Print E-mail
Section 7: The NRC and the Mercury RFD

Selecting an appropriate outcome to focus upon presents a substantial challenge to investigators and those interpreting their studies to use in making public policy.
- Dr. Gary Meyers, Seychelles Children Study investigator

The role of science, from Galileo to Newton and through the centuries, has been to debunk the consensus and move us forward. But now science has been stripped of its basis in experiment, knowledge, reason and the scientific method and made subject to the consensus created by politics and bureaucrats.
- Terence Corcoran

Chicago Tribune:

Solving the mercury problem ultimately will require reducing levels of the pollutant in the environment, according to the National Academy of Sciences, the nation's leading scientific advisory body. For now, though, the academy says consumers can best protect themselves by eating low-mercury fish. (CT - December 11, 2005)

Response:

Unscientific policy proposals

Assertions of a linear improvement in public health from reductions in anthropogenic mercury sources is not supported by the science. Various species of mercury are continually being exchanged and re-transformed among sources from, and within the reservoirs of air, water, soil and terrestrial vegetation without any direct connection to the "pollutant in the environment," by which is meant coal-fired power plant emissions (see Sec. 5 and 21).

Proposals mandating costly 90% mercury emission reductions from coal-fired power plants need to scientifically demonstrate that such actions will yield any measurable improvements to public health (see Sec. 21).

This is particularly required in light of new research coming out of China. China has long been regarded as one of the largest man-made mercury emission sources (from coal combustion in particular). Pacyna and Pacyna (2002) estimated that China’s emissions from coal combustion contributed more than 25% to the total global man-made mercury emissions in 2002. Tang, S., et al. (2007) report that just in Guiyang province in southwest China the annual mercury emissions from coal consumption in power plants will increase to “about 32 tons in 2015, which is comparable to the [current] annual coal combustion mercury emissions in the whole USA.”


The NRC panel

Three pre-natal exposure studies (Faroe Islands, New Zealand and Seychelles Islands) were included in a review conducted by a U.S. National Research Council panel at the request of EPA (see Sec. 2). The panel’s task was to investigate the toxicological effects of MeHg in relation to the review of MeHg exposure reference dose recommendations.

The panel rejected the NZ endpoints and the Connors Continuous Performance Test before settling on the Boston Naming Test. Deeming the Boston Naming Test as sufficiently reliable, the panel preferred an estimated benchmark dose lower limit (BMDL) of 58 ppb. The panel also recommended an uncertainty factor of no less than 10.

The EPA incorporated these recommendations in its MeHg oral reference dose assessment (U.S. EPA, 2001) against the strong objections of others, including members of the U.S. Senate, other relevant agencies and Dr. Michael Dourson and Dr. Kenneth Poirier, former co-chairs of EPA’s own RfD/Reference Concentration Work Group.

"Solving the mercury problem" is a policy judgment and was certainly not the charge of the NRC committee panel on the Toxicological Effects of Methylmercury. In fact, "the panel was not [even] charged to calculate an RfD for MeHg."

One of the reasons that the Boston Naming Test from the highly confounded and non-transparent Faroe Islands study [focused on pilot whale consumption] was selected as a basis for a mercury RfD was because it claimed “subtle” negative findings.

The NRC’s reliance on the Boston Naming Test (BNT) seems little justified :

• The BNT was developed to detect aphasis and brain damage in adults and is not a standard part of child neuropayschological testing.
• No biological reason has been proposed as to why the BNT should be particularly sensitive to prenatal MeHg exposure.
• Halperin et al. (1989) had concluded that, in children, the BNT score relates more to acquired word knowledge than to word retrieval or fluency, and that verbal memory appears to be independent of these linguistic functions.
• In terms of educational findings, the BNT appears to be significantly related to reading comprehension rather than verbal memory or word retrieval.
• Three different seemingly correct interpretations by professional psychologists of the scoring methods publish by Kaplan et al. (1983) for the BNT resulted in large, clinically significant differences in the total score (Lopez et al., 2003).
• Assuming the BNT mean score for U.S. 7-year olds is 38.83 (Halperin et al., 1989), the expected mean score for the approximately 300,000 children exposed per year at the RfD is 38.73 – a decrement of 0.1 point that is clearly not detectable, and not meaningful given the standard deviation of 4 to 6 on BNT scores.

Added to all this, the Faroe investigators themselves acknowledged concerns that PCB concentrations appeared an important predictor for the BNT results; cord PCB concentration was associated with deficits on the BNT.

Even so, the NRC discounted PCB confounding by relying on a statistical analysis performed by the Faroe researchers. Yet, the NRC panel failed to address the most obvious question as to why the Faroe study reportedly observed no significant effects from PCB contamination when PCB exposures were at levels twice as high as the lowest observed effect level for those compounds.

Thus the NRC and EPA avoided having to demonstrate some identifiable health risk to some segment of the population at some defined level of predicted exposure – the precise legislated requirement for regulatory consideration. This plowed a wide field for the future unjustifiable alarmism and exaggerated claims.

Conclusion

If the NRC and EPA had properly included and weighted the Seychelles research findings, a higher RfD would likely have been derived, resulting in no pretext for regulatory action.

By discounting the more appropriate Seychelles findings while resorting to the troubled Faroe paper, the NRC panel seems to have fabricated for EPA a unique sieve that retained dross findings of the Faroe study and allowed the gold of the Seychelles results to slip free. (see discussion in Sec. 1) The consequences are that women and their babies are the poorer for it.
 
< Prev   Next >
[ Back ]