Polychlorinated Biphenyl Exposure and Neuropsychological Status Among Older Residents of Upper Hudson River Communities

Edward F. Fitzgerald; Erin E. Belanger; Marta I. Gomez; Michael Cayo; Robert J. McCaffrey; Richard F. Seegal; Robert L. Jansing; Syni-an Hwang; Heraline E. Hicks


Environ Health Perspect. 2008;116(2):209-215. 

In This Article


Table 1 summarizes the background characteristics of the 253 study participants. Reflecting the study design, age ranged from 55–74 years with a mean of 63.9, and 50% were men. Nearly 60% had some college education. Twenty percent were current cigarette smokers and 27% were past smokers (data not shown). Ninety-seven percent of all participants were white, and 99% were non-Hispanic (data not shown). According to the industrial hygienists, fewer than 5% were occupationally exposed to mercury or pesticides (data not shown). The mean serum total PCB concentration was 3.6 ppb (wet weight) or 537 ppb (lipid basis). The nine dioxin-like PCB congeners were measured in the serum of 232 participants, and the mean TEQ concentration was 34.2 ppt (lipid basis). The means were 4.0 ppb (wet weight) or 738 ppb (lipid basis) for serum DDT + DDE, 3.1 μg/dL for blood lead and 0.30 μg/dL for blood mercury.

Table 2 lists background variables included in the final regression models for each neuropsychological test by domain. For many of the cognitive tests, performance was lower among men, decreased with age, and increased with intellectual function, education, and income. Reaction time decreased with age, and was lower among smokers, arthritics, and persons who reported fewer hours of sleep per week and who did not take nonsteroidal anti-inflammatory drugs (NSAIDs). Older persons and those with arthritis also performed less well on the pegboard test of motor function. BDI scores were higher among persons who reported using antidepressants and gout medications, whereas state anxiety and trait anxiety were higher among those who used sex hormones. Olfactory function was lower among older persons and those who smoked. Scores on the TOMM were within normal limits, indicating that the level of effort of all of the participants was adequate.

Table 3 displays the final multivariate models for the neuropsychological tests with log-transformed serum total PCB concentration after adjusting for the covariates in Table 2 . In the memory and learning domain, performance on the CVLT trial 1 score, declined as log serum total PCB concentration, increased (β = −0.576 per unit change in log-adjusted lipid basis serum PCB, p = 0.035). Specifically, an increase in serum total PCB concentration from 250 to 500 ppb (lipid basis) was associated with a 6.2% mean decrease in performance. The BDI increased with log serum total PCB concentration (β = 1.189, p = 0.007). That is, as serum total PCB concentration increased from 250 to 500 ppb (lipid basis), depressive symptoms increased on average by 19.2%. Only two participants, however, had a score on the BDI that indicated moderate to severe depression; the results were unchanged when the 15 participants who reported taking antidepressants were excluded. The only other significant association was for the WMS test of visual immediate recall (β = 1.016, p = 0.012). In this case, performance improved by 8.9% as serum total PCB concentration increased from 250 to 500 ppb (lipid basis).

To further examine the associations between PCB exposure and the CVLT trial 1 score and the BDI, we divided the study population into quartiles of serum total PCB concentration and compared their adjusted mean test scores ( Table 4 ). The results indicated that the deficits in the CVLT trial 1 score were limited to the highest PCB quartile, whereas scores on the BDI rose in both the third and fourth quartiles. Additional analyses by age and sex (not shown) revealed that the association between serum total PCB concentration and the CVLT trial 1 score was significant among persons 55–64 years of age (β = −0.872 per unit change in log-adjusted lipid basis serum PCB, p = 0.038) but not significant among those 65–74 years of age (β = −0.406, p = 0.311), and similarly significant among men (β = −0.958, p = 0.021) but not women (β = −0.372, p = 0.314). To further evaluate these differences, age was categorized into three groups and analyzed jointly with sex. The results indicated that the PCB–CVLT trial 1 score association was strongest among men 55–60 years of age (β = 1.797, p = 0.006). Effect modification by income was also observed for the CVLT, trial 1 score, with a significant association with serum PCB level only apparent for those with total family incomes of < $45,000 (β = −0.893, p = 0.017 versus β = −0.384, p = 0.403 for those with incomes of ≥ $45,000).

The association between serum total PCB concentration and the BDI was also significant among persons 55–64 years of age (β = 2.934, p = 0.002) and not for persons age 65–74 years (β = 1.069, p = 0.136). The BDI, however, showed a pattern by sex that was opposite that of the CVLT, trial 1 score, with a stronger association among women (β = 2.177, p = 0.009) than men (β = 1.269, p = 0.089). When age was categorized into three groups and examined together with sex, the strongest association was for women 55–60 years of age (β = 1.976, p = 0.169).

Finally, multivariate models were constructed for these two tests with individual serum congener concentrations ( Table 5 ). For the CVLT, trial 1 score, significant negative associations were apparent for PCB-105, 118, 138, 170, 180, and 194. For the BDI, the significant positive associations were observed for PCB-153 through 194. A similar multivariate analysis of PCB TEQs revealed that scores on the BDI increased as PCB–TEQ concentration increased (β = 0.266 per unit change in log adjusted lipid basis TEQ, p = 0.065, data not shown).


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: