Relationship Between the CNTNAP2 Gene Variant and Cognitive and Behavioral Flexibility in Children with Autism Spectrum Disorders

Background: Restricted, repetitive behaviors and interests (RRBI) symptoms are core to autism spectrum disorder (ASD), particularly higher-order RRBI, which are related to cognitive and behavioral inflexibility.  However, identifying risk genetic variation associated with these behaviors has been difficult. A limitation of previous studies is the use of the ADI as the sole basis of phenotypic information. The ADI has only a few items addressing higher order RRBI symptoms and does not provide continuous data.  This qualitative phenotypic data provides considerably less power to identify gene-behavior associations than quantitative data. The Behavior Rating Inventory of Executive Function (BRIEF) is a standardized instrument that provides quantitative measurement of cognitive and behavioral flexibility.  Contactin associated protein-like 2 (CNTNAP2) has been linked to social communication symptoms in ASD, and recent genetic expression analyses suggest that CNTNAP2 is also expressed in regions responsible for flexible behavior. No studies to date have examined the relationship between CNTNAP2 and inflexible behaviors in ASD.

Objectives: Examine the relationship between the single nucleotide polymorphism (SNP) in the CNTNAP2 gene (rs2710102) and higher-order RRBI characteristics as assessed by the BRIEF, Flexibility Questionnaire (FQ), and the compulsive domain from the Repetitive Behavior Scale-Revised (RBS-R) in children with ASD.

Methods: In our preliminary analyses, we have examined 45 children with ASD (80% male; mean age =10.06 years (SD=1.80); mean full scale IQ=107.40 (SD=19.94)). Diagnosis was confirmed with the ADI-R and ADOS (ADI Social interaction mean=20.34 (SD=5.19), Verbal mean=16.33 (SD=4.43), Nonverbal mean=8.79 (SD=3.27), RRBI mean=5.80 (SD=2.66); ADOS, Communication and social interaction total=11.99 (SD=4.63)).  Children were also assessed using the BRIEF (Shift domain mean T-score=68.49, SD=13.43), the FQ (total raw score mean=65.21, SD=20.20), and the RBS-R (compulsive domain severity raw score mean=2.84, SD=3.01).  Children provided saliva samples for genetic testing on the CNTNAP2 gene, and parents rated probands on the three phenotypic measures.  A Spearman’s rho correlation, univariate ANOVA, and Cohen’s d were used to explore the relationship of CNTNAP2 ASD risk allele presence (G) and RRBI behaviors.

Results: Preliminary results showed a significant Spearman’s rho correlation between increasing risk alleles (0,1,2) and difficulties on the Shift scale (rho(N=45)=0.31, p<0.05).  The homozygous non-risk allele group (AA) had the lowest (least impaired) means across all three flexibility measures (Shift scale M=59.86, SD=6.03; FQ Total=52.25, SD=24.24, compulsive scale M=1.88, SD=1.81), the heterozygote (AG/GA) had higher scores (Shift scale M=68.58, SD=15.615; FQ Total=59.68, SD=20.25, compulsive scale M=2.58, SD=2.57), and the homozygous high risk allele group (GG) had the highest scores (Shift scale M=73.17, SD=10.308; FQ Total=68.06, SD=16.611, compulsive scale M=3.11, SD=3.58). The two homozygous groups had medium-to-large effects for differences in flexibility measures (Cohen’s d ranged from 0.46 to 0.99), and the preliminary ANOVA’s yielded marginal findings for the BRIEF.

Conclusions: We find preliminary evidence of an association between CNTNAP2 genetic variation and scores on continuous measures of cognitive and behavioral flexibility, supporting the hypothesis that the risk allele for a SNP in the CNTNAP2 gene may relate to higher parent ratings of RRBI behaviors. Further research between RRBI symptoms and this CNTNAP2 SNP is warranted.