JUNE 1 3 Q G

Number 10


Criteria and Standards Division
Office of Water Regulations and Standards
United States Environmental Protection Agency

Advisory: Aquatic Life

No aquatic life advisory is given because publication of a
final ambient water quality criterion document for aluminum is
scheduled for September, 1986.

Advisory: Human Health

The advisory concentration for aluminum is ambient water for
the protection of human health is estimated to be 4.07 mg/L, when
exposure is assumed to include consumption of 6.5 g of
contaminated fish and 2L of contaminated water per day, or 12.9
mg/L when exposure is assumed to include consumption of
contaminated fish only. These concentrations do not take into
account relative source contributions from other media. Care
should be taken in the application of this advisory, with
consideration of its derivation, as stated in the attached
support document.

human health effects













Table l. Relative Media Contribution of Aluminum 			2

Table 2. Water Quality Advisory for Aluminum (mg/L) 		6



Advisories have been developed to give the best available
scientific information on the aquatic and human health effects of
chemicals in surface waters. They are issued in cases when
information is needed quickly, but where there is not sufficient
data to calculate national ambient water quality criteria.

An advisory concentration for the protection of human health
can be derived from a number of sources. The office of Drinking
Water Health Effects Advisories; Acceptable Daily Intake (ADI)
values from EPA; Office of Pesticides and Toxic Substances risk
assessment; Carcinogen Assessment Group (CAG) cancer risk
estimates; risk estimates; risk estimates derived from the open
literature; or other sources which will be given in the support
document. The advisory concentrations derived from these sources
will vary in confidence and usefulness, based on the amount and
quality of data used as well as the assumptions behind the
original estimates. The user is advised to read the background
information carefully before using the advisory concentrations to
determine the strengths or deficiencies of the values given in
the advisory.


Human Health Section
Water Quality Advisory (Aluminum)

Identifying the most appropriate set of Advisory numbers (fish
only, water only, fish plus water) requires an appreciation of
several parameters:

(1)	Relative source contribution analysis or all sources of
exposure expressed in percent

(2)	Toxicokinetics or the uptake, distribution, retention,
and metabolism of the pollutant

(3)	Health effects including both non-carcinogenic and car-
cinogenic bioeffects

(4)	Quantification of the toxicological effects
A. Relative Source Contribution

The average and average maximum concentrations of
Aluminum (Al) at raw water are approximately 0.1 mg/L
and 1.94 mg/L (ten EPA regions). The National Academy
of Sciences (NAS) ('82) notes a concentration of 0.074
ug/ml and a value of 0.130 mg/L can be calculated from a
paper by Miller et al. ('84). Surface water appears to
have higher concentrations than groundwater. The Al
concentration can exceed 100 mg/L when the pH is 5 and
mining operations can encourage such conditions. The
maximum ambient concentrations by EPA regions have been
determined by Miller et al. ('84) where the average
maximum was 1.94 mg/L and Region VII had the highest
maximum of 5.35 mg/L. The Al detection limit was 0.014

Using a conservative Bioconcentration Factor (BCF) of 10
and assuming that the average surface water concentra-
tion is 100 ug/L and that an adult consumes 0.0065 kg of
fish/day, an Al intake via fish could be calculated to
be 6.5 ug/day via 0.1 mg/L x 10 BCF x 0.0065 kg/day.

In finished community water in the U.S., the average and
average maximum concentration of Al can be calculated
from Miller et al. ('84) to be 0.069 and 1.155 mg/L,
respectively. The use of Al as coagulant appears to
increase the Al concentration in the finished water.

Since Al is a very abundant element and soil concentra-
tions vary considerably, Al concentrations in foods vary
widely. The average daily intake of Al via food is 3 0
mg/L (La 86, NAS 82 and ICRP 75). The intake via air-
borne Al is approximately 0.10 mg/day (ICRP 75).

A relative media contribution analysis for man via
natural sources of Al (less medicines) indicates that

food is the major source of exposure (Table I). The
intake of Al via food is responsible for some 99% while
the remainder is due to air (0.3%) and water (0.5%).
The contribution via water is primarily due to potable
water and only a small amount is contributed by the
consumption of fish. For the purpose of this Advisory,
water will be assumed to be responsible for 1% of man's
intake of Al.

Table 1. Relative Media Contribution of Aluminum
Water 	Human Intake	


Media	Cone. (mg/L) mg/day	%



30	99.2

0.1	0.3

0.15	0.5

O Fish	0.1	(0.0065)	(0.02)

o Potable 0.07	(0.14)	(0.46)

Total 30.25	100

Compromised individuals using antacid medicine
containing Al, people suffering from kidney disorders
and workers exposed to dust containing Al would be
expected to receive Al doses that may be substantial!v
higher than that specified for the normal ?ndividuliY
nation-wide (NAS 82, EPA 84).	aiviauai

B. Toxicokinetics

intake of Al is largely by ingestion and only minor
amounts are inhaled (icrp 75). People normally take in
some 30 mg/day, however, renal-failure	* £

a?* 5_1n 9 Al/day via aluminum hydroxide, an antacid (NAS
82). Normal gastrointestinal absorption rates

approximatelv threeU5fmic P^ients (8-9%) being
approximately three (3) times greater (EPA 84) Th»

normal plasma level of Al is 5 ug/L (EPA 84) Patient

on antacids can have Al plasma levels significantly

higher (40 mg/L) than the normal. According to EPA

(84), neurological effects do not appear until serum

levels exceed 100-200 ug/L. Aluminum gastrointestinal

tract (GIT) absorption was reported to be enhanced by

parathyroid hormone (PTH) (NAS 82).	y


Aluminum is virtually found in all tissues (EPA 84).
Aluminum distributes to the skeleton, muscle, lungs,
brain etc. Aluminum concentration in the brain and
bone are important to the decisions made in this
Advisory. An A1 balance study by the ICRP shows that
excretion occurs mainly via the bile and feces with
minor amounts via sweat, urine, and hair. While the
biological half-life remains unclear, the human body
burden of A1 is estimated by the NAS (82) to range from
50-150 mg where most soft tissues contain 0.2-0.6 ug/g.

Uremic patients are often encouraged to take aluminum
hydroxide to reduce the hyperphosphatemia (NAS 82, La
86). This is accomplished because of the A1 complexes
with phosphate in the gastrointestinal tract, thereby
reducing GIT phosphate absorption. Unfortunately the
resolution of one problem (uremia) .causes a greater
availability of A1 which may cause other potential
health problems like hypercalciuria and osteromalacia.

In 1984 EPA assumed that Aluminum in doses near 100
mg/day was more or less homeostatically controlled (EPA
84). However, some 4% of one's intake is retained and
the* body burden obviously increases with age.

The bioaccumulation of A1 in fish remains unclear (EPA
86). However, using a paper by Buergel and Wrenn (84),
a BCF of ten (10) may be appropriate and will be used in
this Advisory.

C. Health Effects

Aluminum has been associated with encephalopathies known
as Alzheimer's disease and as dialysis dementia (NAS
82). Alzheimer's disease is a neurologic disorder which
occurs in a substantial number of humans after age 40
years. It is slowly progressive and may be influenced
by the ever-increasing A1 body burden. Symptoms include
behavioral change, memory disturbances, spacial
disorientation, agnosia, dysphasia and seizures (NAS
82). Aluminum has been associated with neurofibrillar
tangles in the brain. Plaques and neurofibrillar
degeneration obviously inhibit and prevent proper
conduction of nerve impulses.

Dialysis dementia or dialysis encephalopathy is a rather
rapidly deteriorative neurological disorder which may
cause death within a year (NAS 82). This relentless
progressive neurological disease has often been found in
chronic dialysis patients.

Dialysis dementia is characterized by the onset of
altered behavior, speech disturbances, dyspraxia, tre-
mors, convulsions, personality changes and psychoses
(NAS 82). An epidemiological study in Chicago in the


1970's demonstrated that Al in drinking water greatly-
impacted 20 patients who had been maintained on long-
term dialysis (Du 78). When Al levels were 300-400
mg/L, dementia occurred, and when the Al concentrations
were <14 ug/L to 150 ug/L, dementia was absent. Both Al
concentration and the natural progression of the disease
itself, with or without Al, may be responsible for the
sudden increase in this neurological disorder. From a
conservative viewpoint, Al should be assumed, at this
time, to be an associated etiological factor in this
serious neurological disorder.

A review of Al concentration in dialysis-type water
suggests that 14 ug/L is that critical concentration
with regard to uptake (Hodge 81 and Parkinson et al)
suggested that when dialysis water exceeds 50 ug/L, the
incidence of encephalopathy exceeds 5%.

Aluminum may also be responsible for causing osteomala-
cia (NAS 82). Aluminum may compete with calcium within
the bone, leaving the bone soft and susceptible to
fracturing. Excessive use of aluminum hydroxide could
cause a phosphate absorption deficit which may interfere
with bone mineralization and may lead to direct deposi-
tion of Al in the skeleton (NAS 82). Claims have been
made that Al blocks calcium uptake resulting in hyper-
calcemia (NAS 82).

Aluminum is used to prevent hyperphosphatemia in uremic
patients. Al is used to control phosphorus absorption
(EPA 84). However, reduced phosphorus metabolism
results in significant decreases in serum ATP level.
More ADP and AMP and less ATP could impact the
functioning of other metabolic cycles.

Aluminum does cause pulmonary fibrosis but has been
primarily associated with inhalation exposures via occu-
pational workplace (NAS 82). Pulmonary fibrosis
probably is not caused by Al ingestion.

Aluminum appears not to be mutagenic nor carcinogenic
(EPA 84). One study has reported lymphoma leukemia in
female mice but it would be premature to assume carcino-
genicity in humans. There also has been a claim for
interference with animal reproduction, but again it
would be premature to assume any association with Al at
this time (EPA 84).

D. Quantification of Toxicological Effects

Aluminum is not an essential element (La 86). The ideal
suggested daily intake then obviously would be none.
However, since Al is a very abundant element and people
are exposed to some 30 mg/day, it--is important to
identify that level of exposure where toxicity may

actually start occurring since it simply is not possible
to exclude it. Therefore, a practical ADI must be
determined. Since many people experience neurological
and bone disorders that have been associated with Al, it
rs entirely possible that the ADI is substantially below
our natural level of exposure. Should this be the case,
our only option would be to limit our exposure to the
extent feasible.

Two sensitive indicators of Al toxicity seem to be a
reduction in serum ATP levels or increased serum Al
levels (EPA 84). One may question how important it is
to have reduced energy units or increased Al levels in
the blood. The answers to those questions are uncertain
but they may either serve as threshold markers or
evidence that the speculated association of Al with
Alzheimer's disease, dialysis dementia and osteomalacia
are real.

Utilizing the studies by Ondreick et al. (66) and by
Krasovskii et al. (79) as reviewed by EPA (84), it
seems apparent that 6 mg/kg oral exposure in animals is
a NOAEL since higher acute doses (17,27 + 50 mg/kg)
caused a decrease in ATP. Using the Krasovskii data, an
ADI of 4.2 mg/day can be calculated:

_(6 mg/kg/day) (70 kg) - 4.2 mg/day
100 S.F.

Note: A safety factor (S.F.) of 100 was used because it
was an animal experiment where the NOAEL was identified.

Utilizing an Al balance study (20 days) by Greger and
Baier (83) as described by EPA (84), a NOAEL of 125 mg
Al/day or 1.79 mg/kg/day would seem appropriate since
the gastrointestinal absorption of phosphorus appeared
to begin to be noticeably inhibited and serum levels
were 25 times the norm. The 1.79 mg/kg apparently would
be expected to cause an Al serum level of 125 ug/L or 7 5
ug/L below the 200 ug Al/1 threshold specified within
EPA ('84) Al criteria document. (Note that the normal
serum level is 5 ug/L.) Using the Greger and Baier
study then, an ADI of 12.6 mg/day could be calculated:

1.79 mg/kg/day) (70 kg) -1.3 mg/day
100 S.F.

Note: A safety factor (S.F.) of 100 was used to reflect
uncertainty associated with a short-term, human study
where the NOAEL could be identified.

To avoid having reference doses being dictated by a
single experiment and experimented design, it is often
appropriate to appreciate the dose/response results of


more than one experiment. Therefore, averaging the two
possible values suggests an ADI or RFD of 2.8 mg/day.

Without consideration of other sources of exposure it
is possible to calculate:	'

a.	Lifetime acceptable level in drinking water via ADI/2

b.	Lifetime acceptable level in ambient water if only fish
were consumed but the water was not consumed via ADI/RF

c.	Lifetime acceptable level in which the water was con-
sumed and the fish living within the water were consumed
at the rate of 0.0065 kg/day via ADI/2 plus 10 (0.0065).

The three values for drinking water only, fish only
water plus fish are 1.4 mg/L, 43 mg/L and 1.4 mg/L, '

o Drinking water only:

ADI = 2.8 mg/day = l.4mg/L
2	2 1/day

o Fish only:

ADI	2.8 mg/day	= 43 mg/L

RF	(10 BCF)(0.0065 kg/day)

o Ambient water or fish plus drinking water:

ADI 	2.8 mg/d	=1.4 mg/L

2+RF	2 1/d + (10 BCF)(0.0065 kg/d)

Since water is only 1% of man's A1 intake, all initial
Advisory levels should be modified by 100. However,
since an ADI already has a 100-fold safety factor and
there is a practical limit to which water alone can
control the total Al intake, only an additional factor
of 10 will be utilized to convert initial values to
final water quality advisory levels (Table 2).

Table 2. Water Quality Advisory For Aluminum (mg/L)

Water Use	Relative Source Contribution



Fish plus potable




Fish only



Potable only





Bu 84. Buergel & Wrenn. Health Effects. PB#84-178-44l.
tf.S. EPA.

Du 78. Dunea, Mahurkar,	Mamdani Smith. Role of Aluminum in

Dialysis Dementia.	Ann. Intern. Med. 88:502-504.

/EPA 84. U.S. EPA. Drinking Water Criteria Document for
J	Aluminum. ORD ECAO-Cincinnati, Ohio.

EPA 86. U.S. EPA. Ambient Aquatic Life Water Quality
Criteria for Aluminum. ORD-Duluth, Minn.

Gr 83. Greger and Baier. Excretion and retention of low or
moderate levels of aluminum by human subjects. Fd.

Chem. Toxicol. 21:473-477.

Hodge, Day, OHAIA Research, Ackrill Ralston. Critical

concentration of aluminum in water use for dialysis.
Lancet, pp. 802-803.

ICRP 75. Int. Com. Rad. Prot. Reference Man. (#23), pp.


KR 79. Krasovskii et al. Experimental study of the

biological effects of lead and aluminum following oral
administration. Environ. Hlth. Perspect. 30:47-51.

La 86. Lappenbusch. Contaminated Drinking Water and Your

Health. Lappenbusch Environmental Health, 6480 Overlook
Drive, Alexandria, Virginia 22312, pp. 192.

Miller 84. Miller et al. The Occurrence of Aluminum in
Drinking Water. J. AWWA 5 76(1):84-91.

NAS 82. National Acad. Sci. (SDWC). Drinking Water and
Health, Volume 4, pp. 299.

On 66. Ondreick et al. Chronic toxicity of aluminum in rats
and mice and its effects on phosphorus metabolism. Br.
J. Ind. Med. 23:305-312.

Parkinson, Ward and Kerr. Dialysis encephalopathy bone

disease and anemia; The Aluminum Toxication Syndrome
During Regular Hemodyalysis. Clinical Path., Vol. 34,
pg. 1285-1294.