WHO Food Additives Series, 1972, No. 4
EVALUATION OF MERCURY, LEAD, CADMIUM
AND THE FOOD ADDITIVES AMARANTH,
DIETHYLPYROCARBONATE, AND OCTYL GALLATE
The evaluations contained in this publication were prepared by the
Joint FAO/WHO Expert Committee on Food Additives which met in Geneva,
4-12 April 19721
World Health Organization
Geneva
1972
1 Sixteenth Report of the Joint FAO/WHO Expert Committee on Food
Additives, Wld Hlth Org. techn. Rep. Ser., 1972, No. 505; FAO
Nutrition Meetings Report Series, 1972, No. 51.
MERCURY
Mercury occurs naturally in the environment and is contributed to by
human industrial activity. Because the natural background levels of
mercury vary it is difficult to distinguish natural from artificial
levels in any part of the environment. Annual natural erosion and
weathering is estimated to contribute some 5000 tons to the sea.
Another 4000-5000 tons of mined mercury (about half the annual world
production of 9000 tons) are lost to sea, soil and the atmosphere.
Following absorption mercury accumulates in most living organisms. It
fulfils no known essential function in man or animals.
There is much concern about the potential for human intoxication due
to mercury in foodstuffs particularly fish. The most serious toxic
hazard arises from methylmercury residues in food (Lu et al., 1972).
Sources
(a) Environmental sources
Air. Mercury has a relatively high vapour pressure so that the air
over mercury and its ore deposits generally contains elevated levels
of mercury (0.2-20 µg Hg/m3). Small amounts are also found over
non-mineralized land areas (0.003-0.009 µg Hg/m3), while
insignificant amounts occur over the oceans (0.001 µg Hg/m3)
(Fleischer, 1970; Wallace et al., 1971). If it is assumed that the
average intake of air is 10m3 per day and that the maximum amount of
mercury found over non-mineralized land areas is inhaled this would
amount to 0.1 µg of mercury from this source, probably as elemental or
inorganic mercury. Only limited information is available regarding
tobacco; this suggests that for cigarettes most of the trace mercury
present is retained in the butt and that (on average) not more than
0.1 µg of mercury per cigarette is inhaled.
The combustion of fossil fuels will also liberate traces of mercury
into the atmosphere: these add to the much larger quantities present
in the atmosphere arising from natural vaporization processes from the
earth's surface (Weiss et al., 1971) and may be significant locally.
Water. Although there is considerable evidence that the manmade
pollution of rivers, lakes and estuaries has often increased the
mercury levels in fish, such pollution has not contributed
significantly to the mercury levels found in wide-ranging ocean fish.
Since the mercury content of the ocean has been estimated to be of the
order of 108 tons and the annual industrial output of mercury is
only of the order of 104 tons, the pollution due to man can have had
only local effects, e.g. in lakes, rivers, estuaries and coastal areas
(Miettinen, 1971). On the other hand, the present degassing of
mercury from the earth's crust is estimated to range between 2.5 x
105 and 1.5 X 105 tons per year. Increased flux over the last few
decades may be due to alteration of terrestrial surfaces through the
variety of activities of man. The mercury content in ocean surface
waters may have been augmented by a factor of 0.75 (Weiss et al,
1971).
There is also natural geological contamination of individual lakes or
watercourses due to underlying mineral deposits containing mercury,
which leach into the water under natural circumstances.
The mercury content of streams, lakes and rivers does not normally
exceed 0.1 µg per litre although some water sources located near
mercury ore deposits may contain amounts up to 80 µg/litre (Wershaw,
1970). WHO has recommended that the tentative upper limit for mercury
in drinking water should be 1 µg per litre, this figure being related
to levels found in natural waters used for drinking purposes (WHO,
1971). At the ingestion rate of 2.5 litres of water per day used in
the calculations for the International standards, the upper limit of
mercury intake would be 2.5 µg per person per day, mainly in the
inorganic form.
The levels of mercury found in treated water are lower than those
found in food. Moreover, mercury in water is found in inorganic form.
In terms of methylmercury, therefore, the contribution to the body
load in man from water and air is unlikely to be significant.
(b) Industrial sources
The direct pollution of water by industrial sources is likely to
affect fish more than other foods. Similar pollution will be caused
by dumps of industrial wastes which will transfer mercury into the
soil and, eventually, into water. The transfer into the atmosphere
may form a significant addition locally due to the large amounts
evolved by natural outgassing of surface materials.
Mercury or mercury salts entering water may eventually result
(primarily through microbial but also through other biological and
chemical processes) in the formation of methylmercury compounds which
can be absorbed and accumulated by aquatic organisms. The industrial
sources that cause the transfer of mercury-containing wastes to water
or mud in fishery areas constitute sources of most direct concern.
Furthermore, once mercury has been deposited in watercourses and
estuaries it may remain and give rise to the formation of
methylmercury compounds by bacterial action over a period of many
years (Wood et al., 1968). Thus, both past and present sources of
contamination must be considered. Measures to control industrial
sources can bring about a considerable reduction of new pollution.
However, the problem presented by levels of mercury in fish taken from
already polluted watercourses and estuaries will continue for a long
time. The industrial processes which are likely to introduce
significant quantities of mercury into the environment have been
examined in a number of countries and control measures have been
suggested (Env. Res. 1971). The principal sources are the
chlor-alkali industry using the mercury process, the pulp and paper
industry and other industries using mercury. Effluent from these
industries in watercourses and estuaries has been the cause of much
local pollution in fish. Indeterminate but significant quantities of
mercury originate from mining, smelting and refining of ores, but the
effects are usually localized. Industrial control equipment, hospital
and laboratory instrumentation containing elemental mercury, such as
relays, mercury arc rectifiers and thermometers represent minor
sources when there are losses due to breakages. Laboratory and
hospital wastes, discarded electrical apparatus, fluorescent lamps,
paints, catalysts and other sources of mercury in dumps and sewage
water, through seepage into watercourses, may also cause contamination
of fish.
(c) Agricultural sources
Alkyl, alkoxyalkyl, aryl, and inorganic mercurial fungicides have been
used for seed dressing, as turf fungicides, and in orchards. The
amount of mercury entering the food supply from normal agricultural
use has been usually small compared with that from other sources. On
occasions, the use of alkyl mercury has resulted in the fortuitous
accumulation of mercury by game birds feeding on mercury-treated seeds
in the fields. Although the use of alkyl and aryl mercurial
fungicides has been restricted or prohibited in a number of countries
in recent years, accidents have occurred from the misuse of seeds
treated with alkyl mercury compounds.
(d) Other sources
Additional intake sources are pharmaceuticals containing either
inorganic or organic mercury compounds, and mercurials used as
preservatives in cosmetics and toiletries. Care should therefore be
taken in epidemiological surveys to consider these sources of
exposure, as well.
Transformation of mercury from inorganic to organic
The geochemistry of metallic and inorganic mercury involves numerous
interconversions and transformations. Mercury is present in both
humic complexes and in the relatively unionized and stabilized form of
mercuric sulfide (Wallace et al., 1971). Under oxidizing conditions,
mercuric sulfide is gradually converted to mercuric sulfate, which is
more readily dissociated, releasing ionic mercury in solution.
Conditions of oxidation also promote the conversion of metallic
mercury to ionic mercury. Mercuric compounds located on the earth's
surface will usually degrade to metallic mercury and subsequently
volatilize due to the action of sunlight. On the other hand, most of
the mercury compounds which are carried down into river beds and lake
bottoms will, in the presence of hydrogen sulfide, ultimately be
converted to insoluble mercuric sulfide, which may be, in turn,
converted into methylmercury compounds.
The principal forms of mercury commercially discharged into the
environment are metallic mercury, inorganic mercury, aryl-mercury,
alkyl-mercury and alkoxyalkyl-mercury compounds. The transformation
of inorganic mercury into the organic methylmercury form is promoted
by micro-organisms or other biologically derived alkylating systems
present in the sediments of lakes, rivers and estuaries. These
systems are capable of forming methylmercury and dimethylmercury from
inorganic mercury, both under aerobic and anaerobic conditions (Jensen
& Jernelöv, 1969; Wood et al., 1968).
Both methyl and dimethylmercury compounds appear as initial products
of methylation. High mercury concentrations favour the monomethyl
form while alkaline pH favours the highly volatile dimethyl form,
which decomposes to monomethylmercury at acidic pH.
The formation of methylmercury in bottom sediment under aerobic and
anaerobic conditions was studied using 203Hg labelled mercury. The
reaction proceeds under both conditions, but the rate of reaction
depends upon the number of micro-organisms. No methylation occurs
when the bottom mud is sterilized. The possibility of spontaneous
chemical transformation to methylmercury has also been reported. The
mechanisms of the methylation reaction are not yet fully understood,
although several hypotheses have been advanced.
Methods of analysis
In the estimation of trace quantities of mercury a method is generally
employed which determines the total mercury content. If necessary,
separate analysis for alkyl-mercury compounds should follow. Total
mercury and any mercury compounds are calculated as mercury, and may
be expressed as µg Hg/kg, µg Hg/litre, mg Hg/kg, etc., according to
the substrate examined.
In the basic method for total mercury estimation, the sample is
oxidized with a mixture of sulfuric and nitric acids, converting all
the mercury present into inorganic form. Care is necessary to avoid
losses of mercury by volatilization in this process, and specially
designed apparatus should be utilized. After oxidation, the mercury
is reduced to metallic state and the amount present estimated in the
form of vapour released from the solution in a stream of nitrogen by
flameless atomic absorption spectrophotometry, using the 253.7 nm
line. Due regard should be given to the proportion of mercury
recovered in control experiments.
Similarly, in the determination of methylmercury compounds, these are
extracted with organic solvents (usually benzene), re-extracted into
aqueous cysteine solution and extracted back into solvent after
acidification. Mercury in this solvent is then estimated by gas
chromatography. Thin layer chromatography of the extract may also be
useful as a confirmatory test for establishing the identity of unknown
compounds. Due regard should be given to the proportion of
methylmercury compounds recovered in control experiments.
Neutron activation analysis requiring expensive and highly specialized
equipment has been employed successfully as an alternative technique.
Low temperature combustion has also been used in place of wet
oxidation. When using either of these methods, particular attention
should be exerted to possible losses of mercury, to contamination from
outside sources and to recovery percentages.
The method of the International Union of Pure and Applied Chemistry
for the analysis of traces of mercury in food (IUPAC, 1965) is under
revision and a further study of the Official Method of Analysis (AOAC,
1972) is currently in progress. A basic technique for estimating
methylmercury compounds is described in the method of the Nordic
Committee on Food Analysis (Nordic Committee, 1972).
Levels in food
The amount of mercury compounds found in plant produce is very small
or zero.
Although the use of alkyl mercury fungicides (particularly as seed
dressings) is being discouraged, there may still be possibilities, in
the absence of appropriate regulatory measures, that traces of such
fungicides will get into cereals by accident.
Low levels of total mercury can also occur in meat and dairy produce
and may include some methylmercury compounds, derived presumably from
residues in feeds containing fish-meal or treated cereal grains.
Fish constitute a special problem. Since the tragic accidents at
Minamata and Niigata in Japan, aquatic pollution with mercury and its
resulting uptake and accumulation by edible fish has been given
special attention in many parts of the world. In recent years
thousands of analyses have been performed and comprehensive studies
have been reported. The results show that increased levels are
consistently found in places where the water is contaminated with
mercury from industrial and mining processes. In fish from unpolluted
water, comprising most of the world's total catch of food fish, low
values are generally found. Some of the larger specimens of predatory
species may contain high levels, apparently derived from natural
sources since the same levels are found in museum specimens caught
50-100 years ago.
There is evidence that the level of methylmercury in fish increases
with the age of the fish, e.g., data on swordfish show levels of
mercury clearly increasing with size (age) of the fish. The ratio of
methylmercury to total mercury in edible fish muscle is usually high
(approaching 100%). In shellfish, the proportion is only 50% or less
when mercury is high. However, the high level of inorganic mercury
may be due to the inclusion of viscera in the material analysed.
Information available shows that 99% of the world's commercial catch
has a total mercury content not exceeding 0.5 mg/kg, and 95% probably
contains less than 0.3 mg/kg. In 1971 the United States Food and Drug
Administration conducted a nationwide survey and found less than 3% of
1400 random samples of market fish (mainly deep ocean species) to
contain mercury in excess of 0.5 mg/kg.
Tuna species make up only about 2% of the total world catch of fish,
and almost all of this goes for human consumption. The average
mercury concentrations are below 0.5 mg/kg but old and large specimens
may contain levels above 1 mg/kg. A United States study showed that
95% of the canned tuna contained less than 0.5 mg/kg (U.S., FDA,
1971). Larger specimens of halibut and shark may contain increased
levels of total mercury, up to and exceeding 1 mg/kg (Bligh, 1971;
Faile, 1971; Westöö & Rydälv, 1971).
In some communities in countries where total diet surveys based on
fish consumption of the edible portion of 20 g or less per day have
shown an average daily intake of total mercury of 10 µg per person,
the average amount of methylmercury present has been estimated to be
not greater than 2 µg, depending on the amount and type of fish
consumed (United Kingdom Report).
Freshwater fish in unpolluted waters seem to have total mercury
contents less than 0.2 mg/kg. In some cases, however, fish have been
found to contain values up to 0.5-1 mg/kg in apparently uncontaminated
lakes (Hasanen a Sjöblom; Westöö & Rydälv, 1969; 1971). Unusually
elevated natural mercury concentrations in underlying geological
structures may be responsible for the high levels of methylmercury in
fish from these lakes.
Freshwater fish in mercury-polluted environments has been extensively
studied in some countries. Some larger species, such as pike, perch,
pickerell, bass and burbot have in many lakes been found to contain
1-5 mg/kg of mercury; and in some cases even more.
Freshwater fish in unpolluted waters seem to have total mercury
contents less than 0.2 mg/kg. In some cases, however, fish have been
found to contain values up to 0.5-1 mg/kg in apparently uncontaminated
lakes (Hasanen & Sjöblom, 1969; Westöö & Rydälv, 1969; 1971). In some
cases, sources of pollution have been revealed later on (Westöö, 1972)
e.g. hospital sewage or mercury airborne by prevailing winds from
industrial sources. However, in a large number of cases of bodies of
water in isolated areas, there can be no man-made source of mercury
pollution and in perhaps one half or more of these areas, there is
reason to believe there may be natural occurrences of mercury often
associated with methyl sulfide ore deposits (Tam & Armstrong, 1972).
Other significant factors influencing the mercury content found in
fish are shallow lakes, oligotrophic conditions (organisms other than
fish "competing" for the mercury), pH of the water (low pH favours
formation of methylmercury).
Fish, as well as shellfish and crustacean and eels (anguila species),
that constitute the local population in polluted estuaries may contain
high values of mercury and as they may represent a valuable local fish
supply, they should be more extensively surveyed.
Levels in oysters and shellfish can be controlled by antipollution
measures. Mercury levels in fish from certain lakes in Sweden and
certain rivers in Japan have also fallen after the elimination of
contamination sources.
Fish caught elsewhere have not shown such a response with respect to
mercury levels, but the period of observation was relatively short.
BIOLOGICAL DATA
Effects of mercury
The chemical form of mercury determines the biological effects which
occur. Mercurials fall into a number of well defined classes,
distinguished by differences in absorption, biotransformation,
deposition, retention and excretion. These metabolic features largely
determine the toxic effects produced. Mercurials can be divided into
the following groups: (a) elemental mercury, (b) inorganic mercury
compounds, (c) aryl and alkoxyalkyl mercury compounds, (d) alkyl
mercury compounds.
(a) Elemental mercury
The major route of human exposure is by inhalation of mercury vapour.
The vapour diffuses rapidly (75-85%) across the alveolar membrane as
shown by studies on four volunteers exposed to 50-350 µg/m3 metallic
mercury. This absorption is reduced by the ingestion of alcohol
(Kudsk, 1965).
Occasional instances of acute poisoning from vapours have been
reported. Major symptoms are acute pneumenitis ("metal fume fever"),
gastrointestinal disturbances, bloody diarrhoea and severe kidney
injury with uraemia and anuria. Chronic exposure produced tremor and
psychological disturbances, (erethismus mercurialis) and occasional
proteinuria which may progress to nephrotic syndrome (Kazantzis et
al., 1962; Joselow & Goldwater, 1967). Gingivitis, stomatitis,
excessive salivation, occasional dermatitis, mercurialentis,
non-specific fatigue, weight-loss and pallor have been reported
(Bidstrup, 1964).
(b) Inorganic mercury compounds
Following the absorption of a soluble inorganic mercury salt the
mercuric ion is transported in approximately equal proportions bound
to plasma proteins and haemoglobin in the red cell. The biological
half-life in man has been determined at 42 ± 3 days for the whole body
and 16 days for the RBC. Only 0.2-0.4% of the initial dose appeared
in the whole blood. The RBC/plasma activity ratio was 0.4. Ionic (in
water) or protein bound mercury (calf liver) was handled metabolically
the same way (Rahola et al., 1971). The immediate effects of acute
poisoning are due to irritation, coagulation and superficial corrosion
of exposed tissues. Chronic effects include kidney damage and
intestinal ulceration with haemorrhage (Bidstrup, 1964).
(c) Organomercurials
The biological effects of organomercurials are much influenced by the
particular organic group present. They enter the body by inhalation,
by skin absorption though only slowly, or by oral ingestion. They are
volatile to a variable degree. Aerosol inhalation leads quickly to
solution of the unchanged compound in body fluids and distribution to
the blood. The compounds at present in use have the general formula
R-Hg+-X(-) and biotransformation follows two general pathways: (a)
splitting of the C-Hg bond to liberate Hg++ or (b) modification to
R'-Hg+.
(i) Aryl mercury compounds
Man rapidly metabolises phenylmercury compounds to inorganic mercuric
ions which are excreted in the urine. The RBC/Plasma ratio is about
1. Urine levels of mercury could be used to indicate the degree of
exposure. Human chronic exposure by inhalation at air levels of up to
5.1 mg Hg/m3 had not produced clinical manifestations: blood levels
reached 0.52 µg/ml and urinary excretion 2.8 µg/ml (Ladd et al., 1964;
Goldwater et al., 1964). No systemic injury was noted by exposure
from 0.2-3.2 mg/m3 With urinary excretion of 0-6 µg/ml. However,
chronic exposure increases urinary protein level (Joselow & Goldwater,
1967). In exposed groups of seed dressers mercury levels in urine
were 0.035 µg/ml, blood levels O-0.21 µg/ml and urinary protein was
166 mg/24 hours compared with 70 mg/24 hours in matched controls. No
correlation was established between exposure and urinary mercury
excretion. Serum glutathione reductase was unaffected but serum
phosphoglucoseisomerase was low in the exposed group.
The biological half-life of diphenylmercury in an individual
accidentally exposed to this compound was found to be 14 days (Morsy &
El-Assaly, 1970). There is some evidence that phenylmercury compounds
produce chromosomal abnormalities in Drosophila melanogaster and
possibly teratogenic effects in mice. However, studies with 203Hg
phenyl acetate in pregnant mice showed that although the placenta
takes up large amounts of mercury, only traces were present in the
foetus (Berlin & Ullberg, 1963; Suzuki at al., 1967).
(ii) Alkoxyalkyl mercury compounds
Methoxyethyl mercury compounds are rapidly metabolized to inorganic
mercuric ions and excreted in the urine. The RBC/Plasma ratio is
about 1. In acute poisoning the organs most affected are the kidneys.
There may be intestinal symptoms, headache and weight loss. Exposure
may result in mercury blood levels of 0.3-1.1 µg/ml and mercury levels
in urine of 0.2-1.2 µg/ml without clinical injury (Berlin, 1970).
(d) Alkyl mercury compounds
The distribution of alkyl mercury in the body tissues is very
different from aryl and inorganic mercury. However, absorption,
distribution and excretion are closely similar for the hydroxide,
cyanide, dicyandiamide and propanediol mercaptide. Distribution and
elimination after subcutaneous injection are similar for the hydroxide
and dicyandiamide (Ulfvarson, 1962). MethyImercury compounds are
absorbed via the skin, the respiratory tract in rat, mouse and monkey
(Hunter et al., 1940; Swensson, 1952) and there is almost complete
absorption from the digestive tract (Berglund, 1969; Ekman et al.,
1968a; Clarkson, 1971). Elimination is very slow especially in
primates, with consequent risk of considerable accumulation. It seems
likely that the compounds dissociate in the body: the methylmercury
group attaches to the functional groups of other substrates, e.g.
proteins, being then similarly metabolized. The methylmercury group
distributes rather evenly in the body but it accumulates in the
erythrocytes; the distribution between plasma and erythrocytes varies
with species. In man, 90% of the total blood mercury appears in the
cellular elements. Blood/brain ratios differ from species to species.
Methylmercury easily penetrates the placenta in animals. It also
accumulates in hair. The most part is excreted in the faeces and some
in the urine. Forty to fifty per cent. of the faecal excretion is in
the inorganic form. The stability of alkyl mercury compounds favours
their accumulation in the CNS.
Observations in man
Orally administered CH3203HgNO3 in three male volunteers was almost
completely absorbed (over 90%). Erythrocytes took up the 203Hg within
15 minutes, the RBC/plasma ratio being 10. Maximum blood levels
occurred three to six hours after ingestion. Fifty-five per cent. of
the total 203Hg-activity was found localized in the liver, 12% in the
head. Less than 1% of mercury passed through the gut unabsorbed.
Faecal excretion totalled 34% in 49 days (initial 3% of body burden
falling to 1% of body burden/day). Urinary excretion amounted to 3.3%
in 49 days. No mercury was found in sperm, but some appeared in hair
(0.12% of the dose at 50 days). The epidermis was not considered a
significant route of excretion. The biological half-life was 70-74
days. At steady state equilibrium one unit dose CH3203HgNO3 per
week was calculated to give a total body burden of 15.2 unit doses
(Ekman et al., 1968a; 1968b; Åberg et al., 1969; Falk et al., 1971).
Nine male and six female volunteers were given orally about 2 µg of
methyl-203Hg which had been incorporated into fish muscle protein by
feeding the fish 10 days earlier with labelled methylmercury
proteinate. Most of the activity was excreted in the faeces (6% of
body burden in three to four days, then 1.9% of body burden/day); very
little was excreted in the urine (no activity until the one hundredth
day). Activity in RBC's decreased rapidly, the RBC/plasma ratio being
10. The biological half-life was 50 ± 7 days for RBC and 76 ± 3 days
for whole body activity (Miettinen et al., 1971). When a
methylmercury-free diet followed a high methylmercury-fish diet the
biological half-life for RBC activity was 70 days and for plasma
activity 80 days (Birke et al., 1972). Urinary excretion of mercury
cannot be used to evaluate exposure to alkyl mercury compounds
(Lundgren et al., 1967).
Alkyl mercury poisoning causes sensory nerve damage (paraesthesia of
face and extremities), cerebellar damage (ataxia, incoordination,
dysarthria), calcarine cortical damage (concentric visual field
constriction), temporal lobe damage (hearing defects). All are
associated with raised mercury levels in certain brain areas, liver
and kidney. Patient response is very variable, the Hunter-Russel
syndrome having been defined after industrial exposure (Hunter et al.,
1940; Edwards, 1865; Hill, 1943). Diffuse encephalopathy in 168 cases
with 52 deaths with concentric visual field loss, ataxia and
peripheral neuropathy secondary to ingestion of mercury contaminated
fish and shellfish has now been reported in Minamata and Niigata.
The works of Irukayama (1968), Takeuchi (1968, 1970) and others (Study
Group of Minamata disease) have established the etiologic role of
methylmercury in Minamata disease. The mercury content of organs of
human autopsy cases of Minamata disease (12 victims, age five to 62
years), in which the course of the disease was 100 days or less was
liver 22-70 µg/g, kidney 22-144 µg/g, brain 2.6-25 µg/g. The
brain/iiver ratio was 12:50, and brain/kidney 7:37 (Takeuchi, 1970).
Similar symptoms have been reported in numerous instances following
accidental ingestion of food treated with alkyl mercuric compounds
(ethylmercury and methylmercury derivatives), e.g. ingestion of
treated seeds or of meat from animals fed treated grain (Engleson &
Herner, 1952; Haq, 1963; Ordońez et al., 1966; Jalili & Abbasi, 1961;
Dahhan & Orfaly, 1964; Mukhtarova, 1970; Gis'IU & Pozner, 1970;
Snyder, 1971; Storrs et al., 1970; Curley et al., 1971; Mnatsakonov et
al., 1968; Slatov & Zimnikova, 1968; Sandi, 1970; Okinaka et al.,
1964).
In the Minamata accident 23 infants with cerebral involvement (palsy
and retardation) were born to mothers who had ingested mercury
contaminated fish and shellfish indicating transplacental foetal
damage. The mothers of these patients had shown no manifestations of
the disease or had only slight complaints such as numbness of fingers
and fatigue. In both adult and foetal cases there were degenerative
and regressive alterations of the cerebral cortical nerve cells. In
addition, hypoplasia was observed in the foetal cases. Cerebral
cortical cell involvement was more diffuse in foetuses than in adults
(Matsumoto et al., 1965).
Methylmercury crosses the placental barrier and achieves a higher
concentration in foetal RBC than in maternal RBC. The foetal plasma,
however, has a lower concentration than does maternal plasma. The
difference may be due to differences in the binding of methylmercury
by maternal and foetal haemoglobin. The mercury content of blood of
mothers and newborn infants in Sweden was reported as (average values
ng/g):
Infants RBC 11.5 Plasma 1.9
Mothers RBC 8.6 Plasma 2.5 (Tejning, 1970)
Mercury levels in maternal blood, umbilical cord blood and placental
tissue in Tokyo were reported as: (ng/ml)
Umbilical cord RBC 30.8 Plasma 11.2
Maternal blood RBC 22.9 Plasma 12.4
Placenta 71.5 ng/g (Suzuki et al., 1971)
It has been reported that accidental exposure of women to
ethylmercuric chloride resulted in the accumulation of mercury in
newborn infants because of the presence of mercury in the mother's
milk but the analytical method was not stated (Bakulina, 1968).
Chromosome analysis of cultured lymphocytes from normal controls
(0.004-0.011 µg/g RBC) and individuals with high RBC mercury levels
0.03-0.23 µg/g (from contaminated fish containing 1.7 mg Hg/kg
indicated increased chromosome breakage, dose related to mercury
levels. The significance for man is unknown (Skerfving et al., 1970;
Sumari et al., 1971).
Animal studies
Biotransformation of methylmercury in the rat was studied using
detection of cleavage of the C-Hg bond as indicator. After exposure
to single intravenous dose of 1 mg/kg methylmercury chloride, 80% of
the total mercury excretion was in the faeces. Forty per cent. of the
total mercury excreted was found to be inorganic mercury. Fifty per
cent. of the faecal excretion and 5.20% of the urinary excretion was
inorganic mercury. Most of the faecal inorganic mercury was derived
from faecal bacterial processes. Ten per cent. of the methylmercury
chloride was excreted in the bile within 24 hours, falling to 5% on
day 10. Only 3% was excreted daily in the faeces. Methylmercury
cysteine, or more probably a complex with a sulfydryl-containing
dipeptide molecule, was the major biliary metabolite together with
protein-bound mercury. Methylmercury cysteine was rapidly and nearly
completely absorbed in the small gut, the methylmercury protein
representing 15% of the total faecal mercury and the mercury protein
complex representing another 15%. Seventy per cent. of the total
mercury was in the intestinal cell, 95% as methylmercury.
Methylmercury cysteine accumulated in the kidneys more than
methylmercury chloride; and this could be inhibited by bile duct
ligation despite rising blood and plasma levels (Norseth. 1969, 1971;
Norseth & Clarkson, 1971).
When polystyrene with thiol ligands was fed at 1% of the diet to
groups of five mice, given intraperitoneally CH3203Hg, the mercury
faecal excretion was three times that of controls, whole body count
was reduced, tissue levels after 22 days treatment were lower, brain
levels were 1/5 of controls, kidney levels 1/10 and the blood was free
from methylmercury (Clarkson et al., 1972).
Methylmercury chloride was mixed with soluble swine brain protein to
form methylmercury proteinate. On dialysis against a variety of
thiocompounds much mercury was removed. Similarly,
thioclycollate-treated hair powder reduced mercury content. If mice
were given intramuscularly 200 or 400 µg methylmercury chloride and
fed either normal diet or diet with 3% treated hair powder for 10 or
20 days then the total body mercury of treated mice was lower than
untreated controls (Hirayama & Takahashi, 1971a; 1971b).
CH3203Hg was injected intraperitoneally into a guinea-pig two days
before and three days after parturition. Milk levels were 5-10% of
blood concentration. The blood levels of young mothers (given two
days before parturition) were the same. Rapid transplacental transfer
but difficult transfer to milk was the conclusion (Trenholm et al.,
1971).
Rats were maintained on diets containing 2 mg/kg methylmercury for 12
weeks. Examination of kidney tissue by electron microscopy indicated
extrusion of numerous cytoplasmic masses from the proximal tubular
cells of the pars recta segment in the kidney. Effects were more
prominent in females than in males (Fowler, 1972).
Rats were administered per os 1 mg methylmercuric sulfide per day
for 20 days. Peripheral nerves were examined on the seventh and two
hundred and fiftieth day after withdrawal of the test compound. On
day 7 mitosis of Schwann cells were observed and on day 250 myelinated
fibres were noticeably decreased in numbers, with regeneration of
nerve fibres and an increase in collagen in the intercellular spaces.
Posterior nerve root ganglion cells were, however, intact.
Regeneration of peripheral nerve cells was not a very marked feature
(Miyakawa et al., 1970).
In another study, rats were administered per os 1 mg methylmercuric
sulfide per day for 20 days. Examination of heart muscle by light
microscopy, indicated a few degenerated muscle fibres. Electron
microscopy indicated that Z-band and every other band were indistinct,
and that the myofilaments and the sarcoplasmic reticulum were
disorganized, degenerated or disappeared (Miyakawa or al., 1970).
Japanese quail given 20 mg/kg of mercury as methylmercury in diets
containing 17% (by weight) tuna, survived longer than quail given this
concentration of methylmercury in a corn-soya diet. When rats were
maintained on a purified basal diet containing 20% casein or the same
diet supplemented with 0.5 mg/kg selenium, with various concentrations
of methylmercury in the drinking water all rats fed 10 mg/kg mercury
without selenium died, but those fed diet with selenium were still
alive at the end of week 6 (Ganther et al., 1972).
Squirrel monkeys were given 0.5-1 mg/CH3203Hg/kg body-weight via the
diet. Pairs of monkeys were fed for 21 days (followed by 85 days
withdrawal), 28 days (one having a 37 day withdrawal period prior to
sacrifice) and 36 days (one having a 6 day withdrawal period prior to
sacrifice). After 121 days feeding, the total body burden of mercury
was 2.8 and 3.6 mg Hg/kg. Neither animal showed symptoms, but at
sacrifice the animal with the greater body burden showed slight
changes in the calcarine cortex. After 28 days feeding, body burdens
were 4.3 and 5.5 mg Hg/kg. The animal with the lower body burden of
mercury showed slowly progressing visual disturbance, and possibly
ataxia. At sacrifice, severe damage to the entire occipital cortex
was observed. The animal with the higher body burden was blind, and
possibly ataxic at termination of dosing when it was sacrificed.
Severe calcarine cortex damage was noted. After 36 days feeding,
monkeys showed 5.0 and 5.5 mg Hg/kg body burden. Both animals were
blind, and, at autopsy, damage to the calcarine cortex was noted. The
severity was greatest in the animal permitted a six day withdrawal
period. Blood brain ratio was 0.2 (Nordberg et al., 1969).
Groups each of five male monkeys (3 Macaca mulatta, 2 Macacus
irus) were administered dietary methylmercury chloride, which had
previously been mixed with egg and the freeze dried product added to
the normal diet. Dose levels were 0, 0.01, 0.03, 0.1 and 0.3 mg Hg/kg
body-weight/day. At the highest dose level, onset of neurologic
symptoms were observed at days 50-73 of test, i.e. after a total
intake of mercury of approximately 60-145 mg. Other groups appear
normal after 140 days of test although a light depression in
body-weight has been observed. Mercury content of hair of test groups
after two months averaged 0.3, 4.8, 19, 44 and 202 µg/g, for control
and respective test groups in order of dose level. The animals
continue on study (Ikeda & Tobe, 1972).
Reproduction studies
Oral administration of 0.1, 1, 2.5 or 5 mg/kg methylmercury chloride
from pregnancy day 6-17 to mice, reduced litter size at the highest
level. All pups died at 2.5 mg/kg with retarded cerebellar
development on seventh to fourteenth day post-partum. 0.1 mg/kg was
the no effect level (Clegg, 1971).
Methylmercuric sulfide labelled at the mercury and the sulfur atoms
was given to pregnant rats in daily oral doses of 0.1 mg/kg
body-weight. Autoradiographs demonstrated that the sulfur and the
mercury were present inside the placenta and in the maternal milk.
Neurological symptoms in the young gave evidence of the transfer to
the foetus. Evidence is presented to show that the mercury-sulfur
bond is dissociated in the animal's body soon after oral ingestion and
that a mercury compound without sulfur is transferred to the young
(Fujita, 1969).
Mutagenic studies
In a dominant lethal test performed in mice there was no effect 24
hours after exposure but this may have been too late as regards
mating. When 0.1 mg Hg/kg was injected intraperitoneally prior to
mating reduced pregnancy incidence was noted but no litter size
reduction. In another experiment, male rats received by injection for
seven days, 1, 2.5 or 5 mg Hg/kg and were then mated with untreated
females over a period of four weeks. In matings 10-15 days after
injection there was reduced litter size at all dose levels, the
severity being dose related. Thus a positive response at all levels
was noted (Clegg, 1971).
ASSESSMENT OF HAZARD TO MAN
In view of the large differences in metabolism of methylmercury
compounds in various species of animals, toxicological evaluation
based on animal data was not attempted and only human data were used.
Data were of three types: (1) levels of mercury in hair and whole
blood of patients with methylmercury poisoning in Niigata; (2) levels
of mercury in hair, whole blood and blood cells in subjects without
any symptoms of methylmercury poisoning in Japan, Sweden and Finland;
(3) calculated relationships between intake of methylmercury from the
diet and levels of mercury in hair, whole blood and bLood cells in
subjects in Japan and Sweden.
From these data a minimum toxic dose of methylmercury compounds in
fish was calculated, and a proposal was made for a provisional
tolerable weekly intake of mercury from food. Theoretically, a more
direct approach to the establishment of the minimum toxic dose would
have involved the precise estimation of the dietary intake of fish by
the cases of poisoning in Niigata or Minamata coupled with accurate
analyses of the mercury levels in the fish of Niigata or Minamata
consumed at the time of poisoning, but available data did not permit
this approach. It should be realised therefore that with the
alternative procedure used, each of the above steps embodied factors
of uncertainty, one common factor being the accuracy of the analysis
of the mercury. For these reasons, the proposed tolerable weekly
intake must be considered as provisional for the time being.
Indices in man of exposure to methylmercury compounds.
Methylmercury and other short chain alkylmercury compounds exert their
main toxicological effects on the nervous system and the most relevant
index of exposure would therefore be the concentration of mercury in
relevant areas of the central nervous system. In man, methylmercury
levels in blood cells and in hair provide the best indices of exposure
of the nervous system to methylmercury compounds. If exposure to
other mercury compounds is minor compared with exposure to
methylmercury, analysis of total mercury may be used instead. A
strict correlation between levels in the nervous system, blood and
hair can only be expected when a steady state has been reached, i.e.
during long-term exposure to constant levels of methylmercury
compounds. Blood levels of mercury reflect more accurately the intake
from recent exposure to methylmercury, while hair levels of mercury
reflect the average intake over a more prolonged period. Since
mercury combines with protein components in hair over the period of
its formation, the mercury content in successive segments of hair
gives an indication of the levels of past absorption of mercury
compounds. The estimation of mercury in hair, provided external
contamination can be excluded, may therefore be of value in
epidemiological investigations.
Levels of mercury in whole blood, blood cells and hair. The
relation between mercury levels in blood and hair has been studied in
Sweden, Finland and Japan in persons exposed to different amounts of
mercury through the consumption of contaminated fish. The results are
summarized in the following table:
No. of Blood (x) Hair (y) Linear References
subjects Hg )19/9 Hg µg/g regression
12 .004-.65 1-180 y = 280X - 1.3 Birke et al.,
1967 (Sweden)
51 .004-.11 1- 30 y = 230X + 0.6 Tejning, 1967
(Sweden)
30 .005-.27 0 - 56 y = 140X + 1.5 Sumari et al.,
(1969 (Finland)
45 .002-.80 20-325 y = 260 x + 0 Tsubeky, 1972
(Japan)
The data presented by Tsubaki included 45 persons from the Niigata
area with more subjects in the very high range (seven subjects above
180 µg/g hair). Hair levels at a steady state were approximately 250
times as high as levels in whole blood.
Levels of mercury in hair and whole blood of patients with
methylmercury poisoning in Niigata were determined in a follow-up
investigation extending over many months after the onset of clinically
overt poisoning (Tsubaki, 1972). The decrease of mercury levels in
hair of eight patients studied corresponded to a biological half-life
of 50-108 days (median 66 days), and in the blood of seven patients
studied of 35-137 days (median 55 days). These figures are of the
same order of magnitude as those reported from experiments with
CH3Hg203 in man in healthy subjects Ekman et al., 1968a, 1968b;
Miettinen et al., 1971; Birke et al., 1967; Tejning, 1969a, 1969b,
1969c) and from data on persons in Sweden who stopped eating
methylmercury-contaminated fish (Berglund et al., 1971).
Extrapolation of the mercury levels in the Niigata patients to the
time of onset of symptoms indicated that poisoning occurred at mercury
levels in the hair between 200 and 1000 µg/g but in one case the hair
level was as low as 50 µg/g. Similar extrapolation of the blood data
indicated that poisoning occurred at mercury levels in whole blood
between 0.2-2.0 µg/g.
More than 100 persons with hair levels of 50 µg/g have been found in
Japan, in areas where fish was contaminated by methylmercury. Of
these persons, six in Minamata and 17 in Niigata had blood levels
above 0.2 µg/g. In Sweden and Finland mercury levels in blood above
0.2 µg/g were found in five persons, two of which had O.6-0.7 µg/g.
None of these persons had any detectable evidence of methyl-mercury
poisoning.
Relationship to intake of mercury from fish
The intake of mercury through the consumption of fish and levels of
mercury in blood cells or hair has been studied in subjects without
clinical evidence of poisoning and the following relationships
established.
No. of Hg intake Hg in blood Hg in hair Reference
subjects mg/day (x) Cells µg/g (y2)
µg/g (y1)
6 + 26* 0-0.8 Y1 = 1.4X + - Birke et al.,
0.003 1967
139 + 26* 0-0.4 y1 = 0.6x + - Tejning, 1967;
0.011 1969a, b, c
735 0-0.8 - Y2 = 150X + Kojima & Araki,
1.66 1972
* 26 cases from Tejning's non-fish consumers (1969a, b, c; 1970).
Since mercury in fish is largely in the form of methylmercury, the
proportion of the total body burden of mercury present as
methylmercury also increases with the degree of exposure.
Conclusions
The most reliable data for the toxicological evaluation of mercury in
fish derive from adults with neurological involvement. The lowest
mercury levels associated with the onset of clinical disease have been
reported to be 50 µg/g in hair and 0.2 µg/g in whole blood,
corresponding to 0.4 µg/g in blood cells. These levels would seem to
represent individuals on the lower end of a distribution curve of
methylmercury levels in blood and hair in the population exposed in
Niigata. These individuals were considered to be more sensitive to
the action of methylmercury and this assumption is supported by the
observation of a number of persons in Japan and other countries with
higher mercury levels in hair or blood but without clinical features
of intoxication. On the other hand, the presence of methylmercury
poisoning can only be diagnosed clinically because there are no
specific laboratory tests for its early detection. The occurrence of
prenatal intoxication also calls for caution. Clinical data from
Japan indicate that the foetus is more sensitive than the mother. The
significance of the morphological changes seen in chromosomes of
circulating lymphocytes in a few individuals with a high methylmercury
blood level is not known at present.
Due to the uncertainty and indirect nature of relationships calculated
between intake of mercury through fish, levels of mercury in
"indicator organs" and mercury levels when poisoning develops, any
evaluation must be provisional at present.
EVALUATION
Lowest Hg levels at onset of poisoning in adults with neurological
involvement: 50 µg/g hair and 0.4 µg/g blood cells.
Estimated intake of mercury, mainly as methylmercury compounds,
causing poisoning; 0.3 mg/day over prolonged periods.
PROVISIONAL TOLERABLE WEEKLY INTAKE: 0.3 mg total mercury, of which no
more than 0.2 mg should be present as methylmercury, CH3Hg+,
(expressed as mercury).
This corresponds to a provisional tolerable weekly intake of 5 µg/kg
body-weight of total mercury, of which no more than 3.3 µg/kg
body-weight should be methylmercury compounds (expressed as Hg).
This evaluation is based on a high total mercury intake in the diet,
due to the consumption of fish containing high levels of methylmercury
compounds. Where high total mercury intakes occur for other reasons,
(for example due to inorganic mercury as a consequence of local
natural conditions) the evaluation does not necessarily apply and the
whole situation has to be reconsidered on its merits. When the total
mercury intake in the diet is found to exceed 0.3 mg per week it would
be advisable to undertake analyses for methylmercury. Other suitable
investigations should be instituted and all possible steps taken to
keep the levels of methylmercury as low as possible.
The provisional tolerable weekly in-take, as given above, is probably
not exceeded at present by the population in general in any country,
but it is exceeded by fish-eating minorities of some countries. In
view of the international importance of fish and fishery products in
the diet every effort should be made to prevent further pollution by
mercury.
REFERENCES
Åberg, B., Ekman, L., Falk, R., Greitz, U., Persson, G. & Snihs, J. O.
(1969) Arch. Environ. Hlth., 19, 478
AOAC (1972) Official methods of analysis, Washington D.C.
Bakulina, A. V. (1968) Sov. Med., 31, 60
Berglund, F. (1969) Nord. Hyg. T., 50, 118
Berglund, F., Berlin, M. & Birke, G. (1971) Nord. Hyg. T.
(Supplementum 4)
Berlin, M. (1970) Paper submitted to WHO
Berlin, M. & Ullberg, S. (1963) Arch. Environ. Hlth., 6, 40
Bidstrup, P. L. (1964) Toxicity of mercury and its compounds. In:
Browning, E., ed., Elsevier monographs in toxic agents, Elsevier
Publishing Co., Amsterdam, London & New York, 75-112
Birke, G., Johnels, A. G,, Plantin, L, O., Sjöstrand, B. & Westermark,
T. (1967) Lakartidningen, 64, 3628
Birke, G., Johnels, A. G., Plantin, L. O., Sjöstrand, B., Skerfring,
S. & Westermark, T. (1972) Arch. Environ, Hlth., 25, 77
Bligh, E. G. (1971) Proceedings of the Symposium Mercury in Man's
Environment, Ottawa, February 15-16
Clarkson, T. W. (1971) Soc. Tox., Washington, March 7-11, Abstract No.
125, p. 97
Clarkson, T. W., Small, H. & Norseth, T., Fed. Proc. (in press)
Clegg, D. J. (1971) Proceedings of the Symposium Mercury in Man's
Environment, Ottawa, February 15-16
Curley, A., Sedlak, V. A., Girling, E. F., Hawk, R. E., Barthel, W.
F., Pierce, P. E. & Likosky, W. H. (1971) Science, 172, 65
Dahhan, S. S. & Orfaly, H.,(1964) Amer. J. Cardiol., 14, 178
Edwards, G. N. (1865) Two cases of poisoning by mercuric methide, St.
Bartholomew's Hospital Reports, 1, 141
Ekman, L., Greitz, U., Persson, G. & Aberg, B. (1968a) Nord. Med., 79,
450
Ekman, L., Greitz, U., Magi, A., Snihs, J. O. & Aberg, B. (1968b)
Nord. Med., 79, 456
Engleson, G. & Herner, T. (1952) Acta Paediat. (Uppsala), 41, 289
Environmental Research (1971) Hazards of mercury, 4, 26
Faile, S. P. (1971) Environment, 13, 51
Falk, R., Snihs, J. O., Ekman, L., Greitz, U. & Aberg, B. (1971) Acta
Radiol., 9, 55
Fleischer, M. (1970) U.S. Geological Survey, Profes. Paper 213, p. 6
Fowler, B. A. (1972) Science, 175, 780
Fujita, E. (1969) J. Kumamoto Med. Soc., 43, 47
Ganther, H. E., Goudie. C., Sunde, M, L., Kopecky, M. J., Wagner, P.,
Oh, S. W. & Hoekstra, W. G. (1972) Science, 175, 1123
Gis'IU, P. & Pozner, Z. A. (1970) Pediatr., 49(3), 80
Goldwater, L. J., Ladd, A. C. & Jacobs, M. B. (1964) Arch. Environ.
Hlth., 9, 735-741
Haq, 1. U. (1963) Brit. Mod. J., 1, 1579
Häsänen, E. & Sjöblom, V. (1969) Nord. Hyg. T., 50, 27
Hill, W. H. (1943) Can. J. Publ. Hlth., 34, 158
Hirayama, K. & Takahashi, H. (1971a) Rinshe Byori, 19(1), 51
Hirayama, K. & Takahashi, H. (1971b) Nature (Lond.), 232, 201
Hunter, D., Bomford, R. R. & Russel, D. S. (1940) Quart. J. Med., 9,
193
Ikeda, Y. & Tobe, M. (1972) Two-year study, Interim report submitted
to WHO
Irukayama, K. (1968) Animal experiments with substances obtained by
various treatments of the poisonous fish and shellfish, Minamata
disease. Study Group of Minamata Disease, Kumamoto Univ., Japan
IUPAC (1965) Int. Union Pure Appl. Chem. Inform. Bull. Append., 10, 71
Jalili, M. A. & Abbasi, A. H. (1961) Brit. J. industr. Mod., 18,
303-308
Jensen, S. & Jernelöv, A. (1969) Nature (Lond.), 223, 753
Joselow, M. M. & Goldwater, L. J. (1967) Arch. Environ. Hlth., 159 155
Kojima, K. & Araki, T., Paper submitted to WHO
Kudsk, F. N. (1965) Acta Pharmacol. Toxicol., 23, 250
Kazantzis, G., Schiller, K. F. R., Asscher, A. W. & Drew, R. G. (1962)
Quart. J. Med., 31, 403
Ladd, A. C., Goldwater, L. J. & Jacobs, M. B. (1964) Arch. Environ.
Hlth., 9, 43-52
Lu, F. C., Berteau, P. E. & Clegg, D. J. (1972) International Atomic
Energy Agency, Vienna, Techn. Rep. Ser. 137
Lundgren, K. D.,, Swensson, Å. & Ulfvarson, U. (1967) Scand. J. Clin.
Lab. Invest., 20, 164
Matsumoto, H., Koya, G. & Takeuchi, T. (1965) J. Neuropath. & Expl.
Neurol., 24, 563
Miettinen, J. K. (1971) IAEA Panel on Nuclear Techniques for Studying
Pesticide Residues in Food & Environment, October 25-29, Vienna
Miettinen, J. K., Rahola, T., Hattula, T., Rissanen, K. & Tillander,
M. (1971) Am. Clin. Res., 3, 118
Miyakawa, T., Deshimaru, M., Sumiyoshi, S., Teraoka, A. & Tatesu, S.
(1970) Acta. Neuropath. (Berl.), 17, 6 and 80
Mnatsakonov, T. S., Mamikonyan, R. S., Govorkyan, G. G. & Nazaretyan,
K. L. (1968) Gig. Truda. prof. Zabol., 12(7), 39
Morsy, S. M. & El-Assaly, F. M. (1970) Health Physics., 19, 769
Mukhtarova, N. D. (1970) Gig. Truda. prof. Zabol., 14(9), 27
Nordberg, G. F., Berlin, M. H. & Grant C. A. (1969) 16th Int. Congr.
Occ. Hlth., Tokyo
Nordic Committee on Food Analysis (1972) Methylmercury compounds
determination in fish, No. 82
Norseth, T, (1969) Studies on the biotransformation of
methylmercury/salts in the rat, Ph.D. thesis, Univ. of Rochester,
N.Y., 98
Norseth, T. (1971) Acta Pharm. Toxic., 29, 375
Norseth, T. & Clarkson, T. W. (1971) Arch. Environ. Hlth., 22, 568
Okinaka, S., Yoshikawa M., Mozai, T., Mizuno, Y., Terao, T.,
Watanabe, H., Ogihara, K., Hirai, S., Yoshino, Y., Inose, T., Anzai,
S. & Tsuda, M. (1964) Neurol., 14, 69
Ordońez, J. V., Carillo, J. A., Miranda, C. M. & Gale, J. L. (1966)
Bol. Ofic. Sanit. Panamer., 60, 510
Rahola, T., Aaran, R. K. & Miettinen, J. K. (1971) IAEA/WHO Symposium
on Assessment of Radioactive Bodyburden, Stockholm, November 22-26
Sandi, E. (1970) Poisoning by dressed maize seeds in Akatsi, Ghana in
May 1967, personal communication
Skerfving, S., Hansson, K. & Lindsten, J. (1970) Arch. Environ. Hlth.,
21, 133
Slatov, I. V. & Zimnikova, Z. I. (1968) Gig. i sanit., 33 (4-6), 275
Snyder, R. D. (1971) New Engl. J. Med., 284, 1014
Storrs, B., Thompson, J., Fair, G., Dickerson, M. S., Nickey, L.,
Barthel, W. & Spaulding, J. E. (1970) Morbidity and Mortality Weekly
Report, U. S. Publ. Hlth. Serv., 19, January 24 and 31
Sumari, P., Backman. A. L., Karli, P. & Lahti, A. (1969) Nord. Hyg.
T., 50, 97
Sumari, P., Partanen, T. & Hietala, S. (1971) Private communication to
WHO
Suzuki, T., Miyama. T. & Katsunuma, H. (1967) Ind. Hlth., 5, 290
Suzuki, T., Miyama, T. & Katsunuma, H. (1971) Bull. Envir. Contam. &
Toxic., 5, 502
Swensson, Å, (1952) Acta. Med. (Scand.), 143, 365
Takeuchi, T. (1968) Pathology of Minamata disease, Minamata Disease,
Study Group of Minamata Disease, Kumamoto Univ., Japan
Takeuchi, T. (1970) International Conference on Environmental Mercury
Contamination, Ann Arbor, Michigan, September 1970
Tam, K. C. & Armstrong, F. A. J. (1972) Manuscript Report Series No.
1167, Fisheries Research Board of Canada
Tejning, S. (1967) Reports 67-02-06, 67-02-10, 67-08-31, Dept. of Occ.
Med., S-221-85 Lund, Univ. Hospital
Tejning, S. (1969a) Report 69-04-20, Dept. of Occ. Med., S-221-85
Lund, Univ. Hospital
Tejning, S. (1969b) Nord. Hyg. T., 50, 103
Tejning, S. (1969c) Annual Report to the Swedish Medical Research
Council, May 27, 1969
Tejning, S. (1970) Report 67-08-31, Dept. of Occ. Med., S-221-85 Lund,
Univ. Hospital
Trenholm, H. L., Paul, C. J., Baer, H. & Iverson, F. (1971) Soc.
Toxic., Washington, 7-11 March, 1971, Abstract No. 126, p. 97
Tsubaki, T. (1972) Paper submitted to WHO
Ulfvarson, U. (1962) Int. Arch Gewerbepath. Gewerbehyg., 19, 412
U.S, Food & Drug Administration (1971) The Mercurial Pesticides
Registration Review Panel
Wallace, R. A., Fulkerson, W., Shults, W. D. & Lyon, W. S. (1971) Oak
Ridge National Laboratory Report, ORNL-NSF-EP-1
Weiss, H. V., Koide, M. & Goldberg, E. G. (1971) Science, 174, 692
Wershaw, R. I. (1970) U.S. Geological Survey, Profes. Paper 213,
p. 29
Westöö, G. (1972) Personal comunication
Westöö, G. & Rydälv, M. (1969) Vår foda, 3, 19
Westöö, G. & Rydälv, M. (1971) Vär foda, 7-8, 179
Wood, J. M., Scott Kennedy, F. & Rosen, C. G. (1968) Nature (Lond.),
220, 173
WHO (1971) International Standards for Drinking Water (Third Ed.)