Prenatal Air Pollution Exposure and Autism: Preliminary Findings from Two Prospective High-Familial Risk Birth Cohorts

Oral Presentation
Thursday, May 10, 2018: 2:09 PM
Arcadis Zaal (de Doelen ICC Rotterdam)
H. E. Volk1, B. Y. Park2, F. Lurmann3, H. Minor3, R. McConnell4, S. Ozonoff5, L. A. Croen6, M. D. Fallin1, I. Hertz-Picciotto7 and C. J. Newschaffer8, (1)Wendy Klag Center for Autism and Developmental Disabilities, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, (2)Mental Health, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, (3)Sonoma Technology, Inc., Petaluma, CA, (4)University of Southern California, Los Angeles, CA, (5)Psychiatry and Behavioral Sciences, University of California at Davis, MIND Institute, Sacramento, CA, (6)Division of Research, Kaiser Permanente, Oakland, CA, (7)University of California at Davis, Davis, CA, (8)AJ Drexel Autism Institute, Philadelphia, PA
Background: Prenatal air pollution exposure has been associated with autism spectrum disorder (ASD) in case-control studies, but not yet examined in a prospective cohort in the US.

Objectives: To examine the relationship between prenatal air pollution exposure and risk of ASD and non-typical development in the Early Autism Risk Longitudinal Investigation (EARLI) and Markers of Autism Risk in Babies, Learning Early Signs (MARBLES) cohorts.

Methods: EARLI and MARBLES enrolled pregnant mothers who already have a child with ASD and followed the new infant longitudinally through 36 months of age. Maternal residences and corresponding dates from pregnancy to birth were collected via questionnaire and extracted from prospective follow-up contact reports. Monthly near-roadway air pollution (NRAP) estimates were assigned to geo-coded locations using the CALINE line-source dispersion model, for California locations, or the MOVES model for locations outside of California. Weekly measures of exposure to criteria air pollutants (nitrogen dioxide (NO2), ozone, and particulate matter less than 2.5 or 10 microns in diameter (PM2.5, PM10)) were assigned to each location using inverse distance weighting from the nearest Environmental Protection Agency (EPA) monitor. ASD diagnosis at 36 months was determined based on Autism Diagnostic Observational Schedule (ADOS) score and by meeting DSM-IV-TR criteria for Autistic Disorder or PDD-NOS. Typical development was defined as not meeting ASD criteria and having Mullen Scales of Early Learning Scores within 1.5 SD of the mean, and ADOS more than 3 points below the ASD cut-off. Separate logistic regression models were used to examine the association between NRAP exposure during pregnancy and ASD risk. Distributed lag models (DLM) were used to identify windows of susceptibility to NO2, ozone, PM10, and PM2.5 exposure. All models were adjusted for recruitment site / parent study, child gender, child race, ethnicity, and birth year, maternal age at birth and education level. Analyses were conducted on 114 children from EARLI and 96 from MARBLES.

Results: Average NRAP and NO2 exposure was elevated for US East Cost EARLI study locations (Maryland and Pennsylvania), compared to West Cost EARLI and MARBLES study locations (Northern California). PM10 and PM2.5 exposure was similar across study locations. Ozone level was lower at the Kaiser Permanente EARLI site, compared to others. After adjustment for confounders, increased prenatal NRAP exposure was associated with increased ASD risk, compared to TD, in EARLI (OR=2.2 per 1 SD (~3ppb) change in exposure, 95% Confidence Interval (95%CI) 1.2-4.3), but not in MARBLES (OR=0.85, 95%CI 0.4-1.6). DLM results from MARBLES suggest that increasing exposure to ozone during mid-gestation and to NO2 and PM2.5 during mid-late gestation may be associated with modestly increased ASD risk OR per five unit changed ranged from OR=1.1-1.5. Windows of susceptibility were not identified in EARLI.

Conclusions: Preliminary findings suggest that prenatal exposure to air pollution may be associated with increased ASD risk. However, variation in air pollution exposure based on geographic location, that could suggest different patterns of regional confounding, pollution source, and background susceptibility from other environmental and genetic factors all need additional exploration.

See more of: Prenatal Autism Risk Factors
See more of: Epidemiology