Heritability of quantitative autism spectrum traits in adults: A family‐based study

Abstract

Autism spectrum disorder (ASD) comprises a multi‐dimensional set of quantitative behavioral traits expressed along a continuum in autistic and neurotypical individuals. ASD diagnosis—a dichotomous trait—is known to be highly heritable and has been used as the phenotype for most ASD genetic studies. But less is known about the heritability of autism spectrum quantitative traits, especially in adults, an important prerequisite for gene discovery. We sought to measure the heritability of many autism‐relevant quantitative traits in adults high in autism spectrum traits and their extended family members. Among adults high in autism spectrum traits (n = 158) and their extended family members (n = 245), we calculated univariate and bivariate heritability estimates for 19 autism spectrum traits across several behavioral domains. We found nearly all tested autism spectrum quantitative traits to be significantly heritable (h2 = 0.24–0.79), including overall ASD traits, restricted repetitive behaviors, broader autism phenotype traits, social anxiety, and executive functioning. The degree of shared heritability varied based on method and specificity of the assessment measure. We found high shared heritability for the self‐report measures and for most of the informant‐report measures, with little shared heritability among performance‐based cognition tasks. These findings suggest that many autism spectrum quantitative traits would be good, feasible candidates for future genetics studies, allowing for an increase in the power of autism gene discovery. Our findings suggest that the degree of shared heritability between traits depends on the assessment method (self‐report vs. informant‐report vs. performance‐based tasks), as well as trait‐specificity. Lay Summary We found that the scores from questionnaires and tasks measuring different types of behaviors and abilities related to autism spectrum disorder (ASD) were heritable (strongly influenced by gene variants passed down through a family) among autistic adults and their family members. These findings mean that these scores can be used in future studies interested in identifying specific genes and gene variants that are associated with different behaviors and abilities related with ASD.

Sleep and Activity Patterns in Autism Spectrum Disorder

“…Each ASPE participant completed a 1-hour phone interview to collect general demographic information (see Supplementary Table 1) and assess psychiatric and developmental history, social behavior and autism spectrum traits to determine if they met criteria for autism spectrum disorder (ASD) as defined in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) 1 . This assessment identified probands and their family members who met the criteria for ASD according to a DSM-5 diagnostic checklist or other psychiatric disorders, as well as family members who did not meet criteria for such disorders 44 . All probands and family members were also assessed using informant-reporting (most often parent) for the Social Responsiveness Scale (SRS) 48,49 , the Broad Autism Phenotype Questionnaire (BAPQ) 50 , and the Behavior Rating Inventory of Executive Function (BRIEF) 51 .…”

Allelic contribution of Nrxn1α to autism-relevant behavioral phenotypes in mice

“…To determine whether a CNV in an intron near the 3' end of the Nrxn1 gene elicits behavioral phenotypes, we also generated a mouse model carrying a ~20 kb deletion at intron 17 of Nrxn1α (Nrxn1 ΔIntron17/+ or Nrxn1 ΔIntron17/ΔIntron17 ; subsequently referred to as ΔIntron17/+ or ΔIntron17/ΔIntron17) that is similar to a CNV identified in an individual on the autism spectrum (in the Autism Spectrum Program of Excellence (ASPE) cohort [27]). Upon examining the expression levels of all three different Nrxn1 isoforms in this mouse model, we found that both Nrxn1α and Nrxn1γ transcript levels were unaffected in these mice (Fig 1F and 1H).…”

“…The deletion of exon 9 was confirmed by PCR analysis using primers flanking the deleted region and within the exon 9 sequence. To study a CNV identified in an individual on the autism spectrum [27], mouse homologue of the ~20 kb deleted region intron 17 of Nrxn1α was identified and deleted using the CRISPR/Cas9-medicated genomic editing approach. Two sgRNAs (5' AATATGTGGGCAAGCTGGGT TGG 3' and 5' GAAATGGTACCTTTGATCTA AGG 3') flanking the deletion region in intron 17 of Nrxn1α were injected together with Cas9 protein into 1-cell zygote of C57BL/6J/SJL genetic background.…”

What are quantitative traits and how can they be used in autism research?

“…As noted above, utilization of quantitative trait measures in this manner also improves statistical power, because these measures do not rely on categories or counts of people, but instead provide continuous values for analysis on everyone in a study sample. Thus, studies seeking to detect associations and identify novel factors and genetic variants related to likelihood of autism may benefit from the use of quantitative traits (Sagiv et al, 2015; Taylor et al, 2021).…”

“…Commonalities or differences across these outcome metrics may also inform our understanding of specificity of associations and/or of underlying genetic and neurobiologic pathways. Furthermore, comparability of associations with contributing etiologic factors across quantitative trait measures and diagnosis, as has been seen for some quantitative traits related to autism (Lundstrom et al, 2012; Lyall et al, 2022; Taylor et al, 2021), would argue in favor of such quantitative traits capturing pathways involved in the diagnosis. More of this type of comparative work within the same study populations is needed.…”