For most individuals with autism who had an IQ in the average range (i.e. ≥ 70) as children, childhood IQ proved a reliable predictor of cognitive functioning well into mid- to- later adulthood. However, a significant minority was no longer testable on standard tests as adults. Their current very low levels of functional ability were generally associated with severe behavioural disturbance and persisting and severe language impairment; 50% of these individuals had also developed epilepsy, pointing to the role of organic brain dysfunction. Implications for early intervention are discussed.

A systematic review was conducted of outcome measures used in treatment trials for older adolescents and adults with autism spectrum disorders (ASDs). Of 818 titles only 30 articles (19 of which involved pharmacological treatments) were identified that met inclusion criteria (sample size > 5; mean age of group > 15 years; mean IQ > 30; ASD diagnosis confirmed; use of objective ASD outcome measures; focus on symptoms core to or typically associated with ASDs). Selected studies included randomized and placebo-controlled trials, retrospective assessment studies, case series and open label or case-control trials. Use of outcome measures varied with frequent use of non-standardized assessments, very little use of measures designed specifically for individuals with ASD or of instruments focusing on core ASD deficits, such as communication or social functioning. Most commonly used were the Clinical Global Impression (CGI) rating scale and the Yale-Brown Obsessive Compulsive Scale (Y-BOCS). The strengths or deficiencies of the outcome measures used were not systematically evaluated. Although there are now many well controlled treatment trials for children with ASDs, adult intervention research is very limited. The lack of valid and reliable outcome measures for adults with ASDs compromises attempts at treatment evaluation.