Neuropsychological measures of attention and impulse control among 8-year-old children exposed prenatally to organochlorines

Abstract

Background: We previously reported associations between organochlorines and behaviors related to attention deficit hyperactivity disorder among boys and girls at 8 years of age using a teacher's rating scale for a birth cohort in New Bedford, Massachusetts (USA).

Objectives: Our goal was to corroborate these findings using neuropsychological measures of inattentive and impulsive behaviors.

Methods: We investigated the association between cord serum polychlorinated biphenyls (PCBs) and p,p'-dichlorodiphenyl dichloroethylene (p,p'-DDE) and attention and impulse control using a Continuous Performance Test (CPT) and components of the Wechsler Intelligence Scale for Children, 3rd edition (WISC-III). Participants came from a prospective cohort of children born during 1993-1998 to mothers residing near a PCB-contaminated harbor in New Bedford. Median (range) cord serum levels for the sum of four prevalent PCBs [congeners 118, 138, 153, and 180 (ΣPCB4)] and p,p'-DDE were 0.19 (0.01-2.59) and 0.31 (0-14.93) ng/g serum, respectively.

Results: We detected associations between PCBs and neuropsychological deficits for 578 and 584 children with CPT and WISC-III measures, respectively, but only among boys. For example, boys with higher exposure to ΣPCB4 had a higher rate of CPT errors of omission [rate ratio for the exposure interquartile range (IQR) = 1.12; 95% confidence interval (CI): 0.98, 1.27] and slower WISC-III Processing Speed (change in score for the IQR = -2.0; 95% CI: -3.5, -0.4). Weaker associations were found for p,p'-DDE. For girls, associations were in the opposite direction for the CPT and null for the WISC-III.

Conclusions: These results support an association between organochlorines (mainly PCBs) and neuropsychological measures of attention among boys only. Sex-specific effects should be considered in studies of organochlorines and neurodevelopment.

Figure 1

Unadjusted scatterplots and adjusted negative binomial regression lines of the association between cord serum levels of ΣPCB₄ and errors of omission on the CPT by sex for 8-year-old children born in New Bedford, 1993–1998, adjusted for child’s age at examination, sex, and birth year; maternal age at birth and prenatal smoking; and maternal and paternal education. Adjusted rate ratios: females, 0.95 (95% CI: 0.86, 1.05); males, 1.11 (95% CI: 0.98, 1.27).

Figure 2

Unadjusted scatterplots and adjusted linear regression lines of the association between cord serum levels of ΣPCB₄ and processing speed on the WISC-III by sex for 8-year-old children born in New Bedford, 1993–1998, adjusted for child’s age at examination and sex, maternal age at birth and prenatal smoking, and maternal intelligence and maternal education. Adjusted β-values: females, 0.0 (95% CI: –1.0, 1.1); males, –2.0 (95% CI: –3.5, –0.4).