A recent systematic review reported on how differences in diagnostic methods can influence estimated adult persistence of childhood ADHD. The findings revealed a considerable variation in reported estimates of ADHD persistence across the published literature, and this high variability was due to differences in the diagnostic procedure used.
A systematic database search was conducted to identify studies published in English from 1992 to 31 May 2016 that reported on ADHD persistence rates. Studies were included if ADD or ADHD diagnosis in childhood was based on Diagnostic and Statistical Manual of Mental Disorders (DSM) or International Statistical Classification of Diseases and Related Health Problems (ICD) criteria, or research diagnostic protocols that matched DSM-III or DSM-IV criteria,* and if participants’ mean age at baseline and adult follow-up diagnosis was <12 years and ≥18 years, respectively. A total of 17 studies met the inclusion criteria, and these studies reported ADHD persistence data from 12 different longitudinal samples.
Reported ADHD persistence rates ranged from 4.0–77.0%. By changing the diagnostic method used, multiple persistence rates were derived from five multi-estimate samples. The range of persistence rate estimates for adults aged <23 years was 5.0–73.6%; for those aged 23.0–29.9 years was 4.0–77.0%; and for those aged ≥30 years was 6.0–31.9%. For those adults who were assessed at multiple age points in multi-estimate studies, self-reported ADHD symptoms increased with age, leading to higher persistence rates at older ages.
Persistence estimates based solely on self-report ranged from 4.0–64.0%; those based on parent-report alone ranged from 12.0–76.0%; and those based on collateral-report (parents and close relations) ranged from 26.0–54.0%. When diagnosis relied on self-, parent- and collateral-report combined, persistence estimates ranged from 19.0–77.0%. Estimated persistence rates were lowest when derived from self-report and highest when derived from combined sources.
When used alone for adult ADHD diagnosis, structured interviews, semi-structured interviews and rating scales yielded estimate ranges of 5.0–76.0%, 4.0–77.0% and 6.0–73.6%, respectively. In one study, a persistence rate of 49.6% was estimated using a combination of measures.
When DSM-III-Revised (DSM-III-R) symptom threshold was applied, persistence estimates ranged from 4.0–48.0%, while implementation of DSM-IV or DSM-IV-Text Revision symptom threshold yielded estimates of 6.0–76.0%. The persistence estimate based on a normative cut-off derived from a non-ADHD sample for symptom threshold was 12.0–73.6%. Persistence was higher when estimated according to DSM-IV compared with DSM-III-R, and lower when inferred from any DSM edition relative to a norm-based criterion.
The estimated prevalence rate range was 5.0–73.6% when impairment was not considered as a diagnostic criterion, and 4.0–76.0% when impairment was included in diagnosis. Multi-estimate studies generated mixed results on the influence of the impairment criterion on prevalence rates.
Interpretation of the current findings may be limited by the exclusion of literature published in languages other than English, and by a focus on studies of predominantly white males and middle-class clinical cohorts who were not monitored beyond young adulthood. However, the results support the persistent nature of ADHD from childhood into adulthood. Caution should be observed when relying solely on self-reports or on a strict threshold of six DSM symptoms, as these led to very low persistence estimates. To minimise false classifications, recommended methods for determining adult persistence of ADHD include collecting self- and informant-ratings, requiring the presence of impairment, and using an age-appropriate symptom threshold.
*Symptom threshold; presence of impairment; chronicity; pervasiveness; and ruling out other disorders
Sibley MH, Mitchell JT, Becker SP. Method of adult diagnosis influences estimated persistence of childhood ADHD: a systematic review of longitudinal studies. Lancet Psychiatry 2016; 3: 1157-1165.