Executive Function Abilities in Individuals Receiving an Autism Spectrum Diagnosis in Adulthood

Background: Little is known about cognition in autism spectrum disorder (ASD) across adulthood, especially in later-life. In typical ageing, executive functions are particularly sensitive to age-related decline. Difficulties in some domains of executive function have been observed in ASD, but it is unclear whether older adults with ASD will show the same pattern of age-related change as typical adults.

Objectives:  To examine executive function abilities in adults with ASD, including both gender differences and associations with age.

Methods:  We examined the profile of cognitive abilities and autism traits in a group of 134 adults (97 males) receiving a first diagnosis of ASD in adulthood according to DSM IV or V criteria. Participants were assessed in a specialist adult diagnostic clinic, were aged between 18 and 75 years and all had abilities within the normal range on neuropsychological assessment. Executive function abilities were measured using age-normed scores for: Digit Symbol and Digit Span (Wechsler Adult Intelligence Scale-III or IV), Brixton, Hayling, Trails A and B, Key Search and Zoo Map (Behavioural Assessment of the Dysexecutive Syndrome). A composite score for executive function (Mean EF) was computed. Autism traits were measured using the self-report Autism Spectrum Quotient (AQ), Empathy Quotient (EQ), and Systemising Quotient (SQ).

Results:  Few gender differences in ASD were observed across the measures. Women scored significantly higher than men on the Digit Symbol task, and reported higher AQ scores but this did not reach statistical significance. Correlations between age and executive function measures were calculated using age-corrected test scores; therefore one would not expect a significant correlation with age if individuals with ASD show the same age-related differences as typical adults. Results demonstrate a significant positive correlation between age and Digit Symbol as well as both Trails A and Trails B, indicating that as age increases performance on these measures is improving beyond typical adult performance. Age correlated significantly with self-rated scores on both AQ and SQ, indicating that self-rated scores increased with age.

Conclusions:  In a group of adults without intellectual disability receiving a first ASD diagnosis in adulthood, few gender differences in executive function were observed. In some but not all executive function tasks, significant correlations were observed between age and performance. Results suggest that for some abilities relying on speed and sequencing, late-diagnosed individuals with ASD may demonstrate better performance than typical age-matched peers. Although these results are in keeping with ASD being protective against typical age-related decline, longitudinal studies are required to investigate this fully.