United States
         Environmental Protection
Drinking Water Advisory:
Consumer Acceptability Advice
and Health Effects Analysis
on Sulfate

Drinking Water Advisory: Consumer Acceptability Advice and
               Health Effects Analysis on Sulfate

                  EXTERNAL REVIEW DRAFT
                   Contract Number: 68-C-01-002
                  Work Assignment Number: B-02
                          Prepared for:

                U.S. Environmental Protection Agency
                         Office of Water
                Health and Ecological Criteria Division
                      Washington, DC 20460
                           Prepared by:

                     Sciences International, Inc.
                   1800 Diagonal Road, Suite 500
                    Alexandria, VA 22314-2808
                        EPA 822-R-02-033
                            April 2002
                        Printed on Recycled Paper






      2.1    Water	4
      2.2    Soil	6
      2.3    Air	6
      2.4    Food	7
      2.5    Summary	7


      4.1    Absorption	9
      4.2    Distribution	9
      4.3    Metabolism	11
      4.4    Excretion	...:	11

      5.1    Human	13
            5.1.1  Short-Term Exposure Studies	13
            5.1.2  Long-Term Exposure Studies	15
      5.2    Animal	16
            5.2.1  Short-Term Exposure Studies	16
            5.2.2 Long-Term Exposure Studies	17
            5.2.3 Reproductive and Developmental Studies	18
            5.2.4 Cancer Studies 	18


      7.1    Hazard Characterization 	20
      7.2    Characterization of Organoleptic Effects 	21
      7.3    Dose-Response Characterization	21

8.0    REFERENCES 	23
                        External Review Draft — Sulfate — April 2002                       111


CDC        Centers for Disease Control and Prevention
CSF        cerebrospinal fluid
g           gram
kg          kilogram
L           liter
m3          cubic meters
mg         milligram
min         minute
mM        millimolar
mmol       millimole
NTP        National Toxicology Program
OST        OfiBce of Science and Technology
OW        Office of Water
PAPS       3 '-phosphoadenosine-5 '-phosphosulfate
ppm        parts per million
PWS        public water system
RfD        Reference Dose
SDWA      Safe Drinking Water Act
SDWIS/FED Safe Drinking Water Information System/Federal
SO42        sulfate
SMCL      secondary maximum contaminant level
UCM       unregulated contaminant monitoring
\ig         microgram
umol       micromole
                          External Review Draft — Sulfate—April 2002


The Drinking Water Advisory Program, sponsored by the Health and Ecological Criteria
Division of the Office of Science and Technology (OST), Office of Water (OW), provides
information on the health and organoleptic (taste, odor, etc.) effects of contaminants in drinking
water.  The Drinking Water Advisory documents are a component of the OW Health Advisory
Program. Drinking Water Advisories differ from Health Advisories because of their focus on
aesthetic properties (e.g., taste, odor, color) of drinking water. A Drinking Water Advisory is
prepared when contaminants cause adverse taste and odor influences at concentrations lower
than those for adverse health effects.

A Drinking Water Advisory is not an enforceable standard for action.  It describes nonregulatory
concentrations of the contaminant in water that are expected to be without adverse effects on
both health and aesthetics. Both Health Advisories and Drinking Water Advisories serve as
technical guidance to assist Federal, State, and local officials responsible for protecting public
health when emergency spills or contamination situations occur. They are not to be construed as
legally enforceable Federal standards. They are subject to change as new information becomes
available.  This draft supersedes any previous draft advisories for this  chemical.

The Advisory discusses the limitations of the current database for estimating a risk level for
sulfate in drinking water and characterizes the hazards associated with exposure. The Drinking
Water Health Advisory value was* developed by a panel of experts through a workshop held on
September 28,1998, and sponsored by the Centers for Disease Control (CDC) and the United
States Environmental Protection Agency (U.S. EPA). The experts who participated in the
workshop were:

    Charles Abemathy, Ph.D. - U.S. EPA
    David Cole, M.D., Ph.D. - University of Toronto
    Marie Cassidy, Ph.D. - George Washington University
    Marilyn Morris, Ph.D. - State University of New York: Buffalo
    Guillermo Gomez, Ph.D. - North Carolina State University
    Lorraine Backer, Ph.D. - National Center for Environmental Health at CDC

A workshop report was prepared that summarized the data considered at the workshop and its
findings. This report was published by EPA as Document Number 815-R-99-002 in January
1999 (EPA, 1999b). The workshop report was externally peer reviewed by the following
scientists (EPA, 1999c):

    Laurence L. Brunton, Ph.D. - University of California San Diego
    Paul E. Brubaker, Ph.D. - Brubaker Associates, New Jersey.
    Michael L. Dourson, Ph.D. -  Toxicology Excellence for Risk Assessment, Ohio
                           External Review Draft — Sulfate — April 2002

 This page intentionally left blank.
External Review Draft — Sulfate—April 2002                              VI

                               EXECUTIVE SUMMARY

The EPA Office of Water is issuing this advisory to provide guidance to communities that may
be exposed to drinking water contaminated with high sulfate concentrations. The advisory
provides an analysis of the current health hazard information and an evaluation of available data
on the organoleptic (i.e., taste and odor) problems associated with sulfate-contaminated water,
because organoleptic problems will affect consumer acceptance of water resources.  This
advisory does not recommend a Reference Dose (RfD) because of limitations of available data
for assessing risks. However, the advisory does provide guidance on the concentrations at which
health and organoleptic problems would likely occur. This Drinking Water Advisory does not
mandate a standard for action; rather it provides practical guidelines for addressing sulfate
contamination problems and supersedes previous draft advisories for sulfate

Conclusion and Recommendation

In order to enhance consumer acceptance of water resources, this advisory recommends reducing
sulfate concentrations to or below 250 mg/L, the EPA's Secondary Maximum Contaminant
Level (SMCL) for sulfate. The SMCL is based on taste considerations. It is not a federally
enforceable regulation, but is intended as a guideline for States. States may establish higher or
lower levels depending on the local conditions, such as unavailability of alternate source waters
or other compelling factors, provided that public health and welfare are not adversely affected.

A health-based advisory for acute effects (absence of laxative effects) of 500 mg of sulfate/L is
recommended. This value depends on the absence of other osmotically active materials in
drinking water, which could lower the sulfate level associated with a laxative effect. Where the
water contains high concentrations of total dissolved solids and/or other osmotically active ions,
laxative-like effects may occur if the water is mixed with concentrated infant formula or a
powdered nutritional supplement. In such situations, an alternate low-mineral-content water
source is advised. Infants are more susceptible than adults to diarrhea! water loss because of
differences hi gastrointestinal structure and function.

The soft stool or diarrhea that results from sulfate is an osmotic diarrhea; that is, it happens when
the osmolality (number of dissolved particles) in the intestinal contents exceeds that of the body
fluids.  When this occurs, water is drawn from the body fluids into the intestines, increasing the
moisture content and volume of the fecal matter. Whether or not diarrhea or soft stools occur
depends on the amount of sulfate and other osmotically active materials that are present in the
intestines; these materials include magnesium, sodium, and some sugars. An osmotic-induced
diarrhea ceases once the osmotically active gastrointestinal contents are excreted. In the case of
sulfate, adults appear to adapt within 1 or 2 weeks and are no longer affected by the sulfate in
their drinking water supply. Infants, however, may be more sensitive.

Sulfate in the Environment

Sulfates are naturally occurring substances that are found in minerals, soil, and rocks.  They are
present in ambient air, groundwater, plants, and food.  The principal commercial use of sulfate is
in the chemical industry. Sulfates are discharged into water in industrial wastes and through
atmospheric deposition.  Sulfate concentration in seawater is about 2,700 milligrams per liter

                            External Review Draft — Sulfate—April 2002                            1

 (mg/L).  It ranges from 3 to 30 mg/L in most freshwater supplies, although much higher
 concentrations (* 1000 mg/L) are found in some geographic locations. In the United States, the
 median concentration for a 20-State cross-section was 24 mg/L; the 99th percentile value was 560
 mg/L. In general, food is the principal source of exposure.  However, in areas with high sulfate
 concentrations, exposure from water can exceed that from food.

 Studies of Sulfate Effects           •

 Long-term and short-term exposure studies to determine a hazard assessment for sulfate are
 currently available in humans and animals. The findings from cancer, noncancer, and taste and
 odor studies are discussed below.

 Cancer Studies. There has been no traditional NTP oral cancer bioassay for inorganic sulfate as
 the ion of interest.  In an 8-month preliminary study, no tumors were observed in Wistar rats
 after intramuscular injection of 0.7 mg sodium sulfate every other day for 4 weeks. However, in
 this study the sodium sulfate treatment was used as a control against which to evaluate the
 effects of nickel sulfate and nickel hydroxide.  Accordingly, the present database is of limited
 value for evaluating the tumorigenicity of sulfate. After reviewing toxicity data on sulfates food
 additives, the Select Committee of the Life Sciences Research Office concluded that there was
 no evidence that sulfuric acid or ammonium, calcium, potassium, and sodium sulfates present a
 hazard to the public health when they are used at levels that are current or that might reasonably
 be expected in the future.

 Noncancer Studies.  The collective evaluation of the noncancer data in humans and animals
 suggests that acute exposures to sulfate exert a laxative effect (loose stool) and sometimes
 diarrhea (unusually frequent or unusually liquid bowel movements) following acute exposures to
 high concentrations.  However, these effects are not observed for longer term exposures. This
 may be because of acclimation to sulfate over time.

 No adverse developmental effects were observed following the administration of 2,800
 mg/kg/day of sulfate to pregnant ICR/SIM mice on gestation days 8 to 12. No reproductive
 effects were observed following the ingestion of drinking water containing up to 5,000 mg/L  of
 sulfates by ICR/SIM mice and 3,298 mg/L of sulfates by Hampshire x Yorkshire x Duroc pigs .
 On the basis of these studies, sulfate does not appear to be a reproductive or a developmental

 Studies on Taste and Odor. Few studies are available that report on the organoleptic properties
 (i.e., taste and  odor) of sulfate. None of the studies reported an odor threshold; therefore, all of
 the reported values are based on taste thresholds. It is not possible to precisely identify a
 specific taste threshold for sulfates in drinking water because the taste threshold concentration
 varies among individuals. In addition, the associated cations, different water matrices, and
 temperatures also influence taste. On the basis of the available data, no significant taste effects
have been found to occur at sulfate concentrations of about 200-300 mg/L.
                            External Review Draft—Sulfate — April 2002

Characterization Summary

The data from short-term studies suggest that a mild laxative response can occur at sulfate
concentrations greater than 500 mg/L, especially if there are other osmotically active substances
present in the water.  In the absence of other osmotically active materials, the laxative effects are
unlikely to be observed at concentrations up to about 1,000 mg/L sulfate. These effects are
exhibited as an increase in stool volume, moisture, and/or increased intestinal transit time rather
than frank diarrhea.

Where drinking water contains high levels of sulfate or total dissolved solids, it should not be
used in the preparation of powdered infant formula or nutritional supplements.  An alternate low-
mineral water source should be used.  Because laxative effects have not been observed with
long-term exposures to sulfate-containing water, the data suggest that acclimatization occurs as
exposures continue.

The available database does not permit EPA to construct a quantitative dose-response
assessment for the laxative effects of sulfate. The current SMCL of 250 mg/L should protect
almost all consumers from the esthetic effects of sulfate, and the health-based advisory value of
500 mg/L will protect against sulfate's laxative effects in the absence of high concentrations of
other osmotically active chemicals in the water.
                            External Review Draft —Sulfate —AprU 2002


This advisory provides information to Tribes, States, local drinking water facilities, and public
health personnel on the health and taste effects resulting from sulfate contamination of potable
water.  There are limited scientific data on the health effects of sulfate, but adequate data are
available on the range of exposure concentrations that may pose a concern.

Sulfates occur naturally and are abundant in the environment, generally originating from mineral
deposits, soil, and rocks, or the combustion of sulfur-containing fuels. Sulfate, a soluble,
divalent anion (SO42~), results from the oxidation of elemental sulfur, sulfide minerals, or organic
sulfur (Alley 1993, Field 1972, Wetzel 1983).  The anion is often associated with alkali, alkaline
earth, or transition metals through ionic bonds (Field 1972).

Sulfates are used in mining, pulping, metal and plating industries, water and sewage treatment,
and leather processing, and in the manufacture of numerous chemicals, dyes, glass, paper, soaps,
textiles, fungicides, insecticides, astringents, and emetics (Greenwood et al. 1984). Various
sulfate salts are used in foods (FDA 1999), and ammonium sulfate is used in the fertilizer

Sulfur is the 14* most abundant element in the earth's crust, and the 8* or 9th most abundant in
sediments (Kaplan 1972). It is constantly transferred among compartments by the sulfur cycle
and is ubiquitous in the environment. Anthropogenic sulfur emissions have a significant impact
on the sulfur cycle, with at least 80% of global sulfur dioxide (SO2) emissions and more than
45% of river-borne sulfates traceable to human activity (Moore 1991).

2.1  Water

Sulfate is found almost universally in natural waters at concentrations ranging from a few tenths
to several thousand milligrams/liter (mg/L).  The highest concentrations are usually found in
groundwater and are considered to be a mixture of sulfates from both atmospheric and biogenic
sources. Approximately 30% of sulfate in groundwater may be of atmospheric origin, and the
remaining 70% is of biogenic origin. Sulfates are also discharged into surface water through
industrial wastes and atmospheric deposition of sulfur dioxide.

The sulfate concentration in seawater is about 2,700 mg/L (Hitchcock 1975) and ranges from 3
to 30 mg/L in freshwater lakes (Katz 1977).  Sulfate content in drinking water ranges from 0 to
1,000 mg/L in the United States (Trembaczowski 1991). In a survey of rivers in western
Canada, sulfate concentrations ranged from 1 to 3,040 mg/L, with concentrations generally in
the range of 1 to 580 mg/L (Environment Canada 1984).

Sulfate has been monitored under the Safe Drinking Water Act (SDWA) Unregulated
Contaminant Monitoring (UCM) program since 1993 (57 FR 31776). Monitoring ceased for
small public water systems (PWSs) under a direct final rule published January 8,1999 (64 FR
1494), and ended for large PWSs with promulgation of the new Unregulated Contaminant

                           External Review Draft — Sulfate — AprU2002                           4

Monitoring Regulation (UCMR) issued September 17,1999 (64 FR 50556) and effective
January 1,2001.

The Safe Drinking Water Information System (SDWIS/FED) is a database of analytical data on
the concentrations of contaminants in drinking water. Sulfate levels reported in the SDWIS/FED
database were analyzed from a 20-State cross-section (U.S. EPA 2001). The median
concentration of all PWS samples was 24 mg/L and the 99th percentile concentration of all PWS
samples was 560 mg/L. Minimum reporting levels varied from system to system and State to
State. The 99th percentile concentration is a summary statistic to indicate the upper bound of
occurrence values because maximum values can be extreme values (outliers) that sometimes
result from sampling or reporting error.

Additional sulfate occurrence data submitted for EPA's Chemical Monitoring Reform (CMR)
evaluation by the States of Alabama, California, Illinois, Montana, New Jersey, and Oregon
augment the SDWIS/FED 20-State cross-section analysis (U.S. EPA 2001). Five of these CMR
States are  not represented in the cross-section. Data from the CMR States show that
concentrations are generally similar to those found in the 20-State cross-section (Table 1). Even
in states such as Montana, where the 99th percentile concentration is substantially greater than
that of the 20-State cross-section, the median concentration is still quite similar to the median of
the cross-section States (U.S. EPA 2001).

                  Table  1.  Median and 99th percentile concentrations
                              for sulfate in CMR States
New Jersey
8.1 mg/L
33 mg/L
60 mg/L
22 mg/L
15.9 mg/L
5.1 mg/L
99th percentile
72 mg/L
523 mg/L
760 mg/L
1,200 mg/L
260 mg/L
79 mg/L
Although 88% of the 16,495 systems included in the 20-State cross-section reported sulfate at
concentrations greater than the minimum reporting level, only about 5% exceeded the 250 mg/L
SMCL. Within the twenty individual States, the number of systems that exceeded the SMCL
ranged from 0 to approximately 11%. A greater percentage of surface water systems exceeded
the 250 mg/L threshold; however, groundwater systems do generally show the highest sulfate
concentrations (U.S. EPA 2001).
                           External Review Draft—Sulfate—April 2002

A survey of more than 900 community water supplies found that 25 (-3%) had sulfate
concentrations above 250 mg/L (McCabe et al. 1970).  Another survey of approximately 650
rural water systems reported sulfate present in 271 of 495 groundwater supplies (55%), with a
mean sulfate concentration of 98 mg/L (a range of 10 to 1,000 mg/L). Sulfate was found in 101
of 154 surface water supplies at a mean concentration of 53 mg/L for those systems that detected
sulfate (a range of 15 to 321 mg/L) (U.S. EPA 1994).

Several surveys have been conducted in Canada on the occurrence of sulfates in drinking water.
A study of 17 drinking water supplies in Ontario from 1985 through 1986 found mean sulfate
concentrations of 22.5 and 12.5 mg/L in treated and untreated water, respectively (Ontario
Ministry of the Environment 1987). A study of 78 municipal drinking water supplies in Nova
Scotia between 1987 and 1988 found mean sulfate concentrations of 14.2 mg/L in treated water
(Nova Scotia Department of Public Health 1988). Sulfate concentrations were significantly
higher in Saskatchewan, with median concentrations of 368 and 97 mg/L (range of 3 to 2,170
mg/L) reported for treated ground and surface waters, respectively (Saskatchewan Environment
and Public Safety 1989).  Based on the mean sulfate concentration measured in Ontario (22.5
mg/L) and an average daily water consumption of 2 L per day for an adult, the average daily
intake from this source would be 45 mg. However, in areas with high sulfate levels in drinking
water, such as Saskatchewan, daily sulfate intake could be more than 4,000 mg (WHO 1996).

2.2  Soil

 Sulfate can be formed from the oxidation of elemental sulfur, sulfide minerals, or organic sulfur
 (Alley 1993, Field 1972, Wetzel 1983). It is one of the predominant anions in soil but is not
 highly mobile.  Sulfate anion is often associated through ionic bonds with alkali, alkaline earth,
 or transition metals (Field 1972). Sulfur can be retained in soil through biochemical processes,
 such as incorporation into the soil organic pool as sulfate esters of humic material or other
 complex organic molecules. Sulfur can also be retained by adsorption onto soil particles, such as
 hydrous iron and aluminum sesquioxides. The average sulfur (total) concentration in soils in the
 United States is 1,600 parts per million (ppm or mg/kg; a range of <800 to 48,000 mg/kg)
 (Shacklette and Boemgen 1984).  The determinant in soil for the adsorption of sulfate is the
 content  of hydrous sesquioxides and organic matter.  A strong correlation was found between pH
 and the  ability to remove sulfates from soil solutions; therefore, factors affecting soil acidity
 would also affect sulfate retention (Patil et al. 1989).

 2.3  Air

 The main sources of atmospheric sulfate are sulfur oxides, which are primarily emitted to the
  atmosphere from sulfur-containing fuel combustion.  Total global sulfur dioxide production
  continually increased from 1930 to 1980 (the years when data are available; Moore 1991).
  Global production of sulfur dioxide increased from 49 * 10* metric tons per year to 126 x 10*
  metric tons per year between 1930 and 1980.

  Sulfur dioxide (SO2) emissions have become a major concern for industrialized nations.  Sulfur
  dioxide interacts with atmospheric water and oxygen to produce sulfuric acid (H2SO4), causing
  acid rain (Moore 1991, Wetzel 1983).  In addition, SO2 is converted to  sulfate in the atmosphere
                              External Review Draft—Sulfate—April 2002

 and deposited on soils.  This can lead to the acidification of soil solutions and elevate sulfate
 concentrations in terrestrial waters (Drever 1988).

 Limited information is available on the concentration of sulfates in ambient air. In a study
 consisting of 23,000 samples from 405 sites in 49 States, sulfate concentration was estimated to
 range from 0.5 to 228.4 ng/m3.  The median exposure concentrations (0.7 to 19.5 pg/m3) were
 considered to be more representative of exposures than were the mean values.  Using the upper
 median exposure concentration (19.5 ng/m3) and assuming an inhalation rate of 20 m3/day for a
 70-kg adult, the daily dose due to sulfate in ambient air would be 6 ug/kg/day (Abemathy et al.

 2.4  Food

 In foods, sulfate is present as the salts of sodium, calcium, iron, magnesium, manganese, zinc,
 copper, ammonium, and potassium (FDA 1999). Sulfate salts are used in the food industry in a
 wide variety of products, such as dietary supplements, breads, preserved fruits and vegetables,
 gelatins, and puddings.  The average daily intake of sulfate in food in the United States has been
 estimated to be 453 mg, based on data  on food consumption and the reported usage of sulfates as
 additives (FASEB 1975).  Many sulfate compounds in food are "Generally Regarded as Safe"
 (GRAS) by the U.S. Food and Drug  Administration (FDA 1999).

 2.5  Summary

 Average daily intake of sulfate from drinking water, air, and food is approximately 500 mg, with
 food being the major source.  However, in areas with high sulfate concentrations in the drinking
 water supplies, drinking water may constitute the principal intake source (WHO 1996).


 Sulfate (SO.2') is a soluble, divalent anion; common salts include sodium, potassium,
 magnesium, calcium, and barium sulfate. The majority of sulfate salts are soluble in water, the
 exceptions being the sulfates of lead, barium, and strontium. The chemical and physical
properties for the common sulfate salts are detailed in Table 2 (NIOSH 1981, Budavan 1996,
 Scofield and Hsieh 1983).
                           External Review Draft - Sulfate—April 2002

            Table 2. Chemical and physical properties of snlfates
Chemical Name
Chemical formula
Molecular weight
Physical state
Boiling point PC)
Melting point (°C)
Density (g/raL)
Vapor pressure (mmHg)
Specific gravity
Water solubility
(e/100 mL)
Taste threshold In water
Odor threshold (air)
Sulfuric acid
Oil of vitriol; Hydrogen
Colorless to dark
brown, oily, odorless
< 0.001
1.84 .
Sodium sulfate
(decahydrate is
Glauber's salt)
White powder or
bipyramidal crystals
Potassium sulfate
Tartarus vitriolatus;
Arcanum duplicatum
Colorless or white
rhombic or hexagonal
Magnesium sulfate
Bitter salts;
heptahydrate; epsom
Colorless rhombic
bitter crystal
400 - 600
Calcium sulfate
Anhydrous gypsum;
Anhydrous calcium
Orthorhomic crystal
of variable color
250 - 900
Barium sulfate
Barite; Blanc fixe;
Fine, heavy, odorless
powder or
polymorphous crystals
4.2 - 4.5
4.2 - 4.5
Practically insoluble,

            NA - Not available.


4.1  Absorption

Estimates of sulfate absorption are derived indirectly from reported excretion data. Absorption
of sulfate from the intestine depends upon the amount of sulfate ingested.  For example, Bauer
(1976) reported that when radiotracer doses of 60-80 |aCi35S-sodium sulfate were administered
orally to humans (eight adults), 80% or more of the radioactivity was recovered in the urine at 24
hours, suggesting that at least 80% of sulfate must have been absorbed.  However, the high doses
of sulfate that induced catharsis exceeded intestinal absorption capacity, and thus were excreted
in the feces.

The type of cation associated with sulfate may also influence absorption. Morris and Levy
(1983b) reported 30.2 ± 17.2% of the dose was excreted in 24-hour urine after oral
administration of magnesium sulfate (5.4 g sulfate) to seven healthy individuals, as compared to
43.5 ± 12.0% excreted after oral administration of sodium sulfate (5.4 g sulfate) in five healthy
men (Cocchetto and Levy 1981). This finding suggests that magnesium sulfate is absorbed to a
lesser extent than sodium sulfate. Morris and Levy (1983b) indicated that the comparison of
these two salts is limited in that data were obtained in separate experiments from a small number
of subjects rather than from parallel experiments with the same pool of subjects.

Florin et al. (1991) performed sulfate balance studies in normal subjects and subjects with
ileostomies (subjects with ileum removed surgically).  The study authors wanted to determine
how much sulfate reached the colon and the extent to which diet and endogenous sources
contributed to colonic sulfate concentration. All subjects were fed diets containing between 1.6
and 16.6 mmol SO4/day (0.15-1.6 g/day). Sulfate was measured in the diet, urine, and feces to
determine sulfate balance (using anion exchange chromatography). There was a net absorption
of sulfate with a plateau at 0.48 g/day in the ileostomy patients. Sulfate absorption in normal
subjects did not plateau even at the highest dietary concentration examined. The dietary
contribution to the colonic sulfate pool was determined to be up to 0.86 g/day because linearity
was observed between diet and upper gastrointestinal loss for intakes greater than 0.67 g/day.
The study authors concluded that diet and intestinal absorption were the principal factors
affecting the amount of sulfate reaching the  colon. This study also suggests that the upper
digestive tract is primarily responsible for sulfate absorption.

4.2  Distribution

Inorganic sulfate is freely distributed in blood and does not accumulate in tissues. Most sulfate
found in human tissues is biosynthetically incorporated into macromolecules and is organic. The
normal serum level of sulfate found in humans (0.3 mmol/L or 29 mg/L) is lower than that in
rodents (1  mmol/L or 96 mg/L) or in other animal species (Krijgsheld et al. 1980, Cole and
Scriver 1980).

A circadian variation of serum inorganic sulfate levels has been demonstrated in humans.
Hoffman et al.  (1990) fed seven male volunteers an identical diet, including fluids, a parameter
that is not  typically included in dietary studies. Blood samples were collected in a total of 10
intervals over a 24-hour period. Average serum inorganic sulfate levels were lowest in the

                             External Review Draft — Sulfate — April 2002                           9

 morning (302 pmol/L or 29 mg/L) and highest in the early evening (408 pmol/L or 39 mg/L).
 This difference is statistically significant (p < 0.005). The average 24-hour level was 360
 umol/L, or 35 mg/L. There was considerable variability among subjects. The authors speculated
 that a portion of the variation in serum sulfate could be due to variation in the dietary sulfate

 It is reported that dietary protein could have major influence on the serum sulfate levels. For
 example, Cole et al. (1991) studied 12 fasting subjects who were randomly fed an isocaloric
 meal containing either high or low protein content and monitored their serum sulfate levels up to
 3.5 hours after feeding.  Serum sulfate levels increased from a baseline value of 276 to 314
 (xmol/L (27 to 30 mg/L) at 2.5 hours after the low-protein meal and returned to the baseline by
 3.5 hours.  Serum sulfate levels increased from baseline values of 253 to 382 ^imol/L (24 to 37
 mg/L) at 3 hours and remained significantly elevated at 3.5 hours after the high-protein meal.
 The increase in inorganic sulfate was attributed to the oxidation of sulfur containing amino acids.
 The reason for the differences in the time to reach peak sulfate levels between the low- and high-
 protein groups is not clear. The inorganic sulfate excreted in the urine was not measured in this

 Ingesting drinking water containing high sulfate concentrations has only slight effects on serum
 sulfate levels. Hindmarsh et al. (1991) compared serum levels of inorganic sulfate in 14 healthy
 volunteers at 2 locations (Saskatoon: 8 men; Rosetown: 4 men and 2 women) using municipal
 drinking water with varying sulfate concentrations. In Saskatoon the sulfate concentration was
 77 ppm (mg/L) and in Rosetown the sulfate concentration was 1,157 ppm (mg/L). The cations
 associated with the sulfate were not reported.  Baseline serum levels of inorganic sulfate were
 0.35 ± 0.06 mmol/L (34 mg/L) in the Saskatoon subjects and 0.50 ± 0.11  mmol/L (48 mg/L) in
 the Rosetown subjects. These findings suggest that a 15-fold increase in sulfate concentration in
 drinking water will result in only a 1.4-fold increase in  serum sulfate.  The reasons for observing
 only a slight increase in serum sulfate in Rosetown subjects could be homoeostatic control
 mechanisms and dietary differences.

 Serum inorganic sulfate concentrations are reported to be higher in infants and young children
 than in adults. For example,  Cole  and Scriver (1980) compared serum inorganic sulfate
 concentrations in subjects under 3  years old (n = 46), children between 3 and 4 years old
 (n = 27), adolescents in a hospitalized population (10 to 18 years old) (n = 12), and healthy
 adults (n =10). On the first day of life, mean sulfate concentrations were 0.47 mmol/L (45
 mg/L)  (95% confidence limits of 0.29 to 0.95 mmol/L), and by 3 years of age, concentrations
 dropped to 0.33 mmol/L or 32 mg/L, (95% confidence limits of 0.22 to 0.67 mmol/L), which
 was comparable to the adult levels 0.33 mmol/L or 32 mg/L (95% confidence limits of
 0.22-0.43 mmol/L). The study authors state that the increased concentration in newboms might
 be a result of lower glomerular filtration rates, increased resorption, and/or developmental needs.

 Serum levels of inorganic sulfate were found to be increased in pregnant women in the third
trimester (0.434 ± 0.006 mmol/L [42 mg/L] compared to a nonpregnant control value  of 0.328 ±
0.010 mmol/L [31 mg/L]) (Cole et al. 1985). It was suggested mat the fetus may need to
concentrate inorganic sulfate to support its development. Similarly, Cole et al. (1992) reported
that inorganic sulfate levels in amniotic fluid were increased in the third trimester compared with
the second trimester. Levels  in amniotic fluid were 0.317 ± 0.022 mmol/L (30 mg/L)  in the

                            External Review Draft — Sulfate—April 2002                          10

second trimester, but were 0.693 ± 0.042 mmol/L (66 mg/L) in the third trimester. The sulfate
levels correlated with the creatinine and uric acid levels in the amniotic fluid, suggesting that
renal excretion by the fetus may be the major source of the inorganic sulfate in the amniotic fluid
in the late stages of gestation.

Levels of inorganic sulfate in the cerebrospinal fluid (CSF) were reportedly lower than those in
serum. Cole et al. (1982) measured inorganic sulfate in CSF from 25 infants and children.
Mean CSF sulfate was 0.170 mmol/L (16.3 mg/L) in children less than 3 years of age and 0.095
mmol/L (9.1 mg/L; range 0.059 to 0.165 mmol/L) in children older than 3 years of age. Similar
to CSF concentrations, the serum sulfate concentration decreased from 0.5 mmol/L (48 mg/L) in
the newborn to 0.3 mmol/L (29 mg/L) in children older than 3 years of age. Because both CSF
and serum sulfate levels decrease in a similar fashion, the ratio of sulfate CSF/serum remains
constant at 0.33 in infants and children.

The levels of inorganic sulfate in human colostrum and milk were 0.066 ± 0.021 mmol/L (6.3
mg/L) and 0.029 ± 0.006 mmol/L (2.8 mg/L), respectively (McNally et al. 1991). The mean
sulfate level in the saliva in fasting adults (n = 17) was 0.072 ± 0.004 mmol/L, or 6.9 mg/L (Cole
and Landry 1985).

4.3  Metabolism

Inorganic sulfate is incorporated into several types of biomolecules, such as glycoproteins,
glycosaminoglycans, and glycolipids (Brown et al. 1965, Daughaday 1971, Morris and Sagawa
2000). Inorganic sulfate enters a metabolic pathway as an activated nucleotide intermediate, 3'-
phosphoadenosine-5'-phosphosulfate (PAPS), which serves as a substrate for a number of
relatively specific sulfotransferases. Sulfotransferases are enzymes that catalyze the sulfation
process (or sulfamation in the case of aromatic amines). They are found in the intestinal
mucosa, liver, and kidney (Bostrom 1965, Mulder and Keulemans 1978) and in human platelets
(Rein et al. 1981). Sulfate plays an important role in the detoxification and catabolism of
various endogenous (catecholamines, steroids, bile acids) and exogenous (acetaminophen and
other drugs) compounds.  Sulfate combines with several of these compounds to form soluble
sulfate esters (Mulder and Keulemans 1978, Weitering et al. 1979, Sipes and Gandolfi 1991).

Because inorganic sulfate is used in the metabolism of several compounds, sulfate levels could
be affected by the presence of drugs in the body.  In humans, the effect of acetaminophen
administration on the serum sulfate pool has been studied (Morris and Levy 1983a). In eight
human volunteers orally administered 1.5 g of acetaminophen, the mean serum concentration of
inorganic sulfate was significantly (p < 0.001) reduced from 0.410 to 0.311 mmol/L (39 to 30
mg/L) within 2 hours. Similar findings were observed in other human studies (Hendrix-Treacy
et al. 1986), as well as in animals (Morris et al. 1984), after acetaminophen administration.

4.4  Excretion

Sulfates are usually eliminated by renal excretion in free unbound form or as conjugates of
various chemicals. At high sulfate doses that exceed intestinal absorption, sulfate is excreted in
                            External Review Draft — Sulfate—April 2002                         11

Cocchetto and Levy (1981) conducted a study with five male volunteers who were orally
administered 18.1 g NajSO4-10H2O (5.4 g sulfate) in SO mL of water as a single bolus dose or as
four equally divided hourly doses. Urinary excretion of inorganic sulfate was measured at 24,
48, and 72 hours. Prior to dosing, the baseline excretion of inorganic sulfate (free form) was
measured several times for each individual. The normal baseline excretion of inorganic sulfate
ranged from 13 to 25 mmol/24 hours (1.3 to 2.4 g/24 hours).  To calculate the amount of the
exogenous dose that was excreted, the individual baseline values were subtracted from the total
urinary inorganic sulfate. Over 72 hours, a mean of 53.4% of the single dose or 61.8% of the
divided dose was recovered in the urine.  In general, at least two-thirds of the 72-hour sulfate
excreted appeared in the urine in 24 hours. The divided dose was fairly well tolerated, but the
single dose caused severe diarrhea.

The excretion of inorganic sulfate in humans is dependent on the cation. Morris and Levy
(1983b) reported that orally administered magnesium sulfate in humans was absorbed less
completely and more variably than sodium sulfate. Seven male volunteers received 13.9 g of
MgSO4-7H2O (5.4 g sulfate) in four 1.4-g portions in  100 mL of water over  a 4-hour period.
Based on the concentration of free sulfate in 72-hour urine, at least 37% of the dose was
excreted. An average of 30.2% ± 17% of the sulfete was excreted in urine over 24 hours for
magnesium sulfate  as compared with 43.5% ± 12% for sodium sulfate. There was high
intersubject variation.  Compared to sodium sulfate administered in an identical fashion, the
magnesium salt was less bioavailable.

Bauer (1976) reported that when doses between 60 and 80 ^Ci 35S (as sodium sulfate) were
administered orally to eight humans, 80% was recovered in the urine at 24 hours. For
comparison, 86% of the same dose given intravenously was excreted in the  urine in 24 hours.

The kidney regulates and T"aintaim serum sulfate levels through a capacity-limited reabsorption
mechanism. In humans,  the maximum rate of transport is 0.11 mM/min. If intestinal absorption
is slow or saturated, sulfates are eliminated in the feces. As a result, sulfates do not accumulate
in the body even after consumption of high levels (Morris and Levy 1983a,  Cole and Scriver


Data are available on the short- and long-term effects of sulfate in humans and animals. In
general, a laxative effect is the most common manifestation of exposure to high concentrations
of sulfate.  The soft stool or diarrhea that results from sulfate is an osmotic diarrhea, i.e., one that
results when the osmolality of the intestinal contents  exceeds that of the interstitial fluids. When
this occurs, water is drawn across the intestinal membrane into the lumen, increasing the
moisture content and volume of the fecal matter. This leads to an increased intestinal peristalsis
and evacuation of the intestinal contents. Poorly absorbed dissolved materials such as
magnesium sulfate, sorbitol, or lactulose are often associated with osmotic-induced diarrheas.
An osmotic-induced diarrhea ceases once the osmotically active gastrointestinal contents are
excreted (Stipanuk 2000).
                            External Review Dnjt—Sulfate - April 2002                          12

The osmotic diarrheal response to sulfate is influenced by the total temporal osmolyte load, as
well as the sulfate concentration. For example, if water contains magnesium ion as well as
sulfate ion, the diarrheal response will be increased because both ions are osmotically active.
This complicates interpretations of some of the ecological sulfate studies where data are reported
for sulfate concentrations but not for the presence of other drinking water constituents that may
be osmotically active.

5.1 Human

5.1.1   Short-Term Exposure Studies

Sulfate salts are known to have laxative properties in humans (Schofield and Hsieh 1983). A
dose of 15 g of magnesium or sodium sulfate will produce catharsis within 3 hours, but lower
doses can also produce this effect.  For example, approximately 5 g of magnesium sulfate was
reported to cause significant laxative effects when administered in a dilute solution to a fasting
man (Fingl 1980). Cocchetto and Levy (1981) reported that a single dose of 8 g of anhydrous
sodium sulfate induced severe diarrhea lasting up to 24 hours in five human subjects.  However,
the same amount of sodium sulfate taken at four equal hourly doses produced either no diarrhea
or mild diarrhea of short duration.

Humans appear to develop a tolerance to water containing high sulfate concentrations (Schofield
and Hsieh 1983).  Although the rate at which acclimation occurs has not been determined, it is
generally considered to occur in adults within one to two weeks (U.S. EPA 1999b). No specific
data on the length of time necessary for humans to acclimatize to the  cathartic effects of sulfates
were identified.

The sulfate salt is important in determining the extent and nature of any laxative effect. Morris
and Levy (1983b) reported that when humans were given the same millimolar dose of sulfate  as
either magnesium or sodium sulfate, the magnesium sulfate induced more adverse effects,
ranging from upset stomach to diarrhea, than did  sodium sulfate.

Chien et al. (1968) reported three cases of diarrhea in infants in Saskatchewan, Canada,
attributable to exposure to high sulfate concentrations in the water supply. In each case, local
well water with sulfate concentrations between 630 and 1,150 mg/L and a high total dissolved
solid concentration of 2,424 to 3,123 mg/L was used to prepare infant formulas. The total
dissolved solid concentrations reported in this study greatly exceeded the United States'
secondary standard for total dissolved solids of 500 mg/L.

When milk or water with low sulfate content, and presumably lower total dissolved solids,
replaced the local well water, recovery occurred (Chien et al. 1968).  Diarrhea returned when  the
original well water was re-introduced. The study authors concluded that waters with a sulfate
content higher than 400 mg/L were unsuitable for consumption by infants. Interpretation of this
study is limited by the number of the study subjects (n = 3) and by the lack of data on sulfate
concentration in and osmolarity of the infant formula.

In a case-control investigation to assess the association between infant diarrhea and ingestion  of
water containing elevated sulfate levels, Esteban  et al. (1997) reported that no significant

                            External Review Draft — Sulfate—April 2002                          13

association existed between exposure to sulfate from tap water and subsequent diarrhea in
infants. A total of 274 mothers of infants born in 19 South Dakota counties with high sulfate
concentrations in tap water were identified and interviewed using a telephone questionnaire
(n = 262) or in person (n = 12). The mothers were questioned on the frequency and consistency
of the infant's bowel movements and on the amount of water that the infant drank in the previous
7 days.  Diarrhea was defined as three or more loose stools in a 24-hour period. A sample of the
water used in the infant's diet was also submitted by the mother and analyzed for sulfate.  Two
hundred seventy-four infants were included in the study. Cases were defined as infants that
developed diarrhea (as identified by the mothers); controls were defined as infants that did not
develop diarrhea.

The average sulfate concentration in drinking water for cases was 416 mg/L versus 353 mg/L for
controls. The corresponding median sulfate concentration for cases was 289 mg/L (range
0-1,271 mg/L); the median concentration for controls was 258 mg/L (range 0-2,787 mg/L).
The median water intake for the controls was lower than for the cases (0.2 vs. 0.5 L/day). Mean
and median daily sulfate intake from water for all infants in the study was 29 and 17 mg/kg/day,
respectively. Mothers reported diarrhea in 19% of the infants living in households with sulfate
levels in tap water >500 mg/L and in 14% of infants living in households with sulfate levels
<500 mg/L. There was no significant correlation (OR = 1.4; 95% CI = 0.5-4.0) between the
incidence of diarrhea and the level of sulfate in water samples .

Heizer et al. (1997) examined bowel function in healthy adults following exposure to various
sodium sulfate concentrations in drinking water.  In a single-dose study of six adults (three men
and three women), each subject received drinking water with sulfate concentrations of 0 or
1,200 mg/L for two consecutive 6-day periods. A fluid intake of 36 mL/kg/day was maintained
in these subjects. Stool mass, frequency, consistency, and mouth-to-anus appearance time for
colored markers were measured.  When subjects received sulfate at 1,200 mg/L, mean stool mass
per 6-day pool period increased from 621 g to 922 g (p = 0.03). Mean stool mass per hour
increased from 4.8 g to  6.6 g (p = 0.03) with a change hi sulfate concentration from 0 to 1,200
mg/L.  However, stool frequency, consistency, and mouth-to-anus appearance time were not
significantly different for the high-sulfate water.

Heizer et al. (1997) also reported data from a multiple dose study hi four subjects (two men and
two women).  Each subject received drinking water with increasing sulfate concentrations of 0,
400,600,800,1,000, and 1,200 mg/L over six consecutive 2-day periods. In this study, there
was a significant linear trend for decreasing mouth-to-anus marker appearance time with
increasing sulfate concentrations (p = 0.03).  When the 10 subjects (4 subjects from the multiple-
dose study and 6 subjects from the single-dose study reported above) were used to compare
effects of 0 mg/L and 1,200 mg/L sulfate, significant differences hi stool consistency (p = 0.02)
and transit time (p - 0.03) were observed.  However, none of the subjects reported diarrhea or
passed more than three  stools per day.

In order to determine the effects of high sulfate concentrations hi transient populations (students,
visitors, hunters, etc.), the Centers for Disease Control (CDC), with funding from EPA (1999a),
conducted a study in adult human volunteers. The study was designed to determine if any
adverse effects would occur hi persons suddenly changing drinking water sources from one with
little or no sulfate to one with high sulfate concentrations. A total of 105 participants were

                            External Review Draft—Sulfate—April2002                          14

randomly assigned to five sulfate-exposure groups and were exposed to sulfate in bottled water
over a period of 6 days. The participants received water containing sulfate only for days 3
through 5 and were given sulfate-free bottled water for days 1,2, and 6. Subjects were blinded
to the level of sulfate in the drinking water. The number of participants in each exposure
category varied: 0 mg/L (n = 24), 250 mg/L (n = 10), 500 mg/L (n = 10), 800 mg/L (n = 33), or
1,200 mg/L (n = 28).

Several criteria were used to monitor whether the sulfate had a laxative effect.  The subjects kept
track of the number of bowel movements they had on each day of the study.  They were also
asked to rate the quality of their stools according to three definitions of diarrhea. Osmotic
diarrhea was defined as an increase in stool volume, diarrhea 1 was defined as paste-like or
liquid stools, and diarrhea 2 was defined as change in stool bulk and consistency. There were no
statistically significant differences in the mean number of bowel movements among the groups
on days 3,4,5, or 6; or in the mean number of bowel movements when days 3,4, and 5 were
compared with days 1 and 2. There was also no apparent trend in the percentage of subjects that
reported diarrhea during the exposure period as  opposed to the control period (days 1 and 2)
using any of the definitions of diarrhea.

The data were also evaluated using logistic regression analysis to determine the effects of sulfate
based on dose. The dose was calculated for each subject from the amount of water consumed,
the amount of sulfate in drinking water,  and the  body weight of the participant.  Using this
method, there were no statistically significant dose-response associations between sulfate dose
and reports of diarrhea (one-sided/? = 0.099) using any of the three definitions.

The authors also combined the incidence for diarrhea under any of the three definitions for the
lowest three dose groups (0,250, and 500 mg/L) and compared it with the incidences of the 800
mg/L and 1,200 mg/L dose groups. There were no statistically significant differences between
groups for any of the three diarrhea categories.  However, there was a dose-related trend for
increased stool volume (osmotic diarrhea) with increasing sulfate exposure when the three
groups were compared (9%, 15%, and 18%, respectively).

5.1.2  Long-Term Exposure Studies

Peterson (1951) analyzed the data from about 300 questionnaires that had  been collected by the
North Dakota Department of Health as part of its routine monitoring of the mineral content of
groundwater supplies.  The questionnaires had been distributed to private well owners along with
requests for samples of well water for analysis.  The questionnaires solicited information on
odor, taste, effects on cooking, and laxative effects from the water. The questionnaires analyzed
by Peterson (1951) were a subset of about 2,500 submitted to the Department of Health. They
were selected because analytical data for the water were available and the  questionnaire was
complete enough to be used for data analysis. Peterson (1951) plotted the concentration of
sodium sulfate and magnesium sulfate in the water against whether water use was associated
with a laxative effect. Both sodium and magnesium ions were present in most of the water
samples.  He concluded that sulfate was likely to have a laxative effect when the concentrations
exceed 750 mg/L, but was unlikely to have such an effect at concentrations less than 600 mg/L.
As the concentration of the magnesium in the water increased, the sulfate concentration that was
associated with a laxative response decreased.

                            External Review Draft — Sulfate—April 2002                          15

Moore (1952) analyzed data from 248 wells from the North Dakota Department of Health survey
(Peterson 195 1 data). Questionnaires completed by the users of the wells provided a YES or NO
response for laxative effects for 176 of 248 wells. For 69 wells with positive responses for
laxative effects, the mean sulfate level was 1,250 mg/L and the median was 1,090 mg/L. For
107 wells with negative responses, the mean and median sulfate levels were 500 mg/L and 403
mg/L, respectively. When the data were separated into specific ranges, the percent of YES
responses for laxative effects was 22% (10/46) for wells with levels between 0 and 200 mg/L;
24% (9/37) for wells with levels between 200 and 500 mg/L; 33% (13/39) for wells with levels
between 500 and 1,000 mg/L; and 69% (37/54) for levels >1,000 mg/L. Moore (1952)
concluded that sulfate ion concentration is critical at 1,000 mg/L and at more than 2,000 mg/L is
almost certain to produce discernible physiological effects. For a combination of magnesium
and sulfate, effects are likely to occur when the total concentration exceeds 1,000 mg/L. Many
of the wells in this data set also contained high levels of total dissolved solids and magnesium.
Similar results were obtained when Cass (1953) analyzed the data provided in the North Dakota

5.2  Animal

5.2.1  Short-Term Exposure Studies

The oral LDws of ammonium sulfate, sulfuric acid, and potassium sulfate in the rat are
3,000-4,000 mg/kg, 2,140 mg/kg (FASEB 1975), and 6,600 mg/kg (RTECS 2000), respectively.
The oral LDjo of sodium  sulfate in the mouse is 5,989 mg/kg (RTECS 2000).
Adams et al. (1975) supplied groups of three White Leghorn hens with drinking water containing
250 to 16,000 mg sulfate/L (as sodium or magnesium sulfate). At 16,000 mg/L, the hens
exhibited decreased body weight, decreased feed consumption, decreased egg production, and
increased water consumption.  At this dose, mortality (100%) occurred by day 7 for hens
drinking water with magnesium sulfate, and by day 12 for hens drinking water with sodium
sulfate.  Necropsy revealed focal necrosis of individual renal glomeruli, with uric acid
accumulation in both the kidney and gut. Similar, but less severe, histologic changes were seen
in hens receiving water at 4,000 mg/L for up to 3 weeks.

Paterson et al. (1979) administered drinking water containing 3,000 mg/L of added sulfate (as
either sodium sulfate or a 1 : 1 combination of both sodium and magnesium sulfate) for 28 days to
groups of 17 or 18 weanling pigs (7.7 to 8 kg). The control group (n = 16) received water
containing 320 mg/L sulfate. No significant changes (p > 0.05) in average daily weight gain or
feed/gain ratio were observed when the treated group was compared to controls. Fluid
consumption increased for both groups receiving the high concentration of sulfate in their water,
and stools were soft for these animals compared with controls.

Sulfate (sodium and magnesium sulfate in combination or independently) was administered to
young pigs at concentrations ranging from 600 to 1,800 mg/L in drinking water for 28 days.
Weight gain, feed consumption, water consumption, feed conversion, prevalence of diarrhea, and
evidence of common postweaning enteric pathogens were determined. Sulfate did not impair
performance or health of pigs.  However, loose and watery stools appeared to be more prevalent
in the groups receiving 1 ,800 mg/L sulfate (both salts independently and in combination)

                            External Review Draft— Sulfate— April 2002                         16

compared to control, 600 mg/L, and 1,200 mg/L groups (Veenhuizen et al. 1992). Similar results
were observed in an earlier study by Anderson and Stothers (1978) in which groups of nine
young pigs given water containing 600 mg/L total solids as sodium sulfate displayed scouring
(soft stools), primarily during the first week of a 6-week experimental period.

Gomez et al. (1995) used neonatal piglets to study the effect of inorganic sulfate on bowel
function. Two experiments were conducted to evaluate the effect of high levels of inorganic
sulfate on the growth, feed intake, and feces consistency of piglets, and to determine the dose at
which at least 50% of piglets develop an osmotic diarrhea. In each experiment, 40 pigs with an
average age of 5 days were individually caged and reared with an automatic feeding device.
Groups of 10 pigs were fed 1 of 4 liquid diets containing inorganic sulfate (anhydrous sodium
sulfate) at 0,1,200,1,600, or 2,000 mg/L (of diet) for an 18-day study, or 0, 1,800, 2,000, or
2,200 mg/L for a 16-day study. The levels of added sulfate did not affect (p > 0.05) the growth
of piglets or their feed intake. Whereas 1,200 mg sulfate/L had essentially no effect on feces
consistency, concentrations greater than 1,200 mg/L increased the prevalence of diarrhea.
Concentrations greater than 1,800 mg/L resulted in a persistent diarrhea.  The changes in feces
consistency suggest that the level of added dietary inorganic sulfate at which 50% of piglets
develop diarrhea is between 1,600 and 1,800 mg/L.  Analysis of rectal swabs showed no
evidence of E. coli or rotavirus infection.

5.23,  Long-Term Exposure Studies

In a 90-day subchronic study, Wurzner (1979) examined the effects of sulfate in drinking water
in Sprague-Dawley rats (25/sex/group). Treated animals received mineral waters containing low
(<10 mg/L), intermediate (280 mg/L), or high (1,595 mg/L) concentrations  of sulfate.  Control
animals were provided with tap water containing 9 to 10 mg/L sulfate.  No information was
provided on the concentrations of minerals other than sulfate in the three different mineral
waters.  No mortalities or effects on body weight, food consumption, food efficiency (a measure
of food intake versus body weight change), or water consumption were observed. No soft feces
or diarrhea were observed. No effects on hematology or serum chemistry (blood urea nitrogen,
glucose, triglycerides, cholesterol, total protein, and alkaline phosphatase activity) were
observed after 90 days. Organ weights were not affected, and no histologic changes were
observed at any tissue site. Five rats/sex/group continued treatment beyond the 90 days.  Blood
urea nitrogen in the high-dose group tended to be decreased in both sexes, but this occurred only
after 6 months of treatment.

Digesti and Weeth (1976) supplied groups of four weanling Hereford-Angus heifers with
drinking water containing 110 to 2,500 mg/L sodium sulfate. After 90 days of dosing, no overt
toxicity was observed in any animals; feed consumption, water consumption, and growth were
not affected.  Increased levels of methemoglobin and sulfhemoglobin were  observed in the
animals consuming 1,250 and 2,500 mg/L sodium sulfate; this was attributed to the bacterial
reduction of sulfate to sulfide in the rumen.  At 2,500 mg/L, renal filtration of sulfate was
increased by 37.7% and renal reabsorption was decreased by 23.7%.
                            External Review Draft — Sulfate—April 2002                          17

 5.23 Reproductive and Developmental Studies

 Three oral studies on reproductive and developmental effects were identified for sulfate. On the
 basis of these studies, it appears that sulfate does not induce adverse reproductive or
 developmental effects.

 Sodium sulfate was administered by gavage at a concentration of 2,800 mg/kg/day to pregnant
 ICR/SIM mice on gestation days 8 to 12 (Seidenberg et al. 1986). There was no evidence of
 maternal toxicity or increased resorption rate.  Pup survival was 100%, and no adverse
 developmental effects were observed. Neonatal birth weight was significantly increased in the
 treated group compared with controls.

 Six groups of 10 female, randomly bred albino ICR mice were administered sodium sulfate in
 drinking water at dose levels of 0 (distilled water control), 0 (sodium control), 625,1,250,2,500,
 or 5,000 mg/L beginning 1 week prior to breeding (Andres and Cline 1989). The amount of
 sodium in all groups, except the distilled water control, was kept constant by administering
 sodium bicarbonate.  Control mice, receiving only distilled water, consumed significantly less
 (p < 0.05) than mice receiving sulfate treatments, and sodium-control mice drank significantly
 more water (p < 0.05) than mice treated with sulfate.  All mice were carried to term. No
 differences were found in litter size, litter weaning weights, or gestational or lactational weight
 gain of the dams among sulfate treatments. Histopathology evaluations were not performed.
 The authors concluded that water containing up to 5,000 mg/L sulfate is not toxic to the
 gestating mouse.

 Paterson et al. (1979) investigated the effects of water with a high sulfate content on swine and
 their offspring. The pigs, 31 sows and 27 gilts of Hampshire x Yorkshire x Duroc breeding,
 were randomly divided into three groups that received either tap water (320 mg sulfate/L) or
 water with sodium sulfate added at 1,790 mg/L or 3,298 mg/L.  The animals were given access
 to these waters from prebreeding day 30 through lactation day 28. No significant differences in
 gestation or lactation weight gain, number of pigs delivered, or average pig and litter birth
 weights were reported.

 5.2.4 Cancer Studies

 In a study of the toxicity and carcinogenicity of nickel compounds (nickel hydroxide and nickel
 sulfate) in Wistar rats, sodium sulfate (used as  a control) did not appear to be tumorigenic
 (Kasprazak et al. 1980). In this study, Wistar rats (100 males and 10 females) were injected
 intramuscularly every other day for 4 weeks with 0.7  mg sodium sulfate/rat (approximately 2 mg
 SO42Vkg in aqueous solution at pH 5.6). After  8 months, no tumors were observed in either the
 sodium sulfate or nickel sulfate treated rats. However, the value of this study in assessing the
carcinogenic effects of sulfate ingestion is limited because of the route of exposure, the duration
of the study, and the nonstandard protocol.
                            External Review Draft — Sulfate — April 2002                          18


Water contaminated with sulfates may have an unpleasant taste.  Characteristics such as taste,
odor, and color, often referred to as organoleptic properties, are not used by EPA for developing
primary water standards. Organoleptic properties, however, can  be used in the establishment of
secondary drinking water standards.

EPA established a secondary drinking water standard of 250 mg/L for sulfate in 1984 based on
taste properties (U.S. EPA 1984). This value was adopted from the Public Health Service
Drinking Water Standards (PHS 1962).  Secondary standards are not enforceable by the Federal
Government; they are recommended to States as reasonable goals for contaminants, but mere is
no obligation for the States to reach these goals.

There is a paucity of actual experimental data available on the taste threshold for sulfate. Taste
threshold concentrations for several common sulfate salts have been reported. The taste
thresholds varied depending on the type of salt: 170-370 mg/L sulfate as sodium sulfate, 180-640
mg/L as calcium sulfate, and 320-480 mg/L as magnesium sulfate (Lockhart et al. 1955, as cited
in PHS 1962).

The detection of taste differs from the perception of a taste as unpleasant Accordingly, the
results reported by Heizer et al. (1997) on the response of 10 subjects to sulfate in drinking water
are of interest. After completion of the exposure component of this study, 8 of the 10 subjects
rated the taste of 1,200 mg/L sulfate, as sodium sulfate, as neutral to slightly unpleasant.  One
subject rated the water as moderately unpleasant and another as very unpleasant. This study
indicates there is variability in the response to the taste of sulfate, and the threshold for detecting
an unpleasant taste is apparently above the threshold of taste.

In the study of the laxative effects of sulfate that was conducted for EPA by CDC (U.S. EPA
1999a), the subjects were asked if the smell or taste of the water was different from that which
they usually consumed. All subjects received sulfate-free control water on days 1,2, and 6 of
the study and either the sulfate-free water (controls) or a sulfate-containing water on days 3,4,
and 5 of the study. There was a definite increase in the number of subjects who thought that the
water tasted differently on days 3,4, and 5 for the 800 and 1,200 mg/L concentrations, with 79%
and 82% reporting a difference in taste, respectively.  Average daily water consumption also
decreased in these same groups on days 3,4, and 5 when compared with intakes on days 1,2,
and 6. About half of the participants receiving the 250 and 500 mg/L concentrations reported a
difference in taste (57% and 50% respectively) on the days when they were exposed to sulfate.
Water consumption showed a downward trend over this same period as well. Twenty-five
percent of the control group also reported that there was a difference in the taste of the water.
For this group, there was no downward trend in water consumption across the 3-day exposure

The ability to taste differs among individuals, as well as in the same individual at different times.
Temperature and the presence of other dissolved solids in the water also influence taste. Given
the expected variability in taste, as well as the results in Heizer et al. (1997), the U.S. EPA
secondary maximum contaminant level (SGML) of 250 mg/L should be adequately protective
for adverse sulfate taste effects.
                            External Review Draft—Sulfate—April 2002                         19


 7.1  Hazard Characterization
 Some data are available that report human responses to sulfate. Data include those from
 controlled settings (i.e., studies and experimental trials) and uncontrolled settings (i.e., case
 studies from areas with high sulfate concentrations in the drinking water). Most of the available
 data are based on short-term exposure and were obtained from controlled studies.  Reports on
 long-term exposure are based on responses to questionnaires in North Dakota and South Dakota,
 States with high sulfate concentrations in their drinking water supply. In animals, data on
 reproductive and developmental effects are available for short-term and long-term exposures to
 sulfate. There are limited data on the potential carcinogenic effects of sulfate.

 The available data demonstrate that sulfate induces a laxative effect following acute exposures to
 relatively high concentrations (Anderson and Stothers 1978, Fingl 1980,  Schofield and Hsieh
 1983, Stephen et al.  1991, Cochetto and Levy 1981, U.S. EPA 1999a, Gomez et al. 1995, Heizer
 et al. 1997). The concentrations of sulfate mat induced these effects varied, but all occurred at
 concentrations >500 mg/L. However, the severity of the laxative effect that occurs from acute
 sulfate exposures  may be dependent on the sulfate salt, as well as how the dose is administered.
 For example, magnesium sulfate exerts a stronger laxative effect than sodium sulfate. This
 likely occurs because magnesium sulfate is absorbed less completely than sodium sulfate and has
 a more pronounced effect on the osmolarity of the intestinal contents (Morris and Levy 1983b).
 Additionally, a single dose of sulfate that produces a laxative effect does not have the same
 effect when divided and administered in intervals, i.e., a single dose produced severe diarrhea,
 whereas divided doses produced only mild or no diarrhea (Cochetto and Levy 1981).

 Chronic and subchronic exposures to high concentrations of sulfate do not appear to produce the
 same laxative effect as seen in acute exposures. In a 90-day study using Sprague-Dawley rats,
 Wurzner (1979) did not observe soft feces or diarrhea in rats administered mineral waters
 containing up to 1,595 mg/L of sulfate. However, earlier reports indicate that chronic exposure
 to high sulfate concentrations in drinking water resulted in laxative effects in humans (Peterson
 1951, Moore 1952, Cass 1953). These reports used data that were based on questionnaires,
 which may be subject to bias. For example, the questionnaire included an inquiry about the
 laxative effect that requested a YES or NO response. This type of question is subject to the
 respondent's interpretation of what constitutes a laxative effect. In addition,  sulfate was
 probably not the only contaminant found in the drinking water. Chronic exposure to sulfate may
 not have the same laxative effect as an acute exposure because humans appear to develop a
 tolerance to  drinking water with high sulfate concentrations (Schofield and Hsieh 1983). It is
 not really known when this acclimation occurs; however, in adults, acclimation is thought to
 occur between one to two weeks (U.S. EPA 1999b).

No adverse developmental effects were observed following the administration of 2,800
mg/kg/day of sulfate to pregnant ICR/SIM mice on gestation days 8 to 12 (Seidenberg et al.
 1986).  No reproductive effects  were observed following the ingestion of drinking water
containing up to 5,000 mg/L of sulfates by ICR/SIM mice (Andres and Cline 1989) or 3,298
mg/L of sulfates by Hampshire  * Yorkshire * Duroc pigs (Paterson et al.  1979). Based on these
studies, sulfate does not appear  to be a reproductive or a developmental toxicant.
                           External Review Draft — Sulfate —April 2002                         20

No tumors were observed after 8 months in a study using Wistar rats injected intramuscularly
with sodium sulfate every other day for 4 weeks (Kasprazak et al. 1980).  Because of the short-
term observation period, the route of exposure, and the experimental protocol, it is not possible
to draw conclusions on the potential carcinogenicity of sulfate.  Because of the limited data, U.S.
EPA/Office of Water (1993) has classified sulfate as Group D-not classified as to human
carcinogenicity.  This category is reserved for contaminants with inadequate evidence to support
a determination on carcinogenicity.

7.2  Characterization of Organoleptic Effects

The ability to taste differs among individuals, as well as for the same individual at different
times.  The temperature of the water, the companion ion, and the presence of other dissolved
solids impact the taste sensation.  There is also a difference between the concentrations that
impart a taste to water and those that are classified as causing an unpleasant taste. Each of these
factors makes it difficult to define a taste threshold for sulfate.

The experimental data on the organoleptic properties of sulfate in drinking water are limited.  No
studies were identified that were conducted using standard taste-testing procedures.  In the study
by Heizer et al. (1997), 8 of 10 subjects rated the taste of drinking water containing 1,200 mg/L
sulfate as neutral to only slightly unpleasant. Two subjects classified the taste as moderately to
extremely unpleasant. In the study conducted by CDC for EPA (1999a), about 50% of the
participants could not distinguish between the taste of sulfate-free water and water containing
either 250 mg/L sulfate or 500 mg/L sulfate. Even when the water contained 1,200 mg/L sulfate,
20% of the participants could not detect the taste.

Given the variability in the ability of consumers to identify a taste in water that contains sulfate,
the present SMCL of 250 mg/L appears to be adequately protective of the aesthetic taste
properties of drinking water containing sulfate.

73  Dose-Response Characterization

Although several studies (Peterson 1951, Moore 1952, Cass 1953) have examined the effects of
long-term exposure of humans to sulfate in drinking water, none of them can be used to derive a
dose-response characterization. These studies utilized data collected from the North Dakota
Department of Health Survey (Moore 1952). An increasing trend was observed in persons
reporting laxative effects as sulfate concentrations increased (i.e., 22%, 24%, 33%, and 69% for
sulfate concentrations of 0-200,200-500,500-1,000, and > 1,000 mg/L, respectively). However,
the results of these studies cannot be used to derive a dose-response characterization for the
following reasons: (1) the results are based on recall with little scientific weight (i.e., sulfate may
have induced the laxative effects, but it cannot be proven); and (2) the water samples had
varying concentrations of magnesium and total dissolved solids in addition to sulfate.

No laxative effects were observed in fats (Wurzner 1979) or heifers (Digesti and Weeth 1976)
following long-term exposure to sulfate in drinking water. Consequently, these studies cannot
be used for a dose-response characterization.
                            External Review Draft—Sulfate — April 2002                          21

Because sulfate appears to exert its laxative effect with short-term exposures rather than long-
term exposures, several short-term exposure studies were reviewed.  Three short-term studies
were identified that evaluated the effect of various sulfate concentrations on bowel function in a
controlled environment: two in humans and one in animals. In the multiple-dose study by
Heizer et al. (1997), sulfate concentrations of 0,400,600,800,1,000, or 1,200 mg/L were given
to four subjects (two men and two women) for six consecutive 2-day periods (2 days per
concentration). A significant trend was observed for a decreasing mouth-to-anus marker-
appearance time of chemical markers with increasing sulfate concentration.  For a single-dose
study by the same researchers, six adults (three men and three women) received drinking water
with sulfate concentrations of 0 or 1,200-mg/L for two consecutive 6-day periods. Statistically
significant increases in mean stool mass per 6-day pool and in mean stool mass per hour were
observed in the 1,200-mg/L dose group. However, none of the subjects reported frank diarrhea.

CDC conducted a study for EPA (1999a) that examined the effect of sudden changes in sulfate
levels in drinking water in 105 subjects. The participants received water containing sulfate at 0,
250, 500, 800, or 1,200 mg/L from day 3 through day 5, and were given sulfate-free bottled
water for days 1,2, and 6. There were no statistically significant differences in the mean number
of bowel movements in any group or a dose-response relationship between sulfate dose and
reports of diarrhea (one-sided/? = 0.099).  However, when the diarrhea incidence data for the
lowest three dose groups were compared to  the incidence for the 800 mg/L and 1,200 mg/L dose
groups, there was a dose-related trend for increased stool volume (osmotic diarrhea) when the
three groups were compared (9%, 15%, and 18%). The dose-related trend, however, was not
statistically significant.

Neonatal piglets were exposed to various concentrations of sulfate to simulate the effect of
inorganic sulfate on the bowel function hi infants (Gomez et al. 1995). No diarrhea was
observed in any of the piglets at 0 and 1,200 mg/L concentrations; however, concentrations
greater than 1,200 mg/L resulted in an increased prevalence of diarrhea, and concentrations
greater than 1,800 mg/L resulted in persistent, nonpathogenic diarrhea.

These studies as a group suggest that there is a risk for a laxative-type response to sulfate in
drinking water at concentrations greater than 1,000 mg/L (U.S. EPA 1999a,  Heizer et al. 1997,
Moore 1952).  The observed effect is a response to the net osmolarity of the intestinal contents,
and thus is influenced not only by sulfate intake, but also by the presence of other osmotically
active materials in the drinking water or diet, and by the temporal pattern of sulfate ingestion.
The laxative effect of sulfate can be manifest as an increase in stool mass, increased stool
volume, increased stool moisture, decreased intestinal transit time, and/or frank diarrhea.  Frank
diarrhea did not occur in either of the controlled human studies of sulfate exposure (U.S. EPA
1999a, Heizer et al. 1997). There was merely a slight increase in stool mass or stool volume
with sulfate concentrations of 800 to 1,200 mg/L.

At this time, it is not possible to characterize a dose-response relationship for laxative effects due
to short- or long-term exposure to sulfate. A Centers for Disease Control and Prevention (CDC)
panel favored a health advisory for situations where sulfate levels in drinking water are greater
than 500 mg/L (U.S. EPA 1999b).  The most sensitive endpoint was considered by the panelists
to be osmotic diarrhea. The panelists  concluded that the existing literature supports restricting
sulfate exposure, especially for infants, when the advisory value of 500 mg/L is exceeded. The

                            External Review Dn# — SutftOe —April 2002                         22

panelists referred to the study by Chien et al. (1968), which found that sulfate levels *630 mg/L
caused diarrhea in infants.  It should be noted that this effect was observed after the infants had
ingested formula made with water containing sulfate and other osmotically active agents. In
fact, the total dissolved solid concentration of the water used to prepare infant formulas was high
(2,424 to 3,123 mg/L) and in two cases contained substantial quantities of magnesium (124 and
130 mg/L). The CDC panel concluded that 500 mg/L seemed to be a safe level for sulfate
ingestion, as 500 mg/L was shown to be safe in all studies.  For comparison, the osmolarity of
500 mg/L sulfate as sodium sulfate is 15.6 mOsmol/L whereas tbe osmolarity of the ions (Na,
K, Cl, and citrate) in Pedialyte, a preparation used to treat diarrhea and replenish electrolytes in
infants, is 110 mOsmol/L.  When the dissolved sugars in Pedialyte are included the osmolarity
increases to 250 mOsmol/L.

The experimental data on the organoleptic properties of sulfate in drinking water are limited.  In
a study by Heizer et al. (1997), 8 of 10 subjects rated the taste of drinking water containing 1,200
mg/L sulfate as neutral to slightly unpleasant. Only two classified the taste as moderately to
extremely unpleasant. In the study conducted by CDC for EPA (1999a), approximately 50% of
the participants stated that water containing 250 mg/L sulfate and 500 mg/L sulfate tasted
different from the sulfate-free control water.

Given the apparent variability in consumers' ability to identify a taste in water that contains
sulfate, the present SMCL of 250 mg/L appears to be adequately protective of the esthetic taste
properties of drinking water containing sulfate. The health-based advisory value  of 500 mg/L
will protect against sulfate's laxative effects in the absence of high concentrations of other
osmotically active chemicals in the water. In situations where the water contains high
concentrations of total dissolved solids and/or other osmotically active ions, laxative-like effects
may occur if the water is mixed with concentrated infant formula or powdered nutritional
supplements. In such situations, an alternate low-mineral-content water source is advised.
Infants are more susceptible to diarrheal water loss than adults because of differences in
gastrointestinal structure and function.


Abemathy CO, Dooley IS, Wu J. 2000. Sulfate exposure and potential regulation. CRC Crit
Rev Clin Lab Sci 37(4):401-405.

Adams AW, Cunningham FE, Monger LL. 1975. Some effects on layers of sodium sulfate and
magnesium sulfate in the drinking water. Poult Sci 54:707-714.

Alley WM.  1993.  Regional ground-water quality.  New York: Van Nostrand Reinhold, p. 633.

Anderson DM, Stothers SC. 1978.  Effects of saline water high in sulfates, chlorides and nitrates
on the performance of young weanling pigs.  J Anim Sci 47(4):900-907.

Andres CJ, Cline TR. 1989. Influence of sulfate in drinking water on mouse reproduction
during two parties. J Anim Sci 67:1313-1317.
                            External Review Draft —Sulfate —Aprii 2002                          23

Bauer JH. 1976. Oral administration of radioactive sulphate to measure extracellular fluid space
in man. J Appl Physiol 40:648-650.

BostromH.  1965.  Sulphate conjugation and sulphate conjugates. Scand Soc Clin Chem Clin
Pathol 10:33-52.

Brown RG, Button GM, Smith JT.  1%5. Changes in collagen metabolism caused by feeding
diets low in inorganic sulfur. J Nutr 87:228-232.

Budavari S ed. 1996. The Merck Index. An Encyclopedia of Chemicals and Drugs, 12th ed.
Rahway, NJ: Merck and Company.

CassJS.  1953. Report on the physiological effects of some common inorganic salts in water on
man and domestic animals. University of Cincinnati. Prepared for the Ohio River Valley Water
Sanitation Commission. Cincinnati, OH: The Kettering Laboratory.

Chien L, Robertson H, Gerrard JW. 1968.  Infantile gastroenteritis due to water with high
sulfate content. Can Med Assn J 99:102-104.

Cocchetto DM, Levy G.  1981. Absorption of orally administered sodium sulfate in humans. J
Pharm Sci 70(3):331-333.

Cole DEC, Baldwin LS, Stirk LJ.  1985. Increased renal reabsorption of inorganic sulfate in
third trimester high-risk pregnancies. Obstet Gynecol 66:485-490.

Cole DEC, Landry DA. 1985.  Determination of inorganic sulfate in human saliva and sweat by
controlled-flow anion chromatography. Normal values in adult humans. JChromatogr

Cole DEC, Magor B, Oulton M, Stirk LJ. 1992. Increased inorganic sulfate concentrations in
amniotic fluid. J Perinat Med 20:443-447.

Cole DEC, Scrivcr CR. 1980.  Age-dependent serum sulfate levels in children and adolescents.
Clin Chim Acta 107:135-139.

Cole DEC, Shafai J, Scriver CR.  1982. Inorganic sulfate in cerebrospinal fluid from infants and
children. Clin Chim Acta 107:135-139.

Cole DEC, Thurgood AM, Whiting SJ. 1991. Increased serum sulfate after protein loading in
adult humans.  Can J Physiol Pharmacol 69:25-27.

DaughadayWH.  1971. Regulation of skeletal growth by sulfation factor. Adv Intern Med

Digesti RD, Weeth HJ. 1976. A defensible maximum for inorganic sulfate in the drinking water
of cattle.  J Anim Sci 42:1498-1502.
                           External Review Dn^t — Sulfau- April 2002                        24

DreverJI.  1988.  The geochemistry of natural waters, 2nd ed.  Englewood Cliffs, NJ: Prentice
Hall, p. 437.

Environment Canada. 1984. Detailed surface water quality data, Northwest Territories 1980-
1981, Alberta 1980-1981, Saskatchewan 1980-1981, Manitoba 1980-1981.  Unpublished results
provided by Inland Waters Directorate, Environment Canada.

Esteban E, Rubin CH, McGeehin MA, Flanders WD, Baker MJ, Sinks TH.  1997.  Evaluation of
infant diarrhea associated with elevated levels of sulfate in drinking water:  A case control
investigation in South Dakota. Int J Occup Environ Health 3(3): 171 -176.

Federation of American Societies for Experimental Biology (FASEB). 1975.  Evaluation of the
health aspects of sulfuric acid and sulfates as food ingredients. Bemesda, MD: Life Sciences
Federation of American Societies for Experimental Biology. Prepared for Bureau of Foods, Food
and Drug Administration. Contract no, FDA 223-75-2004.

FDA. 1999. Code of Federal Regulations: Food and Drugs Part 182  and 184,  Office of the
Federal Register.  National Archives and Records Administration.  Washington, DC.

Field CW.  1972. Sulfur: Element and geochemistry. In: Fairbridge RW, ed.  The Encyclopedia
of Geochemistry and Environmental Sciences. New York:  Van Nostrand Reinhold, pp. 1142-

FinglE. 1980. Laxatives and cathartics. In: Oilman AG, Goodman LS, Oilman L, eds.
Goodman and Oilman's The Pharmacological Basis of Therapeutics, 6* Ed New York:
Macmillan, pp. 1002-1612.

Florin T, Neale G, Gibson GR, Christl SU, Cummings JH.  1991.  Metabolism of dietary
sulphate: Absorption and excretion in humans.  Gut 32:766-773.

Gomez GG, Sandier RS, Seal E, Jr. 1995.  High levels of inorganic sulfate cause diarrhea in
neonatal piglets.  J Nutr 125(9):2325-2332.

Greenwood NN, Earnshaw A. 1984. Chemistry of the Elements.  Oxford,  England: Pergamon

Hendrix-Treacy S, Wallace SM, Hindmarsh KW, Wyant GM, Daniklewich A. 1986. The effect
of acetaminophen administration on its disposition and body stores of sulphate. Eur J Clin
Pharmacol 39:143-148.

Heizer WD, Sandier RS, Seal E, Jr, Murray SC, Busby MG, Schliebe BG, Pusek SN. 1997.
Intestinal effects of sulfate in drinking water on normal human subjects. Dig  Dis Sci
42(5): 1055-1061.

Hindmarsh KW, Mayers DJ, Wallace SM, Danilkewich A, Ernst A.  1991.  Increased serum
sulfate concentrations in man due to environmental factors: Effects on acetaminophen
metabolism. Vet Hum  Toxicol 33(5):441-445.

                           External Review Dn# — S*y*t — April 2O02                        25

Hitchcock DR. 1975.  Biogenic concentrations to atmospheric sulfate levels. Proceedings of the
Second National Conference on Complete Water Re-use.  Chicago, IL: American Institute of
Chemical Engineers, pp. 291-310.

Hoffinan DA, Wallace SM, Verbeeck RK. 1990. Orcadian rhythm of serum sulfate levels in
man and acetaminophen pharmacokinetics.  Eur J Clin Pharmacol 39:143-148.

Kaplan IR.  1972. The Encyclopedia of Geochemistry and Environmental Sciences. New York:
Van Nostrand Rcinhold, pp. 1148-1151.

Kasprazak KS, Gabryel P, Jarczewska K. 1980. Carcinogenicity of nickel hydroxide and nickel
sulphate in Wistar rats: Preliminary findings. In: Brown SS, Sunderman FW, eds. Nickel
Toxicity. New York: Academic Press, pp. 59-62.

KatzM. 1977. The Canadian sulphur problem. In: Sulphur and its inorganic derivatives in the
Canadian environment. Ottawa: National Research Council of Canada (NRCC No. 15015), pp.

Krijgsheld XRE, Scholtens E, Mulder GJ. 1980. Serum concentration of inorganic sulphate in
mammals: Species differences and circadian rhythm. Comp Biochem Physiol 67A:683-686.

Lockhart EE, Tucker CL, Merritt MC. 1955.  The effect of water impurities on the flavor of
brewed coffee. Food Res 20:598-605.

McCabe LJ, Symons JM, Lee RD, Robeck GG.  1970.  Survey of community water supply
systems. J Am Water Works Assoc 62:670-687.

McNally ME, Atkinson SA, Cole DE.  1991.  Contribution of sulfate and sulfoesters to total
sulfur intake in infants fed human milk. J Nutr 121(8): 1250-1254.

Moore E. 1952. Physiological effects of the consumption of sahne drinking water. A progress
report to the Subcommittee on the Water Supply of the Committee on Sanitary Engineering and
Environment, January 1952. Washington DC: National Academy of Sciences, Appendix B 1-2.

Moore JW.  1991. Inorganic contaminants of surface water, research and monitoring priorities.
Springer series on environmental management  De Santo RS, series ed. New York: Springer-
Verlag, p. 334.

Morris ME, GaUnsky RE, Levy G. 1984. Depletion of endogenous inorganic sulfate in the
mammalian central nervous system by acetaminophen. J Pharm Sci 73:853.

Morris ME, Levy G. 1983a.  Serum concentration and renal excretion by normal adults of
inorganic sulfate after acetaminophen, ascorbic acid, or sodium sulfate. Clin Pharmacol Ther

Morris ME, Levy G. 1983b. Absorption of sulfate from oraUy administered magnesium sulfate
in man. J Toxicol-Clin Toxicol 20(2): 107-114.
                           External Review Draft-Suyate —April 2002

Morris ME, Sagawa K. 2000. Molecular mechanisms of renal sulfate regulation. CRC Critical
RevClinLabMed. 37(4):345-388.

Mulder GR, Keulemans K.  1978.  Metabolism of inorganic sulphate in the isolated perfused rat
liver. Effect of sulphate concentration on the rate of sulphanation by phenol sulphotransferase.
Biochem 1176:959-965.

National Institute for Occupational Safety and Health (NIOSH).  1981.  Occupational health
guideline for sulfuric acid. U.S. Department of Health and Human Services.  Publication No. 8-

Nova Scotia Department of Public Health. 1988. Chemical analysis of municipal water
supplies: Datafile Printout, 1987-1988.  Halifax, Nova Scotia. Unpublished results provided by
Public Health Inspection, Department of Public Health, Nova Scotia, Canada.

Ontario Ministry of the Environment. 1987. Appendix to Annual Report Air Quality in Ontario.
Air Quality Assessment Unit, Etobicoke, Ontario.

Paterson DW, Wahlstrom RC, Libal GW, Olson OE. 1979. Effects of sulfate in water on swine
reproduction and young pig performance. J Anim Sci 49:664-667.

Patil SG, Sarma VAK, vanLoon GW. 1989. Acid rain, cation dissolution, and sulfate retention
in three tropical soils. J Sail Scl 40:85-93.

Peterson NL. 1951. Sulfates in drinking water.  Official Bulletin N.D. Water Sewage Works

Public Health Service. 1962. Public Health Service Drinking Water Standards. U.S. Department
of Health, Education, and Welfare. Public Health Service. Consumer Protection and
Environmental Health Service. Environmental Control Administration. Public Health Service
Publication 956. Rockville, MD.

Registry of Toxic  Effects of Chemical Substances (RTECS). 2000.  Database. Cincinnati, OH:
National Institute  for Occupational Safety and Health. July 3,2000. www.cdc.gov/niosh/rtecs.

Rein G, Glover V, Sandier M.  1981.  Sulphate conjugation of biologically active monoamines
and their metabolites by human platelet phenosulphotransferase. Clin Chim Acta 111:247-256.

Saskatchewan Environment and Public Safety.  1989. Water quality monitoring data (1970 to
October 1989).  Unpublished.

Schofield R, Hsieh D. 1983. Criteria and Recommendations for Standards for Sulfate in
Military Field Supplies.  Livermore, CA: Lawrence Livermore National Laboratory. Contract no.

Seidenberg IM, Anderson DG, Becker RA. 1986. Validation of an in vivo developmental
toxicity screen in the mouse. Teratogen Carcinogen Mutagen 6:361-374.

                           External Review Draft--Sutfile — April 2O02                         27

Shacklette HT, Boerngen JG.  1984. Element Concentrations in Soils and Other Surficial
Materials of the Conterminous United States. U.S. Geological Survey Professional Paper 1270.
Washington, DC: U.S. Government Printing Office.

Sipes IG, Gandolfi AJ. 1991. Biotransformation of toxicants. In: Amdur MO, Doull J, Klaassen
CD, eds.  Casarett and Doull's Toxicology: The Basic Science of Poisons, 4th ed. New York:
Pergamon Press, pp. 88-126.

Stephen AM, Dahl WJ, Morgan DR. 1991. Effect of high sulfate drinking water on
gastrointestinal function and methane production in healthy subjects: a pilot study.  Proc Can
Fed Biol Soc Ann Meeting Pub No. 301 (Abst).

Stipanuk, M.H. 2000. Biochemical and Physiological Aspects of Human Nutrition.  Philadelphia,
PA: W.B. Saunders Co. p. 706.

Trembaczowski A. 1991. Sulphur and oxygen isotopes behavior in sulphates of atmospheric
groundwater system observations and model. Nordic Hydrol 22:49-66.

U.S. Environmental Protection Agency (U.S. EPA). 2001. Contaminant Candidate List
Preliminary Regulatory Determination Support Document for Sulfate. Office of Water,
Washington, DC. EPA/815/R-01/015.

U.S. EPA. 1999a. Health Effects From Exposure to High Levels of Sulfate in Drinking Water
Study.  Office of Water,  Washington, DC. EPA815-R-99-001.

U.S. EPA. 1999b. Health Effects From Exposure to Sulfate in Drinking Water Workshop.
Office of Water, Washington, DC.  EPA 815-R-99-002.

U.S. EPA. 1994. National Primary Drinking Water Regulations—Sulfate; Proposed Rule. Fed
Reg 59(243):65578-65582.

U.S. EPA. 1993.  Draft Drinking Water Health Advisory for Sulfates. Office of Water,
Washington, DC.

U.S. EPA. 1984. National Secondary Drinking Water Regulations. Office of Water, Washington
DC. EPA570/9-76-000.

Veenhuizen MP, Shurson GC, Kohler EM. 1992. Effect of concentration and source of sulfate
on nursery pig performance and health. J Am Vet Med Assoc 201:1203-1208.

Weitering JG, Krijgsheld KR,  Mulder GJ. 1979. The availability of inorganic sulphate as a rate
limiting factor in the sulphate conjugation of xenobiotics in rats? Sulphation and
glucuronidation of phenol. Biochem Phannacol 28:757-762.

Wetzel RG. 1983. Limnology, 2nd ed. Philadelphia: Saunders College Publishing, p. 767.
                           External Review Draft—Sulfate—April 2002                         28

World Health Organization (WHO). 1996. Guidelines for Drinking-Water Quality. Vol. 2:
Health Criteria and Other Supporting Information. Geneva: World Health Organization.

WurznerHP.  1979.  Exposure of rats during 90 days to mineral water containing various
amounts of sulphate. Z Ernahrungswiss 18:119-127.
                           External Review Draft—Sulfate —April 2002                         29