Retrospective Ratings of ADHD Symptoms Made at Young Adulthood by Clinic-Referred Boys With ADHD-Related Problems, Their Brothers Without ADHD, and Control Participants Jan Loney, Johannes Ledolter, and John R. Kramer University of Iowa Robert J. Volpe Northeastern University Retrospective childhood attention-deficit/hyperactivity disorder (ADHD) symptoms are required to diagnosis adult ADHD, but the validity of self-rated symptoms across time is questionable. Here, boys with ADHD-related problems, their brothers without ADHD, and former schoolmates rated themselves during young adulthood for ages 9, 14, and 19. Brothers rated probands retrospectively at the same ages. The young adults referred as children for ADHD (a) acknowledged childhood symptoms; (b) described improvement over time; (c) did not differ from brothers or controls on most self-ratings of young adult symptoms; (d) rated themselves as more symptomatic at age 9, but less symptomatic at age 19, than their brothers rated them; and (e) agreed only to some degree with brothers’ ratings of probands’ aggression (median correlation = .22). Probands’ ratings showed limited agreement with judges’ symptom ratings (median correlation = .16) and young adult follow-up examiners’ ratings (median correlation = .14). These findings are not accounted for solely by changes in informants, nor by the course of ADHD psychopathology. They suggest some stability but limited internal consistency and validity for retro- spective ADHD ratings by probands and brothers. Keywords: retrospective ratings, self ratings, ADHD symptoms, adult ADHD The diagnosis of adult attention-deficit/hyperactivity disorder (ADHD) typically requires that individuals meet both the child- hood and adult criteria for the disorder. Meeting those criteria ordinarily requires a retrospective report of ADHD psychopathol- ogy, most often by the individuals themselves. Several large-scale longitudinal studies now have followed participants with a history of childhood inattention, hyperactivity, and/or impulsivity through early adulthood (Barkley, Fischer, Edelbrock, & Smallish, 1990; Loney, Whaley-Klahn, Kosier, & Conboy, 1983; Mannuzza, Klein, Bessler, Malloy, & LaPadula, 1993; Rasmussen & Gillberg, 2000; Weiss & Hechtman, 1993). Despite differences in partici- pant samples, assessment instruments, follow-up intervals, and other aspects of methodology, there is broad agreement that symp- toms and diagnoses decline from childhood to adolescence. How- ever, investigators differ sharply on the degree to which they report that core symptoms decline further from adolescence to young adulthood. The Wisconsin group (Barkley, Fischer, Smallish, & Fletcher, 2002) argued that the precipitous drop between adoles- cence and adulthood in the prevalence of ADHD diagnoses re- ported in the New York investigations (Mannuzza, Klein, & Moul- ton, 2003) is in part an artifact of switching informants from parent to participant (self) raters. This issue illustrates the degree to which psychiatric findings are dependent on informants’ descriptions of themselves and others. Interviews, checklists, rating scales, questionnaires, and personal- ity inventories all rely ultimately on this process. Because descrip- tion entails human judgment, its accuracy can be compromised in a variety of ways. Three such factors of particular interest to researchers in child psychopathology are (a) limited familiarity with, or understanding of, the person or phenomenon being de- scribed (Jensen et al., 1999; Klonsky, Oltmanns, & Turkheimer, 2002); (b) the tendency to respond in a socially desirable manner (Huang, Liao, & Chang, 1998) or to rate oneself overly positively (Alicke, Klotz, Breitenbecher, Yurak, & Vredenburg, 1995); and (c) retrospective distortions associated with the passage of time, such as forgetting and reconstruction (Hardt & Rutter, 2004; Henry, Moffitt, Caspi, Langley, & Silva, 1994). Given these and other challenges to validity found in commonly used instruments (Mandal, Olmi, & Wilczynski, 1999; Reid & Maag, 1994), it is not surprising that the self-descriptions of children, adolescents, and adults with a history of childhood problems sometimes disagree considerably with reports from other informants, and that retro- spective ratings of past problems often do not correspond closely with ratings made when those problems took place. The latter discrepancy is of particular concern for investigations of adults when recall of childhood problems, such as ADHD symptoms, forms a basis for lifetime diagnosis (e.g., American Psychiatric Association, 1994). In general, available childhood informants (i.e., parents and teachers) attribute more numerous and severe symptoms of ADHD Jan Loney, Department of Pediatrics, University of Iowa; Johannes Ledolter, Department of Management Sciences, University of Iowa; John R. Kramer, Department of Psychiatry, University of Iowa; Robert J. Volpe, Department of Counseling and Applied Educational Psychology, North- eastern University. The research reported in this article was supported by National Institutes of Health Grant MH-062661. Data were drawn from the Iowa Longitudinal Study of ADHD. Correspondence concerning this article should be addressed to Jan Loney, Post Office Box 777, Maysville, KY 41056. E-mail: loney777@gmail.com Psychological Assessment Copyright 2007 by the American Psychological Association 2007, Vol. 19, No. 3, 269 –280 1040-3590/07/$12.00 DOI: 10.1037/1040-3590.19.3.269 269