Abstract
Attention-deficit/hyperactivity disorder (ADHD) is a childhood-onset condition that may continue into adulthood. When assessing adult patients, clinicians usually rely on retrospective reports of childhood symptoms to evaluate the age-of-onset criterion. Since inaccurate symptom recall may impede the diagnosis and treatment of ADHD, knowledge about the factors influencing retrospective reports is needed. This longitudinal study investigated (a) the accuracy of retrospective symptom ratings by adult participants with a childhood diagnosis of ADHD (self-ratings) and parents or significant others (proxy ratings), and (b) the influence of current ADHD symptom severity and ADHD-associated impairments on retrospective symptom ratings. Participants (N = 55) were members of the Cologne Adaptive Multimodal Treatment (CAMT) study who had been referred and treated for ADHD in childhood and were reassessed in adulthood (average age 27 years). Participants’ retrospective self-ratings were substantially lower than, and did not correlate with, parents’ ADHD symptom ratings provided at study entry, while retrospective symptom ratings provided by proxy respondents correlated moderately with parents’ childhood ratings. In addition, participants were more likely to underreport childhood symptoms (79%) and more frequently denied the presence of three or more childhood symptoms (17%) compared to proxy respondents (65% underreporting, 10% false-negative recall). Proxy respondents’ symptom recall was best predicted by childhood ADHD, while participants’ symptom recall was best predicted by current ADHD symptom severity. ADHD-associated impairments were not correlated with symptom recall after controlling for childhood ADHD. Together, these findings suggest a recall bias in adult patients and question the validity of retrospective reports, even in clinical samples.