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


Clinically, patterns of performance on the CVLT are used to detect learning and memory impairments in patients with dementia, brain tumors, and brain injury (Delis et al. 2000). More specifically, the trial 1 score measures the capacity to learn and retain immediate verbal information, an ability that decreased as serum total PCB concentration increased. The stronger association among men is consistent with the fact that women usually outperform men on the CVLT (Ragland et al. 2000). The reason this association was greatest among men 55–60 years of age is unclear. On the one hand, such a finding appears to contradict the notion that older persons are at higher risk of neurotoxicity from PCBs. On the other hand, it may reflect an acceleration of age-related deficits that are normally not observed until 65 years of age (Paolo et al. 1997) in a group that is already at high risk as a result of their sex. The finding that the association was only significant among those whose annual income was < $45,000 is similar to that found in some studies of children that indicate that the effects of low-level exposure to neurotoxins such as lead are greater among those from poorer socioeconomic backgrounds (Bellinger 2004).

The BDI, which measures the occurrence of 21 symptoms of depression, increased as serum total PCB concentrations increased. The association was stronger among women than men, perhaps reflecting the greater risk of depression among women (Culbertson 1997). The sex difference for depression was age dependent, with the magnitude of the association strongest for women 55–60 years of age, again suggesting a greater susceptibility to PCB-related effects in this age group.

The current study was designed to replicate and expand upon the work of Schantz et al. (1999, 2001). In that study, neuropsychological assessments similar to those conducted here were performed on 101 Lake Michigan fish consumers and 78 non-fish-eating controls 49–86 years of age. The results indicated that three tests of learning and memory were negatively associated with PCB exposure. They included the CVLT, trial 1 score and semantic cluster ratio, and the WMS test of logical delayed recall. The first test was also significantly and negatively associated with PCB exposure in the current investigation, whereas the other two tests were not. The discrepancy in results for the latter two tests may be a consequence of differences in current PCB body burden between the two studies. In Schantz et al. (2001), the median serum total PCB concentration was 7.9 ppb (wet weight), with a maximum of 75 ppb. In contrast, the median serum total PCB concentration for the current study was 3.2 ppb, with a maximum of 19.3 ppb. As in the present study, Schantz et al. (2001) found no tests of other cognitive domains or of motor function were associated with PCB exposure.

The current study builds on the work of Schantz et al. (2001) by extending the outcome assessment to include affective state and olfactory function and the exposure assessment to include individual PCB congeners. Regarding the latter issue, the CVLT, trial 1 score and the BDI were associated with multiple serum PCB congeners, particularly the more heavily chlorinated congeners. Similar findings have been reported for neurobehavioral outcomes in PCB-exposed children enrolled in the Oswego Newborn and Infant Development Study (Stewart et al. 2000). These findings contradict animal and in vitro research that suggests that lightly chlorinated congeners, especially those with ortho-substitution such as PCB-28, have the greatest potential for neurotoxicity (Seegal 2001). This apparent discrepancy may be explained by the fact that the concentrations of the more heavily chlorinated congeners such as PCBs 153, 180, and 138 were 10 times greater than PCB-28 in this and most other human studies (Hansen 1998). In addition, congener concentrations in human samples are highly intercorrelated (Longnecker et al. 2000), making it difficult to identify which congeners are responsible for observed associations.

Schantz et al. (2001) did not examine depression as an outcome, but depression has been linked to PCBs in some studies of occupationally or environmentally exposed adults. For example, complaints of depression, memory loss, insomnia, and nervousness have been reported among workers exposed to PCBs at capacitor plants (Fischbein et al. 1979). More recently, Peper et al. (2005) assessed the neuropsychological status of 30 employees exposed to indoor air contaminated with PCBs from elastic sealants in a German school building compared with that of 30 unexposed controls. The exposed group tended to report more symptoms characteristic of depression, such as distractibility, tiredness, slowness, and less emotional well-being. The mean serum total PCB concentration for the exposed group was 4.5 ppb in the German study, a value more comparable to that of the current study than that of the Lake Michigan fish eaters studied by Schantz et al. (2001).

In addition to the CVLT-trial 1 score, serum total PCB concentrations were significantly associated with a second test of memory and learning, the WMS test of visual immediate recall in the current study. In this case, however, the association was opposite to that hypothesized, with performance increasing as serum total PCB concentrations increased. The reason for this unexpected finding is unclear. Interestingly, the exposed group in the German indoor air study also performed better on the WMS tests of visual immediate and delayed recall than did the controls. Peper et al. (2005) did not include the CVLT, so direct comparison of results for that test is not possible. Schantz et al. (2001) also reported an anomalous finding, with performance on the WMS test of delayed logical memory improving with increasing serum DDE concentration.

Regarding mechanisms, the pathways by which PCBs may affect neuropsychological status in humans are not well understood. As noted previously, alterations in brain levels of neurotransmitters such as dopamine and serotonin may be involved. Dopamine depletion has also been hypothesized as the explanation for excess mortality from Parkinson disease among women who have been occupationally exposed to PCBs in the capacitor industry (Steenland et al. 2006). Another possible mechanism is through the thyroid system. PCB exposure has been related to changes in thyroid hormone levels in humans (Hagmar 2003), and alterations in thyroid hormones have been associated with adult neuropsychological deficits (Wilson et al. 1998).

Caution must be exercised in the interpretation of this study's results. First, performance on only 1 of the 10 subtests of the CVLT significantly declined with serum PCB concentrations. Schantz et al. (2001) found significant associations with only 2 CVLT subtests despite the higher level of exposure in that study, but positive results with additional subtests certainly would have helped to strengthen confidence in our findings. Second, contrary to expectations, performance on another test of memory and learning, the WMS test of visual immediate recall, significantly improved with serum PCB concentration. The meaning of this finding is unclear, but it is consistent with a similar observation by Peper et al. (2005). Third, multiple statistical comparisons were performed at a nominal error of 5% per test; therefore, it is possible that the observed associations are due to chance. One solution would have been to adjust the p-value to account for the 34 tests, but many believe that such an approach may be inappropriate (Savitz and Olshan 1995). Another procedure would have been to combine tests across domains and assess global effects. This method was not followed for two major reasons: a) comparability, because Schantz et al. (2001), Peper et al. (2005), and many other investigations have analyzed each test individually; and b) interpretability, because even tests within the same domain do not measure exactly the same function.

Instead, we chose to focus on the concordance of our results with those of other studies to help differentiate valid from spurious associations. The fact that the one subtest of the CVLT that was significant in the current study was also one of the two reported by Schantz et al. (2001) helps to validate that finding. In addition, blood concentrations of PCBs were unrelated to any measures of executive function, visual spatial function, or motor function in either study. Other studies also have found that depressive symptoms are more common among PCB-exposed persons. Although unpredicted, the finding of improving performance on the WMS test of visual immediate recall with serum PCB concentration is also consistent with at least one previous study of PCB-exposed adults.

Strengths of the current study include sensitive and objective measures of exposure and outcome, a focus on long-term residence to maximize the opportunity for exposure, and a comprehensive set of potential confounders. The results are consistent with the only other study of PCB exposure and neuropsychological status in older adults, which also found memory and learning deficits in persons with higher serum PCB concentrations. The data are also consistent with other research reporting a higher prevalence of depressive symptoms among PCB-exposed adults. Given the possibility of spurious associations due to multiple statistical comparisons and the contrary-to-expected finding of a protective effect for one test of memory and learning, additional studies of PCB exposure among older persons are needed to determine more definitely whether they are a sensitive subgroup.


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: