The Association Between Cognitive Ability and Psychiatric Problems in Adults with ASD

Background: Data on rates of mental health problems in adults with autism vary widely, from 20%-25% in some studies to over 70% in others. The relationship with cognitive level is also uncertain, with some studies suggesting mental health problems are more likely to occur in individuals of higher IQ who are more exposed to social pressures and/or are more aware of their social and other difficulties than individuals with intellectual impairments. 

Objectives: The aim of the  present study was to explore the association between mental health and cognitive functioning in a cohort of adults with autism who have been systematically followed up over several decades

Methods: Participants were 58 individuals who were first diagnosed with autism as children (mean age 6 years); current age mean age is 45 years. All were of average non-verbal IQ when seen as children. Adult measures included IQ assessments and self and informant measures of psychiatric difficulties

Results: Most individuals were able to complete formal cognitive tests as adults and in this group cognitive ability had remained very stable since they were first seen in childhood (mean child IQ 85,5, sd 14.2; mean adult IQ 87.3, sd.20.1). However, ability levels in 15 individuals could be assessed only indirectly; in this group cognitive levels (based on Vineland proxy estimates) had declined significantly (mean 20.5; sd 2.1). Among the total cohort, 44% of adults were rated as having no mental health problems but 28% had mild to moderate difficulties and a further 28% had severe psychiatric problems. Overall there was a small but statistically significant correlation between IQ and ratings of adult mental health (r=-.32, p=.015). Significantly more individuals in the group showing a decline in cognitive ability were rated as showing moderate to high levels of psychiatric morbidity than those whose IQ remained within the average range (Fisher exact test .03). Thus, two thirds of individuals in the “cognitive decline” group had moderate to high levels of psychiatric disturbance compared with 28% of those with an IQ in the average range.   

Conclusions: In this cohort of individuals seen over many years, there was no evidence that higher IQ was associated with a higher risk of psychiatric morbidity; in contrast individuals who had shown an apparent decline in cognitive functioning over the years seemed more likely to experience psychiatric disturbance. However, whether the findings reflect a true decline in cognitive ability, or whether poor test performance was due mainly to psychiatric and behavioural problems remains unclear. Other issues related to the assessment of IQ and mental health changes in long-term follow-up studies will also be discussed