United States
                    Environmental Protection
                    Agency
 Environmental Criteria and
 Assessment Office
 Research Triangle Park NC 27711
                    Research arid Development
EPA/600/S8-84/019F Mar. 1987
<>EPA         Project  Summary
                    Mercury  Health  Effects  Update:
                    Health  Issue  Assessment
                      This  document summarizes  the
                    health basis for the national emission
                    standard for mercury, originally set in
                    1973 and currently under review.
                    Mercury is unique among metals as the
                    only metal  in liquid form at room
                    temperature. It exists in three oxidative
                    states—metallic  (Hg°), mercurous
                    (Hgz++), and mercuric (Hg++) mercury—
                    and a wide variety of chemical forms,
                    the most important of which are com-
                    pounds of methyl mercury, mercuric
                    mercury,  and the vapor of  metallic
                    mercury. The global cycle of mercury
                    involves the emission of Hg° from land
                    and water surfaces to the atmosphere,
                    transport of Hg° in the atmosphere on
                    a widespread basis,  possible conver-
                    sion to unidentified soluble species and
                    return  to land  and water through
                    various depositional processes.  The
                    major source of human exposure to
                    methyl  mercury  is by diet through the
                    consumption of fish and fish products.
                      Mercury vapor is inhaled,  whereas
                    uptake  of inorganic and methyl mer-
                    cury compounds is primarily through
                    oral ingestion. Once absorbed, mercury
                    in all forms is distributed by  the
                    bloodstream to all body tissues; how-
                    ever, tissue distribution of methyl
                    mercury is  more  uniform. Mercury
                    vapor and methyl mercury readily cross
                    the blood-brain and placenta! barriers.
                      Chronic exposure to mercury com-
                    pounds primarily affects the central
                    nervous system and kidneys. Depend-
                    ing upon the form of mercury and level
                    of intake, effects on the adult nervous
                    system can range from  reversible
                    paresthesias and malaise to irreversible
                    destruction of neurons in the cerebellar
                    and visual cortices, leading to perma-
                    nent signs of ataxia and constriction of
                    the visual field. The fetus is most
                    sensitive to methyl mercury poisoning,
                    with effects in  infants ranging from
psychomotor retardation to a  severe
form  of cerebral palsy. All prenatal
effects have to date been  found
irreversible.
  This  Project Summary was  devel-
oped by EPA's Environmental Criteria
and Assessment Office.  Research
Triangle Park, NC. in EPA's Office of
Health and Environmental Assessment.
It announces key  findings of the
research project that is fully docu-
mented in a separate report of the same
title (see  Project Report ordering
information at back).
Summary and Conclusions


The National Emission
Standard of 1973
  On April 6,1973, the U.S. Environmen-
tal Protection  Agency promulgated  a
National Emission Standard for Mercury
as 3 Hazardous  Air Pollutant  under
Section 112 of the Clean Air Act (38 FR
8820). The scientific information  which
served as  the health  basis for  the
standard is presented below.
  By 1973, the fact that exposure to
metallic mercury  vapor caused central
nervous system injury and renal damage
was well established. Prolonged expo-
sure to about 100 //g Hg/m3 of mercury
vapor involved a definite risk of mercury
intoxication. To determine the ambient
air level of mercury that did not  impair
health, the  impact of airborne burden
was considered  in conjunction with
water and food burdens. Methyl mercury
compounds were considered the most
hazardous form  of mercury  and  the
overall human body burden was believed
to be  mainly derived from ingesting
methyl mercury in the diet,  particularly
by fish which concentrate this form of
mercury through the food chain.  When

-------
can methylate inorganic mercury, but thlj
most efficient and effective are certain
aerobes and facultative anaerobes. Strict
anaerobes such as methanogenic bac-
teria are less efficient. Both mono- and
dimethyl  mercury  compounds  are
formed, depending  on pH and  other
conditions.  Monomethyl  mercury com-
pounds rapidly diffuse from the microor-
ganisms and rapidly accumulate  in
aquatic food chains. The highest concen-
trations are found in large predatory fish
at the top of the aquatic food chain, e.g.,
trout, pike, and bass in fresh water and
tuna, swordfish, and red snapper  in
oceanic water.
  Many factors influence the methyl
mercury accumulation in fish. These
include not only the species of fish but
also the age of the fish, the  levels of
mercury in sediment, the presence of
zooplankton,  organic content of sedi-
ments and  particulates  in the water
column, water temperature, redox poten-
tial, and dissolved  oxygen  content.
Current studies  suggest  that  long-
distance atmospheric transport and
acidification of rain water  are correlated
with elevated levels of methyl mercury
in fresh water fish.
  Methyl  mercury  compounds
demethylated by microorganisms in
environment. The accumulation  o
methyl mercury in  the food chain ulti-
mately depends upon an ecological
balance  between  methylation  and
demethylation  reactions as well  as
trapping within organisms and chemical
complexes.


Levels of Mercury in Air,
Water and Food

  Concentrations  of  mercury  in the
atmosphere were estimated to be about
20  ng Hg/m3.  However,  more recent
unconfirmed observations indicate that
a more accurate  estimate of  average
atmospheric levels  is in the range of 2-
10 ng Hg/m3. Concentrations of mercury
in fresh water are around 25 ng Hg/l,
most of which is probably in the mercuric
(Hg++) form.  Concentrations of mercury
in fish and fish products (the dominant
food source of mercury in the human diet)
average from 100 to 220 ng Hg/g fish,
almost all (70 to 90  percent)  of which
is methyl mercury. About two-thirds of
all fish consumers  in the U.S. are tuna
eaters. However, on an individual con-
sumer basis, freshwater fish, e.g., pike,
bass, and  trout, have the highest co
sumption rate. The diet greatly excee
the first standard was set, it was con-
sidered  prudent  to assume that expo-
sures to  methyl mercury (diet)  and
mercury vapor (air) were equivalent and
additive.
  Also known by 1973 was the fact that
methyl  mercury compounds,  when
ingested  in  sufficient amounts, could
produce severe and irreversible damage
to the central nervous system both in the
adult  and in the human  fetus.  Non-
specific  symptoms such as paresthesia
occurred at  the  lowest body burden.
These first symptoms of  intoxication
were observed in adults after prolonged
intake of about  300 fjg Hg as methyl
mercury per  70 kg body weight.  It was
assumed that a safety factor of 10 would
provide  satisfactory  protection  against
genetic lesions and poisoning of the fetus
and of children. Thus, it was determined
that the total intake of methyl mercury
should not exceed 30 fjg Hg/day/70 kg
body  weight.  Because the burdens of
mercury vapor and methyl mercury were
assumed to be equivalent and additive,
it followed that daily absorption of both
forms of mercury should not exceed 30
fjg Hg/70 kg  body weight.
  From estimates of average diets, over
a lengthy period, it was also determined
that mercury intakes of 10 jug Hg/day/
70 kg body weight might be expected.
Thus, to restrict total intake to 30 fjg Hg/
day/70  kg body  weight,  the average
mercury intake from air would have to
be limited to 20 jug Hg/70 kg body weight.
To maintain this level, the air would have
to contain an average concentration of
no more than 1 fjg Hg/m3, assuming the
70-kg adult inhaled 20 mVday.
  Before promulgation of the standard,
data on  the environmental  transport of
mercury did not  permit a clear assess-
ment of the impact  of atmospheric
mercury emissions on aquatic  and
terrestrial environments. Consequently,
the standard promulgated in 1973  was
intended to protect public health from the
effects of  inhaled mercury, taking  into
consideration dietary  contributions to
total body burden. It did not, however,
account for the effects of  atmospheric
mercury  on  other environments  that
contribute to indirect exposures to
mercury.

Findings of Current Report

  The findings of this report are based
mainly  on a review  of the scientific
literature published since the promulga-
tion of the standard in 1973. However,
literature  published  before 1973 and
relevant to assessing human health risks
from airborne  mercury has also been
evaluated. The following is a summary
of the full document.

Mercury Background
Information

The Global Cycle
  Despite  intensive  research in recent
years,  many details, both quantitative
and qualitative on the global cycling of
mercury remain  obscure. Elemental
mercury vapor,  Hg°, emitted from land
and  water surfaces,  is  the principal
species in the atmosphere and is respon-
sible for long-distance (100 to  1,000
kilometers) transport of mercury. The
residence time for this form of mercury
is months,  possibly  years. A "soluble"
form of mercury, of  unknown chemical
species, is  also present in the atmos-
phere  but to a lesser  extent.  Soluble
mercury is  returned by precipitation to
the earth's surface. Its residence time in
the atmosphere is believed to be days.
  Widely varying estimates from  10 to
80  percent  have been made  for  the
anthropogenic contribution of mercury to
the atmosphere. The figure for this report
is about 25 percent. Anthropogenic input
of mercury into bodies of fresh water is
believed to have  increased  mercury
levels in waters and sediments by factors
two to  four as  compared to eras before
the advent of man.  Oceanic sediments
are the ultimate depository of mercury
in the form of insoluble mercuric sulfide.
The amount of mercury in oceanic water
is believed to be so large (tens of millions
of tons) that man's impact  has been
negligible.  However,  anthropogenic
sources have had substantial impact on
the levels of mercury in  aquatic orga-
nisms  in  marine coastal waters near
urban centers.

Biomethylation of Inorganic
Mercury
  Within the context of the global cycle,
microbially-mediated methylation and
demethylation reactions of mercury take
place in sediments. These processes also
can take place  within,  and on, organic
paniculate microenvironments  within
the water column, especially in eutrophic
and  hypereutrophic aquatic systems.
These  reactions play a key role  in  the
entry of methyl  mercury into the human
diet. Microorganisms  are  capable of
methylating inorganic mercury by a non-
enzymic reaction with  methyl  cobala-
mines. Many types  of  microorganisms

-------
    her media as a  source of human
   xposure  and  absorption  of  total
  mercury.

  Pharmacokinetics and
  Biotransformation of
  Mercury in Man and Animals—
  Vapor of Metallic Mercury
    Since  1973, new information has
  become available on the pharmacokinet-
  ics of inhaled vapor in man and on the
  biotransformation of elemental mercury
  in animal tissues. Although a complete
  pharmacokinetic model has not yet been
  established,  some  principal  features
  have been identified.
    Mercury vapor is inhaled and adsorbed
  through the skin. The extent of uptake
  through the latter route is still unknown.
  About 80 percent of the inhaled  vapor
  is retained and rapidly transferred  to the
  bloodstream where it is distributed to all
  tissues  in  the  body. Mercury readily
  crosses  the blood-brain and placenta!
  barriers but accumulates to the greatest
  extent in the kidney. It is mainly excreted
  in  urine and feces;  however, it is also
  excreted in sweat, saliva, and, to a small
  extent, expired air. Because of the rapid
  oxidation of  absorbed mercury vapor  to
  Ijvalent mercury in mammalian tissues,
 _ is likely that  most of  the mercury
  accumulated  in the kidney and  other
  tissues and  excreted is in the  divalent
  form.
   The whole body half-time of mercury
  in man is approximately 50 to 70 days.
  A rapid component in blood has a half-
  time of about three days,  and a slower
  component has a half-time of about 30
  days. A rapid component in the brain has
  a half-time of about 21 days.  There  is
  evidence  of a much slower component
  in the brain  with a half-time of several
  years.
   Dissolved elemental mercury, Hg°,  is
  believed to be the most important mobile
  species  in mammalian tissues.  It  is
  oxidized  to Hg++  by the  hydrogen-
  peroxide-catalase complex in red  blood
  cells, liver, kidney and,  probably,  many
  other  tissues. Ethanol, at low,  non-
  intoxicating doses,  can   inhibit this
  oxidation  process and thereby lead to a
  decreased retention of inhaled mercury.
  Inorganic  divalent  mercury,  Hg*+,  is
  reduced to Hg° in fiver and kidney tissues
  and, probably, in other tissues as well.
  The biochemical mechanism  for this
  reduction  has not been  identified  in
  mammalian cells but is well established
^t bacteria to the extent that even the
I^Bhetics, sequence of the enzymes, and
 active site have been determined through
 DNA-sequence analysis.
   Inhaled mercury vapor can induce the
 metal-binding protein metallothioein in
 kidney tissue,  and  mercury-selenium
 complexes may be formed after chronic
 exposure.

 Compounds of Inorganic
 Mercury
   Since 1973, new information has been
 published on the kinetics of tracer doses
 of inorganic mercury in man. However,
 a  complete pharmacokinetic model for
 either mercurous or mercuric forms of
 mercury has not yet been established.
   Although mercurous chloride is slowly
 and incompletely absorbed after oral
 dosing, amounts  sufficient to  lead to
 symptoms of  mercurialism  can be
 absorbed chronically. Once absorbed,
 mercurous mercury is  probably  con-
 verted to the mercuric form. Ingested as
 such, mercuric mercury has a less than
 15 percent rate of absorption following
 oral dosing. However, inferences from
 animal studies suggest that absorption
 of mercuric mercury may be as high as
 50 percent in children and infants.
   Mercuric mercury is distributed by the
 bloodstream to all tissues in the body but
 penetrates the blood-brain and placenta!
 barriers to a much lesser extent (about
 10 times less) than mercury vapor. As
 with mercury vapor, the kidney concen-
 trates inorganic  mercury to a  much
 greater extent than  other tissues and
 excretes  it mainly  through  urine and
 feces.
  The average biological half-time of a
 tracer dose of divalent inorganic mercury
 compounds in man  is 42 days for the
 whole body and 26 days for blood.

 Methyl  Mercury Compounds
   New information  on the kinetics of
 methyl mercury  in  man and  on  the
 mechanisms of biotransformation and
 excretion does not change  the picture
 that existed when the 1973 standard was
 promulgated. The  pharmacokinetic
 model of methyl mercury in man is much
further advanced than models for other
forms of mercury.
  Oral doses of methyl mercury are 95
 percent absorbed and  are distributed
through the blood to all tissues of the
body within a few days. Like other forms
of mercury,  methyl mercury readily
crosses the  blood-brain and placenta!
barriers;  however, unlike other forms,
tissue concentrations of methyl mercury
are much more uniform, with the kidney
having somewhat higher concentrations
than other tissues. The red blood cell-
to-plasma concentration ratio is typically
20:1, the blood-to-brain ratio is approx-
imately 1:5, and the blood-to-hair (newly
formed) ratio is 1:250.
  About  80  percent  of  total methyl
mercury  is excreted through  feces.
Methyl mercury is also secreted in bile
and  reabsorbed back into  the  blood-
stream to form  an enterohepatic cycle.
The remainder is demethylated by flora
in the intestine  and is excreted as
inorganic  mercury. According to exper-
imental studies, a functioning enterohe-
patic cycle  and active  gut flora are
essential  in order for mercury  to be
excreted after doses of methyl  mercury.
Suckling animals cannot excrete  methyl
mercury in bile and, therefore, are unable
to excrete significant amounts of  methyl
mercury from  the  body. How  these
animal data compare to human data is
unknown  since there is no  available
information on human infants.
  The  average  biological half-time  of
methyl mercury in human adults is about
70 days in the whole body and 50 days
in the blood. The biological half-time in
the brain  is probably similar to that of
the whole body or slightly  longer.  For
adults  heavily exposed to methyl mer-
cury, wide range of biological half-times
(up to 120 days  in the whole body) have
been  reported.  The  reason  for  this is
unknown, but in experimental studies,
diet was shown to affect half-times in
the animals.

Toxic Effects of Mercury in
Man and Animals
  The following  section summarizes the
toxic effects  of mercury in man and
animals, with emphasis on effects from
long-term low-level  exposures.  Where
possible,  greater attention was focused
on direct observations of humans.

Vapor of Metallic Mercury
  Since 1973, several clinical and epide-
miological studies have been published
on workers  occupationally exposed  to
mercury vapor. Generally, these studies
support findings that existed when the
mercury standard was promulgated.
  Occupational studies  showed that
chronic exposure to mercury  vapor
affects primarily the  central  nervous
system and the kidneys.  Effects  asso-
ciated with the lowest exposure levels—
below  100 fig  Hg/m3—produce non-
specific symptoms such as introversion,
insomnia,  and  anxiety.  Biochemical

-------
alterations were observed in enzymes of
plasma and red blood cells, and increases
in urinary excretion of specific proteins
and enzymes occurred. Higher chronic
exposures produce more  pronounced
effects in cognitive function, such  as
short-term memory loss and changes in
personality traits (e.g., increased anxiety
and introversion). An overall body tremor,
typical to mercurialism cases, signals the
motor disfunctioning  of  the  central
nervous system. No threshold for these
and other effects has been established,
although effects have not been seen at
air concentrations around 10/ug Hg/m3.
  Renal effects may be mediated through
an  autoimmune  mechanism and may
exhibit wide individual ranges of sensi-
tivity.  Experimental studies on this
sensitivity indicate the possiblity of  a
genetic component related  to the major
histocompatibility complex.
  Effects on both the nervous system and
the kidneys are usually  reversible,
particularly if  the effects are  mild.
Studies  showed that motor effects
reverse more readily than cognitive and
neurotic effects.
  Information is lacking on  reproductive
and developmental  effects of inhaled
mercury vapor.

Inorganic Compounds of
Mercury
  There has been no new information on
the effects of  inorganic compounds of
mercury on  humans since 1973. Two
adult  case  studies showed that  many
years  of chronic oral intake of mercurous
chloride (250 mg/day) resulted in signs
of mercurialism and chronic renal failure.
Chronic oral exposure  to mercurous
chloride has also caused acrodynia or
Pink's disease in children. Differences in
individual  sensitivity were  reported for
such effects, although,  generally, acro-
dynia  associated with  urinary levels of
50  /ug Hg/l is reversible  in children.
Comparison  studies of mercury  vapor
and mercurous  sales  indicate that
chronic exposure  to  compounds  of
mercuric mercury would mainly  affect
kidney function. However, there are no
experimental studies to support this.

Compounds of Methyl  Mercury
  Information,  since 1973,  is now avail-
able on a population exposed to dietary
methyl mercury. Findings from the study
population support evidence that existed
in 1973 and confirm the sensitivity of the
fetus to methyl mercury exposure. Milder
effects at dose levels lower than those
observed in 1973 have been reported for
the first time for exposed fetuses.  The
possibility of delayed effects appearing
in adulthood has been raised.
  Methyl mercury primarily damages the
central nervous system of adults  and
fetuses. Prenatal exposures at the lowest
recorded  levels produce  signs of psy-
chomotor retardation in infants. Recent
studies show that male infants are more
sensitive than females at these low levels
of maternal dietary intake. Substantially
higher prenatal exposures, some only
occurring in the last trimester of preg-
nancy, produce a severe form of cerebral
palsy. Although detailed consequences
of  methyl mercury  poisoning  are
unknown, prenatal  effects appear to be
due to a derangement of the normal
processes of growth and development of
the central nervous system. Ongoing
studies, based on estimated blood  levels
in the mother  during pregnancy,  show
that the fetus is about three times more
sensitive than the adult to methyl
mercury exposures.
  Effects  on the adult  central  nervous
system result in focal damage to specific
areas of the brain, principally the cortex
of the cerebellum and the visual cortex.
The first  symptoms of  methyl  mercury
poisoning in adults are non-specific, e.g.,
paresthesias and malaise. These effects
are believed to have a low frequency
occurrence of  about 5 percent in  the
general population and are seen at blood
levels ranging from 200 to 500 ng Hg/
ml,  which correspond  to  chronic  oral
daily intakes of approximately 3 to 7 fjg
Hg/kg body weight. There is conflicting
evidence  on the degree of reversibility
of these  first  symptoms  of  poisoning.
Dietary intake levels of methyl  mercury
that produce irreversible destruction of
neurons  in the cerebellar and  visual
cortices leading to permanent  signs of
ataxia and constriction of the visual field
are probably twice as high as those levels
causing mild symptoms.

Human Health Risk
Assessment of Mercury in Air

Direct Exposure  Effects
  Because the 1973 standard was based
on  the  direct  exposure  effects from
airborne  mercury, taking  into consider-
ation the contribution of dietary mercury
intake to total  body burden,  a similar
approach was taken here to determine
to what  extent new information  has
changed  the perspective on direct risks
from mercury in air.
1
  A new evaluation confirms that whil
mercury vapor still accounts for the major
fraction of airborne mercury, particulate
forms of mercury do exist in the atmos-
phere. The diet is by far the dominant,
if not the sole source of human exposure
to methyl  mercury  compounds. In addi-
tion, a current evaluation indicates that
the diet is also the dominant source of
compounds  of inorganic  mercury.  In
comparing different routes of exposure,
contribution of airborne  mercury is
between one-tenth  and one-twentieth of
the total  daily amount  of mercury
absorbed into the body.
  An  analysis of  dose-response and
dose-effect  relationships shows that
current levels of mercury in the atmos-
phere, regardless of  chemical  species,
would present a negligible risk of adverse
health effects from direct airborne
exposures. Current atmospheric levels
are believed to be 20 ng Hg/m3 or less.
Effects  of  mercury  vapor on  human
health have not been detected below 1
fjg  Hg/m3,  and serious debilitating
effects have  not  been observed  in
occupational settings  where  workers
have been exposed for months or years
to air concentrations  below 100 /JQ Hg/
m3. Assuming that all the mercury in the
atmosphere  was in the form of methy
mercury compounds,  it would require
atmospheric concentration of 10 A
-------
^Tec
   rcury in fish  high enough to exceed
  ederal guidelines, an  important recent
finding is  that  high  levels  of methyl
mercury are also found in certain lakes
remote from  anthropogenic sources.
Additionally,  the  problem of  long-
distance transport is exacerbated by the
acidification of rainwater in certain areas
of the United States. A statistical corre-
lation was found between the acidity of
lakes  and  elevated levels of methyl
mercury in fish.
  Many  possible mechanisms for this
elevation may exist, but the overall effect
is that methyl mercury levels in fish are,
to some extent,  indirectly affected  by
airborne mercury and  acid deposition.
The questions, therefore, arise as to what
extent changes in current  atmospheric
levels would  affect levels of methyl
mercury in edible freshwater fish, and
what would be the predicted impact on
human health?
  The answer to the first question cannot
be definitively quantified at  this time.
However, if airborne   mercury levels
increased from current levels of 5 to 20
ng Hg/m3 to the 1,000 ng  Hg/m3 level
that  was used in setting emission limits
for source  categories regulated by the
current standard, the  environmental
  insequences  could  be severe. This
  larp rise in atmospheric air concentra-
tions would increase airborne mercury
in fish to unacceptably high levels.
  To express quantitatively the  health
impact this  would have on the United
States would be difficult. Action guide-
lines by the Food and Drug Administra-
tion  and state regulations already exist
that  prohibit fish consumption if methyl
mercury levels  exceed  1 (ig  Hg/g wet
weight. The outcome, therefore, would
probably be the same as  in Scandinavian
countries where elevated mercury con-
centrations in edible tissue  of fish have
resulted in the  banning  of fish  for
consumption from many  freshwater
lakes. Local communities, such as reser-
vations of native American populations
whose main food source is freshwater
fish  may suffer health consequences if
federal  and local regulations  do not
prevent the consumption  of contami-
nated fish in these communities.
  In  conclusion, a more comprehensive
reevaluation of existing information  is
required if  the  potential  for indirect
exposure effects are to be considered.

-------