Autism spectrum disorder diagnosis in adults: phenotype and genotype findings from a clinically derived cohort

Abstract

Background: The past decade has seen the development of services for adults presenting with symptoms of autism spectrum disorder (ASD) in the UK. Compared with children, little is known about the phenotypic and genetic characteristics of these patients.

Aims: This e-cohort study aimed to examine the phenotypic and genetic characteristics of a clinically presenting sample of adults diagnosed with ASD by specialist services.

Method: Individuals diagnosed with ASD as adults were recruited by the National Centre for Mental Health and completed self-report questionnaires, interviews and provided DNA; 105 eligible individuals were matched to 76 healthy controls. We investigated demographics, social history and comorbid psychiatric and physical disorders. Samples were genotyped, copy number variants (CNVs) were called and polygenic risk scores were calculated.

Results: Of individuals with ASD, 89.5% had at least one comorbid psychiatric diagnosis, with depression (62.9%) and anxiety (55.2%) being the most common. The ASD group experienced more neurological comorbidities than controls, particularly migraine headache. They were less likely to have married or be in work, and had more alcohol-related problems. There was a significantly higher load of autism common genetic variants in the adult ASD group compared with controls, but there was no difference in the rate of rare CNVs.

Conclusions: This study provides important information about psychiatric comorbidity in adult ASD, which may inform clinical practice and patient counselling. It also suggests that the polygenic load of common ASD-associated variants may be important in conferring risk within the non-intellectually disabled population of adults with ASD.

Declaration of interest: None.

Keywords: Autistic spectrum disorders; genetics; social functioning.

Fig. 1

Percentage variance explained by PRS at analysed association levels for ASD. Percentage variance at eight association marker levels derived from linkage independent markers in the ASD genome-wide association study. Significance of associations between single nucleotide polymorphisms and ASD range from 0.5 to 1 × 10⁻⁶. Probability of association to be found at each individual variance level is denoted by P value.