Simulation of methylmercury toxicity resulting from the consumption of mediterranean red tuna fish

233

Toxicology Letters, 10 (1982) 233-237 Elsevier Biomedical Press

SIMULATION OF METHYLMERCURY CONSUMPTION OF MEDITERRANEAN

TOXICITY RESULTING RED TUNA FISH

FROM THE

R. MAGNAVAL*, R. BATTI*, A. BOUVILLE, J.P. MAGNAVAL and R.C. SPEAR** Euratom and C.E.A., DPpartement de Protection Sanitaire, BP No. 6 92260 Fontenay-aux-Roses (Franeel, and **Schoo/ of Pubtic Health, University of California, Berkeley, CA 94720 (U.S. A.) (Received July 26th, 1981) (Accepted August 5th, 1981)

SUMMARY The toxicity of methylmercury is sufficiently high to justify an evaluation of the risk from the consumption of contaminated food. In the method of risk assessment proposed here, distribution frequencies are assigned to the parameters in question, rather than mean values. Total body burden is estimated by numerical simulation on a digital computer. Based on recorded cases of incipient intoxication and allowing a safety factor of IO, the critical body burden has been estimated at between 5, 4 and 6 mg total mercury. From an application of this approach to the assessment of risk from Mediterranean red tuna fish consumption, it appears that there may exist high frequency consumers whose body burdens, while not in the overtly neurotoxic range, approach the computed critical level.

INTRODUCTION

The absorption of a toxic compound by man can be simulated by a dynamic model which takes into account the concentration in the environment, the exposure processes by which it reaches man, its absorption and elimination, and the effect of its possible accumulation in the body. Dietary habits are an essential factor if the route of exposure is via the food chain. The phenomenon of concentration of organometallic pollutants in the various food chains has been well established in the case of mercury. The toxicity of methylmercury is sufficiently high at low doses to justify an evaluation of the toxicity hazard related to the various parameters pertaining to the consumption of contaminated food.

*Present addresses: (R.M.) Universite Paris-Sud, CNRS Tour D4, 92290 Chatenay-Malabry (France); (R.B.) Commission des Communautes Europeennes, 200, rue de la Loi, 1049 Bruxelles (Belgique). 0378-4274/82/~-~/$02.75

(9 Elsevier Biomedical Press

234

We have followed

the approach

of Spear and Wei [1] who proposed

listic assessment to estimate the residue of methylmercury earlier study [2], fishery products have been considered

a probabi-

in the body. As in our as the sole vector of

contamination. Total body burdens have been computed following repeated ingestion of tuna fish. Several fish populations were considered ranging from those with an average mercury concentration < 0.7 ppm to red tuna fish from the Mediterranean sea with an average content of 1 ppm [3]. The total range of concentrations was from < 0.1 ppm to > 3 ppm.

MATERIAL

AND METHODS

Distribution functions rather than mean values were assigned to the parameters in question. Body burdens were estimated by numerical stimulation on a digital computer (Monte Carlo method). Observed distributions of mercury content in fish were provided by a 3-year sampling program reported by Cumont et al. [3]. In all cases the methylated fraction was assumed to be 75% of the total mercury content. Two statistical variables describe consumption of fish: the amount consumed and the frequency of consumption. Fish consumption per meal is represented by a normal distribution, with an average value of 200 g and a standard deviation of 40 g. The frequency of fish consumption is introduced in the form of a binominal distribution with p = q = l/2, with the mean value representing the cases of normal consumption (one meal per week) and frequent consumption (two meals per week), on the basis of available food survey data [4, 51. For methylmercury we assume a single compartment metabolic model with an absorption from the gastrointestinal tract of 100%. The biological half-life of the compound takes into account the variability of this parameter - between 35 and 189 days with an average of 72 days - observed in a population group in Iraq [6].

RESULTS

AND DISCUSSION

The variation of the body burden of a given individual was observed over a period of 2 years (Fig. 1). Statistical equilibrium was reached after about 150 days, but at statistical equilibrium the range of temporal variation between minimum and maximum contents is 35%, approx. 10% being due to the dispersion of mercury concentrations in fish. Wei and Spear [7] calculated the total body burden derived from brain concentrations of mercury in 11 Japanese cases of incipient intoxication which subsequently proved fatal (54 mg, 87 mg, 90 mg, 91 mg, 114 mg, 118 mg, 165 mg, 202 mg, 292 mg) and from Swedish estimates they reported that nervous disorders were to be

235

I

Fig. 1. Simulation

of fish

mercury

(B) frequency

consumption

the course

consumption

of the variability

with a median during

600

400

200 Duroton

concentration pattern.

(days)

of the total mercury

body burden

of 0.3 ppm (----) and 1.O ppm ( Note the maximum

mercury

of four subjects consuming fish ) at a high (A) or normal

body burden

of a given individual

(A)

of fish consumption.

expected for an estimated body burden of about 60 mg. Allowing a safety factor of 10 [8], the critical body burden is assumed to be between 5.4 and 6 mg of mercury. In view of fluctuations in the body burden, even at equilibrium (Fig. 1) and of the safety factor of only 10, maximum body burdens of mercury which occurred over a 2-year exposure were selected as the critical variable. A simulated population sample of 1000 people was used. Critical levels of methylmercury are never reached for normal fish consumption. Critical values may, however, result from repeated ingestion of tuna fish from the Mediterranean region (Fig. 2). Our results suggest that in a significant fraction of a population consuming contaminated

fish (median

mercury

concentration

1 ppm) with a high frequency

pattern

(2 meals a week over a 2 year exposure) there would be a safety factor of < 10 over the lowest body burdens estimated for the fatal adult Japanese cases. The variability of the biological half-life of methylmercury is of importance in the probabilistic assessment

of the critical

population

group.

For example,

computations

based on a

constant average half-life value of 70 days result in a critical group estimate of 0.5% while the introduction of excretion rate variability increases the size of this group to 8% of the population. Because the relationship between environmental levels of methylmercury and the human toxic hazard has not been defined unambiguously [9, lo] this method provides a means of estimating the risk to a population although we must keep in mind our lack of knowledge concerning consumption habits and the consequent pitfalls involved in relating simulated mercury body burden to toxic effects in human populations with individual variability [ 111. Our numerical results rest on overestimates of the consumption of red tuna fish, but additional sources of mercury contamination have not been taken into account.

236

Total mercury Fig. 2. Histogram with a median

body burden

of maximum

mercury

body burden tmg)

concentration

Occupational mercury more complete estimate As bottom sediments persistent high level of health hazard [ 161.

distribution

of a population

of 1 .O ppm at a high frequency

of 1000 people consuming

fish

pattern.

exposure in particular [12, 131 should be considered in a of mercury hazard. remain a reservoir of mercury contamination [14, 151 the heavy metals in Mediterranean fish still poses a potential

ACKNOWLEDGMENT

The authors thank Dr. E. Wei for discussion and critical review of the manuscript. R.M. was supported at the Department of Biomedical and Environmental Health Sciences, University of California, Berkeley by a post-doctoral fellowship of the French Ministry of Foreign Affairs. J.P.M. is currently a doctoral student at the School of Dentistry of the Faculty of Medecine, Montpellier (France). This work was partly Commission

sponsored by the Biology of European Communities.

and

Health

Protection

Program

of the

REFERENCES

I R.C. Spear Contr.

and E. Wei, Dynamics

ASME,

2 R. Magnaval,

R. Batti,

methylmercury

toxicity

Mediterranean, 3 G. Cumont,

aspects

A. Guezengar, resulting

toxicology,

ICSEM/UNEP, des poissons

annees,

First Int. Mercury

4 OECD,

Food consumption

5 E. Bacci, C. Leonzio par la consommation

R. Bittel and A. Bouville,

from the consumption Monaco,

G. Gilles, F. Bernard,

la contamination

of environmental

J. Dynamic

Syst..Measur.

(1972) 114.

G. Stephan,

de mer par le mercure

statistics,

assessment

on Pollution

of

of the

1979, p. 663.

M.B. Briand,

Congres,

A probabilistic

of tuna fish, Workshop

Inst. E. Jimeno, OECD,

Paris,

G. Guillou

a I’occasion Barcelona,

and G. Ramonda,

d’un controle

portant

Bilan de sur trois

I (1974) 141.

1973, p. 290.

and A. Renzoni, Etude sur une population humaine exposee au methylmercure de poisson, Rev. Int. Oceanogr. Med. 41 (1976) 127.

237

6 H. Al-Shahristani Arch.

and K.M.

Environ.

Health,

Shihab,

Variation

of biological

half-life

of methylmercury

in man,

28 (1974) 342.

7 E. Wei and R.C. Spear,

The fatal dose of methylmercury

in man,

J. Am. Med. Assoc.,

216 (1971)

1347. 8 WHO,

Mercury

metabolism,

in: Environmental

Health

Criteria,

1 Mercury,

WHO,

Geneva,

1976, p.

77. 9 G. Birke, A. Johnel, exposed

L. Plantin,

to methylmercury

IO G.F. Nordberg polation

and P. Strangert,

for noncarcinogenic

11 T.W. Ciarkson

I2 J.D.

Cross,

I.M.

Lancet

13 A. Bernard,

The toxicity

mercury

sediments

I5 B. Lachet mercure 16 B. Boxer,

and

Health,

Environ.

Health

of methylmercury

(Ed.), Effects

Studies on humans 25 (1972) 77.

relationships

Perspect.,

and their extra-

22 (1978) 97.

in man: dose-response

and Dose-response

relationships

Relationships

for Toxic

1976, p. 559.

Dale, L. Goolvard, Roels,

J.P.

J.M.A.

Buchet

gel electrophoresis

and lead, Toxicol.

14 M.K. El-Sayed,

Environ.

aspects of dose-response

of metals,

in: G.F. Nordberg

and T. Westermark,

Arch.

Lenihan

and H. Smith,

Methylmercury

in blood

of

2 (1978) 311.

H.A.

polyacrylamide

ranean

Fundamental

Elsevier/North-Holland,

dentists,

S. Skerfving

fish consumption,

effects

and D.H. Marsh,

in adult populations: Metals,

B. Sjostrand,

through

and R.R. Lauwerys,

of urinary

proteins

Lett.,

5 (1980) 311.

Y. Halim,

H.M.

Abdel-Kader

around

Alexandria,

R. Magnaval,

et du chlorure Mediterranean

Equation

mercurique action

Egypt,

Comparison

excreted

and M.H.

Moeness,

Mar. Pol. Bull.,

empirique

traduisant

par des sediments,

plan: an interim

dodecyl

exposed

Mercury

sulfate

to cadmium,

pollution

of Mediter-

du chlorure

de methyl-

10 (1979) 84. I’epuration

C.R. Acad.

evaluation,

of sodium

by workers

Sci. Paris,

Science,

289 (1979) 371,

202 (1978) 585.