Determinants of Autism Prevalence in Hispanics

Background:  Hispanic children have a lower rate of autism spectrum disorder (ASD) diagnosis compared to non-Hispanic counterparts according to U.S. surveillance data. Underlying explanations for this significant disparity in ASD prevalence are poorly understood.

Objectives:  This investigation aims to determine whether the lower prevalence of autism among California Hispanics can be explained by specific demographic and / or clinical factors.

Methods:  This retrospective cohort study includes births drawn from the Office of Statewide Health Planning and Development PDD-Birth files from 01/01/1991-12/31/2008 that survived their first year of life. Children diagnosed with autism between 1991 and 2012 by the California Department of Developmental Services (DDS) were identified (n=33,359). The remainder of the birth cohort served as controls (n= 6,897,246). Demographic variables and ICD-9-CM codes for medical diagnoses and procedures during pregnancy were assessed for confounding. Logistic regression models evaluated the relationship between ethnicity and autism, controlling for parental age and education, maternal diabetes and preeclampsia, child sex, gestational age, delivery payer, delivery mode, and multiple gestations.

Results:  While crude analyses found that Hispanic children were 27% less likely to have autism compared with their Non-Hispanic counterparts, autism odds were actually 14% higher for Hispanic children in models adjusted for clinical and demographic factors (OR 1.14, 95% CI 1.11, 1.18). Similarly, crude odds of autism were 35% lower in Hispanic children with foreign-born parents, and again the association reversed with proper adjustment to parallel the primary analysis (OR 1.15, 95% CI 1.11, 1.20).

Conclusions:  Our investigation reinforces the importance of careful adjustment for confounders in assessing epidemiologic risk. Social determinants greatly influence health disparities. The influence of pregnancy complications known to enhance autism susceptibility played a substantial role in this analysis, but more striking still was the impact of low parental educational attainment – a proxy for socioeconomic status – on autism risk in Hispanics. Parents with more restricted education may not fully understand the implications of neurobehavioral differences in their children. Those who seek help may face major obstacles advocating for their child and accessing much-needed health and educational resources. These hurdles may be more daunting for foreign-born Hispanic parents who may also face language barriers and hold accentuated cultural views.