Draft Report of Quantitative Risk and Benefit Assessment of Consumption of Commercial Fish, Focusing on Fetal Neurodevelopmental Effects (Measured by Verbal Development in Children) and on Coronary Heart Disease and Stroke in the General Population: Appendix D, Estimated Exposure in the Daniels et al. (2004) Cohort
January 15, 2009
This information is distributed solely for the purpose of pre-dissemination peer and public review under applicable information quality guidelines. It has not been formally disseminated by FDA. It does not represent and should not be construed to represent any agency determination or policy.
Because we used the Daniels et al. (2004) data in our risk and benefit assessment for fetal neurodevelopment, we were interested in understanding the exposure to methylmercury in this cohort. We obtained the individual subject data for the cohort and were able to estimate exposure as described below.
Daniels et al. (2004) performed a mercury analysis for a subset of 1,054 mother-child pairs. This subset was drawn from a sample of 1,225 children in the ALSPAC study population from whom umbilical cord tissue had been taken at birth. The cord tissue was analyzed for mercury concentration to provide a measure of prenatal exposure to methylmercury for these children.
Using umbilical cord tissue to measure exposure is somewhat unorthodox as compared to more typical measurements involving mercury in hair and blood that have been conducted in the United States and elsewhere. Nonetheless, we can estimate how exposures in this population compare to U.S. exposures by converting mercury in umbilical cord tissue to mercury in blood and hair as follows:
- Cord blood = cord tissue divided by 8.9 (Sakamoto et al., 2007);
- Maternal blood = cord blood divided by 1.7 (Stern 2005b);
- Maternal hair = cord blood divided by 5 (Axelrad et al., 2007; Butz-Jorgensen et al., 2004).
We can then compare the blood and hair estimates to the results from the NHANES database. We do so in terms of percentiles of exposure as shown in Table IIA-3.
Cord Tissue/Hg Levels
Estimated Cord Blood/Hg Levels
Estimated Maternal Blood/Hg Levels
Maternal Blood/Hg Levels*
Estimated Maternal Hair/Hg Levels
Maternal Hair/Hg Levels
|Average||0.023||2.6||1.5||1.6||0.51 ppm||0.47 ppm|
|10th Percentile||0.004||0.5||0.3||0.2||0.10 ppm||0.04 ppm|
|25th Percentile||0.008||0.9||0.5||0.4||0.17 ppm||0.09 ppm|
|50th Percentile||0.013||1.5||0.9||0.8||0.30 ppm||0.19 ppm|
|75th Percentile||0.023||2.6||1.5||1.7||0.51 ppm||0.42 ppm|
|90th Percentile||0.041||4.6||2.7||3.7||0.91 ppm||1.11 ppm|
|95th Percentile||0.066||7.4||4.4||5.6||1.48 ppm||1.73 ppm|
|99th Percentile||0.186||20.8||12.3||12.4||4.17 ppm||3.50 ppm|
*From the 1999-2004 NANES survey years.
** From the 1999-2000 NHANES survey years. These were the only years during which NNANES measured hair mercury levels.
As the table shows, our estimate of exposure in this U.K. study population was similar to exposure in the United States across the spectrum. The U.K. population's exposure to methylmercury was lower than we would expect because per capita consumption of commercial fish in the U.K. is about 2.8 times higher than it is in the United States (NMFS 2005) while methylmercury concentrations in fish available in the U.K. are similar to concentrations in fish in the U.S. marketplace (Montwill 2007). The lower than expected exposure in this U.K. study population might be due to consumption of mostly low methylmercury fish and avoidance of higher methylmercury fish.