Phenotype and Eye Tracking Heritability in Twin Pairs
Studying twin pairs can give valuable insight into genetic and environmental factors contributing to Autism Spectrum Disorder (ASD). Studies examining twins indicate about a 70-95% percent concordance rate for autism in monozygotic twins and about 30 percent concordance for dizygotic, same-sex twins (e.g., Hallmayer et al., 2011; Nordenbaek et al., 2014). In addition, symptom severity within twin pairs, based on parent questionnaires, concluded that extreme social and repetitive behavior symptoms of autism had high heritability (Frazier et al., 2014). Quantitative measures for reciprocal social behavior also indicate greater trait concordance for monozygotic twins than dizygotic twins (Marrus et al., 2015).
To investigate in a sample of twins, referred for diagnostic evaluation, whether there is high concordance in directly assessed, phenotypic profiles of monozygotic and dizygotic toddlers.
Participants included 8 sets of twins ranging in age from 16 to 35 months (mean=28.75 months; SD=6.57) including 3 sets of monozygotic (MZ) twins and 5 sets of dizygotic (DZ) twins. Participants received a clinical evaluation and underwent an eye tracking session through a research study at the Marcus Autism Center. Diagnostic assessments included the Mullen Scales of Early Learning and the Autism Diagnostic Observation Schedule, 2nd Edition (ADOS-2) and a medical/developmental history was collected.
Of the MZ twins, all pairs were diagnosed with ASD. Of the DZ twins, 3 pairs were diagnosed with ASD, 1 pair with language delay, and one where one twin had no clinical features and the other had a subthreshold developmental delay. There was fair agreement on age of parental reported first concerns for MZ and DZ twins (ICC = 0.49; 0.38, respectively). Strong to almost perfect intraclass agreements, for both MZ and DZ twins on the ADOS-2 were found across total scores (ICC = 0.93; 0.91, respectively), social affect (ICC = 0.72; 0.84, respectively) as well as restricted and repetitive behavior (ICC = 0.85; 0.97, respectively). For the developmental measure, MZ twins had almost perfect and moderate agreement for visual reception and receptive language age equivalents respectively (ICC = 0.91; 0.69). DZ twins had moderate and almost perfect agreement for visual reception and receptive language (ICC = 0.63; 0.81). Expressive language levels showed the greatest heritability, with MZ twins showing strong agreement (ICC = 0.83) and DZ twins showing poor agreement (ICC = 0.29).
Overall, in this small, pilot sample, we explored various behavioral traits which could shed light on differences in concordance between MZ and DZ twins. Both MZ and DZ twins showed strong agreement on autism symptomatology, precognitive abilities and receptive language. Expressive language skills show the greatest level of concordance, with MZ twins having significantly greater similarities than DZ twins. This language level difference could account for some of the differences we noted in age of first concerns. As our study is longitudinal, we expect to add about 30 more twin pairs to our sample as well as analyze eye-tracking to have another, objective measure to assess for concordance prior to the annual meeting.