(This is the fourth in a series of five articles on chemical contamination of fish in the mid-Columbia region.)
The second article of this series cited a 2007 report by the Oregon Environmental Council (OEC), “Pollution in People,” on contamination of 10 individuals, all prominent in Oregon politics, culture, or environmentalism, by six classes of chemicals known or thought to be toxic. Here I focus on mercury and polychlorinated biphenyls, the major contaminants likely to be acquired by eating fish. The OEC is careful to point out that its study does not meet the design standards of the scientific research community. Nevertheless, it contains some striking results which at least should motivate a more rigorous follow-up.
One OEC subject was the executive director of the Umatilla Tribes. My previous use of this report was to point out that this participant’s blood-serum mercury and PCB levels were close to the median values for all ten study subjects, contradicting claims by the Umatilla and Oregon Public Broadcasting that, as a result of a fish-rich diet, he had acquired especially high body burdens of these (and implicitly other) toxics. However, the OEC data contain other gems.
If one graphs mercury concentration against subject age, a remarkable relationship is found: mercury level is proportional to age over the range of 22 to 62 years, with the exception of one extreme outlier, a 65-year-old woman with the mercury level predicted, by extrapolation, for a 10-year-old. In a recent interview, she described a dietary history suggesting several hypotheses to explain her relative freedom from mercury contamination. The straight line through the data points (minus the outlier) extrapolates to the origin; newborns should be negligibly contaminated. Most toxicology models would predict this result, increasing the likelihood that the linear relationship between mercury level and age is not a statistical fluke. The mercury assay of the Umatilla Executive Director groups with the results for others of similar age. Human mercury contamination must depend most strongly on other factors than fish consumption. This conclusion is supported by a statement of the OEC subject with the highest mercury level that he did not eat much fish: more evidence that assumptions underlying Indian concerns about mercury poisoning may be unsound.
An important implication of this result is that children, the most sensitive population, appear to be somewhat protected from mercury toxicity by the simple fact that they have not lived long enough to have acquired much of the chemical. The assay results, expressed in parts per billion (micrograms mercury per liter of blood serum), report mercury concentration. Children’s total body burden of mercury is further reduced by the mere fact of their small size.
If one graphs the PCB assays in the 2007 OEC report against subject age (again rejecting one extreme outlier, in this case a young woman who was the only subject not a life-long Oregon resident), a very different pattern is seen. Serum PCB concentration is age-independent. These subjects appear to store PCB in a constant amount rather than at a constant rate as they age, in sharp contrast to how mercury is accumulated. These two chemicals must be metabolized quite differently. The PCB pattern is encouraging from a public-health viewpoint; it suggests that the body naturally resists PCB accumulation. PCBs are less age-selective in toxicity than mercury is, so it is good news that adults do not experience increasing dose-related risk as they age.
The OEC reported median Oregonian mercury contamination about 2.5 times the national median; this fact may be very important for understanding how humans pick up mercury from the environment. Such a large discrepancy is not easily attributed to unusual diet or statistical accident. The first step toward understanding it would be to map human mercury levels across the country, looking only at life-long residents. If mercury accumulation in the Pacific Northwest exceeds the national median, there must be at least one region with below-average accumulation. Mercury contamination may correlate strongly with regional precipitation because the major contamination pathway starts with extraction of circulating atmospheric mercury by rain and snow. Alternatively or additionally, high mercury levels in humans locally may reflect the high mercury content of the volcanic rock and soil of our region. The next and last article in this series will look more closely at where mercury comes from.
The Washington Toxics Coalition (WTC) published in 2006 a very similar study of human contamination, also called “Pollution in People,” also sampling 10 subjects, adult Washingtonians of some political or cultural prominence (pollutioninpeople.org/results/report/toc). Although the WTC report did not give all subjects’ ages, that information can be obtained off the Internet. The WTC PCB data are directly comparable to the OEC results. When added to the graph of serum PCB content versus age, they cluster well with the OEC data and strengthen the hypothesis that PCBs accumulate to a constant amount in adults rather than at a constant rate. The mercury results should not be expected to track the OEC data closely because the subjects appear not to have been screened for long-term residency in the state and also because mercury was analyzed from hair rather than blood serum. However, the group median mercury content was 1.5-2 times the national median for hair samples, supporting the hypothesis that geography significantly determines mercury accumulation by humans.
The regional dependence of human mercury content has an important implication for any future population studies of the problem. Failure to control for the fact that Americans move around a lot could hide any other causal relationship, because of the high variability it can impose on the results. A 20-year-old Oregon native appears to have the mercury content of a 40-something native of a region with human contamination representative of the country as a whole.