Association of Polychlorinated Biphenyls and Organochlorine Pesticides with Autism Spectrum Disorder in Jamaican Children

Background: Polychlorinated biphenyls (PCBs) and organochloride (OC) pesticides are persistent organic pollutants (POPs) of public health concern. They are hypothesized to play a role in disorders such as autism spectrum disorder (ASD) by interfering with the endocrine or immune systems. Jamaicans may be at high risk for exposure to PCBs and OC pesticides as they have been detected in water, sediment, fauna, and shrimp across Jamaica.

Objectives: To examine the associations of PCBs and OC pesticides with ASD among Jamaican children.

Methods: We conducted an age- and sex-matched case-control study in which we enrolled n=141 pairs of Jamaican children 2-8 years old. ASD cases were confirmed using the Autism Diagnostic Observation Schedule, second edition and Autism Diagnostic Interview–Revised. Developmental delay was ruled out in typically developing (TD) controls using the Social Communication Questionnaire. Socioeconomic status and food frequency questionnaires were administered to the parents/guardians of each child. We collected 2-3 mL of whole blood for genetic analysis by the Human Genetics Center in Houston, Texas and 4-5 mL of serum for analysis of PCBs and OC pesticides by the Michigan Department of Health and Human Services. For PCBs and OC pesticides with ≥30% above the limit of detection (LoD), we replaced observations below LoD with 2^-1/2LoD and calculated arithmetic and geometric means for cases and controls, separately. We dichotomized concentrations at the 75^th percentile and used them as independent variables in conditional logistic regression models to assess the associations of these POPs with ASD. Finally, we conducted interactive models to explore possible interactions between POPs exposures and genotypes of three glutathione S-transferase (GST) genes in relation to ASD.

Results: A majority (79.4%) of cases and controls were male. The mean age of cases and controls was 59.6 months and 59.8 months, respectively. Compared to controls, cases had lower geometric mean concentrations of PCB-153 (14.3 ng/g-lipid vs. 16.5 ng/g-lipid), PCB-180 (6.4 ng/g-lipid vs. 7.2 ng/g-lipid), total PCB (26.0 ng/g-lipid vs. 29.4 ng/g-lipid) and 4,4’-DDE (54.7 ng/g-lipid vs. 76.6 ng/g-lipid), however none of these comparisons were statistically significant (all P>0.06). Using concentrations dichotomized at the 75^th percentile, PCB-150, total PCB, and 4,4’-DDE were not associated with ASD in either univariable or multivariable models (all P>0.09). We found a marginally significant interactive effect between PCB-180 and GSTP1 in relation to ASD (interaction term P=0.06); those with the Ile/Ile genotype had a different level of association (adjusted MOR=0.77, 95% CI: 0.53–1.23) compared to those with Ile/Val (adjusted MOR=0.51, 95% CI: 0.25–1.03) or Val/Val (adjusted MOR=0.59, 95% CI: 0.34–1.03) genotypes.

Conclusions: We found that cases had lower serum concentrations of PCB-153, PCB-180, total PCB, and 4,4’-DDE compared to controls, though differences were not statistically significant. These differences may be explained by diet, as one of the major routes of exposure to these POPs is through poultry and red meat; Jamaican children with ASD may eat less poultry and red meat compared to TD children. Furthermore, we reported a possible interactive effect between PCB-180 and GSTP1, which should be replicated in other studies.