DSM-5 Criteria and Diagnoses in Adulthood: Are There “Strong and Stable” Features?

Background: Autism spectrum disorder (ASD) is a lifelong condition, with some individuals not being diagnosed until adulthood. Individuals seeking diagnosis as adults, however, may present with a different profile compared to children. Previous research examining the stability of ASD behaviours over time has revealed inconsistent findings: while all symptoms tend to decrease over time (e.g. Seltzer et al., 2003), some findings indicate that social and communication difficulties reduce more than repetitive behaviours (e.g. Fecteau et al., 2003) while others find the opposite (e.g. Shattuck et al., 2007). It is, therefore, important to identify which behavioural features are the most robust and stable over time and likely to be potential diagnostic markers in adults.

Objectives: To identify robust features of autism in adults based on DSM-5 criteria.

Methods: Participants were 80 adults with clinically diagnosed ASD (mean age=33.94 years; 52 male, 28 female). The majority (53) were recruited as consecutive referrals from an adult specialist ASD national health clinic, while 27 adults with a known diagnosis of ASD were recruited through university recruitment registers. All adults had an IQ over 70. The Diagnostic Interview for Social and Communication Disorders-Abbreviated (DISCO-Abbreviated; Carrington, Kent, et al., 2014), was carried out independently by researchers. The DISCO-Abbreviated includes published items designed to fit the DSM-5 criteria for ASD, based on a child sample (Carrington, Kent et al., 2014). Frequencies of items and sub-domains endorsed by the adult sample were compared with the original, published sample.

Results: The overall profile of behaviour in adults, indicated by the mean number of behaviours exhibited in the subdomains of the DSM-5 criteria, was similar to previously published data in children; items relating to difficulties with social relationships were endorsed the most frequently, followed by deficits in socio-emotional reciprocity, suggesting that these difficulties in social interactions endure in adulthood. Deficits in non-verbal communication were relatively infrequently endorsed in adults. For the restricted and repetitive behaviours subdomains, the least frequently endorsed behaviours in adults were those related to hypo- and hyper-reactivity to sensory input. The subdomain in which the adult and child profile was most similar measured insistence on sameness. There were no gender differences.

The frequency of individual items endorsed in adults was generally lower than in children. Nevertheless, there were several items with high frequencies in adults. Strong social-communication behaviours included awareness of others’ feelings (85%) and sharing of interests (76.3%), which were also highly frequent in children (88.9% and 88.9% respectively). The strongest behaviour for adults in domain B was a limited pattern of self-chosen activities, endorsed by 89% of adults and 97.2% of children. Importantly, those items identified as being most frequent in adults also remained stable over time.

Conclusions: Individuals diagnosed with ASD in adulthood showed a similar profile of behaviours to the published sample of children. Nevertheless, the reduced frequency of behaviours related to non-verbal communication suggest that these behaviours may be less central to diagnosis in adulthood than childhood. Moreover, several items were identified that may be strong and stable diagnostic markers for adults.