A small percentage of individuals with autism spectrum disorders (ASDs) may go on to lose core symptoms of the diagnosis and achieve “optimal outcomes.” Helt et al. (2008) defined an individual with an optimal outcome as having a history of an ASD diagnosis, demonstrating average or above average academic and adaptive functioning, receiving minimal special education supports specific to autism symptoms, and not meeting criteria for a diagnosis of ASD diagnosis as determined by administration of the Autism Diagnostic Observation Schedule (ADOS).
Lovaas (1987) pioneered the study of “recovery” from ASD when he reported strong cognitive and academic outcomes in a small sample of individuals with high-functioning ASD following early, intensive behavioral intervention. Since Lovaas’ study, a number of researchers have carried out further studies of recovery from ASD with stronger experimental designs and more comprehensive measures of outcome (e.g., Fein et al. 2013; Sallows and Graupner 2005).
Helt et al. (2008) estimate that between 3% and 25% of children diagnosed with ASD lose the diagnosis and exhibit average cognitive, adaptive, and social abilities. Helt and colleagues argue that misdiagnosis of ASD does not explain the phenomenon of optimal outcomes. The authors suggest that recovery from ASD could be limited to certain subsets of ASD and is likely strongly tied to early detection and treatment, characteristics of the child (e.g., cognitive abilities), and maturation.
Recent studies have examined specific characteristics of children and adolescents with optimal outcomes using a variety of measures of psychological and neuropsychological functioning, as well as brain imaging. A comprehensive study by Fein and colleagues (summarized in Fein et al. 2013) indicated that a group of children and teenagers with optimal outcomes had similar academic, executive functioning, language, and social skills in comparison to typically developing study participants. When compared to typically developing peers, the optimal outcome participants did display some subtle differences in their pragmatic language. They also presented with a higher incidence of psychiatric concerns than typically developing controls, particularly symptoms of attention-deficit/hyperactivity disorder and specific phobias (Orinstein et al. 2015).
Eigsti et al. (2016) have also explored ways in which brain networks responsible for language may differ for individuals with optimal outcomes. Findings indicated largely similar patterns of activation for individuals with optimal outcomes compared to those with ASD across several brain regions. However, the optimal outcome group also demonstrated “compensatory” activation in numerous regions, differing from both a typically developing control group and the ASD group. Results suggest that early treatment and learning experiences may result in normalized language performance for optimal outcome individuals, but that differing brain activity may underlie these observable behavioral similarities in language processing.
Further research is needed to understand the nature of differences between individuals with optimal outcomes and those with other developmental trajectories. Research suggests that intensive early intervention, specifically applied behavior analysis approaches to therapy, may contribute to better outcomes in ASD (Fein et al. 2013; Orinstein et al. 2015).