PBDE Exposures during Pregnancy and Risk of Autism Spectrum Disorders at 3 Years: Results from the Prospective MARBLES Study

Background:   Polybrominated diphenyl ethers (PBDE) are flame retardants that were used in consumer products such as electronic devices, construction materials, foam furniture, car seats, and home textiles to reduce their flammability. Both federal and state bans on PBDE mixtures (penta and octa-BDE formulations) were enacted in the early to mid 2000s in response to both evidence that levels in humans were rising and results from toxicologic studies showing adverse neurodevelopmental outcomes in experimental animals.  Nevertheless, exposure to PBDE is still widespread in the general population as a result of long half-lives of these compounds and bioaccumulation up the food chain. PBDE can interact with pathways involved in normal brain development, such as thyroid hormone homeostasis and cell signaling.

Objectives:   To study the associations between exposure to four prevalent PBDEs during pregnancy and the risk of having a child with Autism Spectrum Disorders (ASD) or other developmental concerns (ODC), which include speech and language delay, hyperactivity, and broader autism phenotype.

Methods:   Mothers participating in the ongoing MARBLES (Markers of Autism Risk in Babies – Learning Early Signs) Study have previously delivered a child who received an ASD diagnosis, and are either pregnant or planning a pregnancy. Interviews, self-administered questionnaires, and blood specimens are collected at multiple time points from enrollment through the child’s 3^rd birthday. The child is assessed on various instruments by expert clinicians during study visits at the participant’s home (6 and 12 months) and the UC Davis MIND Institute (24 and 36 months). At 36 months these include the Mullen Scales of Early Learning, Vineland Adaptive Behavior Scales, SRS, ADOS, and ADI-R, which are used to assign final diagnoses: ASD (n = 26), ODC (n = 29) or typically developed (TD, n = 77), based on a consensus clinical best estimate from two independent clinicians and, for ASD only, results of the ADOS and ADI-R. We measured four PBDEs using gas chromatography/ mass spectrometry/mass spectrometry, along with lipids in repeated blood samples collected during pregnancy (1 to 3 samples per woman). Adjusted ORs from multiple logistic regression models are reported for a 2 fold increase in PBDE concentrations (ng/g lipids), along with 90% confidence intervals.

Results:   PBDEs were detected in 82% to 99% of maternal pregnancy plasma samples. After adjustment for maternal age, body mass index, maternal education and year of birth, PBDE153 trended to be associated with increased risk of ASD (OR = 1.35, 90%CI: 1.00; 1.81). This association was strengthened when we restricted analysis to boys (OR = 1.53, 90%CI: 1.07; 2.18). The number of girls with ASD was too few for separate analysis (n=3). No other PBDE was associated with the risk of ASD (p-values > 0.26) or ODC (p-values > 0.35), compared to TD. 

Conclusions:   These results from the MARBLES Study indicate PBDE153 is associated with increased risk of ASD, especially among boys. This is, to our knowledge, the first report of an association between clinically confirmed ASD diagnoses, and PBDEs measured prospectively during critical periods during gestation.