Investigating Prenatal Exposure to Groups of Air Toxics and Autism Spectrum Disorder Using Exploratory Factor Analysis

Friday, May 13, 2016: 5:30 PM-7:00 PM
Hall A (Baltimore Convention Center)
S. L. Stacy1, E. Talbott2 and J. Rager3, (1)Epidemiology, Brown University School of Public Health, Providence, RI, (2)University of Pittsburgh, Pittsburgh, PA, (3)University of Pittsburgh GSH, Pittsburgh, PA
Background:  Autism spectrum disorder (ASD) is a neurodevelopmental disorder characterized by impaired social interaction and rigid behaviors and routines. The prevalence of ASD has increased markedly over the last several decades, motivating investigations into possible causes and risk factors. One area of interest is the role of groups of ambient air pollutants on the development of ASD.

Objectives: N/A

Methods:  Estimates of 30 ambient air toxics from the 2005 National Air Toxics Assessment, modeled at the census tract level, were linked to 217 cases of ASD and 224 controls born in southwestern Pennsylvania from 2005 to 2009. An exploratory factor analysis (varimax rotation) was conducted to reduce these 30 pollutants to a set of key predictors. Factor scores were calculated using two methods: index scores based on sums of quartiles of exposure and least squares regression. These scores informed two sets of logistic regression models of ASD risk, adjusted for mother’s age, race, education, smoking, child’s year of birth and sex. The results of each method for calculating factor scores were compared.

Results:  The air toxics loaded onto 7 main factors. Regardless of the method used to calculate the scores, the factors with the most elevated odds ratios appeared to represent traffic, combustion sources, and plastic and rubber manufacturing.

Conclusions: This type of approach can be used in future studies of ASD and environmental risk factors to identify groups of potentially harmful contaminants and pollution sources.

See more of: Epidemiology
See more of: Epidemiology