An Exploratory Study of Socioeconomic Status, Air Pollution, and 13 Other Variables as Predictors of U.S. State Autism Spectrum Disorder Rates.

This exploratory study determined which of 14 selected predictors of autism spectrum disorder (ASD) found in individual-level research are predictive of ASD prevalence rates in the 50 U.S. states without and with statistical control for socioeconomic status (SES). Analyses used 2017 state ASD rates and SES, race, Big Five personality, IQ, urban population percent, air pollution, health care providers per population, physician shortage, per pupil spending, PN-3 policy and strategy, percent without health insurance, Medicaid-CHIP enrollment, maternal age, prepregnant obesity, and low birth weight variables based on various samples largely from 2017. ASD rates correlated significantly with each of the 15 potential predictors except for percent uninsured, Medicaid-CHIP enrollment, and air pollution (p = .051). However, when each of the 14 potential predictors entered alone on the second step of a regression equation with SES controlled, only race, personality, urbanization, air pollution, PN-3 policy and strategy, and maternal age were significant predictors. Additionally, an equation with these six predictors entered simultaneously on a second step showed that only SES and air pollution were significant. In another equation with only SES and air pollution entered as predictors, they jointly accounted for 55.7% of the variance in state ASD prevalence rates. Both higher SES and greater air pollution were associated with higher ASD prevalence. There was no evidence of multicollinearity or spatial autocorrelation in the 15 regression equations. These results suggest that considering SES and air pollution could prove beneficial in aggregate-level or individual-level analysis of factors associated with an autism diagnosis.