Central Coherence in the Broader Autism Phenotype

Background:

Relatives of individuals with Autism Spectrum Disorders (ASD) with the Broader Autism Phenotype (BAP) show similar but milder cognitive and behavioural difficulties.  The family history interview (FHI) is a measure of the BAP, and is conducted on the subject (FHI-S), and with an informant (the subject’s spouse or parent) about the subject (FHI-I).  We previously reported significant correlations between social difficulties as measured by the FHI and cognitive performance on Theory of Mind tasks (White et al., IMFAR 2008).  Evidence that weak central coherence (CC) is also a part of the BAP is less striking, but has been reported, particularly in fathers, using tests such as the Embedded Figures Task (EFT) and Block Design (BD) (Bölte & Poustka, 2006; Happé et al., 2001; Baron-Cohen & Hammer, 1997).  In the general population, males tend to perform better than females on both BD (Lynn & Mulhern, 1991) and the EFT (Witkin et al., 1971).  It remains unclear, however, as to whether those relatives with higher total scores on the FHI show evidence of weak CC on the EFT and BD tasks.

Objectives: (1) To identify whether gender differences in CC task performance (Block Design; Embedded Figures Task) reported in the normal population extend to relatives of people with ASD. (2) To assess whether weak central coherence correlates with FHI scores, establishing whether this cognitive function can be related to observations of the BAP.

Methods: One hundred and fourteen parents and siblings from the International Molecular Genetics of Autism Consortium (IMGSAC) collection of multiplex families were assessed.  FHI-S and FHI-I were conducted in participants’ homes, however only the FHI-I was used in analysis as its test-retest reliability tends to be higher (Bolton, 1994).  The EFT and the Block Design task were administered at a later date with other cognitive tests.  For the EFT, participants were given 60 seconds to find the embedded figure then trace it with a stylus.  BD was scored as part of the WASI IQ and t-scores were obtained.

Results: EFT mean solution time and BD t-scores were strongly correlated (p<.001). Contrary to our hypothesis of poorer performance by women on CC measures, there were no significant differences between the performance of male and female relatives of individuals with ASD on either test of CC.  Lastly, there was no correlation between either the EFT or BD and the total score from the FHI-I.

Conclusions: Female relatives of people with ASD do not show the usual decreased performance in CC tasks relative to males, which may suggest an underlying difference in the way they process visual information compared with females in the population.  The failure to find correlations between the FHI-I and CC task performance suggests that this aspect of cognition may not be directly related to aspects of the BAP, or that these behavioural and cognitive measures are independent.