Emerging Divergence between Cognition and Adaptive Socialization Skills in Infants Who Develop ASD
Individuals with ASD exhibit delays in adaptive skills in comparison to age and IQ, and this gap appears wider in older versus younger individuals of average intelligence (Klin et al., 2007; Kanne et al., 2010). Since the majority of individuals with ASD do not have cognitive impairment (CDC, 2016) but still fail to achieve levels of independence in adulthood (Howlin et al., 2004; 2013), it is imperative to determine when this divergence emerges.
This study investigates the discrepancies between adaptive socialization and cognitive skills from 12 to 36 months in infants who develop ASD.
Participants included 171 infants (60.2% male; 90.1% White; 2.9% Black; 7% Other) participating in a longitudinal study on ASD from birth through age 3 with the following outcome diagnoses: ASD=21; Broader Autism Phenotype (BAP)=19; High Risk Unaffected (HR-UA)=29; and Low Risk Typical Development (LR-TD)=102. Infants were assessed at 12, 24, and 36 months (12 and 24m for LR) with the Mullen Scales of Early Learning and the Vineland Adaptive Behavior Scales, Second Edition, Survey Form. Mullen Visual Reception scores were used as a proxy for nonverbal cognition (See Table 1). Mixed effects linear growth regression models were conducted to compare linear trends between cognitive and socialization skills from 12-36 months across diagnostic groups. Linear trend estimates derived from these models were compared across groups using slopes with the associated 95% confidence intervals.
Regression slopes for adaptive socialization scores declined significantly from 12-36 months for the HR-ASD (t(118)=-3.27, p=0.001) and HR-BAP (t(107)=-3.70, p<0.001) infants (Figure 1). In contrast, socialization skills did not decline for HR-UA or LR-TD infants. Socialization scores for HR-ASD infants were significantly lower than LR-TD infants starting at 12-months, and lower than both LR-TD and HR-UA infants at starting at 24-months. In contrast, cognition for all groups remained stable from 12 to 24/36 months. Regarding the gap between adaptive socialization and cognitive skills, a significant interaction emerged for HR-ASD (t(99.3) = 2.95, p=0.004), HR-BAP (t(89.4)=3.68, p<0.001), and LR-TD (t(301)=2.21, p=0.028) infants, but slope trajectories for each domain for the LR-TD group did not differ significantly from zero (socialization p=0.109; cognitive p=0.145).
Adaptive socialization skills decline between 12 and 36 months for infants with both ASD or shadow symptoms of the disorder in comparison to both unaffected HR and LR infants, with differences evident as early as 12 months between infants who develop ASD and their unaffected peers. Though our sample has unusually high nonverbal cognition (which accounts for the significant gaps between cognition and adaptive skills in the LR-TD group), the gap between adaptive socialization skills and cognition widens over time for infants with ASD symptomatology, whereas it remains stable for infants unaffected by ASD. Given the pronounced deficits in adaptive behavior that are observed in older individuals with ASD, these findings underscore the need to assess for adaptive behavior early and repeatedly for infants with and at risk for the disorder.