Broader Autism Phenotype in the Mother Is Associated with Discordance in the Social Communication Questionnaire Compared to Best Clinician Estimate

Background:

Parent-based screening for autism spectrum disorder (ASD) is a common, efficient way to identify children needing further evaluation for ASD. To reduce misdiagnosis, it is important to identify factors that influence parent ratings of ASD characteristics. One bias that could increase misreporting is presence of the broader autism phenotype (BAP) in the informant (usually the mother). BAP reflects sub-clinical ASD traits (including pragmatic and broad communication difficulties and poor social skills) reported more frequently in families of children with ASD.

Objectives:

To determine whether BAP in mothers is associated with their reports of at-risk ASD characteristics for their children on an ASD screening instrument when discordant with best estimate clinical judgment. We assess whether discordance is due to over- or under-reporting on the screening instrument compared to the clinician’s best estimate.

Methods:

We used data on children aged 3-5 years with ASD or developmental disabilities potentially related to ASD, identified through health clinics, early intervention and special education programs, who participated in the Study to Explore Early Development. Upon enrollment, children were screened using the mother-completed Social Communication Questionnaire (SCQ). To increase sensitivity, SCQ scores ≥ 11 were considered positive for ASD risk. These children underwent full developmental assessments, which included the Autism Diagnostic Observation Schedule (ADOS) and Autism Diagnostic Interview–Revised (ADI-R). Children who had previous ASD diagnosis or clinician suspicion of ASD underwent full developmental assessment, even if SCQ scores were <11. Best clinical estimate of ASD was determined using the Ohio State University Autism Rating Scale (OARS). A study clinician condensed information from the ADOS, ADI-R, SCQ, and clinician impression to complete a five point rating scale of certainty that a child has ASD; scores ≥4 were considered “positive”. Screening discordance was defined as “over-reporting” (a positive SCQ with negative OARS) or “under-reporting” (negative SCQ and positive OARS). The mother’s spouse, close family member, or friend completed a standardized questionnaire, the Social Responsiveness Scale-Adult (SRS-A), to describe mother’s social traits. BAP was defined as a SRS-A T-score ≥ 60. Regression analyses calculated associations between maternal BAP and risk of screening discordance.

Results:

Our sample included 819 mothers with data on relevant instruments. Seventy-nine mothers were BAP positive (9.6%) and 308 (37.6%) had discordance with clinical judgement. Mothers with BAP had 1.24 times the risk of discordance compared to mothers without BAP (95% confidence interval (CI): 0.96, 1.61, P= 0.1). Risk of over-reporting in mothers with BAP was 1.43 times that of mothers without BAP (95% CI: 1.08, 1.90, P=0.01). Under-reporting was less likely in the BAP group compared to the non-BAP group (RR=0.53; 95% CI: 0.17, 1.60, P=0.3), but not statistically significantly. Effect estimates were similar when a positive SCQ was defined as ≥15.

Conclusions:

Mothers with BAP were more likely to over-report child ASD symptoms on the SCQ and presence of BAP in the mother was not associated with under-reporting of ASD symptoms in the offspring. Further work will examine differences by child ASD severity and whether specific SRS-A domains are associated with screening discordance.