Correlates and Turning Points of Adaptive Functioning Trajectories & Longitudinal Associations with Autism Symptoms from Early Childhood to Adolescence

Abstract

Background: Previous research has shown that heterogeneous trajectories of adaptive functioning in autism are associated with baseline characteristics such as autism symptom severity and IQ. However, evidence remains limited regarding to what extent the developmental variability of adaptive functioning can be accounted for by family characteristics and autism symptoms beyond baseline. It is also unknown whether turning points are present along the trajectories reflecting potential risk or opportunities for change. Objectives: We aimed to examine 1) the parallel-process trajectories of communication, daily living, and social skills from ages 2 to 17 years in an inception cohort of autistic children, and 2) the developmental variability of adaptive functioning and associations with child/family characteristics and autism symptom trajectories. Methods: The current sample (N=406; all diagnosed with autism at ages 2-5 years) was drawn from Pathways in ASD, a large Canadian prospective study. Children’s adaptive functioning and autism symptoms were respectively assessed with Vineland Adaptive Behavior Scales (VABS) and Autism Diagnostic Observation Schedule (ADOS) across 4 to 6 visits, with data restructured by chronological age for parallel-process latent growth curve modeling. Upon deciding the optimal functional form, latent class growth analysis was performed to identify VABS trajectory subgroups. Child (sex, age of diagnosis, baseline nonverbal IQ) and family characteristics (household income, caregiver’s education, race/ethnicity, and nativity), and ADOS growth parameters were included to examine their associations with VABS trajectory subgroup membership. Results: The piecewise latent growth model best described VABS trajectories with two turning points identified at transitions into school age (~6 years) and youthhood (~10 years). We parsed four VABS trajectory subgroups (entropy=.92; Figure 1) that vary by functioning level, change rate for certain VABS domains, and developmental periods segmented by the turning points. Household income and nonverbal IQ, but not autism symptom severity at baseline, remained significant covariates of VABS trajectory membership when adjusting for other covariates. About 16% of our sample was in Class 4, which is characterized by notable early growth across VABS domains and less decline in social functioning upon transitioning into youth. Children in Class 4 had higher yet highly variable nonverbal IQ (M=82.4, SD=25.8) at diagnosis, showed more reductions in social-affect symptoms across childhood (Figure 2), and were more likely to come from a higher-income family (b=.37 to .52, all SE=.13, p<.01). In contrast, the lowest functioning group (Class 1; 20% prevalence) was marked by an early decline across VABS domains despite reductions in autism symptoms, followed by late improvement (particularly in daily-living skills) during school age and then decline in adolescence. Conclusions: This study demonstrated the heterogenous pathways of adaptive functioning domains across developmental stages in autism. We identified key individual and family correlates of developmental variability (e.g., nonverbal IQ, household income) as well as crucial transition periods for functional declines or improvement, which have important implications for understanding risk and resilience processes leading to various developmental outcomes. Our findings highlight the significance of taking developmental and multidimensional approaches to studying autism-related outcomes and thus informing timely and tailored supports for individuals on the spectrum.

Yun-Ju Chen 陳韻如

Postdoctoral Fellow