Methylmercury in Blood and Hair

Why Blood and Hair Concentrations Matter

The best evidence for the toxicity of methylmercury to humans comes from poisoning epidemics in Japan and Iraq.  However, in both those epidemics, the exact amount of methylmercury consumed by the people who were poisoned was unknown.  In Japan, by the time methylmercury was established as the cause of the disease, it was too late to figure out, at least on an individual basis, what the amount of methylmercury ingested was.  However, in Iraq measurements of mercury in blood and hair were taken in order to gauge how much people were exposed to afterward.  That can be done because mercury is slowly (months) removed from the blood, and mercury in hair can stay there for years.  Similarly, blood and or hair measurements have been used in epidemiology studies involving populations that consume large amounts of fish to gauge the extent of exposure of different individuals to methylmercury.

Predicting Blood MethylMercury Based On Dietary Exposure

Biomarkers are useful for characterizing the relationship between exposure and the toxic effects, but since most exposure to methylmercury is usually from fish, it leaves a question hanging:  What is the relationship between consuming fish on a regular basis and levels of methylmercury in blood and hair?  Since this is a very important issue, a study with four different controlled doses of methylmercury from fish in twenty human subjects over a ninety day period (Sherlock et al, 1984).  This study was conducted in the UK over thirty years ago, and because of current restrictions on the use of human subjects probably couldn’t be done today.  Since the change in blood concentrations relative appears to be linear (i.e. has the same proportion regardless of dose), the following analysis assumes linear in order to focus on two other issues, namely the impact of body weight and other unattributed sources of variation.

While recommended dosages of drugs are often prescribed without consideration of body weight, toxicologists usually presume that the internal dose (i.e the concentration in blood) will be directly proportion to body weight.  But contrary to either of those traditional approaches, Sherlock et al (1984) suggested that a correction factor of body weight to the one-third power is the most appropriate.  Since individual subject data were published in the paper, we can look for ourselves. 

The grey squares in the graph above show the uncorrected data from all 20 subjects.  There is clearly a correlation in the relationship between incremental methylmercury levels and body, indicating that some correction for body weight is necessary.  However, correcting the values by assuming proportionality (the black triangles) seems to overcorrect since it results in a trend going the other way.  The correction suggested by Sherlock et al, 1984 of body weight to the one-third power (black diamonds) does work well.  The best correction factor of all seems to of body weight to the power 0.44 (open squares); since the black regression line is completely flat, it corrects for the influence of body weight as well as possible.

However, there is still a considerable amount of variation that is not accounted for.  Using several alternative statistical distributions to describe the additional variation found with corrected values for a 70 kg person yields the following:

Predicting Hair MethylMercury Based On Blood Concentrations

Since most studies use hair as a biomarker, it is also necessary to relate hair methylmercury to dietary exposure from fish.  The a chronic study used n Sherlock et al (1984), also measured hair values (reported in Hislop et al. (1983), but only for the data from are the most relevant to a chronic exposure assessment.  However, hair values were only measured for five of the 20 subjects in the study, all of whom were male.  In addition, only the ranges for the hair-blood ratios are reported.  Other studies have more individual data points and are therefore potentially more useful at characterizing the full range of pharmacokinetic variability.  However, there are a number of other problems with these data. First, blood measurements fluctuate and are dependent on the time since the last fish meal, and as a result, measurements made at a single point in time may not accurately reflect long-term exposure.  Second, since inorganic mercury was not measured independently in hair, it is also possible that there is some contamination of hair from inorganic mercury – perhaps from environmental sources.  Third, errors in the chemical analysis are more likely to be substantial at lower concentrations in blood or hair (i.e. near the limit of detection), resulting in either unrealistically high or low ratios.  Regardless of the explanation, actual pharmacokinetic variation in the studies reporting single measurements of blood and hair is almost certainly narrower that the apparent distribution. 

The following figure shows summary data using a lognormal distribution to represent population variability with uncertainty distributions for the parameters.  The values were chosen to be centered on the values from Hislop, but to also encompass some of the variation from the other studies as well.

Software

Combining the results of the preceding analysis allows prediction of blood and hair levels, albeit with more than a little uncertainty.  Although the underlying functions are statistical descriptions of what happens in a population, they can also be used to predict what will happen in an individual if the population variability is treated as an additional uncertainty.  In that vein, a simple simulation for estimating personal concentrations for methylmercury in blood and hair is presented below.  It also includes a distribution intended to represent other exposures to methylmercury that is based on results from a survey of blood values in the U.S. (EPA, 2013).

The simulation is written in Microsoft Excel and has VBA macros, so macros need to be enabled and you are going to have to trust me as a source.  Sorry.

MeHg_Hair_and_Blood_Simulation.xls

References

Budtz-Jørgensen, E., Grandjean, P., Jorgensen, P.J., Weihe, P., Keiding, N. (2004).  Association between mercury concentrations in blood and hair in methylmercury-exposed subjects at different ages.  Environmental Research, 95, 385-393.

Centers for Disease Control and Prevention. (2005). National Center for Health Statistics, National Health and Nutrition Examination Survey, 2003-2004 data files.  There are more recent values, but they haven’t changed very much.

Hislop, J.S., Collier, T.R., White, G.F., Khathing, D.T., French, E. (1983).  The Use of Keratinised Tissues to Monitor the Detailed Exposure of Man to Methyl Mercury from Fish.  Chemical Toxicology and Clinical Chemistry of Metals, edited by Brown, S.S. and Savory, J.  Academic Press, New York, 145-148.

Sherlock, J., Hislop, D., Newton, G., Topping, G., Whittle, K. (1984).  Elevation of mercury in human blood from controlled ingestion of methylmercury in fish.  Human Toxicology 3:117-131.

U.S. Environmental Protection Agency (2013).  Trends in Blood Mercury Concentrations and Fish Consumption Among U.S. Women of Childbearing Age NHANES, 1999-2010.  Final Report July 2013 EPA-823-R-13-002.   

U.S. Food and Drug Administration (2014). Quantitative Assessment of the Net Effects on Fetal Neurodevelopment from Eating Commercial Fish (As Measured by IQ and also by Early Age Verbal Development in Children).   Additional technical details of the analyses described above can be found in Appendix C, section (a)(3). 

Software

A simple simulation for estimating personal concentrations for methylmercury in blood and hair.  It has VBA macros, so they need to be enabled and you are going to have to trust me as a source.  Sorry.

MeHg_Hair_and_Blood_Simulation.xls

Official Post Soundtrack

Supertramp (1974).  Bloody Well Right.  In: Crime of the Century, Track 2.

Post Notes

Thesis Post #55.  This is the first one with quantitative analysis, which is from the FDA fish risk benefit report.  I tried to make the explanations less technical, but I suppose that my success in that regard is pretty marginal.  l plan several on more, which will in the end develop into a personal risk assessment model that will deviate somewhat from what is in the report.  What is posted here just covers methylmercury pharmacokinetics.  A personal fish consumption module and dose-response functions for both mercury risks and fish benefits will be added later.  Lame live soundtrack is the best I could do.