individuals with comorbidities. In the ALSPAC (Cooper et al., 2018) 74.7 percent of the late‑onset ADHD cases had shown prior childhood symptoms below a clinical cut‑off while roughly 0.4 percent of the total sample matched the operational category of genuine late‑onset. In the MTA LNCG (Sibley et al., 2018) 0.8 percent of young adults met clinical criteria for ADHD yet no documentation of a childhood history appeared after rigorous screening. Conclusions Among these five cohorts, estimates of apparent late-onset ADHD varied substantially according to how childhood symptoms, impairment, comorbidity, and longitudinal assessment were defined. Absence of a documented childhood diagnosis should not be interpreted as definitive evidence of adult onset, while the available evidence also does not establish that all apparent late-onset cases represent previously unrecognized childhood ADHD. Developmental history and consideration of alternative explanations remain important when evaluating adults presenting with ADHD symptoms.