Is Cognitive Variability a Viable Candidate Endophenotype for Autism? Results from a Broader Autism Phenotype (BAP) Study

Background:  

Conflicting accounts of the cognitive profile of autism have long presented a challenge for theorists seeking to explain the neuropsychological processes underlying this complex disorder. Although cognitive impairments among individuals with ASD have been widely reported, as well as superior performance on some measures, there are many inconsistencies in the literature (see Hill, 2004, and Happe & Frith, 2006). It has recently been suggested that cognitive variability, resulting from disruption to the normal processes of brain specialisation (Gilbert, 2008), may in fact be a core feature of ASD. Using multiple-case series analyses, a number of recent studies have reported a highly significant pattern of within-participant cognitive variability among individuals with autism, with evidence of both sub- and supra-normal performance (Towgood, Meuwese, Gilbert, Turner, & Burgess, 2009) particularly on measures of rostral pre-frontal cortex (rPFC) functioning (O' Mahony, 2009).

Objectives:  

This study aimed to evaluate whether cognitive variability represents a viable candidate endophenotype for autism through investigation of cognitive profile of the broader autism phenotype (BAP). Specifically, the study sought to investigate whether fathers of children with ASD would show greater variability in their cognitive profile, including higher incidence of a pattern of both impaired and superior performance, relative to controls. The study also examined whether cognitive variability among fathers of children with ASD would correlate with self-rated social, communication and behavioural difficulties. Finally, the study investigated whether fathers of children with ASD would show greater impairments on measures of rPFC functioning than controls.

Methods:  

The performance of twenty-four fathers of children with ASD and twenty control fathers was compared using a test battery that predominantly assessed rPFC functioning. Measures of IQ, social cognition and information processing were also administered. Group-difference and multiple-case series analyses were used to explore patterns of performance between groups and within each individual's cognitive profile.

Results:  

Multiple-case series analysis indicated that fathers of children with autism did not display significant cognitive variability relative to controls and there was no evidence of greater incidence of a pattern of both impaired and superior performance. Analysis at the group level found no significant difference between the groups in relation to measures of rPFC functioning. A significant group difference emerged in relation to Theory of mind, but this did not survive Bonferroni correction.

Conclusions:  

The results suggest that cognitive variability is not a feature of the broader autism phenotype and that it is unlikely to represent a candidate endophenotype for ASD.  While individuals with ASD may show high levels of variability in their cognitive profiles, rather than being a core feature of ASD,  it may reflect a secondary feature of the disorder.