Expression of the Broad Autism Phenotype in First Degree Relatives from the Simons Simplex Collection

Background: In order to better understand the biological pathways that contribute to aspects of autism, researchers are examining the milder expression of underlying genetic liability for autism that is manifest in non-autistic relatives, defined as the Broad Autism Phenotype (BAP). There has been some evidence of the BAP in parents and siblings from simplex and multiplex autism families.

Objectives: To extend the existing research by examining differences in measurement of the BAP across multiple instruments and across informants, and the degree to which parental scores predict sibling scores.

Methods: Data were collected via the Simons Simplex Collection (SSC). The SSC is a large, multi-site study that utilizes rigorous phenotyping procedures to study families in which one child (aged 4 to 17) has been diagnosed with an Autism Spectrum Disorder (ASD) but neither parent nor any sibling has been determined to be on the autism spectrum. To assess the BAP, parents completed the Broad Autism Phenotype Questionnaire (BAPQ) about themselves and the Social Responsiveness Scale: Adult Research Version (SRS:ARV) about their partners. In families where one member had an elevated SRS:ARV score (above 70), the Family History Interview Interviewer Impressions were also completed. Families were excluded if scores on both instruments were elevated. Parents also completed the Social Responsiveness Scale on unaffected siblings (the sibling closest in age to the proband).

Results: 787 parents completed the BAPQ, and 844 parents completed the SRS:ARV. SRS scores were attained for 385 siblings. On the BAPQ, fathers had elevated scores when compared to mothers for Aloof Personality (p<.001), Pragmatic Language Deficits (p<.001), Rigid Personality (p<.001), and Total Score (p<.001). Relative to published norms, 278 parents (35%) exceeded BAPQ self-report cutoff scores for Aloof Personality; 210 parents (27%) exceeded cutoffs for Pragmatic Language deficits; 340 parents (43%) exceeded cutoffs for Rigid Personality, and 195 parents (25%) exceeded cutoffs for the total score, indicating that they had multiple features of the BAP. SRS:ARV mean scores were not elevated in comparison to an epidemiologic (previously published) reference sample, and there were no differences between mothers and fathers for overall mean scores. Correlations between the subscales of the BAPQ and the SRS:ARV were significant (p<.001), but weak (i.e. r<.50). Sibling mean SRS scores were also not elevated. Maternal scores on the SRS and the BAPQ were more strongly correlated with sibling SRS scores compared to paternal scores on either instrument, though correlations were also weak (r=.26 or less).

Conclusions: Results regarding the presence/absence of traits related to the BAP differed depending upon the method and instrument that was utilized to assess the BAP.  Importantly, the results revealed that the two commonly used self- and informant- report instruments are weakly correlated and may be assessing different constructs. The SRS:ARV allows one to control for total severity of autistic impairment, while the BAPQ assess constructs that the SRS:ARV does not measure. While these results assist in defining intermediate phenotypes and informing models of genetic transmission in simplex families, resolving differences in measurement of the BAP will be important for future studies.