Associations between Socioeconomic Status, Language Ability, and ASD Symptoms in Toddlers with Autism

Background: Autism Spectrum Disorders (ASDs) present major personal, community, and public health challenges due to their associated functional impairments and high cost of care. In addition, socioeconomic status (SES) is a critical, broad-reaching developmental factor that directly impacts language and social development in early childhood (e.g., Hart & Risley, 1995; Johnson et al., 2016). While the field of developmental research focused on ASDs has made strides in characterizing the unfolding of the disorder beginning in early childhood, less is known about the deleterious effects of low-SES on early development in ASDs. It remains unknown to what extent low-SES confers additional developmental risks, above and beyond the vulnerabilities associated with ASDs, in young children during critical periods for language development. Knowledge thereof will have the potential to inform targeted interventions for low-SES children with ASDs.

Objectives: Examine associations between socioeconomic status, language skills, and autism symptoms in toddlers with autism.

Methods: 45 toddlers with autism (13 females), ages 17- to 34-months (mean age = 24.7, standard deviation = 4.4 months), completed developmental and diagnostic assessments as part of their participation in an ongoing study of early brain development in autism. Participants completed the Mullen Scales of Early Learning, the ADOS-2, the Vineland Adaptive Behavior Scales, and the Child Behavior Checklist. Additionally, caregivers provided demographic information, including income, education, and racial/ethnic identity. Participants' postal codes were used to extract median income by postal code, an area-based measure of SES. Spearman's correlation analyses were conducted to test for associations between socioeconomic indices and outcome variables of interest (language, autism symptoms). Minimally verbal children with ASDs (lower quartile of sample on Mullen expressive language scores) were also compared to their peers with relatively more intact language capabilities (upper quartile) with regard to SES, using independent samples t-tests.

Results: Toddlers with autism showed negative associations between neighborhood SES (postal-code-based) and autism symptoms (ADOS-2; r = -.31, p = .08), and household income and parent-rated autism symptoms (CBCL ASD r = -.39, p = .08), though not significantly. Toddlers with ASDs showed positive associations between neighborhood SES, and expressive (r = .36, p = 0.04) and receptive (r = .34, p = 0.06) language, and overall development (Mullen Early Learning Composite; r = .32, p =.07). In comparison to their peers with relatively higher language skills, children with minimal language abilities were from lower income neighborhoods (t_(1,17) = -2.2, p = .04).

Conclusions: Young children with ASDs showed negative associations between SES and autism symptoms, and positive associations between SES and language abilities. Participants with lower language abilities (lowest quartile of sample; T-score ≤ 20) also came from lower-income neighborhoods, compared to their peers with relatively higher language skills (highest quartile of sample, T ≥ 40). Future analyses will include a larger sample size, with increased power to detect associations between SES and relevant outcome measures, stratifications by ethnicity and bilingual status, relations between SES and measures of functional brain connectivity, and associations between SES and language measures in our control sample of TD young children.