Executive Function Deficits Are Associated with ADHD Symptoms Whilst Theory of Mind Is Associated with ASD Symptoms in Adolescents with ASD
Objectives: We build a working model of the associations among EF, ToM and symptoms of ASD and ADHD in adolescents with ASD.
Methods: We investigated a population-based cohort of 100 adolescents with ASD (the Special Needs and Autism Project – SNAP), aged 14-16 years and full scale IQ≥50. Structural equation modelling was used to model the relationships between EF, ToM and ASD or ADHD symptom levels. Measures of inhibition (TEA-Ch Opposite Worlds), planning (drawing task), switching (Trail Making Test and card sorting task), and working memory (Backward Digit Span) indexed EF; while measures of mentalising (the Frith-Happe Animated Triangles, Strange Stories, Reading the Mind in the Eyes) indexed ToM. All neurocognitive assessments took place at 14-16 years. ADHD symptoms were indexed by the parent-rated Strengths and Difficulties Questionnaire (SDQ) hyperactivity subscale (14-16 years) and teacher-ratings of SDQ (10-12 years), parent-rated Profile of Neuropsychiatric Symptoms (PONS) ADHD (14-16 years), and DSM-IV ADHD symptom counts from structured diagnostic interviews with parents (10-14 years). ASD symptoms were indexed by the algorithm scores of Autism Diagnostic Interview-Revised (ADI-R) and Autism Diagnostic Observation Schedule (ADOS), both from 10-12 years, and the parent-reported Social Responsiveness Scale (SRS) from 14-16 years. To account for shared sources of information, a parental-rating latent factor was constructed using a subset of the ASD and ADHD symptom measures (ADI-R, SRS, parent-rated SDQ, PONS, and ADHD symptom counts).
Results: The final model reached acceptable fit (χ2 = 101.7, p = .06; CFI = .94; TLI = .92; RMSEA = .051). EF deficits were associated with ADHD symptoms (standardized coefficient = .47), and not with ASD symptoms, while ToM deficits were associated only with ASD symptoms (standardized coefficient = .74). The EF and ToM latent factors were significantly correlated (standardized coefficient = .56). Prior to including a parent rating factor, a significant correlation was found between ASD and ADHD symptoms (standardized coefficient = .63). However, this model had an overall poorer fit than the present and the correlation between the ADHD and ASD factors became non-significant after including the parent rating factor, which suggests that the association between the two symptom domains was due to a shared parental rating effect.
Conclusions: ADHD symptoms in individuals with ASD are specifically associated with EF deficits. This is consistent with the finding of EF deficits in those with ADHD alone. Our findings would benefit from replication in samples of individuals with pure ASD, ADHD and combined ASD-ADHD profiles to better understand whether this represents a shared underlying factor for observed co-occurrence.