TR-540-61F
FINAL DRAFT FOR THE DRINKING WATER
CRITERIA DOCUMENT ON FLUORIDE
April 9, 1985
Prepared Under Contract 66-01-6750
bv
ICAIR
LIFE SYSTEMS, INC.
Cleveland, OH 44122
for
Criteria and Standards Division
Office of Drinking Water
U.S. Environmental Protection Agency
Washington, DC 20460
-------
TABLE OF CONTENTS
PACE
LIST OF FIGURES iv
LIST OF TABLES t . v
X. SUMMARY 1-1
II. PHYSICAL AND CHEMICAL PROPERTIES II-1
A. Physical rnd Chemical Properties 11-1
B. Manufacture and Uses 7.1-2
C. Summary 11-5
III. TOXICOKINETICS III-l
A. Absorption . , III-l
B. Distribution 111-5
C. Metabolism III-8
D. Excretion III-8
E. Bloaccumulation and Retention III-l1
F. Summary 111-19
IV. HUMAN EXPOSURE IV-1
A. Exposure Estimation IV-1
B. Drinking Water Exposure IV-2
C. Dietary Exposure IV-5
D. Air Exposure ..« IV-7
E. Summary IV-7
V. '&Ytit&lVWALS .................. V-l
A. Acute Tbxlcity ..................... V-l
B. Chronic Toxlcity .................... V-4
1. Bone ....................... V-6
2. Teeth ....................... V-10
3. Reproduction ................... V-15
4. Growth ...................... V-23
5. Kidney ...................... V-24
6. Cardiovascular System ............... V-25
7. Thyroid ...................... V-27
C. Teratogenlcity ..................... V-27
D. Mutagenicity .......... . ........... V-27
continued-
i-b
-------
Table of Contents - continued
PAGE
E. Carcinogenlclty V-31
F. Other V-33
C. Summary V-34
VI. HEALTH EFFECTS IN HUMANS " VI-1
A. Beneficial Effects VI-1
1. Teeth VI-1
2. Bone . VI-3
3. Cardiovascular VI-7
4. Hearing VI-8
5. Other VI-9
B. Acute Toxicity VI-10
C. Chronic Toxicity VI-11
1. Sensitivity to Fluoride VI-12
2. Bone VI-13
3. Teeth ' . . VI-17
4. Kidney VI-27
5. Growth . VIrSO
6. Cardiovascular System VI-30
7. Thyroid VI-32
D. Teratogenicity 4 VI-32
E. Mutagenicity VI-33
F. Carclnogenicity VI-33
G. Epidemiclogical Studies VI-34
1. Mortality Studies VI-34
2. Skeletal Effects Vl-36
3. Effects in Children VI-43
4. Other Studies VI-46
H. Summary VI-47
VII. MECHANISMS OF TOXICITY VII-1
A. Acute Effects VII-1
B. Skeletal Effects VII-1
C. Dental Effects VII-2
D. Summary VII-3
continued-
ii
-------
Table of Contents - continued
PAGE
VIII. QUANTIFICATION OF TOXICOLOGICAL EFFECTS VIII-1
A. Non-Carcinogenic Effects VIII-5
1. Short-Tenn Exposure VIII-5
2. Long-Term Exposure ..... VIII-7
B. Quantification of Non-Carcinogenic Effects VIII-20
1. ' One-Day and Ten-Day Health Advisory VIII-20
2. Adjusted Acceptable Daily Intake VIII-20
C. Carcinogenic Effects VIII-23
D. Existing Guidelines and Standards VIII-25
E. Special Considerations VIII-31
1. High Risk Populations . . . . VIII-31
2. Beneficial Effects VIII-31
3. Interactions VIII-37
4. Relative Source Contribution VIII-37
F. Summary VIII-38
IX. REFERENCES IX-1
ill
-------
LIST OF FIGURES
FIGURE PAGE
III-l Fluoride Btlances in Men During Five-Day Experimental
Period III-4
III-2 Concentrations of• F and Na in Blood after Intravenous
and Oral Administration of Radioactive Sodium Fluoride to
Lambs III-6
III-3 Relation Between Fluoride Concentrations in the Urine
of Humans and That in the Water Supplies Used . Ill-10
III-4 The Effect of Age on the Rate of the Increase of Fluorine
Concentration In the Femur of the Rat 111-15
V-l Calving Rate of Cows on Three Levels of Fluoride Intake . . V-18
VI-1 Relationship Between Fluoride Concentration of Municipal
Waters and Fluorosis Index for Communities with Mean
Annual Temperatures of Approximately 50*F (Midwest) and
70"F (Arizona) VI-22
iv
-------
LIST OF TABLES
TABLE PAGE
II-l Selected Fluoride Compounds and Properties 11-3
III-l F Distribution in a Lamb Killed Two Hours After
Ingestion 111-7
1II-2 The Effect of Added Dietary Increments of Fluoride Ion
(N*F) on Soft Tistut Fluoride Concentrations in Dairy
Cows 111-13
111-3 The Effects of Added Dietary Increments of Fluoride Ion
(NaF) on Bone Fluoride Concentrations in Dairy Cows .... III-l4
III-4 Fluoride Concentrations (Expressed as ppm in Ash) in
Dentine and Enamel at Different Levels of Fluoride
Ingestion . 111-17
IV-1 Estimated Population Exposed to Fluoride in Drinking Water
at Various Concentration Ranges IV-3
IV-2 Estimated Intake of Fluoride from Drinking Water IV-4
IV-3 Reported Daily Dietary Intake of Fluoride (exclusive of
water) IV-6
IV-4 Estimated Intake of Fluoride from the Environment by Adult
Males IV-8
IV-5 Estimated Intake of Fluoride from the Environment by 5-13
Year-Old Children '. IV-10
V-l Effects of Age on Toxiclty of Sodium Fluoride in Rats . . . V-5
V-2 Effects of Ingested Fluoride on Dairy Cattle Fed Various
Levels of Sodium Fluoride From 4 Months to 7.5 Years of
Age V-8
V-3 The Effects of Fluoride Fed as NaF on Various
Physiologic Responses in Heifers V-ll
V-4 Fluoride Concentration in Bones and Teeth of Sheep
Drinking Fluoride-Supplemented Water V-l6
V-5 Breeding Efficiency Over Four Breeding Seasons of Five
Groups on Different Levels of Fluorine Intake, With and
Without Added Dafluorinated Superphosphate V-17
V-6 Reproductive Performance of Hereford Heifers Exposed to
Dietary NaF for Nine Years V-21
V-7 Effect of Fluoride Exposure on Reproductive Performance
of Sheep V-22
V-8 Bone Fluoride and Chromosomal Aberrations in Bone Marrow
and Testis Cells in Mirj Receiving Water with Different
Fluoride Levels V-30
V-9 Mutageniclty of Sodium Fluoride in Microblal Systems:
Number of Responses Per Plate V-32
VI-1 Food and Nutrition Board Estimated Adequate and Safe
Safe Intakes of Fluoride VI-2
Vl-2 Correlation of Osteosclerotic Phases and Fluoride in
Bone Ash VI-16
VI-3 Dental Fluorosis Classification by H. T. Dean - 1934 . . . VI-19
VI-4 Dental Fluorosis Classification by H. T. Dean - 1942 . . . VI-20
continued-
-------
List of Tables - continued
TABLE
PAGE
VI-5 Percentage of Children by Fluorosls Diagnosis for
Each Fluoride-Temperature Zone VI-23
VI-6 Relationship Between Fluoride Levels in Drinking Water
and Incidence of Moderate and Severe Dental Fluorosis
in Texas Children (Age 7 to 18 years) VI-25
Vl-7 Relationship of Drinking Water Fluoride Levels to
Dental .Fluorosis and Caries Reduction in Illinois
Children VI-26
VI-8 Incidence of Abnormal Clinical Findings, 1943-1953 .... VI-38
VI-9 Prevalence of Abnormal Laboratory Findings, 1943-1953
(Participants Residing ir Study Area for the Ten-Year
Period) s . tfI-39
VIII-1 Summary of Moderate and Severe Dental Fluorosis in
Children VIII-11
V1II-2 Maximum Contaminant Levels VIII-26
VIII-3 Food and Nutrition Board Estimated Adequate and Safe
Intakes of Fluoride VIII-27
vi
-------
I. INTRODUCTION
The fluoride ion (F ) is ubiquitous and occur* In igneous and sedimentary
rock, soil, surface water, sea water and air. Elemental fluorine (F ) is a
pale yellow, acrid gas at 20°C, with a freezing point of -219.6'C and a
boiling point of.-188.2°C. Fluorine Is highly reactive; however, the fluoride
ion occurs naturally in combined, mineral forms such as fluorspar,
fluorapatite and cryolite as well as in aluminum, calcium and magnesium salts.
Industrially, -fluorspar is treated with sulfuric acid to produce hydrofluoric
acid (HF). the intermediate from which other fluorine compounds, such as
sodium fluoride (NaF), are prepared. Sodium fluoride is used commercially in
fluxes, for drinking water fluoridation, in tablets and topically applied
preparations for the prevention of dental caries and for scrubbing HF from
fluorine. Sodium fluoride is occasionally used as an insecticide and as a
wood preservative.
Approximately 97Z of ingested fluoride Is rapidly absorbed from the
gastrointestinal tract of the rat and the human. The absorbed fluoride is
distributed throughout intracellular and intercellular spaces by the blood.
Although appreciable quantities of fluoride are not stored in soft tissue, its
rapid uptake and bioconcentration in bone and teeth tre functions c^ both
concentration and duration of exposure. Concentrations in bone increas.e with
increasing age. Absorbed fluoride is usually excreted in urine or deposited
as fluorapatite in calcified tissues. Under steady-state intake conditions,
the urinary concentration of fluoride in adults tends to approximate the
1-1
-------
fluoride concentration of the drinking water. Fluoride nay be excreted
through perspiration In hot environments.
Drinking water, food and air are the main sources of fluoride exposure
for hunans. In general, a 70-kg adult male takes In up to 1.2 vg F/day In
air, 0.2 to 0.8 mg F/day in food and 0.2 to 2.0 mg F/day in water. On a per
body weight basis these values are 1.7 x 10~ ng/kg/day for air, 2.9 x 10~ to
1.1 x 10~2 ag/kg/day for food and 2.9 x 10"3 to 2.9 x 10"4 mg/kg/day for
water. Thus, compared to food and water, the contribution of fluoride by air
is negligible. Under typical exposure conditions H.O eg F/L), adult males
consume 72Z to 91Z of their fluoride intake via drinking water; for five- to
thirteen-year-old children the range of fluoride intake via drinking water is
64Z to 97Z. (At exposures greater than 2.0 mg F/L, drinking water accounts
for over 90Z of the exposure for both groups). On a per body weight basis,
five- to thirteen-year-old children consume 1.4 times as much fluoride via
drinking water as adult males, and newborn, formula-fed infants consume more
than eight t* .es as much as adults.
Acute lethality of NaF in animals varies with rout* of administration,
age and sex. In mice the oral LD.Q was 46.1 mg F/kg compared to an
intravenous LD5Q of 23.0 mg F/kg. The approximat • intraperitoneal LD,. of NaF
in adult rats is 26 mg F/kg. Young rats (less than seven months of age) and
specifically young male rats appear to be resistant to NaF toxicity. Acutely
toxic -doses of fluoride in rats occasionally resulted in fatal polyuria, but
100 mg F/L in drinking water did not cause renal injury.
1-2
-------
Cattle tolerated 27 ppm (equivalent to 0.64 mg F/kg/day) in the diet
without observable deleterious effect. Cattle chronically exposed to 49 ppm
fluoride in the ration (equivalent to 1.17 mg F/kg/day) shoved skeletal and
dental fluorosis. Sheep were less sensitive than cattle to the chronic
effects of fluoride. Growth in most species was unaffected by dietary
concentrations of 100 ppm or less; however, growth in cattle appeared to be
slightly affected at 40 ppm. Cardiovascular effects were observed in dogs at
9 mg F/kg or higher. At concentrations of 50 mg/L or below of fluoride in
drinking water, no structural or functional changes in the thyroid have been
observed in animals. No conclusive evidence has been found to indicate that
fluoride is mutagenic or carcinogenic either in vitro or In vivo.
The beneficial effects of fluoride on human health have been demonstrated,
both in terms of general health and in the treatment of specific diseases.
Fluoride Ingested during childhood results in a marked reduction of dental
caries. Similarly, fluoride has found application in stimulating substituted
bone growth in patients with osteoporosis. The daily intaka levels considered
to be protective against both dental caries and possibly osteoporosis are
established by age category with 1.5 to 4.0 mg/day (0.7 to 2.0 mg/L in
drinking water) the range for adults. Fluoride has also been suggested to
have beneficial effects on the cardiovascular system (reduced aortic
calcification) and hearing (stabilization of patients with active
otospongiosis).
Incidences of human poisoning from NaF have been reported. The estimated
lethal dose for humans is 70 to 140 mg/kg. Hypothetical relationships between
1-3
-------
mongolism and sensitivity to fluoride, *|fwjll as exposure to fluoride and
Ct-ncer Incidence have been reported but not substantiated. Persons with renal
.insufficiency nay be at increased risk to the toxic effects of fluoride.
Delayed skeletal maturation has been reported in children exposed to veter
containing as little as 3.6 ng F/L. These data, however, were derived front a
study of 11- to 15-year-old Tanzanian girls and several confounding factors
(i.e., warm climate, drinking water intake, nutritional status, incidence of
other diseases, etc.) prevent drawing any conclusions from this study for
application to the U.S. situation. Skeletal fluorosis (as measured by
increased bone density) has been observed in populations using drinking water
containing from 4 to 8 mg F/L. Severe skeletal fluorosis has occurred in both
adults and children who consumed drinking water containing 10 or more mg F/L.
Dental fluorosis occurs during the developmental period of enamel
formation. Epidemiological studies have shown that dental fluorosis is a
function of fluoride concentration, age, duration of exposure and possibly
ambient temperature (as related to water consumption). In nearly all
epidemlological evaluations* Including warm climates, objectionable (moderate
and severe) dental fluorosis is generally not observed in a significant
percentage of the population at drinking water concentrations below
2.0 mg F/L. There is no evidence of adverse health effects in humans
resulting from properly controlled fluoridatlon of domestic water supplies.
Fluoride interacts with bones and teeth by replacing hydroxyl or bicarbon-
ate radicals in hydroxyapatite to foim fluorohydroxyapatite. The presence of
1-4
-------
fluorohydroxyapatite increases the crystalline structure of the bone and
reduces its solubility. This is believed to increase the resistance of teeth
to dental caries and, possibly, to decrease the incidence of osteoporosis. As
bone crystal growth continues, fluoride is incorporated into the inner layers
of the crystals as well as on Che surface. The available evidence suggests
that dental fluorpsis results from the effects of fluoride on the epithelial
enamel organ. Specifically, several studies have shown that ameloblasts are
susceptible to fluoride. Dental staining ofteu accompanies fluorosis but does
not determine the degree of the fluorosis.
Populations that appear to be at increased risk from the effects of
fluoride are individuals that suffer from diabetes insipldus or some forms of
renal impairment. These high risk populations represent a relatively small
segment of the general population.
There is a general absence of suitable experimental or clinical data
following short-term oral exposure to fluoride for the derivation of one-day
or ten-day Health Advisory values for children and adults. The dose-response
for dental fluorosis, while subject to considerable variation at different
locations and in different populations, represents a steady Increase in
moderate and severe dental flaorosis with increasing fluoride concentration in
the drinking water. It is generally observed that the incidence of moderate
and severe dental fluorosis begins to impact a marked segment of the
population when the drinking water concentration approaches and exceeds 2.0-srg
F/L. One recent study of children suggested that the maximum protection from"
dental caries was achieved when drinking water contained approximately 2 mg F/L.
1-5
-------
Various studies and reviews indicate that the no-effect levels for the
initial signs of skeletal fluorosis (increased bone density) in adults appear
to be at drinking water concentrations between 3.0 and 8.0 mg F/L. Protection
of human health from this effect is believed to be achieved at 4.0 mg F/L with
an adequate margin of safety. There is no valid evidence to classify fluoride
as a potential carcinogen.
The National Academy of Sciences has estimated an adequate and safe total
intake of fluoride ranging from 0.1 to 0.5 mg/day for infants (less than six
months old) to 1.5 to 4.0 mg/day for adults. These estimates are considered
protective gainst dental caries ac.d possibly osteoporosis.
1-6
-------
II. PHYSICAL AND CHEMICAL PROPERTIES
A. Physical and Chemical Properties
Elemental fluorine is highly reactive. Fluorine is a pale yellow, acrid
gas with a freezing point of -219.6°C and a boiling point of -188.2'C (Weast
1980). However, fluorine is widely disseminated in ionic or combined forms.
The terms "fluorine" and "fluoride" are both used in the general literature to
refer to combined forms of fluorine.
The fluoride-containing minerals fluorspar, fluorapatlte and cryolite are
essentially insoluble in water. They have very high melting aud boiling
points and very low vapor pressures (Drury et al. 1980).
There are hundreds of ionic compounds of fluorine. Some commercially
important ionic fluorides are the sodium, calcium, aluminum and magnesium
salts; these have characteristically high melting and boiling points. Sodium
fluoride is a white crystalline powder with a melting point of 993°C and a
boiling point of 1695°C. This compound is only minimally soluble in water
(4.22 g/100 mL at 18°C) (Veast 1980). Aluminum, calcium and magnesium
fluorides are also only sparingly soluble in water.
Hydrogen fluoride is a colorless liquid or gas with a boiling point of
19.5°C and a fieezing point near -83°C (Weast 1980). Hydrogen fluoride is
highly soluble in water and fumes strongly in contact with the atmosphere.
II-l
-------
Hydrogen fluoride has a high vapor pressure (17.8 psla at 25°C) and its liquid
density is 0.9576 g/cn3 at 25'C (Gall 1966).
Silicon tetrafluoride is a colorless gas with a melting point of -90'C, a
boiling point of -86*C (Veast 1980) and an odor reminiscent of hydrogen
chloride (Windholz et al. 1976). It reacts with vater to form fluorosiliclc
acid (HjSiFg), which Is very soluble in water.
Fluorine also combines covalently with organic compounds. There are
thousands of known fluorine-containing organic compounds, but few of these
occur naturally. The chemical and physical properties of many of these
compounds differ greatly from their hydrocarbon counterparts, mainly because
of the stronger carbon-fluorine bond (Drury et al. 1980).
Table II-l summarizes the properties of most of the naturally occurring
fluorine compounds and various fluorine compounds used industrially.
B. Manufacture and Uses
Of the three major fluoride-containing ores (fluorspar, phusphate rock
and fluorapatite), only fluorspar is used corjmercially as a source of fluorides.
Generally, fluorspar is treated with sulfuric acid to produce hydrogen fluoride.
Hydrogen flvoride is the most important manufactured fluoride and is the
intermediate from which other fluorine compounds are prepared. About 292,000
metr.ic tons of HF were produced in the United States in 1977 (Drury et al.
1980). Approximately 40Z was used to manufacture aluminum, 37Z was converted
II-2
-------
Table ll-l Selected Fluoride Compounds and Properties
Snhatance
fluorine
Mineral Fluarldea
Pluorapar
Cryolite
Pluorapatlte
Ionic Fluor Idea
AltMlnuM Pliinrlde
Calcltm Fluiirlde
NaRnealim Fluoride
Sodlim Fluoride
Covalent Fluoride*
Hydrogen Flunrlde
Silicon TetraMuoride
Fluoroalllrlc Arid
Omanlc Fluarldea
Trlchlornf luorovethan*
Dichlnrodif Iuoro«*rh*ne
Tet ra f luoroiwt hane
Fluoronefhane
Pluororrhan*
Pluoroethene
(vinyl (luorlde)
Tet ra f luoror thv I ene
He*a f 1 iniropropene
Pormila
',
CaP,
)NaF Air.
3Ca*P04)J-C«F3
AIF.
C«F*
MgF,
NaF?
nr
SIP
M S»P
* n
CCI.P
CCljF
cr*
CH!F
r. * F
CjH*F
* f
C2%
e 9
C3F*
Color
yellow
white
white
..—
white
white
white
white
colnrlean
cnlnrleaa
—
— .
-—
...
...
— -
MeltlnR
point ro
-219.62
1*07
1009
1*30
10*0 .
1*02
1225
•03
-83.7
-90
-Ill
-158
-in*
-U2
-1*3
-160
-1*2.5
-156.2
Roll ln«
Point (*C)
-1HB.I*
7511
1791
7SI3
72*0
1*9%
19.5
-86
ION. 5
MOT no In)
23.0
-29. R
-17*. 0
-7"./ '
-17..
-77.0
-76.3
-29.*
Oenalt*
fa/ea)1)"
1.90 (aolld)
1.IH (aotld)
7.97 (aolld)
1.1ft (not Id)
1.07
l.lft
1.0
7.55ft
1.0015 fO'C)
l.ft* (-95*C)
1.77* (Z5'C)
(10Z aoln)
.««7
.111
.117 (-We)
.5«7 (-71T)
.519 r-7*'O
0.675 (in'C)
.519 (-76*C)
.5113 (-*«"C)
Solublllrv
In Hater
(•/ion Mi)
react
0.001* (In'C)
0.0*7 (25*C)
Inaoluble
0.559 (25'C)
o.nni* dft'o
0.007* (I8'C)
4.22 (|*"C»
•leclblc
react*
verv inluhlr
0.011 vtX (25*C)
5.7 (26'C)
0.0015 wrr (75'c)
— — —
--.
0.9* (BO'C. 500 pala)
:::
Vapor
Preaenre
r— Mel
7*« C-M7.9)
7.* (7IOO'C)
1.9 (|009'O
7*0 (>%)7)
7.* (?IPO*C)
760 (19. T)
7*0 (-9i.il)
7201 (75'CI
(101 no In)
7*0 (71.7)
7*0 -79.11)
7*0 -127. 7)
7*0 7*. 7)
7*0 -17.0)
760 -72.7)
716.2 pala (O.'C)
*Denelty Riven at 25"C unleea noted otherwtee.
Adapted tiam I>mr» rt al. 19110 fbitiird on Ho: ton
Perry ami Chi Icon 1973.
1961; Utndholt et al. 147*; Bvrnt 19*6; Vr
-------
into fluorocarbon compounds, 7Z was used in processing uranium, 5Z was used in
alkylatlon catalysts in petroleum refining, 4Z was used in manufacture of
fluoride salts and 4Z was used in stainless steel pickling operations.
Smaller quantities were used as fluxes in metal casting, welding and brazing
operations; as etching agents in glass and ceramics industries; as cleaners in
metal finishing processes; in pesticides; in fluoridation of water supplies
and in toothpaste and other dental preparations (D- .ry et al. 1980).
The fluorination of organic compounds amounted to about 108,000 metric
tons in 1977 (Drury et al. 1980) and is the greatest single end use of
fluorides. Hydrogen fluoride is used in the synthesis of dichlorodifluoro-
methane, trichlorofluoromethane, tetrafluoromethane, tetrafluoroethylene,
•
vinyl fluoride and hexafluoropropene. These compounds are used chiefly for
aerosol propeHants, refrigerants and fluorinated plastics. Small quantities
of other fluorocarbons find specialized uses as inhalation anesthetics, fire
extinguishing agents, cleaners and degreasers (Drury et al. 1980).
Sodium fluoride is widely used in fluxes, for fluoridation of water
supplies, in dentifrices and other dental preparations and for scrubbing HF
from fluorine. It is also occasionally used as an insecticide and a wood
preservative (Drury et al. 1980).
Fluorosilicic acid (H2SiFfi) is sometimes used in hardening cement,
preserving timber, manufacturing enamels and preserving oil pigments. A small
amount of sodium fluorosilicate is used as insecticide (Drury et al. 1980).
11-A
-------
Fluorides are emitted into the atmosphere as a result of the manufacture
of fertilizer and phosphorus from rock phosphate, the operation of aluminum-
and steel-producing furnaces, the manufacture of brick and tile products and
the combustion of coal. The fluorides are generally in the form jf HF,
fluorine, boron trifluoride, H-SiF,, sodium fluorosilicate, aluminum fluoride,
calcium fluoride, lead difluoride, fluorapatite, silicon tetrafluoride and
fluoride partlculates (Drury et al. 1980).
Liquid wastes containing HF or fluoride ion (F~) are generated in
appreciable quantities by glass manufacturers,• pesticide and fertilizer
producers, steel and aluminum makers, met* processing industries and •
inorganic chemical producers (Drury et al. 1980).
G. Summary
Fluorine is highly reactive. It usually occurs as ionic or covalently
#
bonded fluoride. The most common chemical forms are fluorspar, fluorapatite,
cryolite, HF, H.SiFg and fluorocarbons. Most naturally occurring forms of
fluoride are insoluble or only slightly soluble in water.
The main industrial'source it fluoride is.the mineral fluorspar.
Hydrofluoric acid is made from fluorspar and is used primarily in the produc-
tion of aluminum and fluorocarbon compounds.
II-5
-------
Fluoride occurs in most rocks and soils in concentrations of 200 to 1,000
ppm. It is a normal constituent of most natural waters in concentrations up
to 0.3 ppm. The water supplies of most major United States cities naturally
contain 0.02 to 0.1 ppo fluoride. Groundwater fluoride concentrations vary
with the type of rock the water flows through but usually do not exceed 10
ppn.
Fluorides occur in the atmosphere from natural and industrial emissions.
Host atmospheric fluorides are washed out by rainfall which may contain 0.02
to 14 ppm fluorid*..
11-6
-------
III. TOX1COKINETICS
A. Absorption
Soluble fluorides are rapidly absorbed from the gastrointestinal tract of
animals and humans. Zipkin and Likins (1957) demonstrated that when rats were
administered, by intubation. 0.2 mg F as NaF in solution, 862 of the dose was
absorbed in 90 min. In a more definitive experiment by Zipkin and Likins
(1957), single doses of 1.7 to 1.8 mg F/kg of body weight were administered by
stomach tube to groups of ten male rats weighing 110 to 120 g. After 30
minutes the rats were sacrificed and the percentage of the dose remaining in
the gastrointestinal tract was determined. The percentage of the dose
absorbed was estimated by subtracting the percentage of the dose remaining
from 1002. The fluoride was administered in several chemical forms. The
readily ionizable compounds NaF, Na.SiF,, Na.PO.F and SnF. (all administered
in solution) were absorbed to the extent of 50Z, S1Z, 432 and 502,
respectively during the 30 minute period. Compounds not releasing ionic
fluoride were absorbed more extensively. Absorption of KPF, and KBF, was 772
and 762, respectively.
18
Ericsson (1958) adminir.tered 0.08 mg F (labelled with F) per kg of body
weight to groups of six male rats weighing 269 to 289 g. The fluoride was
administered by stomach tube as 5 mL of an aqueous solution containing 1 ppm F
(1 mg F/L). The percentage of the dose per mL of heart blood was maximal
(approximately 0.182/mL) 45 minutes after administration. Analysis of the
gastrointestinal tract for fluoride remaining after eight to ten hours showed
that 892 to 902 had been ab.'.orbed.
-------
A number of studies are available describing Che absorption of fluoride
in humans. For example, Ekstrand et al. (1977) measured plasma fluoride
concentrations after oral administration of 4.5, 6.0 or 10.0 mg F to eight
subjects (six males and two females) 23- to 29-years-old. The fluoride was
administered as tablets or In gelatin capsules with water. In all cases, the
maximum plasma fluoride concentrations occurred within 30 minutes of
administration. In a similar study, 0.5 mg F as NaF tablets was ingested with
water by five children three- to four-years-old and weighing 15.5 to 17.8 kg
(Ekstrand et al. 1983). Plasma fluoride concentrations were measured at 0, 30
and 60 minutes after administration. As in adults, maximum plasma fluoride
concentrations were observed 30 minutes after ingestion.
The concordance among these studies suggests the rat is an adequate model
for the short-term pharmacokinetics of fluoride in humans. In summary,
soluble fluoride ingested by the human is absorbed from the gastrointestinal
tract at least to the extent of 97Z. Absorption is rapid with maximum plasma
fluoride concentrations attained in approximately 30 minutes in the human as
well as in the rat.
Carlson et al. (1960a) studied the absorption of fluoride in humans.
18
Subjects consumed 1 mg fluoride (as NaF contain! .g F) in 250 mL water.
Maximum plasma concentrations (0.13 to 0.17 mg/L) were reached within 60
18
minutes. At 150 minutes f was no longer present in the stomach. The
gastrointestinal absorption of fluoride in five men 19- to 27-years-old was
studied by McClure et al. (1945). Fluoride balances were determined over
five-day periods while their normal diets (containing 0.50 to 0.90 mg
III-2
-------
fluoride/day, or 0.007 to 0.013 ng/kg/day for a 70-kg individual) were supple-
mented with NaF in water, NaF in food, CaF_ in water, CaF. in food, bone meal
in food or cryolite in food. Figure 111-1 shows the fluoride balances
associated with the various forms of fluoride exposure. Fecal excretion
suggests that NaF in water and food and CaF. in water are extensively
absorbed, while fluoride in bonemeal and in mineral cryolite is less well
absorbed. The role of gastrointestinal secretion of fluoride was not
determined in this work.
The results of McClure et al. (1945) are similar to those found by
Largent (1960). Largent studied the gastrointestinal absorption of soluble
fluorides in human subjects. Soluble fluorides were administered in the
following manner:
NaF, 2 to 4Z in aqueous solution.
CaF. in aqueous solution and as the dry salt in capsules.
- Bone meal as a slurry in an aqueous mediup. and as the dry ma! eiial
in capsules.
Cryolite as a solid in capsules.
Finely powdered fluorapatite (rock phosphate) in capsules.
Complete fluoride balance data were collected. Total fluoride intake
during these studies Tanged from 3.49 to 22.3 mg/day (0.05 to 0.32 mg/kg/day
for a 70-kg individual). Normal dietary intake of fluoride during these
studies ranged from 0.4 to 0.8 mg/day (0.006 to 0.012 mg/kg/day). Approxi-
mately 962 to 972 of the fluoride from aqueous solutions containing NaF
III-3
-------
FLUORIDE
EXPOSURE
NaP - URBANA WATER •
NaF - IN FOOD
CaF2 - URBANA WATER3
MINERAL CRYOLITE
BONE MEAL
*
U.Q
ELIMINATION OF INGESTED FLUORIDE
PERCENT ELIMINATED VIA:
FECES URINE PERSPIRATION
'/////A
'//////////.
4.65
'//////
3.88
~y///,
PERCENT 0 10 20 30 40 50 60 70 8U 90 100 110
•Urbana water was the local tap water.
Adapted from McClure et al. (1945).
Figure III- 1 Fluoride Balances in Men During Five-Day Experimental Period
III-4
-------
or Ca?2 was absorbed. Absorption of fluoride fron calcium fluoride, bone
meal, cryolite and rock phosphate, administered as solids, was approximately
62Z, 37Z, 772 and 87Z. respectively.
B. Distribution
Fluoride added intentionally to drinking water supplies exists in solu-
tion as fluoride ion. (Feldman et al. 1957). Perkinson et al. (1955) stated
that the rate and pattern of removal of fluoride from the blood is similar to
— +2
that of ions such as chloride (Cl ) and calcium (Ca ), in contrast to sodium
(Na ), which soon reaches an equilibrium value (see Figure III-2). The
authors found that the initial rates of removal of fluoride from sheep and cow
blood (expressed as percentage of dose per minute), were 41Z and 32Z, of the
intravenously administered dose respectively. These data suggest a rapid
distribution of fluoride among the tissues of the body. In a lamb killed two
18
hours after ingestion •»£• NaF containing F, the absorbed fluoride was in fact
found to be widely distributed (Table II1-1).
j
Carlson et al. (1960a) indicated that distribution of fluoride is also
18
rapid in humans. Subjects consumed NaF containing F in water (250 mL at
1 mg/L). Epigastric (abdominal) counts were monitorcJ with a porcable
scintillation counter. One hundred and fifty.minutes after dosing, the
remaining epigastric counts were attributable to fluoride in the spine.
Counts in muscle (contracted biceps) started to decline 50 minutes after
Ingestion, until at 250 minutes they were nearly zero. In contrast, counts in
the femur had declined only 15Z from their maximum value (at 50 minutes) by
250 minutes.
1II-5
-------
0.1
100 200 300 400 500
TIME (minutes)
Disappearance ol "F and "Ni (ram Mood ol lamb alter
intravenous admlnliilralion
Adapted from Perklnson et al. (1955).
z
Ul
10
Ul
a.
10
200 400
TIME (minutes)
GOO
Blood "F and "Na alter oral administration ol radioactive
sodium lluorlde lo lamM
IB 22
Figure III-2 Concentrations of F and Na in Blood after Intravenous
and Oral Administration of Radioactive Sodium Fluoride to Lambs
-------
18
Table III-l F Distribution in a Lamb Killed Two Hours After Ingestion8
Blood
Bile
Muscle
Spleen
Pancreat
Lymph noae
Liver
00061
00050
00015
0.00016
0.00033
0.00028
0.00033
Rib epiphysis 0.0048
Rib shaft 0.0018
Femur epiphysis 0.0046
Femur shaft 0.0008
Angle of mandible 0.0045
Molar tooth 0.0010
values in percentage of dose per gram fresh weight.
Adapted from Perkinson et al. (1955).
III-7
-------
C. Metabolism
Bone is formed when calcium and phosphorus are deposited on a collagen
matrix (Kay et al. 1964). The resultant mineral phase is known as hydroxy-
apatite and has the formula Ca.Q(PO,),(OH)?. Fluoride is believed to replace
the hydroxyl ion. (OH~) and possibly the bicarbonate ion (HCO ~) associated
with normal hydroxyapatite (Neuman et al. 1950, McCann and Bullock 1957). The
resultant material is called fluorohydroxyapatite, or simply fluorapatlte.
Kay et al. (1964) analyzed the crystal structure of hydroxyapatite using
neutron diffraction and X-ray diffraction techniques. Their data indicated
that the OH of hydroxyapatite is in a less stable configuration than the F~
of fluorapatite. This might explain the ability of fluoride to harden bone
and to increase the resistance of teeth to caries.
D. Excretion
The principal route of excretion of ingested fluoride is via the urine.
.t
as has been demonstrated in a variety of species. However, in species other
than man, there is little published data relating fluoride concentrations in
drinking water and in urine over prolonged periods of time. Several studies
on livestock b-ive been reported. For example, Shupe et al. (1963) fed pairs
of dairy cattle rations containing 12 (normal), 27, 49, or 93 ppm fluoride on
a total dry matter basis from about 4 months to 7.5 years of age. Each pair
of animals also received one of two levels of calcium-phosphorus mineral and
one of two levels of concentrate mix. The total population consisted of
TTT_R
-------
16 pairs of cattle. Urinary fluoride excretion was measured for each pair of
animals at intervals from 89 to 2,396 days on experiment. The data indicated
that urinary fluoride concentrations were highly related to the fluoride
ingested. Also, as time on experiment increased, and therefore skeletal
stores of fluoride increased, the proportion of absorbed fluoride deposited in
the bone decreased and the proportion excreted in urine increased. Therefore,
for a given level of fluoride Intake, urinary concentration of fluoride
increased with increased duration of intake.
Figure II1-3 illustrates the strikingly linear relationship between the
concentration of fluoride in drinking water and that in urine when individuals
are constantly exposed to fluoride. Zipkin et al. (1957) demonstrated the
rapidity of urinary excretion of ingested fluoride. The authors demonstrated
that when 5 mg F (as NaF) was ingested in a glass of water, 20* (1.6 mg F) of
the fluoride appeared in the urine within three hours. After eight hours, the
rate of urinary excretion of fluoride returned to the pre-exposure level.
Machle and Largent (1943) studied the excretion of fluoride in a human
subject; 6 to 19 mg fluoride/day was added to the diet (equivalent to 0.19 to
0.6 mg/kg/day for a 70-kg adult). Over this range of intake it was found that
about half of the absorber, fluoride was excreted in the urine. 'Jsing fluoride
labeled with F, Carlson et al. (1960a) demonstrated that in two human
subjects 51Z and 63Z of the fluoride filtered by the kidney was reabsorbed.
In contrast, at least 99.5Z of filtered chloride is reabsorbed in a normal.
individual. The relative inefficiency of the human kidney in reabsorbing
filtered fluoride accounts for the rapid urinary excretion of fluoride.
II1-9
-------
02 4 6 8 10 12 14 16 18
FLUORIDE CONCENTRATION IN URINE (ppm)
oMcClure&Kinser(1944)
x Urgent (1961)
A Ukins, McClure & Steere (1956)
o.+ Zipkin et al. (1956)
Adapted from WHO (1970)
20
Figure III-3 Relation'Between Fluoride Concentrations in the
Urine of Humans and That in the Water Supplies Used
ITI-10
-------
In climates with warn temperatures a significant fraction of total
fluoride excretion may be via perspiration. McClure et al. (1945) measured
excretion of fluoride In the perspiration of individuals maintained for eight-
hour periods in "comfortable" conditions (temperature 84 to 85°F, relative
humidity 491 to S2Z) and In "hot-moist" conditions (temperature 100-101*F,
relative humidity 66Z to 70Z). Under the "comfortable" conditions about 25Z
of the fluoride excreted per day appeared in perspiration. Under the
"hot-moist" conditions up to 46Z of the excreted fluoride was In perspiration.
Hodge et al. (1970) have pointed out that the importance of this route of
excretion under different climactic conditions cannot yet be stated due to a
lack of information.
•
£. Bloaccunulatlon and Retention
In the body, the only significant covalent interaction of fluoride is
vith the hydroxyapatite in bones and teeth. Consequently, soft tissue
concentrations of fluoride rise transiently after ingestion of fluoride
(Carlson et al. 1960b, Hein et al. 1956), but long-term retention and
accumulation are confined to calcified tissue (Wagner et al. 1958). Suttle et
al. (1958) measured soft tissue concentrations of fluoride in 20 heifers
exposed to 0 to 50 ppm (equivalent to 1..4 mg/kg/day) added iluoride in their
ration for 5.5 years. Control animals (0 ppm added fluoride in their ration)
had soft tissue fluoride concentrations from 2.1 ppm (thyroid) to 5.3 ppm
(adrenal) and an average whole blood fluoride concentration of 0.34 ppm.
Heifers exposed to 50 ppm (equivalent to 1.41 mg/kg/day) added fluoride in
TTT-11
-------
their ration had soft tissue fluoride concentrations from 4.2 ppo (pancreas)
to 19.3 ppm (kidney), dry weight, and an average whole blood fluoride
concentration 0.67 ppn. It should be noted that the kidney Is an important
route of excretion for fluoride. Data on soft tissue storage of fluoride are
summarized in Table I11-2.
Twenty heifers exposed to 0, 20, 30, 40 and 50 ppm added fluoride
(ingested as NaF; equivalent to 0, 0.53, 0.86, 1.03 and 1.36 mg/kg/day,
respectively) in their ration for 5.5 years showed progressive increments in
bone fluoride concentration corresponding to the amounts of added fluoride
(Suttie et al. 1958). Some samples of bone from animals exposed to 50 ppm
added fluoride (equivalent to 1.36 mg/kg/day) in their ration contained more
than 8,000 ppm fluoride on a fat-free dry weight basis. The data from this
report are summarized in Table III-3.
The deposition of fluoride in the skeleton of female Roltzman rats was
studied by Suttie and Phillips (.1959). Three age groups, weanlings, young
adults (10-weeks-old) and mature rats (18-weeks-o.ld). were started on a diet
containing 0.1Z NaF. Rats were sacrificed at various times up to 113 days
after the start of the exposure tc fluoride. At sacrifice, femurs were
removeJ and analyzed for fluoride. The resu'ts of this study are summarized
in Figure III-4. These data show that -fter an Initial phase of rapid .uptake
of fluoride into bone, the rate of upta!.e gradually diminishes. Moreover, the
concentration of fluoride in bone at the end of the experiment was inversely
correlated with the initial ages of the rats. The authors believed that this
111-12
-------
Table 111-2 The Effect of Added Dietary Increments of Fluoride Ion
(NaF) on Soit Tissue Fluoride Concentrations in Dairy Cows
Tissue Levels
Lot F Added Cow Heartb Liverb Kidneyb Pancreasc Thyroid0 Adrenal0 Bloodd
(ppn) Ho.
1 0
t
11 20
111 30
IV 40
V 50
VI 50 +
CaCO,
3
2
3
4
Av.
5
6
7
8
Av.
10
11
12
Av.
13
14
15
16
Av.
17
19
20
Av.
22
23
24
Av.
1.8
3.3
1.7
2.3
2.7
4.4
2.5
4.0
3.4
2.7
4.7
3.0
3.5
3.3
5.5
4.5
2.9
4.0
5.0
3.2
6.3
4.6
3.8
4.6
5.6
4.6
3.1
1.9
1.9
2.3
2.4
2.1
2.3
3.9
2.7
2.5
4.6
5.1
4.1
5.6
3.3
3.4
3.0
3.8
2.3
2.1
4.8
3.6
3.5
2.7
2,8
3.0
3.1
2.9
4 4
3.5
7.3
6.0
8.3
12.8
8.6
12.5
9.1
10.5
10.7
19.7
20.4
7.8
16.0
16.0
13.7
15.4
28.9
19.3
11.1
7.7
8.4
9.0
1.7
4.0
2.0
2.8
1.4
2.6
3.2
1.7
2.2
2.0
5.0
3.5
3.5
'3.0
5.1
-
3.4
3.8
4.5
4.1
4.0
4.2
3.6
3.5
3.0
3.4
0.6
2.7
2.9
2.1
2.9
7.0
6.6
11.2
6.9
2.4
4.1
4.1
3.5
4.9
5.0
—
7.6
5.8
5.2
12.2
4.4
7-3
4.2
4.9
9.0
6.0
2.0
11.9
2.2
5.3
3.8
3.3
3.5
3.3
3.3
3.3
2.5
3.5
-
8.8
4.7
8.7
4.1
6.4
4.2
4.1
4.1
0.49
0.32
0.22
0.34
0.34
0.59
0.84
0.39
0.54
0.45
0.89
0.30
0.55
0.94
0.69
-
0.31
0.66
4 0.38
0.78
0.84
0.67
1.10
0.68
0.88
0.89
f All values in ppm.
.Dry weight.
, fat-free weight.
ole blood.
Adapted from Suttie et al. (1958).
111-13
-------
Table 111-3 The Effects of Added Dietary Increments of Fluoride Ion
(NaF) on Bone Fluoride Concentrations in Dairy Cows
II
20
III
30
IV
40
50
Bone Concentrations
Lot
I
F Added
(ppm)
0
Cov
No.
2
3
4
Av.
Meta-
carpal
593
878
463
645
Meta-
tarsal
482
647
436
522
Frontal
647
701
592
645
12th
Rib
694
635
703
677
5
6
7
8
Av.
10
11
12
Av.
13
14
15
16
Av.
17
19
20
Av.
2720
2770
2170
2660
2580
4180
3780
4120
4030
5520
5840
4510
3740
4900
7610
5470
4050
5710
2610
2990
2610
3030
2810
4090
3310
4200
3870
5280
5380
4710
3180
4640
7800
4920
4360
5690
3200
3340
3110
3430
3270
4540
4920
4800
4750
6180
6010
4730
5640
8260
6800
6700
7250
3290
4770
4030
3910
4000
4900
5970
5280
5380
7030
7070
4100
4860
5770
9000
8100
6870
7990
fConcent-ations expressed as ppm flouifide.
Dry fat-free weight.
Adapted from Suttie et al. (1958).
111-14
-------
10*.
UJ
z
0.
0.
Z
LU
CO
O A
A
• = WEANLINGS'
O = YOUNG ADULTS*
A - MATURE RATS'
10
5 ^ 10 20
40
60
80
100
120
DAYS
•Eath point is the mean of three animals
Adapted from Suttie and Phillips (1959)
Figure III-6 The Effect of Age on the Rate of the Increase of Fluorine
Concentration in the Femur of the Rat
111-15
-------
difference was related to Che surface area per mass of bone which could be
reached by body fluids. Thus, adult bone is more fully mineralized while the
infant bone is new, hydrated and available for fluoride exchange.
Studies of humans have also shown that soft tissues are not important
sites of storage for fluoride. For example, Smith et al. (1960) examined 122
tissue samples from autopsies of 23 individuals who had lived in an area where
drinking water contained from 1.0 to 4.0 ppm fluoride. No significant accumula-
tion of fluoride in heart, liver, lung, kidney or spleen was found. Fluoiide
concentration in the aorta did increase with age, probably because of
increased calcification of the aorta with age rather than increased exposure
to fluoride.
In contrast to soft tissues, teeth (McClure and Likins 1951) and bone
(Smith et al. 1953, Suttie et al. 1958) readily take up fluoride. Table III-4
summarizes data from several reports on fluoride concentrations in human
teeth. Fluoride concentrations in teeth are a function of dose and duration
of exposure. Jackson and Weidmann (1959) determined that the rate of increase
of fluoride concentration in human teeth decreases with increasing age. This
study demonstrated that in West Hartlepool. an English city with a drinking
water fluoride concentration of 2.0 ppm, th.- fluoride content of tooth enamel
varied with age in the following manner: 5- to 11-years-old, 17±0.9 mg
fluoride/100 g enamel; 20- to 35-years-old, 32+0.28 mg/100 g; 50- to
73-years-old, 37±7.5 mg/100 g.
111-16
-------
Table 111-4 Fluoride Concentrations (Expressed as ppm in Ash) in Dentine and
Enamel at Different Levels of Fluoride Ingestion
Fluoride in Ash (ppm)
Reference
McClure & Likins
-(1951)
Jackson & Weidmann
(1959)
Jenkins & Speirs
(1953)
Brudevold ,
Steadman &
Smith (1960)
?A11 data for humans
Age Dose
Adult 0.1
7.6
20-49 yr <0.5
20-35 yr 1.2
20-35 yr 1.9
Adult <0.25
1.4
2.0
20-29 yr 0.1
1.0
3.0
5.0
and lifelong exposures.
surface
—
—
590
960
1310
571
889
1930
3370
Enamel
interior whole
86
658
108
180
320
80
110
270
48
129
152
570
Dentine
whole
332
1958
508
922
1290
—
-- r
Adapted from WHO (1970).
111-17
-------
Zipkin et al. (1956) studied fluoride concentrations in the bones of
humans exposed to fluoride in drinking water at concentrations from 0.1 ppm
(Nev York City) to 4.0 ppm (Lubbock, Texas). The authors found a linear
relationship between the concentration of fluoride in water and the concen-
tration in bone. This study was not analyzed to account for the amount of
time that individual subjects had lived in the designated areas.
The concentration of fluoride in human bone also increases with duration
of exposure. Smith et al. (1953) fo'-nd a linear relationship between age in
years and concentration of fluoride in bone ash from lifetime residents of an
area with a drinking water supply containing approximately 0.06 ppm fluoride.
Fluoride concentrations as high as 1,300 ppm were observed in bone ash.
Jackson and Weidman (1958) analyzed levels of fluoride in bone from residents
of three English cities with different concentrations of fluoride in their
drinking water: West Hartlepool, 1.9 ppm; South Shields, 0.8 ppm; and Leeds,
less than ~0.5 ppm. In each case, a plateau in bone fluoride concentration
appeared at about age 55. The parameters influencing whether or not
concentrations of fluoride in human bone plateau with increasing age are not
understood (NAS 1971). In view of the large intersubject variability, there
may not be a true plateau. Also, at only 0.1 ppm fluoride in Che water, a
plateau tar/ not be reached in a lifetime.
Machle and Largent (1943) showed that when adult humans absorbed up to 18
mg of fluoride p.er day, about half of this amount was deposited in the skeleton,
The fraction of the absorbed.dose of fluoride deposited in the skeleton
111-18
-------
of younger persons is somewhat greater. For example, Zlpkin et «1. (1956)
measured concentrations of urinary fluoride In children and In adults before
and after fluorldatlon of a community water supply. In the adults, urinary
fluoride concentration equaled that of the fluoridated drinking water after
one week. In the children, three years passed before the fluoride
concentration in- urine approximated that of the ingested drinking water.
In quasi-steady state conditions of fluoride intake, a corresponding
skeletal.concentration of fluoride is reached which then continues to increase
slowly vith tine. The skeletal concentration is related directly to the level
of steady state Intake. The rate of uptake and retention in the bone declines
vith age, but whether or not concentrations in the bone reach a plateau
commensurate with the daily intake cannot yet be stated with certainty. When
intake is elevated above "normal" amounts, either briefly or perhaps over
several• weeks, approximately half of the additional absorbed fluoride will be
deposited in the bone. Upon reestablishing the "normal" steady state, the
excess fluoride retained in the bone also declines. There is no significant
accumulation or retention of fluoride in soft tissues.
F. Summary
Following ingestiou, soluble fluorides are rapidly absorbed from the
gastrointestinal tract at least to the extent of 97Z. Absorbed
fluoride is distributed throughout the tissues of the body by the blood.
Fluoride concentrations in soft tissues fall to pre-exposure levels within a
111-19
-------
few hours of exposure. Fluoride exchanges with hydroxyl radicals of hydroxy-
apaclte (the inorganic constituent of bone) to fora fluo:onydroxyapatlte.
Fluoride that is not retained is excreted rapidly in urine. In adults under
steady state intake conditions, the urinary concentration of fluoride.tends to
approximate the concentration of fluoride in the drinking wacer. This
reflects the decreasing retention of fluoride '^primarily in bone) with
increasing age. Under certain conditions perspiration may be an important
route of fluoride excretion. The concentration of fluoride retained in bones
and teeth is a function of both the concentration of fluoride intake and the
duration of exposure. Periods of excessive fluoride exposure will result in
increased retention in the bone. However, when the excessive exposure is
eliminated, the bone fluoride concentration will decrease to a concentration
that is again reflective of intake.
ITT-20
-------
IV. HUMAN EXPOSURE
Humans can be exposed Co fluoride in drinking water, food and air
(Letkiewicz 1984). This section summarizes available pertinent information in
order to assess the relative source contribution of fluoride from drinking
wat^r, food and -air.
A. Exposure Estimation
This analysis is limited to drinking water, food and air since these
media are considered general sources of fluoride for all individuals. Some
individuals may be exposed to fluoride from other sources, notably in
occupational settings and from the use of consumer products containing
fluoride. In limiting the analysis to these three sources, it must be
recognized that individual exposure can vary widely based on several
uncontrollable factors. Life style, food consumption and physiological
characteristics (age, sex and health status) can all affect daily exposure and
intake. Individuals living in the same neighborhood or even in the same
household can experience vastly different exposure patterns.
Data and methods to e timate exposure of identifiable population
subgroups from all sources simultaneously have not yet been developed. To the
extent possible, estimates are provided giving the number of individuals
exposed to each medium at various fluoride concentrations,
IV-1
-------
B. Drinking Water Exposure
It is estimated that over 862 of the 195,595.000 people using public
water supplies are exposed to fluoride at levels of 1.0 mg/L or less; most
(77.6?) are receiving water containing fluoride at levels of 0.1 to 1.0 mg/L.
Approximately 835,000 people In the U.S. are exposed to drinking water levels
exceeding 2.0 mg/L. Approximately 90% of those exposed to fluoride at levels
above 2.0 mg/L receive their drinking water from groundwater sources (Table
IV-1).
%
Table IV-2. presents the estimated daily intake of fli ride from drinking
water for three population groups (ai'.ult males, 5- to 13-year-old children and
newborn formula-fed infaats) as a function of the fluoride levels in drinking
water. The data indicate that, on a per body weight basis, the drinking water
intake of fluoride by children in the 5 to 13 age g"oup is approximately.1.A
times that of the adult male. The intake of newborn formulr -fed infants is
more than eight-times that of adult males. The drinking water intake
calculations used here do not include the factor for afr temperature that is
allowed for in the existing EPA and PHS standards. The basis for that
relationship has recently been questioned (Coniglio 1984) and revised drinking
water regulations are not expected to incorporate such a factor in the Maximuc
Contaminant Level (HCL).
IV-2
-------
Table IV-I Estimated Population Exposed to Fluoride In Drinking Water at
Various Concentration Ranges
System Type
Groundwater
Surface water
Total
(Z of total)
Number of
people served
in U.S.
(thousands)
69.239
126,356
195.595
(100Z)
Number of people (thousands) exposed to fluoride at concentrations (mg/L) of:
<0.1
3.067
13,548
16,615
(8.5Z)
0.1-
1.0
61,552
90,132
151,684
(77. 6Z)
>1 .0-
2.0
3,872
22,590
26,462
(13. 5Z)
>2.0-
3.0
408.9
71.8
481
(0.2Z)
>3.0-
4.0
163.1
6.0
169
(0. 1Z)
>4.0- >5.0-
5.0 6.0
105.5 56.1
5.1 0.04
111 56
(<0.1Z) (<0.1Z)
>6.0-
7.0
9.3
2.7
12
(<0.1Z)
>7.0-
8.0 >8.0
2.0 3.2
0.4 0.08
2.4 3.2
(<0.1Z) t<0.1Z)
-------
Table IV-2 Estimated Intake of Fluoride from Drinking Water
Fluoride concentration
in drinking water
(mg/L)
a
b
c
d
0-2
> 2-4
> 4-6
> 6-8
> 8
Calculations based on
uses 10 mg/L for the
Calculation based on
Calculation based on
Approximate percent of popu- Drinking water intake per individual8
lation exposed to fluoride (ng/kg/day)
in drinking water at the . Adult.
indicated concentration range males
99. 6%
0.3Z
< 0.1%
< 0.01Z
« 0.01Z
2.9 x 10~2
8.6 x IO"2
1.5 x 10"1
2.0 x 10"'
2.9 x JO"1
the midpoint of the indicated concentration
drinking water level; assumes 100% absorption
an adult male weighing 70 kg consuming 2 L of
a 10 year-old child weighing 33 kg consuming
Children Newborn Formula-
(5-13 year-old)0 fed Infants
4.2 x
1.3 x
2.1 x
3.0 x
4.2 x
10"2 2.4 x JO"1
10"1 7.3 x IO"1
10"' 1.2
IO"1 1.7
IO"1 2.4
range except for > 8 mg/L range, whicl
•
water per day.
1.4 L of water per day.
-------
C. Dietary Exposure
Several estimates have been made of the daily dietary intake of fluoride
in the United States (exclusive of drinking water). These are shown in
Table IV-3. These estimates generally place fluoride dietary intake in the
range of 0.2 to 0.8 mg/day.
In contrast, Osis et al. (1974) reported a higher daily dietary intake of
fluoride at 1.6 to 1.9 mg over a 6-year period in an area with a fluoridated
water supply. Kramer et al. (1974) reported fluoride dietary intakes of 1.7
to 2.4 mg/day in 12 cities using fluoridated water. In four cities using
nonfluoridated water the intake .was 0.8 to 1.0 mg/day. Both of these studies
used a method of analysis reported by Singer and Armstrong (1965). However,
Singer et al. (1980) indicated that the method used in those studies would
lead to an overestimation of fluoride. While these .values may not be
quantitatively valid, it is interesting to note that Kramer et al. (1974)
provided useful data on the correlation of dietary levels observed in the
various cities with the level of fluoridation of drinking water. While no
direct correlation was observed for individual cities, the mean dietary level
in fluoridated cities was about three times that of the nonfluoridated cities,
and the mean fluoride cbn'.ent of the drinking water in fluoridated cities was
also about three times that of non-fluoridated cities.
IV-5
-------
Table IV-3 Reported Daily Dietary Intake of Fluoride
(exclusive of water)
Source
WHO (1970)
NAS (1980)
Undervood (1973)
Hodge and Smith (1970)
Category of
Age 1 -
4 -
7 -
10 -
Adult
Adult
Adult
Individual
3
6
9
12
Dally Intake (mg)
0.027 - 0.265
0.036 - 0.360
0.045 - 0.450
0.056 - 0.560
0.2 - 0.3
0.3 - 0.5
0.3 - 0.8
Singer et al. (1980)&
Young adult male
(age 16 - 19)
0.333 (San Francisco)
0.378 (Buffalo)
0.587 (Atlanta)
0.368 (Kansas City)
Excludes all beverages.
IV-6
-------
D. Air Exposure
The information on levels of fluoride in air suggest that airborne
fluoride contributes little to total daily intake. Assuming that an adult
male inhales 23 m* air/day and absorbs 100% of the Inhaled fluoride, airborne
2
fluoride present at the usual limit of detection (0.05 ug/m ) would contribute
about 1.2 ug/day to an individual's intake. This is to be compared to the
estimated values of 200 to 800 ug/day for food and drinking water.
E. Summary
Table IV-4 shows the relative source contribution of food, air and
drinking water for the daily fluoride intake for an adult male in the United
States. The predominant sources of fluoride for the adult male in the United
States are food and drinking water. Typical air levels of fluoride are
extremely low. Most fluoride air levels are below the limits of detection,
3 3
usually 0.05 ug/m . Fluoride at 0.05 ug/m would, with 100Z absorption,
contribute only 1.2 ug/day to an adult male's daily intake (23 m /day
respiration volume is assumed). For an adult, male weighing 70 kg, the
corresponding air dose is 1.7 x 10 mg/kg/day. The air contribution t
to be negligible except when food and drinking water doses -re zero.
The food intake shown was derived from the data presented in Letkiewicz
(1984), which suggested that the daily dietary intake was 0.2 to 0.8 mg/day.
Assuming 100Z absorption for a 70-kg adult male, these values correspond to
2.9 x 10"3 to 1.1 x 10~2'mg/kg/day.
IV-7
-------
Table IV-4 Estimated Intake of Fluoride from the Environment by Adult Males
00
Fluoride
concentration
In drinking
water (mg/L)
0-2,
> 2-4
> 4-6
> 6-8
> 8
Estimated percent of the popu-
lation exposed to Indicated
fluoride concentration range
from public vater supplies
(Z of total)
99. 6*
0.3Z
< 0.1Z
c 0.01Z
« 0.01Z
Drinking water
Intake per Indi-
vidual (mg/kg/day)
2.9xlO"2
8.6xlO"2
1.5xlO~'
2.0x10"'
2.9x10"'
Total Intake per Indlvld'ial
In mg/kg/day (Z from drinking water)
Food intake per
Individual
(mg/kg/day): 2.9xlO~3 7.lxlO"3
3.2xlO"2(9IZ) 3.6xlO~2(8IZ) 4.
8.9x!0"2(97Z) 9.3xlO"2(92Z) 9.
l.5x!0"'(98Z) l.6xlO"'(95Z) I.
2.0xlO~'(99Z) 2.1xlO"'(97Z) 2.
2.9xlO~'(99Z) 3.0xlO~'(98Z) 3.
l.lxlO"2
Ox!0"2(72Z)
7x!0"2(89Z)
6x!0"'(93Z)
lxlo"'(95Z)
Ox!0"'(96Z)
8 Dally Intake from air (estimated to be less than 1.7x10 mg/kg/day) considered negligible relative to food and
drinking water.
Calculation baaed on an adult male weighing 70 kg consuming 2 L of water per day and using the midpoint of the Indicated
concentration range except for the > 8 mg/L range, which uses 10 mg/L for the drinking water level.
c Based on data showing the dally.adult dietary intake of fluoride ranging from 0.2-0.8 mg/day and a 70-kg adult (2.9xlO~3
mg/kg/day - 0.2 mg/day; 7.1x10 mg/kg/day - 0.5 mg/day; 1.1x10" mg/kg/day - 0.8 mg/day).
-------
Under the typical drinking water exposure condition of about 1.0 mg/L,
drinking water accounts for an estimated 727. to 91* of total fluoride intake
for the adult male, with food contributing the remainder. Where drinking
water levels exceed 2 mg/L, the contribution from drinking water is generally
expected to exceed 90Z of total intake.
A similar comparison of relative source contribution is shown in Table
IV-5 for 5- to 13-year-old children. As in the case of the adult male, the
contribution from air is negligible relative to drinking water and food.
Under the typical drinking water exposure condition of about 1.0 mg/L,
drinking water accounts for 64Z to 972 of total fluoride intake. Where
drinking water levels exceed 2.0 mg/L, drinking water accounts for more than
90Z of total intake.
IV-9
-------
Table IV-S Estimated Intake of Fluoride from the Environment by 5-13 year-old Children
Estimated population of
Fluoride 5-13 year-old chlldrei. exposed
concentration to Indicated fluoride concen- Drinking water Food intakedper
In drinking tratIon range from public Intake per indl- individual
water (mg/L) water supplies
Total Intake per Individual
in mg/kg/day (X from drinking water)
All kai*K |*« * *••*•* <»••*• «bv •••*•*••.
vldual0 (mg/kg/day) (mg/kg/day): I.1x10
-3
1.3x10
-2
2.4x10
-2
0-2
> 2-4
> 4-6
> 6-8
> 8
26.039
86
22
1
.600
.800
.300
.900
400
4
1
2
3
4
.2xlO~2
.3x10"'
.1x10"'
.0x10"'
.2x10"'
4
1
2
3
4
.3xlO"2
.3x10"'
.1x10"'
.0x10"'
.2x10"'
(97X)
(99X)
(99X)
(IOOX)
(100X)
5.5xlO~2
1.4x10"'
2.2x10"'
3.1x10"'
4.3x10"'
(76X)
(91X)
(94X)
(96X)
(97X)
6.
1.
2.
3.
4.
6x!0"2(64X)
5x10"' (84X)
3xlO~'(90X)
2x10"' (93X)
4xlO~'(96X)
a Dally intake from air (estimated to be less than 2.3x10 mg/kg/day) considered negligible relative to food and
drinking water.
Baaed on 1981 data provided In Statistical Abstract of the United States 1982-83 showing I3.37X of the total U.S.
population falling In the 5-13 year-old age group.
c Calculation based on 10 year-old child weighing 33 kg consuming 1.4 L of water per day using the midpoint of the
indicated concentration range except for the > 8 mg/L range which uses 10 mg/L for the drinking water level.
d Baaed on data showing the dally dietary intake of fluoride for children ages 4-12 to range from 0.036-0.56 tog/day and
assuming a body weight of 33 kg (I.lxl0~ mg/kg/day - 0.036 mg/day; 1.3x10* mg/kg/day - 0.30 ing/day; 2.4x10 mg/kg/day
- 0.56 mg/day).
-------
V. HEALTH EFFECTS IN ANIMALS
A. Acute Toxicity
The toxicologlcal effects of fluoride in animals are summarized below.
To provide a common basis for comparison of individual studies, all dosr
values have been expressed in terms of milligrams fluoride per kilogram body
weight (mg/kg). Where the literature provided dose information in alternative
units (i.e., ppm in drinking water), the dose in terms of mg/kg has been
calculated and shown parenthetically in the text.
Leone et al. (1956) described the acute and subacute physiological and
pathological effects of fluoride (as sodium fluoride) administered
intravenously and orally to male and female dogs. When fluoride was infused
intravenously in four dogs at the rate of 5.4 mg F/min, the mean acute lethal
dose was 36.0 ± 0.5 mg F/kg with death occurring after 59 to 64 minutes of
infusion. The principal effects observed were a progressive decline in blood
pressure, heart rate, central nervous system activity (pupil size, response to
light, tendon reflexes) with vomiting and defecation. All effects became
(a) Equation for conversion of dose values from ppm in drinking water to
mg/kg:
ppm in
drinking x 1 mg/L x Dally Water
_ ., water 1 ppm Consumption, L
Dose, mg/kg - (Animal Weight. Kg) *
V-l
-------
evident vh"n the infused dose reached approximately 20 tng F/kg. At a mean
dose of 30.6 ng F/kg, the respiratory rate was depressed and electrocardio-
graphic changes indicated a conversion to atrioventricular nodal or ventricular
rhythm. The terminal cardiac event was either ventricular fibrillation or
asystole.
In a separate study by Leone et al. (1956), dogs were Infused at the race
of 5.4 mg F/min to total doses of 25, 20 and 15 mg F/kg and the animals
observed until death or sacrifice. The number of dogs at each dose was 3, 2
and ft, respectively. Dogs receiving 25 mg F/kg died within 1 to 31 hrs after
Infusion. One of the dogs administered 20 mg/kg died after 36 hours and the
second died after seven days. None of the dogs receiving 15 mg F/kg died.
These animals were sacrificed 36 hrs, 7 and 16 days after infusion. The
approximate LD.Q from this study was estimated to be 20 mg F/kg. The major
effects observed were vomiting, defecat.ion and central nervous system
depression.
In a third study, two dogs were administered a dose of 5 mg F/kg, infused
at the rate of 1 mg F/min daily for 23 days. There were no deaths nor
evidence of toxic effects or weight loss, and the electrocardiograms were
normdl. However, blood pressure and respJ.atory rate were not measured. Four
dogs were also evaluated in a very limited oral study. Each dog was
administered a single oral dose of sodium fluoride. The doses administered
were 38, 81, 260 and 3,100 mg F/kg. The principal effects observed were
vomiting .and frequent defecation. Each dog appeared to recover completely
within 18 to 24 hours.
V-2
-------
During these studies by Leone et al. (1956) serum calcium levels were
measured in dogs receiving total doses of 36, 25. 20 or 15 mg F/kg
intravenously and in the one dog administered 3,100 mg F/kg orally. In one
instance the calcium concentration increased slightly, in one instance it was
unaltered, and in the remaining dogs it was lowered slightly from predosing
levels. No statistically significant conclusions can be drawn from these
observations. Microscopic examination of tissue reactions from .all dogs dying
after fluoride administration, and in one animal sacrificed 36 hours after
receiving 15 mg/Ug total dose, showed generalized hypereraia and acute focal
hemorrhages. All other animals showed some focal hyperemia and focal
hemorrhages, but these were no more severe than were seen in the control dogs.
In essence, the findings of Leone et al. (1956) provide no evidence of
cumulative effects following daily administration of sublethal doses of
fluoride for up to 3 weeks. The physiological effects and pathological
changes seen in dogs resemble those reported for humans (Lidbeck et al. 1943).
The pathological studies performed did not identify a specific mechanism of
death, though direct toxic cellular effects cannot be discounted.
The acute toxicity of NaF in fasted male white mice of uniform weight (10
grams) was also studied b/ Leone et al. (1956). The oral LD__ and intravenous
LD,n were evaluated in groups of ten or more mice. The arbitrary endpoint was
24 hours after administration. The oral LD5Q with standard error was 46.Oil.6
mg fluoride ion/kg compared with an intravenous LD$Q of 23.0+0.9 mg/kg. Mice
dying within two hours after injection developed, successively: cyanosis,
v-3
-------
dilation of the ear vessels, depression of respiration, tremors, clonic
convulsions, paralysis of the hind legs, loss of righting reflex, depression,
respiratory arrest and death. Those with longer survival periods (2 to 24
hours) vent through similar but less severe stages, progressing to a long
terminal depression.
Maynard et al. (1951) studied the effects of age and sex on the acute
lethality of NaF in rats. The rats were given an intraperitoneal dose of
26 mg NaF/fcg, the approximate LD5_ for animals weighing 200 to 300 g. At less
than seven months-of-age both sexes seemed to be resistant to NaF toxiclty,
with the females less resistant than the males. At seven months or more,
there were no differences between the sexes. These data are summarized in
Table V-l.
These studies of acute fluoride toxicity are a representative sample of
those available. They illustrate the essential characteristics. Other
published studies of acute fluoride toxicity do not differ significantly in
their content.
B. Chronic Toxicity
In practical terms, chronic effects of excessive exposure to fluoride
have been most important in domestic animals, especially cattle (MAS 1971).
For this, reason, the chronic toxicity of fluoride has been studied mainly in
cattle and in sheep.
V-6
-------
Table V-l Effects of Age on Toxiclty of Sodium Fluoride in Rats
Hale Albino Rats
Age
(months)
1
2
3
4
7
12
Av. Wt.
(grams)
90
229
288
297
361
336
Mortality
No. rats in 26 hr
used (Z)
25
25
25
25
25
25
0
0
8
12
84
92
Female Albino Rats
Age Av. Wt.
(months) (grams)
1 79
2 164
3 174
4 190
7 213
12 219
No. rats
used
25
25
25
25
25
25
Mortality
in 24 hr
(Z)
0
16
4
28
80
92
, 26 mg NaF/kg; Concentration, 20 mg/mL Water. No mortality vas produced at
a dose of 20 mg NaF/kg in comparable animals.
*
Adapted from Maynard et al. (1951).
V-5
-------
The studies discussed In this section are a representative sample of
those available. The conclusions from the studies selected are consistent
with those not discussed herein. The material presented was selected for the
quality of the research effort that generated It and for its illustration of
the essential characteristics of fluoride toxlcity.
1. Bone
Suttie et al. (1961) exposed Holstein calves to dietary NaF. At the
start of the experiment the calves were 6- to 27-weeks-old. Sodium fluoride
was added to their diet to supply 1.0, 1.2, 1.6, 1.6 and 2.0 mg fluoride/
kg/day. The majority of the cattle were removed from the experiment either
during or at the end of the second lactation period. Length of exposure in
calendar time was not specified and varied from animal to animal. Severe
fluorosis (characterized by rapid veight loss, general deterioration of
condition, intermittent lameness and stiffness) was consistently associated
with a skeletal fluoride concentration greater than 5,500 ppm. This
concentration was reached by the first lactation in cows receiving 2.0 mg
fluoride/kg/day and by the second lactation in cows receiving 1.6 mg
fluoride/kg/day. The authors stated that a fluoride level in bone in excess
of 5,500 ppm is one of the most reliable i-dices of fluoride toxicosis.
Shupe et al. (1963) studied the effects of dietary fluoride on 32
Holstein-Friesian dairy heifers. The animals received 12 (normal). 27, 4? or
93 ppm fluoride (equivalent to 0.30, 0.64, 1.17 and 2.08 mg/kg/day) on a total
V-6
-------
dry matter basis in their diet from age three to four months until 7.5 years.
Eight animals were used in each of the four dose groups. Changes, in bone were
narked at 93 ppm (2.08 mg/kg/day), moderate at 49 ppm (1.17 mg/kg/day) and
very slight at 27 ppm (0.64 mg/kg/day). There were no discernible effects on
bone at the 12 ppm level (0.30 mg/kg/day). These data are summarized In Table
V-2.
Affected bones appeared chalky white and had roughened, irregular per-
iosteal surfaces. They were also larger and heavier than normal. Puffy
joints and intermittent lameness developed in some cows in which osteofluorotic
lesions were palpable. Shupe et al. (1963) considered lameness and stiffness
to be inconclusive measures of fluoride toxicity. Bone lesions were scored
according to the scheme in the legend of Table V-2.
Fluoride concentrations in dry. fat-free rib biopsy samples increased
with increasing time of exposure for all dose groups. After 7.3 years (2,663
days) the fluoride concentration was approximately 900 ppm in animals on the
normal diet. At this same time, the rib fluoride concentrations were
approximately 2,500, 5,500 and 8,200 for the cattle receiving 27, 49 and 93
ppm fluoride in the ration, respectively. The rate of increase with time was
greatest in those cattle idministered 93 ppm fluoride. The first clinically
discemable bone lesions appeared on the medial surface of the proximal third
A
of the metatarsal bones and were bilateral. These effects were observed after
1.5 to 2 years in cattle on the 93 ppm fluoride ration and after 3.5 to 4
years in cattle on the 49 ppm fluoride ration. As the degree of' osteofluorosis
V-7
-------
Table V-2 Effects of Ingested Fluoride on Dalrv Cattle Fed
Various Levels of Sodium Fluoride From 4 Months to
7.5 Years of ARC*
Chronic fl unroll is
Average
F In Boiiture-fref diet
(pp»)
Teeth classification
(incisors)
Teeth clasfif ication
(swlars)
F In bone (ppi<)
F in urine (pps)
F in nilk fppa)
F in blood (ppo)
Average F in ioft tissue's
(pp»)
Perlosteal hyperostosis
Age
fvear*)
2
4
ft
2
4-
6
2
4
6
2
4
6
2
4
6
2
4
6
2
4
6
2
4
6
2
4
6
Noraal
condition*
Up to IS
Up to IS
Up to IS
0-1
0-1
0-1
0-1
0-1
0-1
401-714
706-1.139
653-1.221
2. 27-3. 78
3.54-5.3
3.51-6.03
Up to 0 12
Cp co 0.12
Up to 0.12
Up to 0.30
Up to 0.30
Up to 0.30
Up to 1.20
Op to 1.20
Dp to 1.20
0
0
0
No adverse
effects
15-30
15-30
15-30
0-2
0-2
0-?
0-1
0-1
0-1
714.1.605
1.138-2.379
1.221-2.794
3.78-8.04
5.3-10.32
6. 03-11. 29
Up to 0.12
Up to 0.12
Dp to 0.1?
Cp to 0.30
Up to 0.30
Up to 0.30
Up to 1 . 20
Up to 1.20
Up to 1.20
0-1
0-1
0-1
Borderline
30-4n
30-40
30-40
2-3
2-3
2-3
0-1
0-1
1-7
1.605-2.130
2.379-3.138
2.794-3.788
8. 04-10. St
10.32-13.31
ll.29-li.78
0.08-0.15
0.08-0.15
0.08-0.15
0.15-0.40
0.15-0.40
0.15-0.40
Up to 1.20
Up to l> 20
Up to 1.20
0-1
0-1
0-2
Moderate
40-f.O
4o-*n
40-60
3-4
3-4
3-t
0-1
1-2
1-3
2.130-3.027
3.138-4.504
3.788-5.622
10.54-14.71
13.31-18.49
14.78-20.96
0.15-0.25
0.15-0.25
0.15-0.25
0.30-0.50
0.30-0.50
0.30-0.50
Up to 1.20
Dp to 1.20
Up to 1.70
0-2
0-3
0-4
Severe
60-109
60-109
60-109
4-5
4-5
4-5
0-3
1-4
1-5
3.027-4.206
4.504-6.620
5. 622-8.676
14.71-19.86
1*. 49-75. 63
20.96-30.09
0. 15 and above
0. IS and above
0. 15 and above
0.50 and above
0.50 and above
0.50 and above
Up to 1 . 20
Up to 1.20
Up to 1 .20
0-3
0-4
0-5
Secondary changes
occur
All Absent Absent Occasionally Present
noticed
Present
* Data are base* on controlled experiments, but also
extensivel* studied and evaluated.
> an be correlated with nuocrous field cases that h
-------
increased, palpable hyperostoses appeared in the rani of the mandihular bones,
and the 7th through 12th ribs became wider and thicker. The degrees of
periosteal hyperostosis were classified as 0 * normal, 1 - questionable, 2 -
slight, 3 « moderate, 4 * marked and 5 • excessive. Cattle on the normal diet
were scored as normal through 6 years of age. Those cattle on 27 ppm ration
were scored 0 to 1 through 6 years; those on 49 ppm ration were scored 0 to 2
at 2 years, 0 to 3 at 4 years, and 0 to 4 at 6 years; and those on 93 ppm
ration were scored 0 to 3,' 0 to 4 and. 0 to 5 at 2, 4 and 6 years,
respectively. Radiographs taken at age 7.5 years (approximately 7 years on
fluoride) showed increased coarsening and thickening of the trabecular pattern
with a ground glass appearance for cattle on the rations containing 49 and 93
ppm fluoride. Periosteal hyperostosis, subperiosteal increased density in
some cases, endosteal and cortical porosity, and mineralized spurs at points
of attachment of tendons to leg bones were also observed at these dose levels.
In the study-of Shupe et al. (1963), no effects at any exposure level on
hoofs, soft tissues or blood were found. Milk production was only affected
a
after clinical signs of skeletal fluorosis, lameness and molar abrasion had
developed. These observations imply the animals were not able, because of
their advanced skeletal and dental fluorosis, to maintain a nutritional level
.adequate for milk production. Effects on milk production were found in the
cows exposed to 49 and 93 ppm (1.17 and 2.08 mg/kg/day) fluoride. Suttie et
al. (19S7b) also found that the effects of fluoride on milk production were
secondary to other clinical signs resulting in curtailed feed intake.
V-9
-------
2. Teeth
Suttle et al. (1957a) added NaF to the ration of 24 Holstein heifers
divided into six groups of four each. The animals were approximately two-
years-bid at the start of the experiment and exposure to fluoride was
maintained for 5.5 years. Fluoride was mixed with the diet so that the cattle
received 20, 30, 40 or 50 ppm added fluoride per day on the basis of dry feed
weight (four animals per dose). These concentrations were equivalent to 0.53,
0.86, 1.03 and 1.36 mg/kg/day, respectively. The earliest observable indica-
tion of excessive exposure to fluoride was a mottling of the growing teeth.
Mottling was scored according to the scheme of Hobbs et al. (1954), as
follows:
Classification Description
1 Normal Tooth
2 Questionable Effect
3 Marginal Effect
4 Definite Effect
5 Severe Effects
The latter two ratings are characte.ized by varying degrees of hypoplasia
of the enamel and tooth. Slight mottling and wear (scored 3, marginal) was
observed on the fourth incisors of the animals ingesting 0.53 mg/kg/day.
Animals ingesting 0.86 to 1.36 mg/kg/day had teeth which were scored 4 or 5.
These data, reflecting the results following 5.5 years of fluoride exposure,
are summarized in Table V-3.
V-10
-------
Table V-3 The Effects of Fluoride Fed as NaF on
Various Physiologic Responses in Heifers
Fa
Lot Added
I
11
111
IV
V
^Values
0
20
30
40
50
are
Av.
Fecal Fa
(dry wt.)
14
•
~~
27
30
37
expressed as
Av. Calf
Bone Fa
(dry fat
free vt.)
11
j
86
136
104
140
ppm fluoride.
Av. Score of
Incisors
Milk F
.(whole milk)
0.16
0.31
0.29.
0.30
0.38
—
0.27
0.44
Pair
3
1
1
2
4
4
1 tnr
Pair Av. No. of
4 Services
1 2.1
3 1.5
4 2.1
5 2.3
5 2.2
r» nn r* an f1 4 nn
Adapted from Suctie et al. (1957a).
V-ll
-------
Subsequent to the appcnrance of effects on dentition, a total of four
animals, two In lot IV and two in lot V, developed svmptoi of fluorosis.
These symptoms and their sequence were:
1. A refusal of fluoride-supplemented foods.
2. Excessive weight loss.
3. Stiffness in the legs with resulting lameness.
These effects were sufficient to debilitate the animals within several
weeks. Such symptoms of fluorosis were not seen in the groups exposed to 30
ppm fluoride or less.
In the study of Shupe et al. (1963), pairs of heifers (three- to four-
oonths-old) received 12. (normal), 27, 49 or 93 ppm of fluoride (equivalent to
0.30, 0.64, 1.17, and 2.08 mg/kg/day) on a total dry matter basis in the
ration. In addition, each ration was supplemented with one of two levels of a
concentrate or one of two levels of a Ca-P mixture. The experiment,
therefore, included 32 animals divided into pairs among 16 treatment groups.
The. experiment continued until the animals were 7.5 years-old.
Der-.nding on the amount of fluoride ingested, affected teeth erupted with
different degrees of mottling, staining, hypoplasia and hypocalcification.
Dental fluorosis was scored according to the scale described in Table V-2,
which is essentially the same as that of Hobbs et ai. (1956) and the following
•v
•
tooth classifications were established:
V-12
-------
(0) Normal: smooch, translucent, glossy white enamel; tooth normal
shovel shape.
t
(1) Questionable effect: slight change, exact cause not determined; may
have enamel flecks; cavities may be unilateral or bilateral but with
the absence of mottling.
(2) Slight effect: slight mottling of enamel; may have slight staining
but no wear; teeth normal shovel shape.
(3) moderate effect: definite mottling and staining of enamel; coarse
mottling (large patches of chalky enamel); teeth may have slight
signs of wear.
(4) Marked effect: definite mottling, staining and hypoplasia; may have
pitting of enamel; definite wear of teeth; enamel may be a pale
cream color.
(5) .Excessive effect: definite erosion of enamel with excessive wear of
teeth; staining and pitting of enamel may or may not be present.
In cattle consuming the highest dose of fluoride (i.e., 93 ppm in the
ration) the incisors were classified as A to 5, beginning as early as 2 years
of age. The molars were classified as 0 to 3 at 2 years-of-age, 1 to 4 at 4
years and 1 to 5 at 6 years. For cattle at the dose of 49 ppm, the incisors
were scored as 3 to i beginning at 2 years. In these same Animals, the molars
were scored as 0 to 1 at-2 years, 1 to 2 at 4 years, and 1 to 3 at 6 years.
In cattle administered 27 ppm fluoride, the incisors were scored as 0 to 2
through 6 years-of-age and the molars were scored as 0 to 1 through 6 years.
Incisors and molars of cattle administered the normal ration (12 ppm fluoride)
were scored 0 to 1 throughout the 6 years. Thus, the cattle administered the
V-13
-------
diet containing 27 pptn (0.64 mg/kg/day^ fluoride nearly represents a no effect
level (NOAEL).
The effects of water-borne fluoride on sheep have been studied by Peirce
(1959). Mature ewes (total of 150) were divided into three groups: Group A
(control) was given drinking water having 0.3 ppo fluoride; fluoride was added
to Che drinking water of groups B and C so that they were exposed to 10 and 20
ppm fluoride, respectively, ^hese ewes were mated over a period of six weeks
and all. the offspring were weaned when the youngest lamb was three-nonths-old.
At this point the mothers were removed from the experiment. The experiment
was continued until the younger animals (group A, 21 wethers and 11 ewe lambs;
group B, 17 wethers; group C, 20 wethers and 10 ewe lanbs) were almost
seven-years-old. During the course of the experiment the animals drank almost
nothing in the winter months and up to 3 to A L/day in the summer. For the
whole experiment, mean intakes of fluoride were 0.25 mg/kg/day for group B and
0.48 mg/kg/day for group C.
4
Weight gain was not significantly affected by treatment with fluoride at
any point in the experiment. Wool production was reduced in the groups
receiving added fluoride in their drinking water. Sheep in groups B and C
shoved characteristic signs of dental fluurosis. These included mottling of
incisors and molars, selective abrasion of the molars and wear of various
types and degrees of severity on the incisors. The degree of mottling and
erosion was slight in the sheep of group B but was severe in those of group C".
Only one animal in group C appeared to'be unaffected. Selective abrasion of
molars occurred in about 252 of the sheep in group B and in all but two of the
V-14
-------
sheep in group C. In addition, the abrasion was more severe in group C. The
fluoride content of bones and teeth was significantly increased in animals in
groups B and C. See Table V-4 for a summary of these data.
3. Reproduction
The effect of fluoride in drinking water on the reproductive efficiency
of Afrikander heifers was studied by Van Rensburg and De Vos (1966). At the
start of the experiment the heifers were from 2.5- to 2.75-years-old and were
free from tuberculosis, brucellosis and coital diseases. The animals were
divided into five groups of ten animals each and breeding was started nine
months after the start of the experiment. During the first two seasons the
animals were served naturally, while artificial insemination was used in the
last two seasons. Defluorioated superphosphate was added to the drinking
water of some experimental groups (see Table V-5) at the rate of 1 g
phosphorus/gallon. This was done in order to test the hypothesis that
phosphate might retard the action of fluorine.
Table V-5 summarizes the fluoride exposir . data and the breeding records
of these animals. Inspection of these data shows that cows receiving 5, 8 or
12 ppm fluoride in drinlc-ng water (estimated to equal 0.55, 0.88 and 1.32
mg/kg/day, respectively) suffered significant decreases in calving rates
(Figure V-l). The authors state that in Afrikander heifers not exposed to
excessive fluoride, reproductive efficiency normally increased during the
V-l 5
-------
Table V-4 Fluoride Concentration in Bones and Teeth
of Sheep Drinking Fluoride-Supplemented Water
Metacarpus
Experimental
Group
A
B
C
Fluoride
Content
of Water
(ppm)
0.3
10
20
Fluoride
in
Ash
(Z)
0.12
0.35
0.40
Fluoride
in
Bone8
(%)
0.08
0.25
0.29
Rib
Fluoride
in
Ash
(Z)
0.15
0.41
0.46
Fluoride
in
Bone
(%)
0.10
0.26
0.30
Molars
Fluoride
in
Ash
-------
Table V-5 Breeding Efficiency Over Four Breeding Seasons of Five Croups
on Different Levels of Fluorine Intake, With and Without
Added Defluorlnated Superphosphate
Breeding Seasons
Croup
1
9 Heifers
5 ppm fluorine
2
10 Heifers
8 ppta fluorine
3
10 Heifers
5 ppm fluorine ,
plus defluori-
nated super-
phosphate
4
10 Heifers
8 ppm fluorine,
plus defluori-
nated super-
phosphate
5
10 Heifers
12 ppm fluorine,
plus defluori-
nated super-
phosphate
Total for 49
heifers
No. served
No. of services
No. conceived
Services per conception
'Calving rate (Z)
No. served
No. of services
No. conceived
Services per conception
Calving rate (Z)
•
No. served
No. of services
No. conceived
Services per conception
Calving rate (Z)
No. served
No. of services
No. conceived
Services per conception
Calving rate (Z)
No . served
No. of services
*'.o. conceived
Services per conception
Calving rate (%)
No. served
No. of services
No. conceived
Services per conception
Calving rate (Z)
1
9
9
8
1.12
88
10
14
8
1.75
80
10
12
8
1.50
80
10
15
9
1.67
90
9
10
9
1.11
90
48
60
42
1.43
85.7
2
8
8
8
1.00
88
6
6
6
1.00
60
9
10
9
1.11
90
6
6
6
1.00
60
5
5
4
1.25
40
34
35
33
'1.06
67.3
3
8
12
6
2.00
66.7
9
15
6
2.50
60
9
9
7
1.28
70
10
13
5
2.60
50
6
11
2
5.50
20
42
60
26
2.31
53.1
4
6
8
4
2.00
44.4
8
11
5
2.20
50
5
6
3
2.00
30
8
11
1
11.00
10
7
15
1
15.00
10
34
51
14
3.64
28.4
Adapted from VanRensburg. and DeVos (1966),
V-17
-------
uu
§
o
tr
Uf
a.
BREEDING SEASONS
. Adapted from VanRensburg and DeVos (1966)
Figure V-l Calving Rate of Cows on
Three Levels of Fluoride Intake
V-l 8
-------
period over which this study WBF conducted. The deleterious, effects of
exposure to excessive fluoride on reproductive performance preceded the
development of clinical signs of fluorosis. This is significant as it sug-
gests chat the reproductive effects were not a consequence- of ill health
secondary to skeletal or dental fluorosis. However? by the end of the fourth
breeding season, general ill health, loss of appetite, and erosion and
mottling of teeth were increasingly evident.
Defluorinated superphosphate tended to exacerbate rather than mitigate
the toxicity of waterborne fluoride (Table V-5). Van Rensburg and De Vos
(1966) speculated Chat the defluorinating process may not have been completely
effective and that the animals receiving defluorinated superphosphate nw,
actually have been exposed to higher than indicated concentrations of fluoride.
Therefore, the data for exposure groups 3, 4 and 5 should be interpreted with
caution.
Suttie et al. (1957a) added NaF to the diet of Holstein heifers. The
animals were two-years-old at the start of the experiment and exposure to
fluoride was continued for 5.5 years. Fluoride was mixed in the ration so
that the cattle received 0.53, 0.86, 1.06 and 1.36 mg added fluoride/kg/day.
Over the course of this .-xperiment there was no effect on reproduction as
measured by the average number of services per conception. Other indices of
reproduction were not evaluated in this study.
Hobbs and Merriman (1962) studied the effect of NaF on reproductive
perfonsance in Hereford heifers. These animals were free from tuberculosis
V-19
-------
and Bang's disease and were immunized against brucellosis. Sodium fluoride
was added to feed so that over the nine-year period of exposure, groups of
three calves received 0.17. 0.39, 0.59, 0.91. 1.03, 1.24, 1.56 and 1.96 mg
fluorlde/kg/day. The cows were yearlings at the start of the experiment.
They were bred first when two-years-old, then at yearly intervals for nine
years. The breeding records of these animals are sumaarized in Table V-6. It
is apparent that there was some deficit in reproductive performance associated
with exposure to 1.56 and 1.96 mg/kg/day. Exposure to less than 1.56
mg/kg/day did not have an obvious effect on reproductive performance.
The effect of fluoride exposure on reproductive performance in sheep was
studied by Peirce (1959). He determined the percentages of lambs born by two
generations of experimental ewes (for example, 10 ewes giving birth to 10
lambs would be 100%). The first generation was divided into three groups of
approximately 50 ewes/group (group A, control; group B, 10 ppm or 0.25
mg/kg/day; group C, 20 ppm or 0.48 mg/kg/day). The percentages of ewes which
produced lambs were 93%, 982 and 98%, respectively. The actual percentages of
lambs born were 1151, 124% and 121X, respectively. The second generation ewes
retained in the experiment (11 in group A and 10 in group C) were mated first
at age 18 months and thereafter at yearly intervals for a total of six
gestations. The relevant data are summarized in Table V-7. These data show
that provision throughout gestation of drinking water containing 10 or 20 ppc
fluoride had no adverse effect on reproduction in these sheep.
V-20
-------
Table V-6 Reproductive Performance of Hereford Heifers Exposed to
Dietary NaF for Nine Years
Dose
of Fluoride
(mg F/kg/day)
0.17
0.39
0.59
0.91
1.03
1.24
1.56
1.96
No. of Cows
At Start of
Experiment
3
3
3
3
3
3
3
3
No. of Cows
At End Of
Experiment
3
la
3
3
3
3
2
Total
No. Calvings
Over 9 Seasons
17
18
11
23
17
20
12
6
Number Calvings
Expected If 1
Calving/Year/Cow
27
27
16
27
27
27
27
22
Actual
Z of
Expected
Calvings
63
67
69
85
63
74
44
27
3One animal died in first year of experiment; one animal sacrificed in seventh year
.of experiment.
One animal sacrificed in fifth year of experiment.
Adapted from Hobbs and Merriman (1962).
V-21
-------
Table V-7 Effect of Fluoride Exposure on Reproductive Performance of Sheep
Ewes
Year Mated
1952 11
1953 11
1954 11
1955 11
1956 10
1957 10
Total 64
Total as
Percentage
of ewes
mated
Croup Aa
Eves
Lambed
10
11
10
10
10
7
58
91 Z
Lambs Ewes
Born Mated
11 10
13 9
14 9
11 9
11 9
8 8
68 54
1063,
_b
Group C
Ewes
Lambed
8
8
9
8
9.
7
49
91%
Lambs
Born
9
12
13
12
16
8
70
130%
^Control animals (0.3 ppm F in drinking water).
Test animals (20 ppm F in drinking water).
Adapted from Peirce (1959)
V-22
-------
4. Growth
Some species differences are evident in the effect of fluoride on growth.
Growth is not affected in most species at 100 ppra fluoride in the diet (Hodge
and Smith 1965). Cattle, however, appear to be more susceptible. Suttie et
al. (1957a) reported that at SO ppm added fluoride, attainment and maintenance
of the adult body weight was slightly depressed in a study on Holstein heifers
fed rations containing 0, 20, 30, 40 or 50 ppm fluoride. In the study of
Shupe ec al. (1963) (see Section V.B.I.), Holstein heifers were fed rations
containing 12 to 93 ppn fluoride; according to the analysis of Stoddard et al.
(1963), growth was not significantly affected at any of these levels. Hobbs
and Merriman (1962) reported that, in a ten-year study, 10 to 100 ppm
(estimated 0.17 to 19.6 mg/kg) fluoride as NaF in the ration had no adverse
effect on the body weight of cattle, whereas 200, 300 or 600 ppm (estimated
3.92, 5.88 or 11.76 mg/kg, respectively) resulted in lower weight. Peirce
(1959) reported that ingestion of drinking water containing 0.3 to 10 or 20 mg
fluoride/L (estimated 0.08, 0.25 or 0.48 mg/kg/day, respectively) by sheep for
approximately seven years did not ^significantly affect weight gain. Hobbs et
al. (1954) reported that ewes given rations containing up to 100 ppm
(estimated 2.4 mg/kg) fluoride as NaF over a three-year period showed no
differences in final weight conpated to control animals; weight gains were
decreased at 200 ppn (estimated 4.8 mg/kg).
V-23
-------
5. Kidney
Sufficiently high doses of fluoride have been shown to produce a vasopres-
sln-resistant polyuria resembling nephrogenic diabetes insipidus (Rush and
Willis 1982). Investigations into the mechanism of this effect have been the
subject of several reports. Roman et al. (1977) found that rats (male, Fisher
344 strain, weighing 250 to 350 grams), after four dai^y subcutaneous injections
of 7.6 mg fluoride/kg body weight (as NaF). showed statistically significant
(P<0.05) increases in urinary flov, glomerular filtration rate, percent sodiun
excretion and percent water excretion. They suggested that fluoride inhibits
tubular resorption by inhibiting active chloride transport in the ascending
limb of the loop of Henle.
Rush and Willis (1982) have reported that fluoride inhibits sodium
chloride absorption in the. ascending limb of Henle's loop and antidiuretic
hormone-mediated water absorption across the collecting duct. In this study,
rats (male, Fisher 344 strain, weighing 200 to 250 grams) received intravenous
infusions of 5.7, 27.9 or 41.8 ug fluoride/kg body weight/minute for 2.5
hours. Whitford and Taves (1971) reported that in 16 female rats (weighing
approximately 200 g each, strain not mentioned) receiving 0.4 to 4 ug fluoride
intravenously ove.' a two-hour period, plasma fluoride concentrations of 950
ug/L were associated with a definite increase in the r»te of urine flow. From
the above studies and many others with similar results, it can be concluded
that the acute effects of fluoride on the kidney are related to both the peak
blood concentration of fluoride and the length of time the kidney is exposed
•» *
to high concentrations.
V-24
-------
Hodge and Smith (19h5) reviewed a number of chronic studies of fluoride
Ingestlon by various species, and concluded that hlstologlcal and functional
changes can be seen after single doses of 20 to 30 mg/kg and that renal
Injuries do not develop when the drinking water contains less than 100 ppm
fluoride.
6. Cardiovascular System
4.
Leone et al. (1956) investigated the effects on blood pressure and heart
rate of fluoride administered intravenously to dogs. Blood pressure and heart
rate were decreased at doses of 20 to 30 mg fluoride/kg, and respiratory rate
was increased. At 31 mg/kg, atrioventricular nodal rhythm, ventricular
tachycardia and ventricular fibrillation were evident. Caruso and Hodge
(1965) reported that oral administration of 10 mg fluoride/kg in mongrel dogs
(three males and nine females weighing 6.9 to 16.3 kg) did not evoke
hypotension, 15 mg fluoride/kg decreased blood pressure in two of three dogs,
and 23 or 36 mg fluoride consistently depressed blood pressure-in three dogs.
Caruso et al. (1970) summarized published observations on the effects of
fluoride on blood pressure in dogs and concluded that orally, a dose of at
least 9 mg fluoride/kg is required to bring about a hypotensive effect.
Sodium fluoride intravenously increase^ respiratory rate in proportion to the
decrease in blood pressure. Death was caused by respiratory arrest, the heart
continuing to beat for a time after breathing stopped. Ventricular
fibrillation occurred terminally. Strubelt et al. (1982), however, concluded
from their studies with male Wistar rats (weighing 340 to 420 g) that
V-25
-------
cardiovascular failure resulting from cardiodepresslve and vasodilating
effects of fluoride, rather than respiratory depression, was the cause of
death.
Effects on the electrophysiology and histology of the rabbit heart have
been reported by Takamori (1955). In these experiments, rabb-its (mature white
rabbits weighing 2 kg, 37 treated and 16 controls, sex not specified) received
daily oral doses of 6.5, 13.5, 22.5 or 65 mg fluoride/kg (as NaF). Duration
of the study is not clear from the report, though electrocardiograms are shown-
for rabbits receiving 6.5, 13.5 or 22.5 mg fluoride/kg for 62, 20 or 59 days,
respectively, and histological sections are shown for 6.5 mg/kg at 132 days,
for 30 mg/kg at 19 and 51 days, and for 50 mg/kg at 31 and 60 days. The
author indicated that electrocardiograms showed depressed ST, inverted I,
prolonged QT inverval, multi-focal ventricular precature contraction, bundle
branch block and pulmonary P. Histologically, the heart muscle showed regres-
sive degeneration, infiltration of cells, hyperemia, hemorrhages and thicken-
ing of the vessel walls. The effects were stated to be proportional to the
amount of fluoride fed and the duration of the feeding.
Complete histopathologic studies were done by Taylor et al. (1961) on
surviving rats (albin*, male, Rochester strain, 75-days-old, weighing 200 to
270 grams) sacrificed 30. days after receiving single injections of 3.6 to 21.7
mg fluoride (as NaF)/kg given intravenously or intraperitoneally. No effects
were seen in the heart, thyroid, lung, salivary gland, stomach, Intestine,
liver, adrenal; testis or femur.
V-26
-------
In a chronic study reported by Hansen (1978), nice (female, CSE nice. 3
to 4-weeks-old, initially weighing 22.5 to 25.5 grams) were given drinking
water containing 1 to 6 ng fluoride (as NaF)/L for six months. No
histological effects attributable to fluoride were seen in the heart, stomach,
intestines or bones.
7. Thyoid
The fact that iodine is tak-»n up by the thyroid gland has led to a
concern that fluorine, another halogen, might also become concentrated in this
18
organ. Studies with F, however, have shown this not to be the case; concen-
trations do not exceed those found in the blood (Hein et al. 1956). The
available evidence indicates that structural and/or functional alterations in
animals are not produced at or below fluoride concentrations of 50 mg/L in
drinking water (Hodge and Smith 1965, Harris and Hayes 1955).
C. Teratogenicity
\
No information on the teratogenicity of fluoride in animals was located
in the published literature.
D. Mutagenicity
Mohamed and Chandler (1976) studied the clastogenic effects of fluoride
added as NaF to the drinking water of highly inbred male mice. The mice
weighed 20 to 25 grans at the start of the experiment. The- treatment levels
V-27
-------
were 1. 5, 10, 50. 100 and 200 ppm fluoride (estimated to equal 0.2, 1.0, 2.0.
10, 20 and ^0 Bg/kg/day, respectively). A total of 12 groups of nice were
used (72 mice in all). At each dose level, mice were exposed for either three
weeks or six weeks. The authors reported that cytological studies on bone
marrow cells and on spermatogenesis indicated the presence of fragments,
bridges and other chromosomal abnormalities.
The in vitro effects of NaF on mouse, sheep and cow oocytes and the ir±
vivo effects of NaF on mouse oocytes were studied by Jagiello and Lin (1974).
For the in vitro experiments, oocytes were removed from donor animals and
incubated under conditions which stimulated meiosis. Sodium fluoride was
added in fetal calf serum to mouse oocytes at total concentrations from 0.01
to 0.4 mg NaF/mL, and in sheep serum to sheep and cov oocytes at total concen-
trations from 0.01 to 0.2 mg NaF/mL. Observed effects of NaF treatment included
Inhibition of division, atresia and fragmentation of chromosomes. Sheep and cow
oocytes were more sensitive than mouse oocytes to these NaF treatments. For •
in vivo experiments, mice were treated parenterally with NaF. Oocytes were
then harvested for the study of meiosis. Several dosing regimens were used:
A. 500 ug NaF/mouse, intravenous.
B. 500 ug NaF/mouse, subcutaneous.
C. 250 ug NaF/mouse/day for 16 days, subcutaneous.
D. 5 ug NaF/g body weight/day -for 35 days, subcutaneous.
Meiotic abnormalities were teen in 6 of the 28 cells examined froc
regimen A. ' These were cells at metaphase II with fuzzy, indistinct borders,
V-28
-------
and one cell had an abnormal anaphase-1-telophase-I. None of the other
regimens (B, C or D) resulted in the formation of abnormal oocytes.
In contrast to the results of Moharoed and Chandler (1976). Krara et al.
(1978) found no effect of NaF in drinking water on the frequency of sister
chromatid exchange in mice. Twelve-week-old mice were taken from colonies
which had been maintained for at least the seven prior generations on a low-
fluoride diet (estimated to equal less than 0.1 mg/kg/day) or a high-fluoride
diet (50 ppc - estimated to equal 10 mg/kg/day). Sodiun fluoride was added to
the drinking water of the group exposed to 50 ppm fluoride. Sister chromatid
exchange status was identified .in a separate laboratory with no knowledge of
the fluoride status of the animals. No significant differences in sister
chromatid exchange status were found between the low- and high-fluoride
groups.
Data consistent with those of Kram et al. (1978) were obtained by Martin
et al. (1979). Mice were taken from a colony that had been maintained for-at
least five generations on a diet containing 0.5 ppm fluoride (estimated to
equal 0.1 mg/kg/day) and drinking water having 0 to 50 ppm fluoride, (estimated
to equal 0 to 10 mg/kg/day) added as NaF. Testis and bone marrow cells froc
these mice were subjd-Led to cytologi al analysis (number of breaks,
fragments, deletions, multivalents and multiradicals). No deleterious effects
of fluoride exposure on chromosomes from testis and bone marrow cells were
found. See Table V-8, Experiment 1, for a suncary of the Jata.
V-29
-------
Table V-8 Bone Fluoride and Chromosomal Aberrations In Bone Marrow and
Testis Cells in Mice Receiving Water with Different Fluoride Levels
i
W
o
Bone-Harrow Cells
Bone Fluoride
• No. of
Croup Mice
Experiment
Lifetime
0 ppm F
50 ppm F
Experiment
One week
0 ppm F
Six weeks
0 ppn. F
0-ppm F
* TEH
1 ppm F
5 ppm F
10 ppm F
50 ppm F
100 ppm F
1
9
6
2
5
4
10
10
10
5
10
gZ in
Ash
0.0019
O.B6
0.011
0.008
0.01 I
0.021
0.016
0.151
0.2«»5
K
* o.ooor
* 0.01
i 0.001
I 0.001
» 0.001
* 0.001
» 0.002
t O.MI9
• 0.019
No. of
Mice
9
7
R
5
4
9
7
8
5
8
Cells
Scored
427
279
400
250
122
319
350
316
250
333
Testis Cells
Abrrrat ions
No. of Rate
Cells (Z)a
3
1
13
3
18
3
3
1
1
3
0.67
0.2R
3.25
1.20
12.50
0.67
0.86
0.42
n.40
0 75
b
t 0.33
t 0.28
t 1.00
i 0.49
t 1.68
* 0.47
• 0.50
t 0.42
! 0.40
! 0.53
No. of
Mice
8
7
9
5
5
10
10
10
5
10
Cells
Scored
399
350
450
154
240
336
424
428
182
414
No. of
Cells
4
2
3
1
10
5
2
12
1
1
Aberrnt tons
Rate
(Z)'
1.00
0.57
0.31
0.40
4.00
1.0
0.60
2.7
0.40
0.2O
t.
* 0.65b
* 0.57
J 0.17
t 0.40
? 2.61
t 0.80
t 0.27
s 2.0C
* 0.40
t 0.20
BThe aberration rate (Z cells examined with aberrations) was calculated for each mouse and the average f
-------
Martin el al. (1979) also maintained mice for six weeks on drinking water
containing 1, 5, 10, 50 or 100 ppn fluoride (estimated to equal 0.2, 1 0, 2.0,
10 or 20 mg/kg/dayj as NaF- These groups and .their controls did not differ in
average intake of food or fluid or in average weight gain during the course of
fluoride exposure. At the end of the exposure period bone marrow and test Is
*
cells were examined for chromosomal aberrations. No significant differences
between control end exposed groups were found in rates of bone marrow chromo-
somal aberrations (see Table V-8, Experiment 2). Some heterogeneity In
chromosomal aberration rates in testis cells was found. This heterogeneity
was attributable to one animal in the group exposed to 10 ppm fluoride.
Statistical analysis of the data showed no significant effect of exposure to
fluoride on the number of chromosom.il abnormalities in testis cells (see Table
V-8, Experiment 2).
The mutagenicity of NaF was tested in Salmonella typhimurium and in
Saccharomyc.es cerevisiae by Martin et al. (197-9) (see Table V-9). Sodium
fluoride was added to plates at 0.1, 1, 10, 100 and 500 ug/plate, with and
without microsomal enzyme preparations from rats treated with Aroclor 1254.
There was no indication of mutagenic activity in this experiment. One test
which gave an elevated result (TA100) was repeated. There was no repetition
of t'.e elevated result.
E. Carcinogeniclty
No information on the carcinogenic potential of fluoride in animals was
located in the literature. However, The National Cancer Institute initiated
V-31
-------
Tnble V-9 Mutagenlctty of Sodium Fluoride In Mlcroblal Systems: Number of Responses Per Plate
i
OJ
10
Salreonella typhireurium revertants/Plate
S. cerevlalae
tryptopliane *
Convertants/Plate
Test Conditions
No activation
Solvent control
Positive controls
Sodium fluoride
(ug/plate)
0.1
1.0
10.0
100.0
500.0
Activation
Solvent control
Positive controls
Sodium fluoride
(gg/plnte)
O.I
1.0
10.0
100.0
500.0
1000.0
2000.0
TAI535
18
<1000
30
24
20
24
27
29
128
21
36
53
36
j 1
TA1537
21
<1000C
13
18
23
14
14
19
HOOO8
12
14
18
22
19
TA1538
11
< 1 000
10
18
20
7
8
2°h
627°
!4
9
17
5
13
TA98
24
< 1 000
25
22
21
24
8
*8h
>1000
30
37
' 27
20
24
TA100 TA100"
132
795b
56
165
171
147
160
239- 243
148 748
149
209
208
2R7
464 261
259
289
TAIOO"
194
>1000
174
156
192
197
206
158
202
D4
23
103b
15
22
23
16
20
100,
1571
115
91
81
70
61
^Repeat at high levels of NaF with activation.
N-Methyl-N'-nitro-N-nltrosoguanldlne. 10
*jQuInacr.ine mustard, 10 g/plate.
2-Nltrofluorene, 100 g/plate.
eErjuivalent concentration to 25 mg of wet tissue of a 9000 g supernatant fluid prepared from liver of Sprague-Dauley
adult male rat Induced by Aroclor-1254 five days prior to kill was added to each plate.
f2-Anthramlne, 100 ug/plate.
B8-AmlnoquInollne, 100 ug/plate.
'^-Acetylflmlnofluorene, 100 ug/plate.
nlm«*t ''y In 11 rnnnmln* , ill! ; molrn/plat^.
Adapted from Martin ct al. (l')in).
-------
studies, during August 1979. to determine the carcinogenic and/or
lexicological poter.tlal of sodiun fluoride (NaF) in rats and nice. The
National Toxicology Program (NTP) took over the responsibility for oversight
of the studies in November 1982. The studies consisted of three parts: (Da
one-month subchronic study; (2) a six-monti. subchronic study with dosages
based on the previous experiment; and (3) a two-year chronic study based on
data from the six-month subchronic experiment •'maximum doses of NaF which were
not expected to affect the longevity of mice and rats were used). The chronic
study began in December 1981 and terminated in December 1983. Unfortunately,
problems developed seven months into the chronic study. The problems were not
treatment related (some rats in both the treatment and control groups
exhibited toxicollis and ocular lesions), but may have been related to the
diet which was low in several trace elements and vitamins. The validity of
the study was questioned and a new chronic study was scheduled. The Technical
Report from the new study should be issued in 1988.
F. Other
Other manifestations of chronic-fluoride toxicity in various species have
been reported in the literature, but have not been reviewed here because U)
they have been described in only one vf, at be««t, a very few investigations,
(2) the dosages of fluoride employed have been far beyond anything to be
encountered in the use of fluoridated water supply or (3) there are uncertain-
ties about study protocols. Examples of these "or.e of a kind" effects include
production of urinary calculi and effects or. collagen, plasma fibrinogen,
V-33
-------
, adenyl cyclase activity, enzymes, adrenal function, renal
stones, otosclerosis or mineral levels In different organs.
C. Summary
The intravenous LD5Q of fluoride in dogs is approximately 20 mg/kg. Dogs
survived oral doses of up to 3,100 ng NaF/kg. Age and sex influenced the
acute lethality of NaF in rats. In these studies, young rats
(seven-months-old or less) were less sensitive than older rats and young naie
rats were less sensitive than young females.
Chronic exposure of cattle to fluoride added to their ration caused
symptoms of dental and skeletal fluorosis. Fluoride added to the ration at 27
ppe (approximately 0.6& mg/kg/day) on a dry weight basis was tolerated with
only minor effect. Higher concentrations, 49 and 93 ppm (1.17 and 2.08
ng/kg/day, respectively) produced serious symptoms of dental and skeletal
fluorosi's. The appearance of dental fluorosis preceded that of skeletal
fluorosis. Milk production was impaired only after laneness and loss of
appetite were apparent. One study showed that heifers exposed to as little as
5 ppn fluoride in their drinking water suffered impaired reproductive perform-
ance, but other stud es found no effect on heifer reproduction at dier.ary
fluoride concentrations more than twice that amount. Sheep exposed chroni-
cally to 10 or 20 ppm fluoride in their drinking water developed significant
dental fluorosis and produced less wool. Weight gain was not affected and
there was no effect on reproductive perfonr-ance.
-------
Crovch in most species is unaffected by dietary concentrations of
fluoride of 100 ppm or less. Cattle appear to be more sensitive, and growth
has been reported to be affected slightly at 50 ppm. However, one
investigation found no adverse effects at 100 ppm.
The kidney responds to acutely toxic doses of fluoride by failure to
properly resorb water, leading to polyuria. Renal injuries do not develop
when drinking water contains less Chan 100 ppm fluoride.
In dogs, oral doses of 9 mg fluoride/kg have been reported to cause
hypotension, electrocardiogram irregularities .and slowing of the heart.
Structural and/or functional changes of the thyroid gland in animals are
not produced at fluoride concentrations of SO mg/L in the drinking water.
Sodium fluoride in drinking water was reported to be clastogenic for mice
but this result could not be confirmed in at least two other studies. Sodiun
fluoride was not mutagenic in Salmonella typhimuriuc or in Saccharomyces
j
cerevisiae. No information on the teratogenicity or carcinogenicity of
fluoride in animals was found in the literature.
V-35
-------
VI. HEALTH EFFECTS IN HUMANS
A. Beneficial Effects
1. Teech
The principal beneficial effect attributed to fluoride is its role in
prevention of dental caries. A detailed review of the literature in this area
will not be attempted here because it has been adequately addressed elsewhere
in this document. Studies have been reviewed that describe the continuum fror
beneficial effects to adverse dental fluorosis with increased exposure to
fluoride. A summary of the daily fluoride intake levels considered to be
protective against both dental caries and possibly osteoporosis is provided ir.
Table VI-1. A recent study by Driscoll et al. (1983) demonstrated that an
increase in the average drinking water fluoride concentration from 1.0* '.-.
2.08 mg F/L resulted in significant (P<0.05) reduction of dental caries of
school children (8- to 16-years-old) in several Illinois communities. The
authors noted, however, that communities with higher drinking water
concentrations (up to 3.8/ mg F/L) did not result in any additional
significant dental caries reduction at P<0.05 level. Further details of this
study (which aljo evaluated the incidence of dental f^uorosis) are discussed
in Section VI.C.3.
Fluoride is also believed to improve the esthetic appearance of teeth.
As part of the Newburgh-Kingston fluoride demonstration, A. L. Russell
» » ^
recorded the occurrence of .developmental enamel hypoplasias (not related to
fluoride in drinking water) in children 7- to H-years-old (Ast et al. 1956).
-------
Table VI-} Food and Nutrition Board Estimated Adequate and
Safe Intakes of Fluoride
Age
Group
<6 months
fc-12 months
!-:• years
4-6 years
7 years-
adulthood
Adults
Estimated
Weight (kg)
6
9
13 '
20
30a
70
Recommended Intake
of Fluoride (mg/day)
0.1-0.5
0.2-1.0
0.5-1.0
1.0-2.3
1.5-2.5
1.5-4.0
Estimated
Equivalences (mg/kg/day )
0.02-0.08
0.02-0.11
0.04-0.08
0.05-0.13
0.05-0.08
0.02-0.06
Estimated weight for childron seven to ten years old.
Adapted froit NAS (1980).
VI-2
-------
In Kingston, where the drinking water contained 0.05 n-.g F/L, 115 (18.7 percent)
of the 612 children examined showed these nonfluoride opacities. Only 36 (8.2
percent) of 438 children using the fluoridated Newburgh water (1.0 to 1.2 tng
F/L) showed these changes. Ast et al. (1956) suggested that this fluoride
drinking water concentration (1.0 to 1.2 ng F/L) appeared to reduce the
incidence of hypoplastic spots on the teeth. Richards et al. (1967N have
suggested that teeth classified as showing questionable, very mild or mild
dental fluorosis are desirable from an esthetic point of view.
2. Bone
The therapeutic use of sodium fluoride as a means of inducing new bone
growth in cases of osteoporosis is under active investigation. For example,
Jowsey et al. (1972) described the effects in 11 patients with progressive
osteoporosis who were administered 30, 45, 60 or 90 mg of KaF daily. In four
of the patients the dose was increased from 30 to 60 mg daily and in another
patient increased from 45 to 90 mg daily. In one instance the dose was
decreased from 60 to 30 mg per day. The patients, 10 of whom were female,
ranged from 54 to 72 years-of-age. All subjects received vitamin D twice
weeklv and a daily supplement of calcium. Treatment was continued for 12 to
17 months. Bone biopsy samples wer«> taken before, and afcer treatment and
biochemical studies were also performed. The results indicate that
administration of less than 45 mg oT NaF daily does not consistently increase
r
bone formation, but that 60 mg or more resulted in the production of abnormal
bone. Side effects were evident in at least one patient receiving 30 mg NaF.
» • -^
Mild arthralgia and stiffness of the joints were reported by four patients and
VI-3
-------
occasional epigastric dyspepsia was experienced by six patients. Dally
addition of vitamin D and more.than 600 mg Ca appeared to prevent Increased
bone resorptlon and even to decrease resorptlon. The authors concluded that
doses of 50 mg of NaF daily, supplemented with 600 mg or more of calcium dally
and 50,000 units of vitamin D twice weekly should increase skeletal mass
without undersirable skeletal effects. Also, further vertebral fractures
should cease after several years of treatment.
Dambacher et al. (1978) -reated 33 post-menopausal women with 100 mg NaF
daily for two years and another 23 similar patients with 50 mg NaF dally for
two >ears. A decrease of cortical bone was evident at both dose levels.
However, cancellous bone was increased to some extent in half of those
receiving the lower dose, and in over 70 percent of those receiving the higher
dose. The findings also suggested that two years of treatment at the lower
dose or one year at the higher dose avoided new vertebral fractures. Gastro-
intestinal discomfort sometimes combined with nausea was encountered chiefly
at the higher dose, but was of minor clinical importance. Osteoarticular pain
i
was the major side effect of fluoride therapy and was seen in about 60 percent
of the patients at both dose levels. The maximum effect was seen after 6 to
12 months of treatment and then gradually disappeared. In 18 percent of the
pat^encs treatment had to be discontinued.
More recently Rlggs et al. (1982) reported findings with regard to the
occurrence of vertebral fracture in post-menopausal osteoporosis. Five groups
of women, totaling 165 patients, were studied during the period from 1968 to
1980. Treatment regimens Included: (1) controls (placebo or no treatment),
-------
(2) calciun supplement with or without vitamin D, (3) fluoride plus calcium
with or without vitamin D, (4) estrogen and calcium with or without vltanin D
and (5) fluoride, calcium and estrogen with or without vlwanlii D. Fluoride
doses were 40 to 60 mg NaF daily with a total of 61 patients (of 165 total)
receiving fluoride. Of these, 23 (38 percent) developed adverse reactions
which caused five of them to withdraw from the study. Thirteen of the 23
patients had joint pain and swelling or painful plantar facial syndrome; nine
patients had severe nausea and vomiting, peptic ulcer or blood-loss anemia and
one patient had both rheumatic and gastrointestinal symptoms. These effects
were not seen in the control patients or in those treated with calcium alone
or with viamin D, or with calciuc plus estrogen with or without vitanin D.
Among these groups, vitamin D was discontinued or the dose reduced because of
hypercalcenia or hypercalciuria. Thirteen percent of the 60 patients
receiving estrogen required hysterectomy or uterine dilation and curettage,
but none had endometrial carcinoma or vascular thronbotic events.
Among the patients treated with NaF, 60 percent showed radiographically
demonstrable increases in vertebral bone mass. Patients with these changes
showed about one-seventh the fracture rate of the other patients. The
incidence of fractures per 1000 patient-years for patients treated with
fluoride, calciac and estrogen (with or without vitait^n D) was significantly
less than in controls (P<1 x 10~ ), and also was significantly less than in
those treated with fluoride and calcium (with or without vitamin D) (P
-------
inclusion In a combined therapy may not be warranted. The authors believe
vitamin D should not be Included because of the Increased incidence of
hypercalcemia or hypercalciuria or both.
Berstein et al. (1966) compared the Incidence of osteoporosis, reduced
bone density and collapsed vertebrae in two populations using water supplies
with different concentrations of fluoride. In this study, a roentgenogram of
the lateral lumbar area- of the spine and answers to a questionnaire were
obtained for 166 males and for 134 females who were long-term residents of
areas where the water supplies contained to It to 5.8 mg F/L. Similar informa-
tion was obtained for 312 male and 603 female long-tenr. users of water supplies
containing 0.15 to 0.3 mg F/L. More than 50 percent of the participants in
each area had never lived outside their respective areas. The subjects of
each sex in each population were grouped by age into those 45- to 54-years-old,
55- to 64-years-old and 65-years-old and over. Evidence of osteoporosis,
reduced bone density and incidence of collapsed vertebrae were higher in the
low fluoride area in both sexes. For women 55- to 64-years-old and 65-years-old
and older the difference in prevalence of reduced bone density was significant
at the P<0.01 level. In men the difference was significant only for the 55-
to 64-year-old group (P<0.05)- More subjects in the high fluoride area had
normal or increased bone density. T..ere was no significant difference in the
Incidence of collapsed vertebrae among male residents of the two areas. For
women, the greater incidence of collapsed vertebrae in the low fluoride area
was significant at the P<0.05 and P<0.01 levels for the 55- to 64-year-cld and
the 65-year-old and over groups, respectively. The authors concluded that
VI-6
-------
t> to 5.8 mg F/L In drinkinjt water "materially and significantly" reduced the
prevalence of osteoporosis and collapsed vertebrae, and that the effect* were
more pronounced in women than in men.
Using data from the. 1973 to 1977 National Health Interview Surveys,
Madans et al. (1983) compared the incidence of hip fractures as a measure of
osteoporosis in two populations whose water supplies contained different
•concentrations of fluoride. Mere than 80 percent of 30,£.73 females plus
25,997 males used water containing less than 0.7 mg F/L. At least 80 percent
of 21,810 females plus 18,034 males used water with more than 0.7 mg F/L. The
hip fracture hospitalization rate for females in the low and high fluoride
areas were 2.4 per 1000 and 2.2 per 1000, respectively. For males the rates
were 1.0 and 1.1, respectively. The data suggest that a concentration of 0.7
mg F/L is not sufficient to protect against osteoporotic hip fracture. It was
possible to identify 1,2'-2 women and 1,111 men, 40 years of age or over, who
used water supplies containing mo..e than 0.7 mg F/L (specific concentrations
could not be identified). Among these persons there was one case of hip
fracture in the males and none in the females. The authors suggested that the
hypothesis of a protective effect of higher levels of fluoride among women
should not be ignored and that the optimal level exceeds 0.7 mg F/L.
3. Cardiovascular
In the study by Bernstein et al. (1966) the incidence of aortic
calcification (as seen in the X-ray films) was less in residents of the high
•
fluoride area than in those using low fluoride water. The difference was
Vl-7
-------
approximately 4P percent and W«P statistically significant for men In all age
groups'. Women In the 55- to 64-year-old group also showed a statistically
significant difference in the Incidence of aortic calcification. A similar
trend, although not statistically significant, was observed in females
j65-years-old* and older.
I*. Hearing
Shambaugh and Causse (1974) treated more than A,000 patients with active
otospongiosis of the cochlear capsule with sodium fluoride for 1 to 8' years,
using doses of 40 to 60 mg daily with calcium and vitamin C supplements. The
fluoride was administered in enteric coated tablets. In abouc 80 percent of
the patients so t. aated there was a stabilization of the sensorineural
component of hearing loss, with recalcification and inactivation of the
actively expanding demineralized focus of otospongiosis. In a few cases
hearing was improved, while in others the hearing loss continued to worsen.
In a number of instances, cessation of therapy after stabilization of hearing
and recalcif^cation had been achieved was followed (two to seven years later)
by reappearance of a demineralized focus and an increase in the sensorineural
loss. Shambaugh and Causse (1974) reconmend a maintenance dose of 20 mg daily
of sodiur. fluoride after stabilization has been achieved.
Causse et al. (1980) gathered more evidence for the beneficial effect of
fluoride therapy on otospongiotic foci through polytomographic studies,
statistical analysis of 10,441 cases (with a follow-up of three months to ten
VI-8
-------
years) and by comparing trypslr concentration In the perilyu-.ph before and
after NaF therar-v. Trypsln, which Is toxic to hair cells and destroys
collagen fibrils in the bony otic capsule, was significantly (no P value
given) reduced in 66* of cases at moderate NaF (45 ing/day) doses. Fluoride
therapy causes expulsion of cytotoxlc enzymes into labyrinthine fluids and
retardation of sensorineural deterioration. The long-term effect of therapy
Is che reduction of the'bone remodelling activity of the otospongictic focus.
NaF therapy (in patients with cochlear deterioration and progressive cochlear
component) can improve hearing in children but can only arrest deterioration
in older patients. NaF may retard, but cannot release, stapedial fixation.
Fluoride action reduces vertigo as an effect on vestibular function. Dosages
used by the authors range from 3 to 60 mg/day depending on the nature of the
otospongiotic inspairnient (in children only 1.5 to 10 mg/day are prescribed to
avoid stunting growth). The authors observed no fluorosis in more than 10,000
cases.
5. Other
Black et al. (1949) examined the toxicity of sodiuir fluoride in relation
to the beneficial effects of fluoride therapy in the treatment of malignant
neoplasia. They described the effects of fluoride administered to more than
7C patients for periods of 5 to 6 months. Most pf these subjects, suffering
from malignant neoplastic disease, were being treated with metabolic
inhibitors. Some were '.eukemic children 3- tc 6.5-yeara-cid, whilp others
were adults including elderly Individuals. Doses for. the children were 2<* to
50 mg NaF (9.0 to ?2.5 mg F) four times dally. Doses for adults were 80 rag
VI-9
-------
Na* J36.3 mg F) four times dally. The material was administered orally with
ar. antacid containing 4 percent aluminum oxide or as an enteric coated tablet
to avoid gastric Irritation. No evidence of systemic toxicity or of
parenchymatous damage was seen which could be attributed to fluoride, even
though some patients had received more than 27 g of sodium fluoride over a
period of 3 months. Criteria evaluated included growth and development in the
children, mottled enamel, eruption of permanent teeth, hematopolsis, liver
function, albumin-globulin ratio, blood sugar and cholesterol concentrations
and kidney function. Postmortem data from 4 cases showed no parenchymatous
degeneration attributable to fluoride. In hypertensive patients a tendency
was noted for decreased diastollc and systolic blood pressure. In two
patients with functioning colostomles there was no apparent effect of the
fluoride on the exposed mucosa of the colon.
In certain instances, Black et al. (1949) administered fluoride by
intravenous infusion or injection. For example, a 16-year-old female with an
adrenal carcinoma received a total dose of 5600 mg of sodium fluoride
(2,533 mg F) in a period of nine consecutive days. There were no signs of
acute or chronic toxicity. The injection of 400 mg of KaF (180 mg F) was
painful in two of three instances, as was the injection of 800 mg of SaF (360
mg F). However, when infused, this Amount was not painful.
B. Acute Toxicity
Lidbeck et al. (1943) described a mass poisoning in which 17 pounds of
roach powder containing NaF was inadvertently added to a ten-gallon mixture of
VI-10
-------
scrambled eggs. Two hundred and sixty-three cases of acute poisoning resulted
and 47 of these were fatal. The episode Is described by the authors as
follows:
The food was rejected by many of the patients because of a salty or
soapy taste, while others complained of numbness of the mouth.
Extremely severe nausea, vomiting and diarrhea occurred abruptly and
at times simultanr usly, and blood was noted in the vomitus and the
stools In many instances. Soon after the meal there were complaints
of abdominal burning and crampllks pains. General collapse developed
in most instances but at variable periods of time, apparently
depending on the concentration of the poison. This was character-
ized by pallor, weakness, absent or thready pulse, shallow unlabored
respiration, weak heart tones, wet cold skin, cyanosis and equally
dilated pupils. When this picture was pronounced, death almost
invariably occurred. Local or generalized urticaria occurred in
some instances, while in others there was a thick mucold discharge
from the mouth and nose. When death was delayed, and in some cases
in which recovery occurred, there were paralysis of the muscles of
deglutition, carpopedal spasm and spasm of the extremities. -Convul-
sions and abdominal tenderness and rigidity were absent. In the
majority of cases, death occurred between two and four hours after
ingestion of the food, although in a few instances death was delayed
for eighteen to twenty hours.
Hodge and Smith (1965) tabulated numerous reports of accidental and
intentional poisonings with fluoride and concluded that a dose range of 5 to
10 grains of NaF can be cited as a reasonable estimate of a "certainly lethal
[single] dose" for a 70-kg man. They noted that this corresponds to 70 to UO
mg/kg.
C. Chronic Toxieity
Prolonged exposure to excessive fluoride is known to cause skeletal and
dental fluorosis. These effects will be described in detail, however, this
VI-11
-------
section will first address occasional reports which have appeared in the
literature suggesting a wide variety of toxic effects of fluoride exposure.
These include abnormal sensitivity to fluoride (Grlmbergen 1974, Waldbott
1962), mongolism (Rapaport 1959), a decreased margin-of-safety in people with
renal insufficiency (Hanhijarvl et al. 1972, Juncos and Donadio 1972) and
cancer (Yianouyia'nnis and Burk 1977). The reports on mongolism and cancer
will be discussed in Sections VI. D. and VI. F., respectively.
1. Sensitivity to Fluoride
Allergic or idiosyncratic sensitivity to fluoride has been the subject of
a number of reports iGrimbergen 1974, Waldbott 1962, for others see NAS 1977).
These studies contained various weaknesses in experimental design and in
statistical analysis which have been discussed in a report by the NAS Safe
Drinking Water Committee (NAS 1977). Waldbott presented case reports of
people who were allegedly sensitive to fluoride. The NAS notes that this
report has been criticized because Waldbott was, for some time, the only
investigator to have reported this type of sensitivity. The study by
Grimbergen (1974) appears to support the interpretation by Valdbott.
Grlmbergen administered either NaF in water or a placebo in a double-blind
tee..' to subjects who were suspected of being sensitive to fluoride. However,
the statistical analysis of these data has been questioned by the NAS. They
note that when a large number of samples are taken, some positive responses
would be expected by chance. Grimbergen did not address this issue. Doubt
that true sensitivity to fluoride exists has also been expressed by the World
VI-12
-------
Health Organization (WHO 1970). They reason that billions of people worldwide
are regularly exposed to fluoride through tea drinking (brewed tea having more
fluoride than the water from which it is made) or water fluoridation, so any
subpopulatlon that is sensitive to fluoride should be readily apparent.
2. Bone
A number of factors govern the amount of fluoride in the skeleton.
Important among these are (1) previous exposure, especially to a relatively
constant fluoride intake; (2) the dose, which is reflected in the blood
concentration; (3) renal status, which also affects the blood concentration,
and (4) the age of the individual (Hodge and Smith 1981).
Endemic skeletal fluorosis is recognized in several parts of the world;
it was«first described in India (Shortt et al. 1937a). The most severe forms
of generalized osteosclerosls also have been reported from this country. The
4
findings in a 45-year-old Indian farmer suffering from fluorosis have been
described in detail by Singh et al.
-------
some were fused together. Bone ash frop. this subject contained 6,300 mg fluoride/
kg of ash (6,300 ppm). The mechanical properties of the left radius and ulna
of this subject were tested by Evans and Wood (1976). Their results shoved
that tensile strength, strain, energy adsorbed to failure and modulus of
activity were reduced; compressive strength, strain and energy were increased.
Compressive properties exceeded tensile properties; bone density was increased.
It should be recognized that the severe changes reported in areas of
severe fluorosis such as the Punjab are not necessarily seen in all residents.
Factors affecting the incidence of skeletal fluorosis include duration and
level of exposure to fluoride In the environment, nutritional status,
concurrent infections and physical severity of the individual's occupation
(Singh and Jolly 1970).
Roholm (1937) identified three stages in the progression of skeletal
fluorosis. These are (as quoted in Smith and Hodge 1979):
r
Phase I: osteosclerosls in pelvis and vertebral column. Coarse and
blurred trabeculae, diffuse increased bone density to X-ray.
Phase II: increased density and blurring of contours of pelvis, verte-
bral column extended to nos, extremities.
Phase III: greatly increased density of bone; irregular and blurred
contours. All bones affected, particularly cancellous bones.
Extremities thickened. Considerable calcification of ligaments of
neck and vertebral column.
-------
In addition, Fritz (1958) recommended adding two earlier stages of
fluoride-induced changes, the earliest labeled "subtle signs," the second
"phase 0-1." Both are characterized by slight radiological changes, e.g.,
enlargement of trabeculae In the lumbar spine. Both of these classification
schemes have been developed from experience with industrial exposure to
fluorides. Singh and Jolly (1970) point out that Roholm's Phase I.Is hardly
ever seen in endemic fluorosis cases; most of these show the changes of phases
II and III.
Franke et al. (1975) and Schlegel (1974) have attempted to correlate the
concentration of fluoride in bone ash with the various osteosclerotic phases,
as shown in Table VI-2. These data indicate that the early detection of
slight radiological changes, e.g., enlargement of trabeculae in the lumbar
spine, will be associated with bone ash fluoride concentrations of 3,500-4,500
ppm.
There is limited evidence to permit an estimate of the waterborne
fluoride concentration associated with the appearance of fluoride
osteosclerosis. For example, Hodge and Smith (1970) quote evidence that in
the aluminum industry, average urinary excretions of 5 og F/L in randomly
collected samples fc-e not associated with osteoselerosis. Dirman et al.
(1976) indicated that aluminum workers whose average pre-shift urinary
fluoride concentration is less than 4 mg F/L do not show radiographically
demonstrable increases in bone density, altered trabecular patterns or
ligamentous calcification. According to Figure III-3 (see Section III for
greater detail), these urinary fluoride concentrations correspond to
VI-15
-------
Table Vl-2 Correlation of Osteosclerotic Phases and Fluoride in Bone Ash
MB Fluoride/kg Bone Ash (ppm)
Osteosclerotic Phase Franke et al. (1975) Schlegel (1974)
Normal 500-1,000
Fritz
Prestage 3,500-4,500
0-1 5,000-5,500 6,900
Roholm
I • 6,000-7,000 5,200
II 7,500-9,000 7,500
III MO.OOO 8,400
Adapted from Smith and Hodge (1979).
VI-16
-------
waterborne fluoride concentrations of approximately 5 and 4 mg F/L,
respectively. Smith and Hodge (1959) have suggested that, in the human,
osteosclerosis probably will not be seen with skeletal fluoride concentrations
of 4,000 ppm (dry fat-free basis). They also state that effects will be
observed In a small proportion of individuals with skeletal fluoride
concentrations of approximately 6,000 ppm. These skeletal concentrations
correspond to fluoride concentrations in the water of A and 6 mg F/L,
respectively'(Smith and Hodge 1959, Hodge and Smith 1981). It should be
pointed out that at least at levels of intake corresponding to ^ 0.1 ppm
fluoride in the water, skeletal fluoride concentrations may vary up to ±507.
(Smith 1983b).'
3. Teeth
The tendency for excessive exposure to fluoride for prolonged periods
during the time of tooth formation to cause fluorosis of dental enamel is of
concern. Although the causative agent was-not known at the time, a report of
dental fluorosis (then called "mottled enamel") appeared in 1901 - Denti di
Chlaie (Eager 1901). This report described the condition in certain Italian
immigrants. Black and McKay (1916) and Kempf and McKay (1930) reported the
same condition was endemic in parts of the U.S. Experimental data suggesting
the connection between exposure to excessive amounts of fluoride and abnor-
malities in teeth appeared in 1925. McCollum et al. (1925) noted effects of
dietary fluorine on the teeth of white rats and similar findings were reported
by Schultz and Lamb (1925). The connection between fluoride and mottled
enamel was.first recognized by Smith et al. (1931).
VI-17
-------
Dean (1*33) reported on the distribution of mottled enamel in the U.S.,
and in 1934 he published a classification system for mottled enamel (Dean
1934). This classification system is provided in Table VI-3.
Subsequently, Dean published a revised classification system for dental
fluorosis (Dean 1942) which is still in use today. This system comprises six
classifications into which the individual child or tooth may be assigned.
Classification of an individual child is based on the two teeth in the child's
mouth that show the most advanced forms of fluorosis. Since the classification
of the severity of dental fluorosis is critical to the regulation of fluoride
in drinking water, Dean's revised system is given in Table VI-4.
Dean (1942) used this system as the basis for defining a Community
Fluorosis Index (CFI). The CFI is a -means of .comparing one group or popula-
tion with another on the basis of average severity of fluorosis. It is
computed by averaging the numerical fluorosis scores assigned to individual
children within a given population.
Dean and Elvove (1935) defined the permissible maximum level of fluoride
in a domestic water supply (or minimum threshold for dental fluorosis) as the
highest concentration of fluoride iicapable of producing a definite degree of
dental fluorosis in as much as 10Z of the group examined. The group examined
for purposes of defining the CFI should consist of at least 25 children,
9-years-old or older, who, since birth, have continually consumed the water
under investigation (i.e., used the water for both drinking and cooking). The
VI-18
-------
Table VI-3 Dental Fluorosis Classification by H. T. Dean - 1934
Classifica-
tion
Criteria
Normal The enamel presents the usual translucent semivitriforo type of
structure. The surface is smooth and glossy and usually of a
pale creamy white color. For purposes of classification, all
persons showing hypoplasia other than mottling of enamel are
included in this category.
Questionable There are slight aberrations in the translucency of normal
enamel, ranging from a few white flecks to occasional white
spots, 1 to 2 mm in diameter.
Very mild Small opaque paper white areas are scattered irregularly or
streaked over the tooth surface. It is principally observed on
the labial and buccal surfaces, and involves less than 25% of the
tooth surfaces of the particular teeth affected. Small pitted
white areas are frequently found on the summit of the cusps. So
brown stain is present.
Mild The white, opaque areas on the surfaces of the teeth involve at
least half of the tooth surface. The surfaces of molars, bicus-
pids, and cuspids subject to attrition show thin white layers
worn off and the bluish shades of underlying normal enamel.
Faint brown stains are sometimes apparent, generally on the upper
incisors.
Moderate No change is observed in the form of the tooth, but generally all
of the tooth surfaces are involved. Surfaces subject to attri-
tion are definitely marked. Minute pitting is often present,
generally on the labial and buccal surfaces. Brown stain is
frequently a disfiguring complication. It must be remembered
that the incidence of brown stain varies greatly in different
endemic areas, and many cases of white opaque mottled enamel,
without brown stain, are classified as "moderate".
Moderately Kacroscopically, a greater depth of enamel appears to be
Severe involved. A smokey white appearance is often noted. Fitting is
more frequent and generally observed on all the tooth surfaces.
Brown stain, if present, is generally deeper in hue and involves
more of the affected tooth surfaces.
Severe The hypoplasia is so marked that the form of the teeth is at
times affected. The pits are deeper and often confluent. Stains
are widespread and range from chocolate brown to almost black in
some cases.
VI-19
-------
Table Vl-6 Dental Fluorosls Classification by K. T. Dean - 1962
Classifica-
tion
Criteria
Normal
(0)
The enamel represents the usual translucent semivitriform type of
structure. The surface is smooth, glossy, and usually of a pale
cre-aoy white color.
Questionable The enamel discloses slight aberrations from the translucency of
(0.5) normal enamel, ranging from a few wnite flecks to occasional
white spots. This classification is utilized in those instances
where a definitive diagnosis of tne mildest form of fluorosis is
not warranted and a classification of "normal" is not justified.
Very Mild Small, opaque, paper white areas scattered irregularly over the
(1) tooth but not involving as much as 25 percent of the tooth
surface. Frequently included In this classification are teech
showing no more than about 1-2 mm of white opacity at the tip of
the summit of the cusps of the bicuspids or second molars.
•
Mild The white opaque areas in the enamel of the teeth are more
(2) extensive but do not. involye as much as 50 percent of the tooth.
Moderate All enamel surfaces of the teeth are affected, and surfaces
(3) subject to attrition show wear. Brown stain is frequently a
disfiguring feature.
Severe All enamel surfaces are affected and hypoplasia is so marked that
(A) the general fbm of the tooth may be affected. The major
diagnostic sign of this classification is discrete or confluent
pitting. Brown stains are widespread and teeth often present a
corroded-like appearance.
VI-20
-------
authors determined the CFI for four communities: Colorado Springs, Colorado;
Monniouth, Illinois; Galesburg, Illinois; and Pueblo, Colorado. The mean
annual fluoride content of the municipal water supplies and corresponding CFI
for these communities were: Colorado Springs, 2.5 ppji - "slight"; Monmouth,
1.7 ppo - "slight"; Galesburg, 1.8 ppm - "slight"; and Pueblo, 0.6 ppm -
"negative".
Galagan and Lamson (1953) studied the relationship between fluoride
concentration in municipal waters and the CFI for communities with differing
mean annual temperatures. They found that Arizona communities with mean air
temperatures of 70°F had "objectionable" fluorosis (CFI exceeding 0.6) at
about 0,8 mg fluoride/L in their drinking water, while.inidwestern communities
with mean air temperatures of 50°F did not suffer "objectionable" fluorosis
until their drinking water contained about 1.7 mg fluoride/L '.see Figure VI-1),
Richards et al. (1967) established slightly different optimal values for
fluoride, using only three temperature zones, that were generally in agreement
with the earlier studies. The authors pointed out that fluorosis was not
entirely absent at optimum fluoride concentrations in drinking water. Their
goal was to establish the fluoride levels at which "objectionable" fluorosis
was present; objectionable was defined as moderate and severe fluorosis. The
results of thif study are summarized in Table VI-5. .-.t 'should be noted,
however, that a recent study in Canada (EHD 1982) concluded that water
consumption is independent of temperature. Thus, the Agency has concluded
that there is insufficient data to quantitatively incorporate tenperature in
drinking water regulations.
VI-21
-------
2
o
IX)
0.8
0.6
0.4
Mean Annual Temperatures
O Approximately 70* F
• Approximately SO* F
0.4
Objectionable fiuorosis
(Fluoride removal indicated
Borderline
Negative
0.6
0£
1.0
1.4
1.6
1-8
2.0
FLUORIDE CONCENTRATION (ppm)
Adapted from Galagan and Lamson (1953)
Figure VI-1 Relationship Between Fluoride Concentration of
Municipal Waters and Fluorosis Index for Communities
with Mean Annual Temperatures of Approximately 50 F
(Midwest) and 70° F (Arizona).
VI-22
-------
Table VI-^ Percentage of Children by Fluorosls Diagnosis fcr Each
Fluorlde-Tecperature Zone
Fluorosls
diagnosis
and temperature
(Mean maximum)
65T or lover
Normal
Questionable
Very mild
Mild
Modei ate
Severe
66'F - 79°F
Normal
Questionable
Very mild
Mild
Moderate
Severe
«•
80"F or higher
Normal
Questionable
Very mild
Mild
Moderate
Severe
Fluoride concentration
9-0-0.15
Zone 1
(Na - 330)
97.3
2.4
0.3
- —
Zone 2
(K - 707)
96.1
3,5
0.4
___
__.
Zone 3b
(N - 209)
52.6
46.9
0.5
_ —
___
0.2 - 0.4
Zone 4
(K - 169)
71.6
26.0
2.4
Zone 5
(K - 709)
74.2
19.5
6.2
0.1
Zone 6
(N - 335)
32.2
44.8
20.0
3.0
0.5 - 0.7
Zone 7
(N - 340)
44.7
40.9
13.5
0.9
Zone 8
(N - 688)
26.6
42.9
28.6
1.9
Zone 9
(K - 331)
18.1
51.1
26.0
4.2
0.6
In drinking
0.8 - 1.0
Zone 10
(N - 316)
40.0
39.2
18.0
2.8
Zone 11
(K - 548)
22.8
44.3
26.6
5.8
0.5
Zone 12
(K = 350)
18.3
26.0
37.7
15.1
2.9
water fppn)
1.1 - 1.3
Zone 1 3
(N - 302)
33.1
41.1
22.5
3.3
— — -
Zone 14
(N - 508)
26.6
32.7
28.1
9.6
2.8
0.2
Zone 1 5
(N - 310)
8.4
29.0
37.5
17.4
7.4
0.3
1.3
or more
Zone 16
(N - 306)
11.1
23.5
29.5
15.7
15.0
5.2
Zone 1 7
(N - 553)
14.8
18.4
27.8
20.8
12.8
5.4
Zone 18
(N * 229)
8.3
18.8
25.3
27.9
12.7
7.0
*N - number of children diagnosed.
Fluoride concentration =0.2 ppm.
Adapted from Richards et al. (1967).
VI-23
-------
Because the relationship between fluoride concentrations in drinking
water and cotanunity fluorosis indices was established many years ago, a demand
has arisen for evidence confirming or re-establishing the fluoride/fluorosis
relationships. Segreto et al. (1984) investigated the possibility that
significant changes in cultural and dietary pattern nay have altered fluoride
intake patterns from those developed 20 to 40 years ago. They selected 16
Texas communities that obtain their drinking water from local wells and
surveyed children (7- to 18-years-old) who were lifetime residents of each
community'for enamel mottling using Dean's (1942) classification system. The
fluoride levels in the drinking water was expressed by the authors in terms of
the relationship to optimal for prevention of dental caries. Personal
communication with one of the authors (Dr. Edwin M. Collins), however,.
indicated that the actual fluoride levels ranged from 0.2 to 3.2 mg/L (see
Table VI-6). The combined incidence of moderate and severe dental fluorosis
observed ranged from minimal at 0.2 mg F/L to 31.6 percent at 3.2 mg F/L. The
authors, however, reported only one case of severe fluorosis (at 3.2 mg F/L).
The observed variation in the fluo»->sis incidence of different fluoride
drinking water levels (see Table VI-6) could be due to differences in the
lifestyle in the different communities, variation in the susceptibilities of
the children examined or other factors.
Driscoll et al. (1983) reported the results of a cross-sectional survey
of the prevalence of dental fluorosis and dental caries among 807 school
children- (8- to 16-years-old) in seven Illinois communities. Fluoride
concentrations in community drinking water ranged fron 1.06 to 4.07 mg F/L.
The results of this study are summarized in Table Vl-7 and indicate a
VI-24
-------
Table Vl-fc Relationship Between Fluoride Levels in Drinking
Water and Incidence of Moderate and Severe Dental
Fluorosis in Texas Children (Age 7 to 18 Years)
Fluoride level
in drinking water
Relative
to
mg F/L optimum
0.2 0.3 .
0.3 0.4
0.4 0.3
0.8 1.0
1.1 1.3
1.1 1.4
1.1 1.3
1.6 2.5
1.9 2.7
1.9 2.3
2.0 2.3
2.0 2.7
2.3 2.7
2.3 2.9
2.4 3.1
3.2 4.3
Number of
children
exan. .sd
103
126
223
361
211
128'
187
301
170
' 23
109
200
90
67
113
190
Combined incidence of
moderate and severe.
dental fluorosis, Z
0.0
0.0
0.0
0.3
0.9
0.0
1.1
3.3
13.5
13.0
14.7
4.0
6.7
32.8
4.4
31. 6C
Actual concentrations not reported by authors. Values obtained
.through Personal communication with coauthor (Dr. E. M. Collins).
All cases were classified as moderate dental fluorosis except
as noted by footnote.
C0ne case (0.5 percent) was classified as severe dental
fluorosis.
Adapter1 froc Segreto et al. (1964).
VI-25
-------
Table VI-7 Relationship of Drinking Water Fluoride Levels To Dental
Fluorosls and Caries Reduction in Illinois Children
Fluoride level
in drinking
water, mg/L
Number of
children
evaluated
Children with
moderate and
severe dental
fluorosis, 7.
Decrease in
caries score
from 1.06 mg/L
level, Za
1.06
2.08
2.84
3.84
336
143
192
136
2.4
13.3
27.6
30.2
37.3:
55.i:
35.7C
^"Measured as mean DKF surface score.
Significantly different (P<0.05) from score at 1.06 mg/L,
but not from each other.
Adopted from Driscoll et al. (1983).
VI-26
-------
dose-response increase in Che incidence of moderate and severe dental
fluorosis with increased fluoride level in the drinking water. The Incidence
of moderate and severe dental fluorosis ranged from 2.4 percent (of 336
children evaluated) at 1.06 mg F/L to 30.2 percent (of 136 children
evaluated) at approximately 3.84 mg F/L. Concurrent with this increase in
dental fluorosis, the authors observed a significant (P<0.05) decrease in
dental caries (as measured by reduction of mean DMF surface score) in children
of all fluoride levels above 1.06 mg F/L. Unlike the dental fluorosis
results, the dental caries reduction was not observed to exhibit a
dose-response relationship above the level of 2.08 mg F/L in the drinking
water. There was no statically significant (P<0.05) difference in the
reduction of dental caries between children exposed to an average 2.08 mg F/L
through 3.84 mg F/L.
Wenzel and Thylstrup (1982) have suggested that a clinicohistological
classification of dental fluorosis may be more sensitive than that described
by Dean (1942).
4. Kidney
It r'.oes appear that patients with renal ir jairment have a lower
margin-of-safety to fluoride effects than the average person. Hanhijarvi et
al. (1972) measured plasma levels of free ionized fluoride in about 2,000
hospital patients in Finland. In patients with normal creatinine clearance,
plasma fluoride from individuals, living in non-fluoridated areas was about
one-half that of people from fluoridated areas (0.7 uM vs 1.4 uM). The
VI-27
-------
authors also noted a correlation between serum creatlnine and serum fluoride
in renal patients from both the fluoridated and the non-fluoridated areas.
Fluoride increased with increasing concentration of serum creatinine in one
patient from the non-fluoridated area (serum fluoride increased from 0.8 uM to
3.A uM while serum creatinine rose from normal to 1,200 uM). In a patient
from the fluoridated area, serum fluoride rose from 1.4 uM to 5.0 uM and the
corresponding serum creatinine from normal to 700 uM. In renal patients
undergoing dialysis., serum fluoride concentrations as high as 25 uM were
recorded. These results are consistent with the work of Berman and Taves
(1973) who measured renal clearance of serum fluoride in normal and in uremic
patients. Normal fluoride clearance averaged 58 mL/min while uremic subjects
had a mean fluoride clearance of 3.1 mL/min.
A correlation between renal failure, polyuria, polydipsia, and clinical
and rpentgenographic evidence of systemic fluorosis was reported by Juncos and
Donadio (1972). They discussed two case reports. In one, an 18-year-old male
had a daily consumption of about 2 gallons of water from an artesian well
containing 2.6 ppa fluoride. His teeth were mottled, very opaque, and caries-
free. The patient's normal daily urine volume was 5 to 6 L and clinical
indices of renal function were abnormal:- inulin clearance (C. ) was 26 mL/min
in
(vs 120 mL/min normal), PAH clearance (C .) was 118 mL/mi-i (vs 600 mL/min
pan
normal). Roentgenograms showed increased density of bones. The patient's
.intake of fluoride from drinking water was about 0.33 mg/kg/day (based on a
body weight of 57.4 kg). Similarly, a 17-year-old female with significantly
impaired renal function (C. » 19 mL/min; C . , -99 mL/min), a history of
drinking "large amounts of water" and teeth which were opaque with diffuse
VI-28
-------
brownish mottling was found to have marked reduction in renal size, blunting
of the calyces, pyelocaliectasis and ureterectasis. The authors did not know
whether chronic excessive fluoride intake caused the renal damage but did
believe that the systemic fluorosis was due to Impaired renal function.
Oreopoulos et al. (1974) examined the effect of fluoride in the dlalysate
of patients undergoing chronic renal dialysis. In a double-blind study, 20
patients (11 fluoride-exposed and 9 controls) were -nvestigated for an average
period of 20.6 months. Dialysate water was initially deionized and fluoride
(1 tag/L) or chloride (control) was added via coded ampules. At the end of the
study the only difference detected between the control and exposed groups was
a statistically significant (P<0.05) increase in osteosclerosis in the
fluoride-exposed group. No differences in various biochemical, radiological
or other histological parameters were detected.
No injuries to the human kidney from long-term non-occupational exposure
to fluoride have been reported. Geever et al. (1958) did not find an unusual
incidence of renal pathology or renal disease as a cause of death in a
population using water containing 2.5 ppm fluoride. No differences in renal
status were evident between the residents of Bartlett (8 ppn fluoride in the
water supply", and Cameron (0.4 ppm fluoride), Texas (Leone et al. 1954).
Urinary excretion of albumin, sugar, red blood cells and formed elements by
the children from Newburgh (1.2 ppm fluoride) did not differ significantly
from that of the children from Kingston (essentially no fluoride) (Schlesinger
et al. 1956a). Abnormalities In renal function, e.g., decreases in urea
clearance and glomerular filtration rate, have been reported in. Indian
VI-29
-------
subjects with advanced skeletal fluorosis (Shortt et al. 1937a). Water
supplies used by these patients contained up to 10 ppm of fluoride (Shortt et
al. 1937b).
5. Growth
A possible depression in height, weight and chest circumference has been
reported in Japanese children with mottled enamel, compared to control subverts
whose teeth were not mottled (Takamori 1955). Water supplies used by these
children contained as much as 3.4 ppm fluoride. However, the absence of
adequate information on the nutritional status, hereditary background and
general state of health of these children makes it difficult to accept these
findings as valid. Such, observations have not been made in this country. For
example, in the Newburgh-Kingston area in New York State, Schlesinger et al.
(1956b) found no significant differences in height or weight between the
children using fluoridated water for ten years (Newburgh, 1.2 ppm fluoride)
and the control population (Kingston, essentially no fluoride). McClure
t
(1944), in a survey of high school'boys and young adults living in areas where
the water supplies contained up to .6 ppm fluoride, found height and weight to
be unrelated to fluoride exposure.
6. Cardiovascular System
Analysis of the death rates from cardiovascular-renal disease in Newburgh
and Kingston demonstrated no significant difference in this respect between
the two communities (Schlesinger et al. 1956b). Rogot et al. (1978) also
VI-30
-------
found no effect of fluoride in water on heart death rate trends. Geever et
al. (1958) reported a lessor percentage of deaths due to cardiovascular
disease in persons who had lived more than 20 years in a community where the
water supply contained 2.5 ppm fluoride than in persons living 5 to 20 years
in that community, but the difference was not attributed to a protective
effect of fluoride. A lower incidence of deaths due to heart disease was
shown in 20 towns using fluoridated water, compared to 15 towns where the
water was not fluoridated (Taves 1978). Luoma et al. (1973) found an inverse
correlation between the percentage prevalence of heart disease in male
residents of four Finnish communities where fluoride in the drinking water
ranged between 0.05 and 2.57 ppm.
Okushi (1954) and Takamori (1955) described mycardial changes seen in
children and adults using water supplies containing 0 to 13 mg F/L. The
changes described were shown by X-ray or electrocardiography. Unfortunately.
in most instances only ranges of fluoride drinking water concentration are
given and specific concentrations cannot be associated with the observed
changes. However, from a careful examination of the tabular data presented by
Okushi (1954), it appears that the lowest-observed-adverse-effect level
(LOAEL) was 2.5 mg F/L. Changes observed at this dose included myocardial
damage, sinus tachycardia and prolonged P-R ar1 Q-T intervals. One 12-year-
'Old boy. consuming water with 2.5 mg F/L showed.no signs of myocardial injury.
Morever, there may well have been a number of subjects unaffected at this
concentration, in as much as the findings were negative in 14 to 16 subjects
using waters containing 1.9 to 4.8 mg F/L, but for whom specific concentrations
• >
were not identified. Also, positive effects were seen in one child and two
VI-31
-------
adults for whom specific fluoride drinking water concentrations were not
specified. Dr. B. Lawrence Riggs (personal communication) was unable to
observe any electrocardiographical effects in patients receiving 30 to 63
mg NaF/day or 13.6 to 29.5 mg F/day. Dr. Riggs confirmed this statement after
reexamination of the data.
7. Thyroid
No significant effects on the incidence of abnormal clinical findings
related to the thyroid gland were seen in long-term residents of Bartlett,
Texas, where the water supply contained 8 ppm fluoride (Leone et al. 1954).
Geever (1958) examined thyroids t; K;T -•_ autopsy from 728 persons who had used
a water supply containing 2.5 ppm t . joride for periods of less than 5 to
more than 20 years. Prolonged use of this water did not significantly -affect
•
the incidence of pathological findings in this gland.
D. Teratogenicity
\
The study by Rapaport (1959) suggested a dose-related association'between
the number of cases of mongollsm registered in institutions and the concentra-
tions of fluoride in the watr.-. This study lias been criticized by the Royal
College of Physicians (1976). Among the errors cited in the study, the author
based his study on fluoride concentrations in the water of the communities
where the mothers gave birth,-rather than on fluoride in the areas where the
mothers lived during pregnancies. These findings have not been substantiated
by other reports (Berry 1958, Needleman et al. 1974) .
VI-32
-------
E. Mutapenlcit>
No data concerning nutagenic effects of fluoride in humans were found in
the available literature (IARC 1982).
F. Carcinogenicity
Yiamouyiannis and Burk (1977) presented an analysis of mortality data
which showed an increase in the cancer mortality rate among residents of
fluoridated areas. This work has been criticized (Strassburg and Greenland
1979, Oldham and Newell 1977). It was shown that Yiamouyiannis and Burk h<:d
failed to consider the age-sex-race structure of the populations they studied.
Inclusion of these factors in consideration of the data invalidated the
conclusion that fluoridation was responsible for an increase in the cancer
mortality rate. In other studies, Hoover et al. (1976) and the Environmental
Health Directorate of Canada (1977) found no correlation between fluoridation
of water and the cancer mortality rate. In addition, the National Cancer
Institute, whose data were used as the basis of the study by Yiamouyiannis and
Burk, noted errors, omissions, and statistical distortion and stated that,
"results of this analysis fail to support any suspicion of hazard associaced
with fluorliation" (NCI 1975).
The claims of Yiamouyiannis and Burk have also been re-examined by Taves
(1979) and by Kinlen and Doll (1981) and found not to be substantiated by the
data. Kinlen and Doll have obtained additional information on the numbers of
deaths from cancer in the cities concerned, which permits a proper direct
VI-33
-------
ttethod of standardizing cancer death rates. The result*! were shown to be
Identical with the standardized rate determined by the indirect method, and
both methods Indicated less change in cancer mortality rates in the fluoridated
cites than in the nonfluoridated cities during the interval 1950 to 1970.
Cook-Mozaffari et al. (1981) and Cook-Mozaffar! and Doll (1981) examined
cancer mortality in fluoridated and nonfluoridated areas, as well as trends
after fluoridation, and concluded there was no evidence from England, Wales,
the U.S., Australia or New Zealand that addition of fluoride to water supplies
increased the risk of dying from cancer.
The International Agency for Research-on Cancer (IARC 1982) concluded
there was no evidence that an increased level of fluoride in the drinking
water was associated with increased cancer mortality. Similar conclusions had
been reached earlier by Rogot et al. (1978) and by the Governor's Task Force
on Fluorides (Office of Science and Technology, State of Michigan 1979).
G. Epidemiological Studies
1. Mortality Studies
The largest study of overall mortality rates in high-fluoride (0.7 to &.0
ng/L) versus low-fluoride (less than 0.25 mg/L) areas considered 32 paired
cities (Hagan et al. 1954). The high-fluoride areas had a slightly higher
mortality rate than the low-fluoride areas (1,010.6 per 100,000 population
versus 1,005.0 per 100,000, respectively^. The authors state that this
VI-34
-------
difference IP not statistically significant although they did not cite their
criterion for statistical significance.
There have been a number of additional statistical evaluations of death
rates for all causes and death rates from specific causes in high-fluoride
versus low-fluoride areas. The Illinois Department of Public Health (1952)
published data on death rates from heart disease, cancer, nephritis, diabetes
and all causes In populations using low-fluoride (0 to 0.4 mg/L) surface
vaters as compared to populations using veil waters with higher fluoride
concentrations (0.8 to 2.0 mg/L). It was concluded that "mortality experience
in Illinois offers little or no support for claims of adverse effects being
produced by limited ingestion of fluorides."
An extensive mortality study in Colorado Springs, Colorado, provided
information concerning pathological "effects in residents after'prolonged use
of water containing 2.5 mg/L fluoride (Geever et al. 1958). The study was
based on 904 necropsies performed by resident physiciars in training under the
senior author's direct supervision. Necropsy protocols were classified
according to the major cause of death, the contributing causes unrelated to
the major cause and the incidental pathological condition. Comparative
statistical analyses of the pathologic find.ngs revealed no significant
differences that could be related to length of residence in the areas. For
example, there were three deaths attributed to bone cancer in 334 long-term
residents (more than 20 years) and two bone cancer deaths in 188 short-terr.
residents (less than 5 years).
VI-35
-------
The Ministry of Health (England) reported on mortality and morbidity In
high-fluoride (0.4 to 5.8 mg/L) versus low-fluoride (less than 0.2 mg/L) areap
(Hea.sman and Martin 1962). According to the authors, there was no difference
In overall mortality between the two groups (approximately 200,000 in each
area). Notable was the fact that stomach cancer was shown to be no more
prevalent among heavy tea drinkers than among those whose daily consumption of
tea was moderate. (Brewed tea adds about 1 mg/L fluoride to the water in
which it is prepared.)
2. Skeletal Effects
Leone et al. (1956) compared the effects of exposure LO fluoride in
drinking water in a high-fluoride area (Bartlett, Texas; 8 mg/L) and a
low-fluoride area (Cameron, Texas; '0.4 mg/L). Thi? scudy commenced in 1943,
before the practice of fluoridating drinking water was introduced. A total of
116 individuals living in Bartlett were given thorough physical examinations.
As controls, 121 individuals living, in C^neron were also examined. The towns
were similar with regard to geography a.-.J racial composition vlth the
principal occupation in bcth towns being agriculture. In 1943, 57.8 percent
of the Bartlett participants were 55-years-old or older whereas only 47.2
percent of the Cameron parti .ipants were in this age category. In 1953, this
age category accounted for 55.2 and 46.9 percent of the 10-year participants
in Bartlett and Cameron, respectively. Thus, there were more older persons
anor.g the Bartlett participants. The male-* female ratio for both groups was
approximately 1 to 2 in both 1943 and 1953.' At 8 mg F/L, the Bartlett water
VI-36
-------
was about eleven •times ihe currently recognized optlir.uir. for preventing denta]
carles under the prevailing climatic conditions.
Both water supply systems had been in continuous use since the turn of
the century and the selected individuals in each community had been in
continuous residence for at least 15 years at the time of the survey was
begun. In 1953, ten years after the initial examinations, follow-up
examinations were administered. Average length of fluoride exposure In 1953
was 37 years In the high fluoride area and 38 years In the control area (Leone
et al. 1955). All of the original participants were accounted for in the
follow-up study.
The results of these examinations are listed in Tables VI-8 and VI-9.
The comprehensiveness of the physical examinations is evident from these
tables. However, very few statistically significant differences (P-0.05) were
found between the two groups. These were limited to greater incidence rates
of cardiovascular abnormalities and for urinary albumin, and lower rates for
."i
white blood cell counts, neutrophiles and lymphocytes, in the Cameron
residents. The nature of the differences, however, does not necessary
establish a conclusive dose-response relationship associated with fluoride
exposure, 'it should also be noted that the incidence of bone fractures was
greater in the Bartlett residents (Table VI-8). However, this difference was
not statistically significant at the P-0.05 level. The greater number of
older persons (as veil at- accident rates, athletic activity and other
non-fluoride related factors) in 1953 among the Bartlett participants may have
influenced this incidence rate. Dental fluorosis was evident in all Bartlett
VI-37
-------
Table Vl-8 Incidence of Abnormal Clinical Findings, 1943-1953
Characteristic Studied
Number
at Risk
Bartlett
Number
Abnormal
Rate 2
Cameron
Number
at Risk
Number
Abnormal
Rate
Arthritic change
Blood pressure
Sys>. 151 mci/Hg and over
Dlas. 100 mg/Hg and over
Pulse pressure 75 nm/Hg
and over
80
58
73
70
11
18
11
9
13.8
31.0
15.1
12.9
89
81
83
89
13
20
11
16
14.6
24."
13.3
18.0
Bone changes
Density
Coarse trabeculation
Hypertrophic
Spurs
Osteoporosis
Bone, Increased density
(new cases)
Cataract and/or lens opacity
Thyroid
Cardiovascular (except , .
uncomplicated hypertension)
Heating (decreased acuity)
Tumor and/or cysts
Fractures
Drinary tract calculi
Gall Stones
74
74
74
74
74
66
79
74
80
72
80
•80
72
73
7
4
8
1
5
1
8
3
10
14
12
12
14
0
9.5
5.4
10.8
1.4
6.8
1.5
10.1
4.1
12.5
19.4
15.0
15.0
19.4
0.0
81
81
81
81
81
79
85
82
92
78
92
92
76
80
2
2
6
4
10
-
12
6
22
10
10
7
12
1
2.5
2.5
7.4
4.9
12.3
-
14.1
7.3
23.9
12.8
10.9
7.6
15.8
1.2
3Bone changes determined by simultaneous reading of identical views of X-rays taker.
.in 1943 and repeated in 1953.
Bartlett: 4 increased density, 3 decreased density. Caaeron: 2 increased density.
Significant difference between Bartlett and Cameron at p=0.05.
Adapted froa Leone et al. (1954).
VI-38
-------
Table VI-9 Prevalence of Abnormnl Laboratory Findings, 1943-1953
(Participants Residing In Study Area for the Ten-Year. Period)
Laboratory Determination
Hemoglobin
Hematocrit
Red blood cell count
White blood cell count
tH
1
«o Differential count
Neutrophilcs
Lymphocytes
Eosinophiles
Sedimentation rate
Blood calcium
Serology (S.T.S.)
Urine albumin
Urtnc glucose
Year
1943
1953
1943
J953
1*43
1953
1943
1953
1943
1953
1943
1953
1943
1953
1943
1953
1943
1953
1943
1953
1943
1953
1943
1953
Number
Examined
116
79
-
79
116
80
116
78
71
78
71
78
71
78
-
79
-
79
71
84
115
77
115
77
Bartlett
Number
Abnormal
34
20
-.
5
' 25
6
17
11
15
23
2
35
0
6
-
31
-
9
2
2
3
5
2
0
Rate 7.
29.3
25.3
-
6.3
21.6
7.5
14.7
14.1
21.1
29.5
2.8
44.9
0.0
7.7
-
39.2
-
11.4
2.8
2.4
2.6
6.5
1.7
0.0
Number
Examined
121
83
-
82
121
85
121
82
71
82
71
82
71
82
-
83
-
66
71
95
121
85
121
85
Cameron
Number
Abnormal
37
26
-
7
24
2
5
7
6
13
1
36
0
14
_
22
_
7
3
2
10
12
4
1
Rate Z
30.6
31.3
—
8.5
19.8
2.4
4.1
8.5
8.5
15.9
1.4
43.9
0.0
17.1
_
26.5
_
10.6
4.2
2.1
8.3
14.1
3.3
1.2
Significant
Difference
(P - 0.05)
No
No
_
No
No
No
Yes
No
Yes
Yes
No
No
No
No
—
No
No
No
No
Yes
No
No
No
Adapter! from Leone et nl. (1954).
-------
participant? who were born and in continuous residence there during the
formation of the permanent dentition. There was also one instance of dental
fluorosis in a Cameron resident with a history of early fluoride exposure.
Leone et al. (1955) reported a comparison of radiographs taken in 1943
and 1953 of the participants in the Bartlett-Cameron study. The findings
reported were limited to an evaluation of anterior-posterior views of the
lumbar spine, sacrum, pelvis, trochanters and the proximal one third of the
femur. These regions were chosen because the earliest and most definitive
skeletal changes associated with fluoride occur in these areas. In 1943, 16
of the Bartlett residents showed "roentgenographic bone changes in varying
degree . . . considered of interest to this study." Ten years later, nine of
•
these subjects showed no further bone changes., four showed Increased bone
density, three showed a decrease toward a "normal" appearance and one nev case
of increased bone density was Identified.
The radiographs indicated that in persons using the water supply
containing 8 mg/L fluoride about 10 to 15 percent of the population who
resided in the Bartlett area for an average of 37 years experienced an
increased bone density (with or without coarsened trabeculatior.) with a
"ground-glass" appearance. Other observations included coarsened
trabeculation, showing lines of stress without increased bone density and
increased thickening of cortical bone and periosteum with equivocal narrowing
of the bone marrow spaces. These changes are slight, often difficult to
recognize and in most instances equivocal in nature. Apparently, these
changes are not deleterious within the level of statistical significance for
Vl-40
-------
this study because there was no unusual incidence of bone fracture, arthritis,
hypertrophic bone changes or exostoses; no Interference with fracture healing;
no cases of "poker back"; and no evidence of associated functional or systemic
effects. Bone biopsy samples for the determination of fluoride concentrations
were not taken.
The study by Leone et al. (1954) is the only one of its kind available on
U.S. residents who have used high-fluoride waters for a prolonged period. It
is interesting to note that only a small proportion of the study population in
Bartlett was affected to the extent described by Leone et al. (1954). It
should be noted that Singh and Jolly (1970) called attention to the fact that
advanced radiological changes reported from hyperendemic areas of India are
not universally seen in the population as a whole. Singh and Jolly (1970)
have suggested that nutritional status, other sources of fluoride intake and
involvement in heavy labor in a hot climate may influence the incidence of
severe fluorosis in these Indian populations. We have no information on these
factors in the Texas communities studied by Leone et al. (1954). However, the
possibility of hard work in hot temperatures, resulting in the ingestion of
o
large amounts of water, may be a significant factor for these two agricultural
communities.
Leone et al. (1960) compared the radiological findings from 546 residents
between 30- and 70-years-old of Framingham. Massachusetts, (0.04 mg/L
fluoride) with those of residents of Bartlett, Texas, (8.0 mg/L fluoride) and
Caaeron, Texas, (0.4 mg/L fluoride) cited earlier by Leone et al. (1954). The
prevalence of increased bone density and coarsened trabeculation were
VI-41
-------
significantly less in Fratringhair. than In Bartletl, and comnarable to the rates
observed in Caneron. The prevalence of ligamentous calcification (bone spurs)
was higher in Framinghair. Also, there was an unusually high number of cases
of osteoporosis in the Framlngham population. The authors suggest that
deleterious effects on the bone structure of adults may be associated with
prolonged use of low-fluoride waters.
Stevenson and Watson (1957) reported roentgenographic changes typical of
fluorosis. In this study, medical records on file at the Scott and White
Clinic (Temple, Arizona) for the period fros 1943 through 1953 were examined.
Only 23 instances of a roentgenographic diagnosis of fluoride osteosclerosis
were found in a total of approximately 170,000 roentgen examinations of the
spine and pelvis. These cases were associated with individuals (44- to
85-years-old) who resided in Texas or Oklahoma and, in one instance, in
Kansas. Four of the 23 subjects used water containing 8.0 mg 'F/L, three used
water containing 7.6 mg F/L, seven used water containing 5.0 to 5.4 mg F/L and
one used water containing 4.0 mg F/L. Specific water supplies and fluoride
concentrations were not identified for the remaining eight persons, however,
they all lived in areas known to be high in fluorides. The earliest changes
were observed in the pelvis and lumbar spine and consisted of slightly
increased bone density and a slight "ground-glass*1 appearanc . The most
advanced changes encountered were s. chalky white appearance of the vertebral
column and pelvis, and a slightly increased density and coarse trabecular
pattern in the ribs. There was slight roughening of the periosteum of bones
of the forearm or legs in a few patients. Calcification of the sacrospinous
and sacrotuberous ligaments was also observed. No relationship was evident
VI-42
-------
between the roentgenographic findings and the clinical diagnosis of the
patients' condition. Stevenson and Vat sen (1957) concluded that
roetgenographlcally detectable fluoride osteosclerosis was not produced by
drinking water containing less than 4 mg F/L.
Hodges et al. (1941) concluded that prolonged use of water supplies up to
3 mg F/L did not cause radiologically demonstrable sclerosis of the skeleton.
In this study roentgenograns were made of the pelvis end lower lumbar spine of
86 subjects who had used water supplies containing 1.2 to 3.0 mg F/L for none
to 61 years (ages were 7.5 to 71 years), and of 31 subjects 18- to
78-years-old vho had used drinking water containing approximately 2.5 tng F/L
for 18 to 68 years. Generalized skeletal sclerosis was not observed in any of
these 117 subjects.
Dinoan et al. (1976) evaluated 56 persons occupationally exposed to
airborne fluorides1 in the aluminum industry. In this study preshift urinary
fluoride concentrations less than 4 mg F/L were apparently not associated with
increased bone density, alteration of trabecular patterns or ligamentous
calcification, as revealed by X-ray examination. According to Figure III-3 a
urinary concentration of 4 mg F/L in adults corresponds to a concentration of
approximately 4 mg F/L in the drinking vatet ,
3. Effects in Children
One of th« earliest fluoridation -studies involved complete pediatric
examinations during the first ten years of exposur? to elevated fluoride
VI-43
-------
levels. Five hundred children in Newburgh, New York, "the fluoridated city"
(1.2 tng/L fluoride in the drinking water), were compared to 405 children in
Kingston, New York, the control city ("essentially fluoride-free"). The
examinations included roentgenograms (of the right hand and wrist, both knees
and the lumbar spine), blood and urine analyses and general physical
examinations. Smaller groups of children were subjected to special tests
including; visual acuity, hearing levels and additional urine analyses. The
urine analyses were designed to evaluate if fluoride had an irritating effect
on the kidneys. After evaluation of the data from all examinations, it was
concluded that there were no differences of medical significance between the
two groups of children (Schlesir.ger et al. 1956b, Ast et al. 1956).
Specifically, with .regard to bone, there was no evidence of increased bone
density or alteration in rate of skeletal maturation.
McCauley and McClure (1954) compared- radiographs of the right: hand and
wrist for a total of 2,050 children, 7- to 14-years-old, residing in Aciarillo
or Lubbock, Texas» to those in Cumberland, Maryland. The drinking water in
the communities contained 3.5 to 4.4, 3.3 to 6.3 and 0.1 mg/L fluoride,
respectively. Skeletal age and quantitative index' pf ossification were
derived froc the radiographs. The data indicated that calcification of the
carpal bones of the childrf i was not affected by exposure to fluoride, nor was
there any evidence of advanced skeletal maturity and bone development. In as
ouch as development of bones of the hand and wrist parallels that of the rest
of the skeleton, the authors concluded that skeletal development throughout
the body was not affected by fluoride exposure.
VI-4 4
-------
Skeletal fluorsis In children hap not been reported in this country.
However, skeletal effects have been described in Indian children and children
in Tanzania. Teotia et al. (1971) found diagnostic radiological findings In
six Indian children, 11- to 14-years-old. Water from four veils in the
district 01 Rai Barell Uttar Pradesh contained 10.35 to 13.5 mg/L fluoride.
The duration of symptoms was one to ten years. Grossly limited movements of
.the spine, thoracic kyphosis and flexion deformities of the hips and knees
(suggesting crippling fluorosis) were present in four children. Mottled
discoloration of the teeth was present in five cases. Skeletal radiographs
showed osteosclerosis of the spine and pelvis in six cases. Four cases
demonstrated coarsened trabeculation In the knees and elbows and calcification
of the Interosseous membrane of the forearm.
Teotia et al. (1979) examined 550 children, 4- to 15-years-old, from the
same district of India and found diagnostic radiological findings in 200 of
the children. The effects observed included; osteosclerosis (particularly of
the spine, pelvis and thorax), periosteal bone formation, exostoses and
calcification of ligaments, interosseous membrane and muscle attachments.
Roentgenological findings typical of hyperparathyroidism were seen in 43
cases. Of the 200 children with diagnostic radiological changes, 32.5 percent
were sycp on-free, 67.5 percent were symptomatic, 51.5 percent were without
crippling deformities and 16.5 percent were crippled. All of the children
showed mottled discoloration of the teeth. Water from four wells in the area
contained 24 to 26 mg/L fluoride.
VI-45
-------
Wenzel ct al. (1982b) examined the effects of fluoride on dental enamel,
skeletal maturity and bone structure in 11- to 15-year-old Tanzanian girls.
The children were born and raised in areas where the drinking water contained
< 0.2, 1.5, 2.5 or 3.6 mg/L fluoride. Dental fluorosis was positively
associated with fluoride concentration in the drinking water. No conclusions
could be drawn regarding the effect of fluoride in the water on skeletal
maturity. This was attributed to differences among the groups in nutritional
state and exposure to disease. There was, however, a correlation between
retardation of skeletal maturity with increasing dental fluorosis for the
group using water containing 3.6 mg/L fluoride. The authors suggest that
increased fluoride exposure slows skeletal maturation. However, due to the
very warm climate of Tanzania, drinking water, intake would have a significant
impact on the total dose. Such a relationship was not evident in a similar
study of 12- to K-year-old Danish girls whose drinking water contained < 0.2
or 2.4 mg/L. fluoride (Wenzel et al. 1982a).
4. Other Studies
•7
Other epideaiological studies Include one in Russia (Knizhnikov 1958)
where the health of natives of Shchuchir.sk (3. A to 4.0 ng F/L) was cor .pared
with that of natives i'. Kokchetav (0 to 0.9 mg F/L). In add tion to the usual
examinations of bones and teeth, the participants were examined for
hypertension, bradycardia, somnolence, coagulation of blood, parathesls and
urticaria-type rash. All of the illnesses listed were less frequent in the
fluoride area than in the control area with the exception that severe dental
fluorosis was prevalent in the fluoride area. Also, there was ar. unexplained
VI-46
-------
and unusually low incidence of diseases of bones, muscles and Joints in the
fluoride area.
A case-control study in South Carolina concluded that high fluoride
concentrations in drinking water (4.18 mg F/L) exert a protective effect
against the development of primary degenerative dementia (Still and Kelley
1980). The authors hypothesize that fluoride attenuates the neurotoxicity of
aluminum. There are some weaknesses in the study, one of which is that it was
based on hospital admissions rather than on occurrence rates.
H. Summary
Fluoride has been shown to have several beneficial effects, both in terms
of general health and in the treatment of specific diseases. Numerous studies
have documented the benefits provided by fluoride in preventing dental caries
in children. Many of these studies evaluated the transition to adverse dental
effects (fluorosis) at higher dose levels. Conclusions from these studies
indicate that the beneficial effects are obtained and the adverse effects
prevented when the drinking water (in an average temperature climate) is
approximately 1 mg/'L.
Fluoride has been demonstrated to have a positive effect on bone develop-
ment and has found application in stimulating new bone growth in patients with
osteoporosis. To a lesser degree, fluoride has also been suggested to have
possible effects on the cardiovascular system (i.e., reduced aortic
• .
calcification when drinking water contained 4.0 to 5.8 mg F/L) and hearing
VI-47
-------
(I.e., stabilization of the sensorineural component of hearing loss In
patients with active otosponglosis when 40 to 60 eg was administered dally).
A "certainly lethal (single) dose" of NaF for a 70-kg man is estimated to
be 5 to 10 g or 70 to 140 mg/kg. Fluoride may c?use a wide variety of toxic
e' ects in humans. Among these, the claims of allergic or idiosyncratic
sensitivity, mongolism and cancer have not be.en substantiated. On the con-
trary, sound evidence suggests that fluoride does not cause sensitivity,
mongolism or cancer. There is evidence which suggests that persons with
chronic renal insufficiency may have a lower margin-of-safety for the toxic
effects of fluoride.
Chronic exposure to either too low or too high a concentration of
fluoride may have deleterious effects on the skeletal systec. An increase in
the incidence of severe osteoporosis was correlated with use of drinking water
containing 0.4 mg/L fluoride. Severe skeletal fluorosis has been reported in
persons living in areas of naturally high fluoride concentrations (up t.o 14
mg/L). Radiologically detectable osteosclerosis has been observed in about
10 percent of long-term residents using water supplies containing 8 mg/L
fluoride. Retardation of skeletal maturity has beer, observed in children
using a water supply contacting 3.6 mg/L fluoride. In other situations,
skeletal fluorosis has not been described in populations whose water supplies
contained less than 4 mg/L flu'oride.
An important effect of fluoride is dental fluorosis (mottled er.asel).
Numerous studies have examined the relationship between concentrations of
VI-48
-------
fluoride in community drinking water supplies and the occurrence of dental
fluorosis. Some studies have determined that the acceptable level of fluoride
in water varies with the mean annual temperature of the area in question,
because people drink more water when the environment is wanner. In one study,
concentrations of fluoride causing cosmetically "objectionable" dental
fluorosis .varied from 0.8 mg/L at mean temperature of 70°F to 1.7 mg/L at mean
temperature of 50°F (estimated to equal 0.05 mg/kg/day). The Agency, however,
has concluded that there is insufficient data to quantitatively incorporate
temperature in any future drinking water regulations. Concentrations
associated.with intentional fluoridation of drinking water (0.7 to 1.2 mg/L)
have not shown adverse effects on health or longevity. Factors considered
include growth, effects on the kidney, cardiovascular system and thyroid,
teratogenicity and mutagenicity.
VI-49
-------
VII. MECHANISMS OF TOXIC1TY
A. Acute Effects
The mechanism by which fluoride causes acu'.e lethality at high doses has
not been fully defined. Obviously, there is interference wlti the normal
metabolism of cells and essential enzymatic reactions may be blocked. There
may be interference with the origin and transmission of nerve impulses,
perhaps as the result of calcium complex formation. Other metabolic roles of
calcium may be interrupted (e.g., blood clotting and membrane permeability).
Also, there may be severe renal tubular damage and injury to the mucosa of the
stomach and intestine. Vomiting and diarrhea result in appreciable water
loss, electrolyte imbalance and a clinical picture of shock (Hodge and Sulth
1965).
B. Skeletal Effects
Fluoride is involved in bone mineral deposition in several ways. It may
be essential to the precipitation or nucleation of the apatite lattice in an
orderly fashion on collagen fibers. Fluoride frosn extracellular fluid
exchanges with hydrpxyl ions and perhaps bicarbonate ions i.i the surface layer
of hydroxyapatite crystals to form fluorohydroxyapatite. This material shows
an increased crystalline structure and less solubility than does hydroxy-
apatite. Fluoride is incorporated into the inner layers of the crystal
lattice, as well as on the suface of newly formed crystals, by the accretion
VII-1
-------
of new nineral. Osteorlastic resorpfion of old bone and osteoblastlc deposi-
tion of new bone, resulting in continual remodeling of the skeleton, allows
release and re-uptake of fluoride into bone mineral. Fluoride apparently
increases the rare of form&tlon of new bone, Che number of ostenblasts and the
serum activity of the osteoblastic iso-enzyne skeletal alkaline phosphatase.
The effects of fluoride may be modulated by parathyroid hormone and by human
skeletal growth factors (Neuman et al. 1950, McCann and Bullock 1957. Snlth et
al. 1953, Zipkin et al. 1956, Hodge and Smith 1981, Faccini and Teotia 1974,
Farley et al. 1983).
C. Dental Effects
Evidence suggests that dental fluorosis results from effects of fluoride
on the aneloblasts. Developing enamel and enamel-forming cells are the first
to respond when rats are injected with sodium fluoride. The newly formed
enamel matrix is faulty and poorly mineralized. The staining frequently seen
with mottled teeth may be the. result of oxidation of organic material
integrated in the dental structures. It has also been suggested that it may
be related to food pigments which have penetrated the hypoplastic enamel.
Mottling, however, does not determine the degree of dental fluorosis (Schcur
and S-jith 1934, Schour and Poncher 1937, Sh.pe et al. 1963, Gabovich and
Ovrutsky 1969, Dean 1934).
VI1-2
-------
D. Sumaary
The mechanism for acute lethality at high fluoride dose levels is not
fully defined. It is believed that certain essential enzymatic reactions may
be blocked and there nay be Interference with the origin and transmission of
nerve impulses. -The metabolic roles of calcium and physical damage to the
kidney and the mucosa of the stomach and intestine are also believed to be
associated with the acute lethality mechanism. Fluoride interacts with bones
and teeth by replacing hydroxyl or bicarbonate ions in hydroxyapatite to fore
fluorohydroxyapatite. Fluoride cay function as an essential key to bring
about precipitation or nucleation of the apatite lattice in an oriented
fashion on collagen fibers. Accretion of new mineral continues, and fluoride,
brought to the surfaces of nevly formed crystals by the extracellular fluid,
replaces the hydroxyl ion. As crystal growth continues, fluoride is
incorporated into inner layers of the crystals as well as on the surface.
Remodeling of the bone structure takes place by an interplay of osteoclastic
resorption of old bone and osteoblastic deposition of new bone. The presence
of fluorohydroxyapatite increases the crystalline structure of the bone
and reduces its solubility. Available evidence suggests that dental fluorosis
results from toxic effects of fluoride on the epithelial enamel organ.
Specifically, several inve tigators have shown that ameloblasts are
susceptible to fluoride. Dental staining often accompanies fluorosis but does
not itself determine the degree of fluorosis. The staining is believed to be
due to the oxidation of organic material in defective enamel or the
penetration of hypoplastic sections of enamel by food pigments.
VII-3
-------
VIII. QUANTIFICATION OF TOXICOLOCICAL EFFECTS
The quantification of toxicological effects of a chemical consists of ar
assessment of the non-carcinogenic and carcinogenic effects. In the
quantification of non-carcinogenic effects, an Acceptable Daily Intake (ADI)
is calculated. An Adjusted Acceptable Dally Intake (AADI) and Health Advisory
(HA) values for the chemical are then calculated to define the appropriate
drinking water concentrations to licit human exposure. For ingestion data,
this approach is illustrated as follows:
.__ (NOA£L or LQAEL in mg/kg/day) (Body Weight in kg)
AD1 = Uncertainty/Safety Factor
' _ ADI _ /T
" Drinking Water Voluae in L/day " °8/L
where:
NOAEL - no-observed-adverse-effect level.
LOAEL = lowest-observed-adverse-effect level.
Body weight = 70 kg for adult or 10 kg for child.
Drinking water volume • 2 L per day for adults or 1 L per day
for children.
L'nce-iainty/Safety Factor = 10, 1TO or 1,000.,
Utilizing these equations, the following drinking water concentrations
are developed for non-carcinogenic effects:
VIII-•
-------
A one-day HA for 10-kg child.
2. A one-day HA for 70-kg adult.
3. A ten-day HA for 10-kg child.
4. A ten-day HA for 70-kg adult.
5. A lifetime AADI for a 70-kg adult.
The distinctions made between the HA calculations (items 1 through 4) are
associated with the duration of anticipated exposure. Items 1 and 2 assume a
single acute exposure to the chemical. Itecs 3 and 6 assume a limited period
of exposure (possibly 1 to 2 weeks). The HA values will not be used in estab-
lishing a drinking water standard for the chemical. Rather, they will be used
as informal scientific guidance to municipalities and other organizations when
emergency spills or contamination situations occur. The AADI value (item 5)
is intended to provide the scientific basis for establishing a drinking water
standard based upon non-carcinogenic effects.
A KOAEL or LOAEL is determined from animal toxicity data or human .effects
data. For animal data, this level is divided by an uncertainty factor because.
there is no universally acceptable quantitative method to extrapolate from
anleaIs to hucans. The possibility must be considered that huaans are more
sensitive to the toxir effects of chemicals than are animal:. For human data,
an uncertainty factor is also used to account for the heterogeneity of the
human population in which persons exhibit differing sensitivity to toxic chemicals.
AT. Office of Drinking Vater (ODW) codification of the guidelines set forth by
the National Actdeiay of Sciences (NAS 1977, 1980) is typically used in
establishing uncertainty factors as follows:
VIII-2
-------
• An uncertainty factor of JO Is used when good acutp or chronic human
exposure data are available and supported by acute or chronic
toxlcity data in other species.
• An uncertainty factor of 100 Is used when good acute or chronic
toxlcity data identifying KOEL/NOAEL are available for one or more
species, but human data are not available.
• An uncertainty factor of 1,000 is used when United or incomplete
acute or chronic toxicity data in all species are available or when
the acute or chronic toxicity data identify a LOAEL (but not
NOEL/NOAEL) for one or more species, but human data are not
available.
The uncertainty factor used for a specific risk assessment is judgmental.
Factors that cannot be incorporated in the NAS/ODW guidelines for selection of
an uncertainty factor, but muse be considered include: (1) the quality of the
toxicology data, (2) the significance of the adverse effect and (3) the
existence of counterbalancing beneficial effects.
If toxicological evidence requires the checical to be classified ai, &
potential carcinogen (there is insufficient evidence to classify fluoride as a
carcinogen following oral exposure), mathematical.models are used to calculate
the estimated excess cancer risks associated with the ingest ion of the
checical via drinking water. The bioassay data used in these estimates are
VIIT-3
-------
fror anlr.al experiments. In order to predict the risk for humans, these data
cust be converted to an equivalent human dose. This conversion includes
correction for non-continuous animal feeding, non-lifetime studies and for the
difference in size. The factor that compensates for the size difference is
the cube root of the ratio of the animal and human body weights. It is
assumed that the .average human body weight is 70 kg and that the average human
consumes 2 liters of water per day. The multistage model is then fit
to the equivalent human data to estimate the risV. at low doses. The upper 95*
confidence limit of this estimate is used. Excess cancer risks can also be
estimated using other models such as the one-hit model, the Weibull model, the
loplt model and the probit model. There is no basis in the current
understanding of the biological mechanisms, involved in cancer to choose among
these models. .The estimates of lov doses for these models can differ by
several orders of magnitude.
The scientific data base used to calculate and support the setting of
risk rate levels has an inherent uncertainty. This is because the tools of
scientific measurement-, by their very nature, involve both systematic and
random error. In most cases, only studies using experimental, animals have
been performed. There Is thus uncertainty when the data are extrapolated to
husans. When developing ^isk rate levels, several othi.r areas of uncertainty
exist, such as (1) incomplete knowledge concerning the health effects of
contaminants in drinking water, (2) the impact of test animal age, sex and
species and the nature of target orgar. sys'teas examined on the toxicity study
results and (3) the actual rate of exposure of internal targets in test
VIII-A
-------
an1maIF or humans. Dose-response data are usually only available for high
levels of exposure, not for the lower levelc of exposure for which a standard
is being set. When there is exposure to more than one contaminant, additional
uncertainty results from a lack of Information about possible synergistlc or
antagonistic effects.
It has been concluded, however, that the foregoing risk assessment
procedures are not appropriate for application with the available fluoride
data. The typical assessment assumes a higVi to low dose extrapolation will be
made. In the present case, the extensive availability of human data requires
an interpolation rather than extrapolation. Possibly more important is that
the typical assessment procedure does not provide for any quantitative inputs
for a chemical's potential beneficial effects. Fluoride has well documented
beneficial effects that must be addressed (balanced) during the assessment.
Thus, the assessment that will be performed for fluoride must rely largely
upon an Interpolation of the available human data and give due consideration
to balance the required degree of human health protection froir adverse effects
with the documented beneficial effects.
A. Non-Carcinogenic Effects
1. Short-Terc Exposure
;.s;
Acute toxic effects in the human following ingestion of fluoride have
been described by Lidbeck et al. (1943). In this instance, ingestion resulted
from the inadvertent mixing of roach powder containing sodium fluoride with
VIII-5
-------
food being served in an institution, but no reliable measure of the amount of
fluoride ingested was possible. The initial effect of rapidly ingesting large
amounts of fluoride is irritation of the gastrointestinal tract, causing
vor.itlng and diarrhea. Both the vonitus and the feces may contain blood.
These symptoms nay proceed to collapse and eventual death (Lldbeck et al.
1943). No single, target systen appears uniquely susceptible to these acute
effects, suggesting that fluoride acts as a general systemic poison at very
high doses. This explanation is consistent with the ability of high
concentrations of fluoride to bring about a state of bhock, to inhibit
essential enzymatic processes such as cellular respiration and to interfere
with essential roles of calcium (Hodge and Smith 1965).
Black et al. (1949) described the effects of fluoride administered to
more than 70 patients for periods of five to six months. Most of these
subjects, suffering from malignant neoplastlc disease, were being treated with
metabolic Inhibitors. Some were leukemic children 3 to 6.5 years old, while
others were adults including elderly individuals. Doses for the children were
20 to 50 mg NaF (9.0 to 22.5 tag F) four times daily. Doses for adults were 80
mg NaF (36.3 mg F) four times daily. The material was administered orally
with an antacid containing 4 percent aluminum oxide or as an er.teric coated
rablet to avoid gastric irritation. N' evidence of systemic toxicijy or of
parenchymatous damage was seen which could be attributed to fluoride, even
though some patients had received more than .27 g of sodium fluoride over a
period of three months. Criteria evaluated included growth and development in
the children, mottled enaael, eruption of permanent teeth, hecialopoiesis, liver
function, albucin-globulin ratio, blood sugar and cholesterol concentrations
' VIII-6
-------
and kidney function: Postmorter. d*ta fror four cases showed no parenchynatou.=
deceneratlon attributable to fluoride. In hypertensive patient.* a tendency
wns noted for decreased diastolic and systolic blood pressure. Tn two
patients with functioning colostorcies there was no apparent effect of
fluoride on the exposed mucosa of the colon.
2. Long-Tern Exposure
Comprehensive investigations by Shupe et al. (1963) evaluated the effects
of fluoride or. dairy cattle to Include changes observed in the teeth. In this
study pairs of cows were fed rations containing 12 (normal), 27, 69 or 93 ppir,
fluoroxlde on a total dry matter basis. Feeding was continued from 4 months
to 7.5 years of age. Depending upon the amount of fluoride ingested, affected
teeth erupted with different degrees of mottling, staining, hypoplasia and
hypocalcification. The following tooth classifications were established:
. .r;T:e*v
(0) Normal: smooth, translucent, glossy white enamel: tooth normal
shovel shape.
(1) Questionable effect: slight change, exact cause not determined; may
have enamel flecks; cavities may be unilateral or bilateral but with
the absence of mottling.
(2) Slight effect: slight mottling of enamel; may have slight staining
but no wear; teeth normal shovel shape.
(3) moderate effect: definite nettling ar.d staining of enamel; coarse
•• : •*
mottling (large patches of chalky enamel); teeth nay have slight
signs of wear.
VIII-7
-------
(4) Marked effect: definite mottling, Ftalnlng and hypoplasla: trey have
pitting of enamel; definite wear of teeth; ensr.el may be a pa?e
cream color.
(5) Excessive effect: definite erofion of enamel wlih excessive wear of
teeth; staining and pitting of enamel may or may not be present.
In cattle consuming the highest dose of fluoride (I.e., 93 ppm in the
ration) the incisors were classified as 4 to 5, beginning as early as two
years of age. The molars were classified as 0 to 3 at two years of age, 1 to
4 at four years and 1 to 5 at six years. For cattle at the dose of 49 ppn;,
the incisors were scored as 3 to 4 beginning at two years. In these same
animals, the molars were scored as 0 to 1 at two years, 1 to 2 at four years,
and 1 to 3 at six years. In cattle administered 27 ppn fluoride, the incisors
were scored as 0 to 2 through six years of age and the molars were scored as 0
to 1 through six years. Incisors and nolars of cattle administered the normal
ration (12 ppm-fluoride) were scored 0 to 1 throughout the six years.
Richards et al. (1967) indicate that objectionable, dental fluorosi-s
(moderate and severe according to the classification scheme by Dean 1942) in
humans appears with the foil-owing combinations of waterborne fluoride
concentrations and nea- annual temperatures:
• 1.4 to 1.6 ppm fluoride at 65°F or less.
• 1.1 to 1.3 ppr fluoride at '656F to 79°F.
• 0.8 to 1.0 ppir fluoride at 80°V or higher.
VIII-8
-------
These vnlues are similar to those reported by Galagan and l.amson M9531
as shown In Figure VI-1 (see Section VI). In the classical scheme for rating
fluorosis, teeth diagnosed ap normal exhibit no clinically observable evidence
of exposure to fluoride. Richards et al. (1967) supppst that such teeth
should be classified as fluoride deficient rather than normal. Their data
•indicate that as the percentage of children shoving clinical evidence of r.i]d
fluorosis approaches four to six percent, some objectionable (moderate)
fluorosis begins to appear.
Because the relationship between fluoride concentrations: in drinking
water and community fluorosis indices was established many years ago, a demand
has arisen for evidence confirming or re-establishing the fluoride/fluorosis
relationships. Segreto et al. (1984^ investigated the possibility that
significant changes in cultural and dietary patterns may have altered fluoride
intake patterns from those developed 20 to 40 years-ago. They selected 16
Texas communities and surveyed children (7 to 18 years old) for enamel
mottling using Dean's (19412) classification systec. The fluoride levels in
the drinking water were expressed by L^C authors in terras of the relationship
to optimal for prevention of dental caries. Personal communication with one
of the authors (Dr. Edvir. M. Collins), however, indicated that th." actual
fluo- ide levels ranged from 0.2 m^/L to 3.2 mg/L. The combined incidence of
moderate and severe dental fluorosis observed ranged from minimal at 0.2 mg
F/L to 31.6 percent at 3.2 mg F/L. The authors, however, reported only one
case of severe fluorosis (at 3.2 mg F/L). The observed variation in the
»
fluorosis Incidence at different fluoride drinking water levels could be due
VIII-9
-------
to differences in the lifestyles of the different communltles, variation In
the susceptibilities of the children examined or other factors.
Drlfcoll et al. (1983) reported the results of a crosp-sectiona1 survey
of the prevalence of dental fluorosis and dental caries among 807 school
children (8 to 16 years old) in seven Illinois connr.unlt ies. Fluoride
concentrations in the community drinking water ranged from 1.06 to 4.07 mg
F/L. The results of this study indicate a dose-response increase in the
incidence of moderate and severe dental fluorosis with increased fluoride
level in the drinking water- The incidence of moderate and severe dental
fluorosis ranged from 2.A percent (of 336 children evaluated) at 1.06 mg F/L
to 30.2 percent (of 136 children evaluated) at approximately 3.84 mg F/L.
Concurrent with this increase in dental fluorosis, the authors observed a
significant (P<0.05) decrease in dental caries (as measured by reduction of
mean D^fF surface score) in children of all fluoride levels above 1.06 nig F/L.
Unlike the dental fluorosis results, the dental caries reduction did no:
exhibit a dose-response relationship above the level of 2.08 mg F/L in the
drinking water. There was no statistically significant (P<0.05) difference in
t
the reduction of dental caries among children exposed to an average 2.08 mg
F/L through 3.84 mg F/L.
A. surcary of the incidence of moderate and severe dental fluorosis fror
six studies spanning more than 40 years (1937 to 1954) is provided in
Table VIII-1. The data assembled in this table are fror. six different
sources, each with technically sound but varied procedures, analytical methods
VII1-10
-------
Table VIII-1
Summary of Moderate and Severe Denca]
Fluorosls In Children
Fluoride
drinking water Number cf
concentration children
(mg/O evaluated
0.2
0.3
0.4
0.4
0.4
0.5
0.5
0.6
0.7
0.8
0.8
0.9
1.0
1.1
1.1
1.1
1.1
1.2
1.2
1.2
1.2
1.3
1.5
1.6
1.8
1.8
1.9
1.9
103
126
223
82
263
113
403
614
316
95
361
123
50
336
211
187
128
70
633
152
171
447
110
301
57
170
273
170
Dental
fluorosis Incidence','
Moderate
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
2.0
2.0
0.3
0.0
0.0
1.8
0.9
1.1
0.0
13.0
0.0
0.0
0.0
0.0
0.9
3.3
3.5
1.2
1.1
13.5
Severe
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
1.0
0.0
0.0
0.0
0.6
0.0
0.0
0.0
3.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
r
Reference
6
6
6
A
4
L
U
U
U
U
6
6
It
3
6
6
6
U
4
4
4
4
2
6
2
1
1
6
«.c;:tir.ued-
References:
1 * Dean (1942) a? summarized by Albertini et al. (1982).
2 •*. Dean and Elvove (1937) ,as summarized by Albertini et al. (1982).
3 - Driscoll et al. (1983).
4 • Galagan and Laoson (1953) as summarized by Albertini et al. (1982).
5 • Lewis and Faine as summarized by Albertini et al. (1982).
6 - Segreto et al. (1984).
VIII-11
-------
Table VITI-] - continued
Fluoride
drinVlng vater
concentration
(mp/L)
1.9
2.0
2.0 '
2.1
2.2
2.2
2.3
2.3
2.4
2.5
2.6
2.9
2.9
3.2
3.8
3.9
3.9
£.0
4.0
6.0
6.2
A. 4
6.8
5.7
7.6
8.0
14.1
Number of
children
evaluated
23
109
•200
143
179
138
90
67
113
166
404
192
97
190
21
136
289
39.
101
59
39
189
36
38
65
21
26
Dent?]
fluorosis incidence,*
Moderate
13.0
14.7
4.0
8.6
13.6
11.0
6.7
32.8
6.6
16.2
8.9
7.8
23.7
31.1
9.0
7.4
33.9
38.0
60.0
23.7
33.0
66.0
6.0
50.0
10.8
67.6
38.5
Severe
0.0
0.0
0.0
4.9
0.0
0.7
0.0
0.0
0.0
3.4
1.5
8.3
3.1
0.5
0.0
22.8
13.2
6.0
2.0
11.9
3.0
17.9
0.0
39.5
58.5
42.9
53.8
Reference
6
6
6
3
2
1
6
6
6
2
1
3
2
6
5
3
2
5
5
2
5
2
5
2
1
2
1
VIII-12
-------
and sample sizes. Therefore, no effort has been made to merge these findings
into a single dose-response distribution or to perform any statistical
analysis of the assembled data. The table IP provided to supply a
consolidated sampling of the historical data on dental fluorosis incidence and
to conveniently reflect the general dose-response relationship of Increased
dental fluorosis with increased dose. It should be noted, however, that the
Incidence of objectionable dental fluorosis (moderate and severe) does not
generally Impact a significant percentage of the population until the drinking
j
water concentration approaches 2.0 mg F/L.
At the request of the EPA, the U.S. Surgeon General examined the
relationship of fluoride in drinking water and the aspects of dental
fluorosis. The results of that evaluation (Koop 1982, Albertini et al. 1982)
led to the general conclusion that, while not considered ar. adverse health
effect, the undesirable cosmetic effects to teeth could be minimized by
limiting the fluoride concentration to twice the optimum for the reduction of
dental caries. The Surgeon General encouraged communities to limit water to
twice optinuo (about 2 ttg F/L) to provide this protection for children up to
age nine, but emphasized that there is no sound evidence to indicate that
adverse effects on general or dental health (dental fluorosis was not judged
to br an adverse effect) are associated with concentrations of fluoride 'hat
are naturally found in U.S. public water supplies. The Surgeon General
repeated his earlier opinion on limiting fluoride concentrations to twice the
optinum (about 2 eg F/L) in his response to a subsequent EPA request to
:- . :•<;
evaluate the nondental effects of fluoride (Shapiro 1983, Kocp 1984).
VIII-13
-------
The EPA, with the assistance of the National Institute of Mental Health
(SIMK), convened an ad hoc Review Panel of behavioral scientists to
investigate the potential psychological and/or behavioral effects associated
with dental fluorosls. This ad hoc Review Panel reviewed background
information and conducted a meeting on October 31. 1984 in Bethesda, MD to
discuss this issue and determine if consensus opinions could be formulated-.
The conclusions and recommendations of the Review Panel's deliberations were
summarized in a Novenber 17. 1984 report (Kleck 1984) and are repeated below:
"It is concluded that individuals who have suffered impaired dental
appearance as Che result of moderate to severe (dental) fluorosis are
probably at Increased risk for psychological and behavioral problems or
difficulties. Since this conclusion is based on extrapolations from
research on the effects of physical appearance characteristics other than
dental fluorosis, it is suggested that investigations be supported to
directly assess the social, emotional, and behavioral effects of
fluoride-induced cosnetic defects. Finally, the Panel recommends
research be done on the further development of techniques for the
amelioration or removal of the unaesthetic appearance effects associated
with some levels of dental fluorosis."
Skeletal changes in bones of cattle ingesting 12 (normal), 27, 49 or
93 pps fluoride on a total dry matter basis have been described by Shupe et
al. (1963). Fluoride concentrations in dry. fac-free rib biopsy samples
increased with increasing tine of exposure for all dose groups. After 7.3
\ears (2,663 days) the fluoride concentration was approximately 900 pun in
animals on the normal diet. At this sact? time, the rib fluoride
concentrations vere apprcxisately 2,500, 5, SOU and 8,200 for the cattle
receiving 27, 49 and 93 ppr> fluoride in the ration, respectively. The rate of
i
increase with time was greatest in those cattle adnlnistered 93 ppc fluoride.
VIII-1A
-------
The first clinically discernible bone lesions appeared on the medial surface
of the proximal third of the metatarsal boner and were -bilateral. These
effectF were observed after 1.5 to two years in cattle on the 93 ppn fluoride
ration and after 3.5 to four years In cattle on the 69 ppm fluoride ration.
As the degree of osteofluorosis increased, palpable hyperostoses appeared in
the rarcl of the mandibular bones, and the 7th through 12th ribs became wider
and thicker. The degrees of perlosteal hyperostosis were classified a? 0 -
normal, 1 = questionable, 2 » slight, 3 • moderate, 6 • marked and 5 "
excessive. Cattle on the normal diet were scored as normal through six years
of age. Those cattle on 27 ppm ration were scored 0 to 1 through six years;
those on 69 ppm ration were scored 0 to 2 at two years, 0 to 3 at four years,
and 0 to 6 at six years; and those on 93 ppm ration were.scored 0 to 3, 0 to 6
and 0 to 5 at two, four and six years, respectively. Radiographs taken at age
7.5 years (approximately seven years on fluoride) showed increased coarsening
and thickening of Che trabecular pattern with a ground glass appearance for
cattle on the rations containing 69 and 93 ppm fluoride. Perlosteal hyper-
ostosis, subperiosteal Increased density in some cases, endosteal and cortical
porosity and mineralized spurs at points of attachment of tendons to leg bones
were also observed at these dose levels.
Leone et al. (1955) described roc itgenographlcally detectable changes
observed in 10 to 15 percent of persons residing an average of 37 years In
Bartlett, Texas where the water supply contained 8 eg F/L. Observations
' included increased bone density with or without coarsened trabeculation, with
a "ground glass" appearance; coarsened trabeculation, showing lines of stress,
without increased bone density; and increased thickening of cortical bone and
VIII-15
-------
periosteum with equivocal narrowing of bone marrow spaces. Bone biopsy
sample? tor the determination of fluoride concentrations were not taken.
Stevenson and Watson (1957) reported an Increase in bone density and n
definite but slight "ground .glass" appearance in spinal and pelvic roentgeno-
grams of 23 subjects who were long-terra residents of high fluoride areas in
Texas, Oklahoma.and Kansas. For 15 of these individuals, the drinking water
contained 4 to 8 mg F/L. The fluoride content of the drinking water was unknown
for the remaining eight subjects. Calcification of the sacrospinus and sacro-
tu.berous ligaments was also evident in 15 of the 23 subjects. Although a
total of 170,000 X-ray films were examined, the authors were unable to develop
a meaningful incidence rate because information was lacking as to the total
number of films examined for persons exposed to specific levels of fluoride.
Hodges et al. (1941) examined roentgenograms of the pelvis and lumbar
spine of 86 persons (7.5 to 71 years'Old) who had used water supplies
containing approximately 1.2 to 3 mg F/L for up-to 61 years. They found no
occurrence of generalized sclerosis. A second population (ranging in age from
18 to 78 years) which had used a water supply containing.approximately 2.5 mg
F/L for 18 to 68 years was similarly evaluated. Again, no instance of
generalized skeletal Jluorosis was observed.
Wenzel et al. (1982b) observed a significant relationship of dental
fluorosis and reduced skeletal maturity in 11- to 15-year-old Tanzaniar girls
whose drinking water contained only 3.6 mg/L of fluoride. The authors
suggested that Increased fluoride exposure slows skeletal maturation. Due to
VIII-16
-------
the warm climate (I.e., Increased exposure and total dose), dietary and other
factors, the relevance of these results to the U.S. population Is not veil
establIshed.
There Is limited evidence to penalt an estimate of the vaterborne
fluoride concentration associated with the appearance of fluoride
osteosclerosis. For example, Hodge and Smith (19701 quote evidence that in
the aluminum industry, average urinary excretions of 5 r.g F/L in randomly
collected samples are not associated with osteosclerosis. Dlnnan et al.
(1976) indicated that aluminum workers whose average pre-shift urinary
fluoride concentration is less Char. 4 mg F/L do not show radiographlcally
demonstrable increases In bone density, altered trabecular patterns or
ligamentous calcification. According to Figure III-3 (see Section III for
greater detail), urinary fluoride concentration is essentially equal to the
concentration of fluoride in the drinking water Ingested at steady-state
exposure conditions. Thus, the absence of clinically or radiographically
demonstrated osteosclerosis in the studies cited by Hodge and Smith (1970) and
by Dinman et al. (1976) could be estimated to be associated with exposures to
drinking water containing approximately 5 and 4 mg F/L, respectively. Smith
and Hodge (1959) have suggested that, in the human, osteosclerosis probably
will .lot be seen vlth skeletal fluoride concentrations of 4000 ppc (dry
fat-free basis). They also state that effects will be observed in a small
proportion of individuals with skeletal fluoride concentrations of
approximately 6,000 ppm. These skeletal concentrations correspond to fluoride
•» £ •*
concentrations in the water of 4 and 6 mg F/L, respectively (Smith and Hodge
1959, Hodge and Smith 1981).
VITI-17
-------
At the request of the U.S. Environmental Protection Agency (EPA), the
U.S. Public Health Service (PHS) conducted an evaluation of the nondental
health effects of fluoride. At the direction of the Surgeon General an ad hoc
committee was assembled to review the available literature. The corarlttee met
on April 18-19, 1983 in Bethesda, MD and summarized their findings In a report
to the Surgeon General (Shapiro 1983). That report was formally transmitted
to the EPA with a letter from the Surgeon General on January 23. 19P&.
The committee listed the nondental health effects of fluoride as: (1)
death (acute poisoning); (2) gastrointestinal hemorrhage; (3) gastrointestinal
irritation; (6) erthralgias; and (5) crippling fluorosis. Gastrointestinal
effects are not known to occur at fluoride concentrations in drinking water.
In adults, mild osteosclerosis, as opposed to crippling fluorosis, is not
considered an adverse effect.
Based on their review of the available literature the Surgeon General's
ad hoc committee made the following conclusions (Shapiro 1983) :
1. It is inadvisable for the fluoride content of drinking water to be
greater than twice the current optimal level (1.4 to 2.4 mg/L) for
children up to age 9 in order to avoid the (.icosmetic effects of
dental fluorosis. This conclusion coincides with the recommenda-
tions of the Surgeon General relative to the dental effects of
naturally occurring fluorides.
VIII-18
-------
2. The fluoride content of drinking water should not be greater than
four tines the optimal level for any coranunlty water supply. This-
conclusion recognizes that, fluoride Intake fror water between 5.0
and 8.0 mg/L (it tines to 10 tines optimum') has beer associated, In a
very small number of subjects, with the radlologic appearance of
early osteosclerosis, which while not an adverse health effect, Is
however, an indicator of demonstrable osseous changes not to be
anticipated at lower levels (less than four times optimum) of fluoride
3. The difference between 4 tines and 10 times optimum represents an
adequate margln-of-safety unless further research warrants
reconsideration of these levels. There exists no directly
applicable scientific documentation of adverse nedical effects at
levels of fluoride below 8 mg/L (ppm). Therefore, it can be
concluded that four times optimum in U.S. drinking water supplies is
a level that would provide "no known or anticipated adverse effect
with a margin-of-safety."
4. The effects of various levels of fluoride intake on rapidly
developing bone in young children are not well understood. Also,
the modifying effects of to.al intake, length of exposure, other
nutritional factors and debilitating illness, are not well
understood. Therefore, the committee strongly recommends that the
PHS and the EPA join to enlarge the body of information relative to
'skeletal maturation and'growth in children ingesting more than the
recommended daily intake of fluoride.
VIII-19
-------
B. Qyjantificatlon of Non-Carcinogenic Effects
As stated earlier, the extensive amount of health effects Information on
humans and the need to establish a balance between adverse and beneficial
effects prevents use of the typical risk assessment approach to derive
appropriate drinking water concentration values for fluoride. Thus, the
approach selected must rely largely upon an interpolation and direct
application of the available human data on adverse and beneficial effects.
1. One-Day and Ten-Day Health Advisory
There is an absence of appropriate short-term animal or human
experimental or clinical studies on the effects of fluoride following oral
ingestlon frorc which one-day or ten-day Health Advisory (HA) values can be
calculated. The National Academy of Sciences Safe Drinking Water Committee
has reviewed the available literature on fluoride, but did not recommend a
x
suggested-no-adverse-response-level (SNARL) for fluoride (NAS 1977).
2. Adjusted Acceptable Daily Intake
Dental Fluoroslf. As stated earlier, the available date on the incidence
of dental fluorosis In humans (especially children) is extensive. As
summarized in Table VIII-1, the incidence of objectionable (moderate and
severe) dental fluorosis is not consistently observed in a marked segment of
the population until the drinking water concentration approaches 2.0 mg F/L.
This observation is consistent with the Surgeon General's recommendation
VIII-20
-------
(Koop 1982, Albertlnl et al. 1982) that communities limit drinking water to
twice optimum (about 2 mg F/L) to minimize the undesirable cosmetic effects of
dental fluorosls in children.
The Surgeon General's opinion on protecting children from dental
fluorosls v»as clearly presented in the context of ensuring adequate fluoride
exposure to provide reduced dental caries experience. It should be noted that
in the survey by Driscoll et al. (1983), the maximum statistically significant
reduction of dental carles was achieved at a drinking water concentration of
2.08 mg F/L. At 2.84 and 3.86 mg F/L, no statistically significant
improvement in dental carles reduction was obtained although the incidence of
moderate and severe dental fluorosls .increased.
Skeletal Fluorosls. No single human experimental or clinical scudy
provides an adequate basis for developing ar AADI for skeletal effects. It
should be clearly stated that skeletal fluorosis increases in severity with
both dose and duration of exposure to fluoride. In its mildest form, it is
characterized by an increase in bone density (osteosclerosls) that is
detectable only through X-ray examination. The most severe form (crippling
skeletal fluorosis) is characterized by irregular bone deposits. At the
rec- est of the F.PA, the U.S. Suryeon General examined the nondental health
aspects associated with fluoride in drinking water. An ad hoc advisory
con&ittee met In April, 1983 in Bethesda, MD and provided their report
(Shapiro 1983) and a later formal response from the Surgeon General (Koop
1984) to EPA. The Surgeon General concluded that he did not consider changes
in bone density to be an adverse health effect and that adverse effects
VIII-21
-------
(arthraigias) are not likely to occur at human dose levels below 20 mg F/day
(10 mg F/l. for an adult consuming 2 L water/day). The ad hoc committee
concluded that four times the optimal fluoride concentration (approximately
4 mg F/L) in drinking vater should provide an adequate margin of safety for
preventing adverse health effects which were not documented to occur in the
U.S. population below 8 og F/L.
Singh and Jolly authored a review of the skeletal effects of the fluoride
(WHO 1970). Their conclusion stated:
"It is, therefore, possible to conclude that the histopathological
changes of endemic fluorosis occur only at higher levels of intake than
1-4 ppm."
In a more recent survey of fluoride by WHO (1984), it was stated that
"...at 3.0 to 6.0 ng/L skeletal fluorosis may be observed; when 10 mg/L is
exceeded, crippling fluorosis could ensue." It should be noted that both WHO
summaries consider the effects of fluoride on worldwide populations. Thus,
their conclusions may not be directly applicable to the U.S. situation.
Both the Surgeon General's and WKO's evaluations of the available health
effects data on skeletal fluorosis appear to be generally consistent with the
primary published literature. The investigations with human subjects by Leone
et al. (1955). Stevenson and Watson (1957), Hodges et al. (1941), Hodge and
Smith (1970) and Dinman et al. (1976) -provide evidence that the no-observed-
adverse-effect-level (NOAEL) for the initial symptoms of skeletal fluorosis
VIII-22
-------
(increased bone density) is within the range from 3 to 8 mg/L of fluoride In
drinking water. The data compiled by Smith and Hedge (1959) and Hodge and
Smith (1981) indicate that radiologlcally detectable osteofluorosis is not
observed in bones containing approximately 5,000 ppm or less of fluoride on a
dry, fat-free basis. This skeletal fluoride concentration is associated with
a drinking water concentration of approximately 5 aig F/L (Hodge and Smith
1981). Although never observed In the U.S., the apparent NOAEL for crippling
skeletal fluorosis is approximately lO mg F/L (Shapiro 1983, Koop 1984, WHO
1984). Therefore, it is believed that a drinking water concentration of 4.0 mg
F/L will provide protection from crippling skeletal fluorosis with an adequate
margin of safety. Again, this is consistent with the Surgeon General's recom-
mendation to limit drinking water fluoride levels to four times optimum (about
4 mg F/L) to provide protection from crippling skeletal fluorosis (Koop 1984,
Shapiro 1983).
C. Carcinogenic Effects
No valid studies on the carcinogenic potential of fluoride in animals
were located in the literature. However, the National Cancer Institute initi-
ated studies during August 1979 to determine the carcinogenic and or toxico-
logical potential of sodium fluoride (NaF) in rats and mice. The National
Toxicology Program (NTP) took over the responsibility for oversight of the
studies in November 1982. The studies consisted.of three parts: (1) a one-
month subchronic study; (2) a six-month subchronle study with dosages based on
the previous experiment; and (3) a-^two-year chronic study based on data from
the six-month subchronic experiment (maximum doses of NaF which were not
VIII-23
-------
expected to affect the longevity of mice and rats were used). The chronic
study began in December, 1981 and terminated in December, 1983. 'Unfortun-
ately, problems developed seven months into the chronic study. The problems
were not treatment related (some rats in both the treatment and control groups
exhibited toxleollla nnd ocular IsaloniO but may MAVP Ueeti related to the tlloi
vhich was lov in several trace elements and vitamins. The validity of the
study was questioned and a new chronic study was scheduled. The Technical
Report from the new study should be issued in 1988.
Yiamouyiannis and Burk (1977) presented an analysis of mortality data
which they clained shoved an increase in the cancer mortality rate aeon;? resi-
dents of fluoridated areas. Later analyses (Strassburg and Greenland 1979,
Oldham and Newell 1977) have shown that Yiamouyiannis and Burk had failed to
consider the age-race-sex structure of the studied populations. Inclusion of
these factors in consideration of the data invalidated the conclusion that
fluoridatlon was responsible for an increase in the cancer mortality rate.
Other studies by Hoover et al. (1976) and the Environmental Health Directorate
of Canada (1977) found no correlation between fluoridation of water and the
cancer mortality rate. Further, the International Agency for Research on
Cancer (IARC) has performed an assessment of the degree of evidence for the
carcinogenicity of fluoride in humans and in experimental animals (WHO 1982).
This assessment concluded that no evidence could be found in the literature to
indicate that fluoride is carcinogenic.
VIII-24
-------
D. Existing Guidelines and Standards
Protection for the industrial worker against excessive exposure to air-
borne fluoride is achieved by occupational*standards set by OSHA and based on
the American Conference of Government Industrial Hygtenists Threshold Limit
Value (TLV) for airborne fluoride of 2.5 mg/m . This is a concentration which
should not cause an adverse health effect in a person so exposed for eight
hours/day, five days/week (NAS 1971). The Pennsylvania Short-Tenn Limit for
exposure to airborne fluoride is 10 mg/m for 30 minutes. This concentration
is permissible as long as the TLV is observed on .a tlet-weighted basis (NAS
1971).
•
Under the requirements of the National Interim Primary Drinking Water
Regulations of 1975 (USEPA 1976), EPA set the standarc-s (MCL) for fluoride
shown in Table VIII-2. These levels are twice Che concentrations defined as
optimal for the control of dental caries. The EPA (USEPA 1979) defined "un-
reasonable risk to health" as a fluoride concentration producing moderate to
severe fluoi'osls, or specifically, a Community Fluorosis Index exceeding 1.5.
In theory, the Index of 1.5 would correspond to fluoride concentrations exceed-
ing the established MCL for fluoride (twice the optimum for each temperature
zone).
The Food and Nutrition Board of the National Research Council has esti-
mated adequate and safe total intakes of fluoride as shown in Table VIII-3.
These levels are considered to be protective against dental caries and possibly
••
*
against osteoporosis (NAS 1980).
VIH-25
-------
Table VI11-2 Maximum Contaminant ' velpfl
Temperature Concentration
Degrees Degrees
Fahrenheit Celsius Milligrams per liter (p'pm)
53.7 and below 12.0 and below
53.8 to
58. 4 to
63.9 to
70.7 to
79.3 to
58.3. 12.1 to 14.6
63.8 14.7 to 17.6
70.6 17.7 to 21.4
79.2 21.5 to 26.2
90.5 26.3 to 32.5
2.4
2.2
2.0
1.8
1.6
1.4
&Hlghest peralssib.'.e concentration of a contaminant in the water delivered
to the consumer's tap.
Adapted from USEPA (1976).
VIII-26
-------
Table VII1-3 Food and Nutrition Board Estimated Adequate and
Safe Intakes of Fluoride
Age
group
<6 months
6-12 months
1-3 years
4-6 years
7 years-
adulthood
Adults
Estimated
weight (kg)
6
9
13
20
30a
70
Recoismended intake
of fluoride (nig/day)
0.1-0.5
0.2-1.0
0.5-1.0
1.0-2.5
1.5-2.5
1.5-4.0
Estimated
equivalences (mg/kg/day)
0.02-0.08
0.02-0.11
0.04-0.08
0.05-0.13
0.05-0.08
0.02-0.06
aEstinated weight for children seven to ten years old.
Adapted from NAS (1980).
VIII-27
-------
The Association for the Advancement of Medical Instrumentation has sug-
gested a maximum concentration of 0.2 mg F/L for water being used in dialysis.
The specific health effects basis for selection of this value, however, Is not
stated (Association for the Advancement of Medical Instrumentation 1981).
The Canadian Public Health Association (1979) recommended that 1.2 mg F/L
be established as the optlmua. concentration in that country's drinking water.
The World Health Organization (WHO 1970), after an extensive review of
the health effects of fluoride, concluded that: "When nutrition is adequate,
enrichoent of water so that it contains 1.0 to 1.2 ppo is advisable in temper-
ate zones. In warmer regions, the content should be smaller." The derivation
of these suggested levels for.fluoride is not specifically explained. The
suggestion is made after an extensive review of the literature on the
relationship of fluoride levels to dental carles experience and to dental
fluorosis. In the preface to this publication (WHO 1970) it is stated that
"The objective of this monograph is to provide an impartial review of the
scientific literature...It is not intended to be a practical guide to the use
of fluoride as a health measure...."
More recently the World Health Organization (WHO 1984) stated that, "at
(drinking water) levels above 1.5 mg/L, mottling of teeth has been reported
very occasionally, and at 3.0 Co 6.0 mg/L skeletal fluorosis may be
observed...." This review is based largely upon information in the 1970 WHO
Monograph (WHO 1970) and further states that no new evidence has been obtained
to justify modification of the current 1.5 mg/L guideline value for fluoride
VIII-28
-------
IP drinking water. The report caution* that "Iocs] climatic condition? and
Increased water IntnVe should be considered when applying this recommended
guideline value." The WHO reference to occasional skeletal fluorosls with the
consumption of drinking water In the range of 3.0 to 6.0 mg F/L in an estimate
of the lower limit for this effect under a variety of environmental and
nutritional conditions that are not necessarily reflective of the U.S.
situation.
The NAS (1977) has discussed fluoride in their document on drinking water
and health. This work includes several comments pertinent to the estimation
of an MCL for fluoride:
On the basis of studies done 'over 15 years ago, occasional objec-
tionable mottling would be expected to occur in communities in the
hotter regions of the United States with water that contains
fluoride at 1 ppm or higher and in any community with water that
contains fluoride at 2 ppm or higher. However, this may not be
the case today; more liberal provisional limits seem appropriate
while studies are conducted to clarify the subject.
...it was estimated that objectionable fluorosis occurs in the
range of 0.8-1.6 mg/liter fluoride, depending on the temperature.
No recent U.S. surveys or studies of communities have been found
on which a sound decision could be made that greater concentra-
tions are without objectionable effect.
...there is no generally accepted evidence that anyone has been
harmed by drinking water with fluoride concentrations considered
optimal for the annual mean temperature in the temperate zones.
In Gabovich and Ovrutsky (1969), a translated Russian review document, it
is stated that '.'All Union State Standard 2784-54" (USSR) has set by law the
fluorine concentration of 1.5 mg/L as the caximuc permissible amount in tap
water. After discussing the effect of ambient temperature on fluorosis, they
say that .(p. 612), "In year-round fluoridation of the water with a single dose
VIII-29
-------
of fluorine, the Commission on Hygiene of the Water Supply and Sanitary
Protection of Bodies of Water at the Ministry of Health of the USSR recommends
1 tng/L for regions with cold and temperate climates, for warm climates
0.9 mp/L, and 0.7 to 0.8 mg/L for hot climates." Gabovich (1952, cited In
Cabovich and Ovrutsky 1969) made the following cocinents concerning various
concentrations of fluoride in drinking water:
• Up to Ol3 mg/L is a very low concentration. At this concentration the
incidence of caries is high and defects in the mineralization of bones
are most frequently observed.
• Water with 0.7 to 1.0 ng/L has an optimum concentration. Damage by
caries is minimal, signs of dental fluorosis are also minimal.
• Fluoride concentration Is high at 1.0 to 1.5 og/L, but acceptable with
permission of the health authorities. Caries control is good and there
are signs of mild fluorosis. This level is acceptable in the absence of
data indicating an unfavorable influence on the health of the population.
• Concentrations of 1.5 to 2.0 mg/L are higher than the permissible level.
Caries control is good, but fluorosis is objectionable.
• Two to six mg/L is a high concentration. Caries control is not optimal
and fluorosis is objectionable with 10% to 30S having severe fluorosis.
VIII-30
-------
• Six to fifteen mp/l. Is a very high concentration. Tories control 1* not
optlnn] and up to JOOr of the population are afflicted with fluorosl*.
with the predominance of the severe forms.
Cabovlch and Ovrutsky (1969) state that the Indian standard for fluoride
In water is I mg/L (permissible) with 2 rag/L not permissible.
E. Special Considerations
1. High Risk Populations
Relatively small segments of the general population may be at increased
risk from waterborne fluoride. For example, polydipsia and polyurla
associated with diabetes insipidus and some forms of renal Impairment may
result In an excessive intake of drinking water and waterborne fluoride.
Skeletal fluorosis in patients with impaired renal function has been described
by Juncos and Donadio (1972). Patients with impaired renal function have been
shown to have a lesser renal clearance of fluoride than tave normal subjects
(Schiffl and•Binswanger 1980).
2. Beneficial Effects
•
a. Teeth
The principal beneficial effect attributed to fluoride is its role in
prevention of dental caries. A detailed review of the literature in this area
V1II-31
-------
will not hr attempted here becaupr It has been adequately addressed elsewhere
In this document. Studies have been reviewed that describe the continuum from
beneficial effects to dental fluorosip with Increased exposure to fluoride. A
summary of the dally fluoride intake levels considered to be protection
against both dental caries and possibly osteoporosis is provided in
Table VII1-3.
Fluoride.is also believed to improve the esthetic appearance of teeth.
A. L. Russell recorded the occurrence of developmental enamel hypoplasias (not
related to fluoride in drinking water) in children 7 to 14 years old (Ast et
al. 1956). In Kingston, where the drinking water contained 0.05 mg F/L, 115
(18.7 percent) of the 612 children examined showed these nonfluoride
opacities. Only 36 (8.2 percent) of 438 children using the fluoridated
Newburgh water (1.0 to 1.2 mg F/L) showed these changes. Ast et al. (1956)
suggested that this fluoride drinking water concentration (1.0 to 1.2 mg F/L)
appeared to reduce the incidence of hypoplastic spots on the teeth.
b. Bone
Jowsey et al. (1972) described the effects in 11 patients with
progressive osteoporosis who were administered 30, £5, 60 or 90 mg of NaF
daily. The patients, ten of whom were female, ranged from 54 to 72 years of
age. All subjects received vitamin D twice weekly and a dally supplement of-
calciuc. Treatment was continued for 12 to 17 months. The results indicated
that administration of less than 45 mg of NaF daily did not consistently
VIII-32
-------
Increase bone fornntlon, hut that 60 mg or more resulted in the production of
abnormal bone.- Side effects were evident in at least one patient receiving 30
mg NaF. Mild arthralgia and stiffness of the Joints were reported hy four
patients and occasional epigastric dyspepsia was experienced by six patients.
Daily addition of vitamin D and more than 600 mg Ca appeared to prevent
increased bone resorption and even to decrease resorptlon. The authors
concluded that doses of 50 mg of NaF daily, supplemented with 600 mg or more
of calcium daily and 50,000 units of vitamin D twice weekly should increase
skeletal mass wlchout undesirable skeletal effects. Also, further vertebral
fractures should cease after several years of treatment.
Dambacher et al. (1978) treated 33 post-menopausal women with 100 mg NaF
daily for two years and another 23 similar patients with 50 mg NaF daily for
two years. A decrease of cortical bone was evident ac both dose levels.
However, cancellous bone was increased to some extent in half of those
receiving the lower dose, and in over 70 percent of those receiving the higher
dose.. The findings also suggested that two years of treatment at the lower
dose or one year at the higher dose avoided new vertebral fractures.
Gastrointestinal discomfort sometimes combined with nausea was encountered
chiefly at the higher dose, but was of minor clinical importance.
Osteoarticular pain was the major side effect of fluoride therapy and was seen
in about 60 percent of the patients at both dose levels. The "maximum effect
was seen after 6 to 12 months of treatment and then gradually disappeared. In
18 percent of the patients treatment had to be discontinued.
VIII-33
-------
et aK (1982) studied five groups of women, totaling 165 patient*,
durlnp the period from 1968 to 1980. Fluoride was given (1) with cnJclum with
or without vitamin D and (2). with calcium and estrogen with or without vitamin
D. Doses were 40 to 60 mg NaF dally with a total of 61 patients (of 165
total) receiving fluoride. Of these, 23 (38 percent) developed adverse
reactions which caused five of then to withdraw from the study. These effects
were not seen in the control patients or in the other experimental groups
(those treated with calcium alone or with vitamin D, or with calcium plus
estrogen with or without vitamin D).
Among the patients treated with NaF, 60 percent showed radiographically
demonstrable increases In vertebral bone mass; Patients with these changes
showed about one-seventh the fracture rate of the other patients. The
incidence of fractures per 1000 patient-years for patients treated with
fluoride, calcium and estrogen (with or without vitamin D) was significantly
less than in controls (P<1* x 10 ) and also was significantly less than in
those treated with fluoride and calcium (with or without vitamin D) (P
-------
areas where the water supplies contained 4 to 5.8 vtp F/L. Similar Information
was obtained for 312 male and. 403 female long-term users of water supplies
containing 0.15 to 0.3 mg F/L. More than 50 percent of the participants In
each area had never lived outside their respective areas. The subjects of
each sex In each population were grouped by age into those 45 to 56 years old,
55 to 66 years old and 65 years old and over. Evidence of osteoporosis,
reduced bone density and incidence of collapsed vertebrae were higher in the
low fluoride area in both sexes. For women 55 to 66 years old and
65 years old and older the difference in prevalence of reduced bone density
•.
was significant at the P<0.01 level. In men the difference was significant
only for the 55- to 66-year-old group (P<0.05). More subjects in the high
fluoride area had normal or increased bone density. There was no significant
difference in the incidence of collapsed vertebrae among male residents of the
two areas. For women, the greater Incidence of collapsed vertebrae in the low
fluoride area was significant at the P<0.05 and P<0.01 levels for the 55- to
66-year-old and the 65-year-old and over groups, respectively. The authors
concluded that 4 to 5.8 mg F/L in drinking water "materially and
i
significantly" reduced the prevalence of osteoporosis and collapsed vertebrae,
and that the effects were more pronounced in women than in men.
c. Cardiovascular
In the study by Bernstein et al. (1966) the incidence of aortic
calcification (as seen in the X-ray films) was less in residents of the high
fluoride area than in those using low fluoride water. The difference was
VIII-35
-------
approximately 40 percent and wan statistically significant for men In all age
groups. Women In the. 5S- to 64-year-old group also showed a statistical ly
significant difference In the Incidence of aortic caltlfIcation. A similar
trend, although not statistically significant, was observed In females 65
years of age and older.
d. Hearing
Shambaugh and Causse (197O treated more than 6,000 patients with active
otospongiosls of the cochlear capsule with sodium fluoride for 1 to 8 years,
using doses of 40 to 60 mg dally with calcium and vitamin D supplements. The
fluoride was administered in enteric coated tablets. In about 80 percent of
the treated patients there was a stabilization of the sensorineural component
of hearing loss', with recalcification and Inactivation of the actively
expanding demineralized focus of otospongiosls. In a few cases hearing was
Improved, .while in others the hearing loss continued to worsen. In a number
of instances, cessation of therapy after stabilization of hearing and
recalcifIcation had been achieved was followed (two to seven years later) by
reappearance of a demineralized focus and an Increase in the sensorineural
loss. Shambaugh and Causse (1974) recommended a maintenance dose of 20 mg
daily of sodium fluoride after stabilization has been achieved.
Causse et al. (1980) gathered more evidence for the beneficial effect of
fluoride therapy on otospongiotic foci through polytocographic studies,
statistical analysis of 10,441 cases (with a follow-up of three months to ten
VIII-36
-------
years) and by comparing trypsln concentration in the peri lymph before and
after NaF therapy. Trypeln, which IP toxic to hair cells and destroys collagen
fibrils in the bony otic capsule, was significantly (no P value given) reduced
in 66? of cases at moderate NaF (65 ing/day) doses. Fluoride therapy causes
expulsion of cytotoxlc enzymes into labyrinthine fluid? and retardation of
sensorineural deterioration. The long-term effect of therapy is the reduction
of the bone remodelling activity of the otosponglotic focus. NaF therapy (in
patients with cochlear deterioration and progressive cochlear component) can
Improve hearing in children but can only arrest deterioration in older patients.
NaF may retard, but cannot release, stapedial fixation. Fluoride action
reduces vertigo is an effect on vestibular function. Dosages used by the
authors range from 3 to 60 ing/day depending on the nature of the otospongiotic
impairment (in children only 1.5 to 10 mg/day are prescribed to avoid stunting
growth). The authors observed no fluorosis in more than 10,000 cases.
3. Interactions
Aluminum salts inhibit the absorption of fluoride, as has been shown by
Hobbs et al. (1954). Incorporation of aluminum sulfate into the ration of
livestock resulted in an increase of fecal excretion of fluoride, a decrease
in urinary excretion, decreased skeletal storage and lessened mottling of
incisor teeth.
4. Relative Source Contribution
»
Approximately one million persons in the .U.S. use water supplies which
contain more than twice the local optimal concentration for prevention of
VIII-37
-------
dental carl PP. These people live In about 1,200 communities, mostly In the
southwest fSmall 1983). Water supplies used by some of these communities may
contain more than 4 mg F/I.. In general, waterborrie fluoride concentrations
are higher west of the Mississippi River than they arc In the eastern part of
the continental U.S. (Smith 1983a). Further, the highest concentrations in
water supplies are observed in areas where the soil is rich In apatite or
other fluoride minerals and the water is obtained' from veils (NAS 1971).
Frequently these higher fluoride areas are In regions where the annual mean
temperatures are somewhat elevated. As a result, water consumption may be
higher and the fluoride intake by the population increased.
According Co the National Institute for Occupational Safety and Health
(NIOSH 1975, 1976) there were approximately 22,000 workers exposed to hydrogen
fluoride In 57 different occupations and 350,000 workers exposed to inorganic
fluorides In 92 occupations (1976 and 1975 estimates, respectively). These
workers are exposed to airborne fluorides during their worklrg hours, but the
concentrations are limited by Occupational Safety and Health Administration
standards intended to prevent the development of skeletal fluorosis. There is
no evidence tha :. •'• '.duals occupationally exposed to fluorides and using
fluoridated war developed radlographically demonstrable skeletal
fluorosis.
F. Summary
While there is considerable variation in the numerous epidemiological
studies performed, it is believed that the incidence of objectionable
VII1-38
-------
(moderate and severe) dental fluorosis begins to impact a marked segment of
the population when the drinking water concentration approaches 2.0 mg F/L.
The available human data indicate that the no-effect-level for the
initial symptoms of skeletal.fluorosis (increased bone density) is between 3.0
to 8.0 mg F/L. The NOAEL for crippling skeletal fluorosis in the U.S. is
believed to be at drinking water concentrations at or above 10.0 mg F/L.
Thus, a drinking water concentration of 4.0 mg F/L is considered to provide
adequate protection for crippling skeletal fluorosis with a margin of safety.
VIII-39
-------
IX. REFERENCES
Albertlnl T, Bock V, Confrancesco J, Drlscoll V, Small JS, Clark N. 1982.
Ad-hor committee report on der.tal fluororip. Draft Report to the Chief Denta]
Officer, PHS. July 21.
Association for the Advancement of. Medical Instrumentation. 19P1. American
National Standard for Hemodlalysis Systems. Arlington VA: Association for
the Advancement of Medical Instrumentation, p. 3.
Ast DB, Smith D.7, Wachs B, Cantwell KT. 1956. Newburgh-Kingston carles-
fluorine study XIV. Combined clinical and roentpenographic dental findings
after ten years of fluoride experience. J. Am. Dent. Assoc. 52:314-325.
Bennan LB, Taves DR. 1973. Fluoride excretion in normal and ureralc humans.
Clln. Res. 21:100.
Bernstein DS, Sadowskv N, Hegsted DM, Gurl CD, Stare, FJ. 1966. Prevalence
of osteoporosis in high- and low-fluoride areas In North Dakota. JAMA
198:499-504.
Berry WT. 1958. A study of the influence of mongolism In relation to the
fluoride content of water. Aa. J. Ment. Defic. 62:634-636.
Black GV, McKay FS. 1916. Mottled teeth: an endemic developmental imperfec-
tion of the enamel of the teeth heretofore unknown in the literature of
dentistry. Den. Cosmos 58:129-156.
Black MM, Kleiner IS, Bolker H. 1949. The toxicity of sodium fluoride in
man. NY State J. Med. 49:1187-1188.
Canadian Public Health Association. 1979. Criteria document in support of a
drinking water standard for fluoride. Ottawa, Ontario, Canada: Canadian
Public Health Association.
Carlson CH, Armstrong WD, Singer L. 1960a. Distribution and excretion of
radiofluoride in the human. Proc. Soc. Exp. Biol. Med. 104:235-239.
Carlson CH, Singer L, Armstrong WD. 1960b. Radiofluoride distribution in
tissues of normal and nephrector.ized rats. Proc. Soc. Exp. Biol. Med.
103:418-420.
Caruso FS, Hodge HL. 1965. The effect of oral doses of sodium fluoride on
blood pressure in dogs. J. Dent. Res. 44:99-101.
Caruso FS, Maynard EA, DiStefano V. 1970. Chapt. 4. Pharmacology of sodium
fluoride. In: Smith FA, ed. Handbook of experimental pharmacology. Vol. XX.
Pharmacology of fluorides. Part 2. New York: Springer-Verlag, pp. 14*,-!65.
Causse SR, Shanbaugh GE, Causse JB, Bretlau P. 1980. Enzymology of
otospongiosis and NaF therapy. Am. J. Otology 1(4):206-213.
IX-1
-------
Conlgllo W. 1984. Water Intnke and temperature. Memorandum of Januarv 24,
1984. Office of Drinking Water, U.S. Environmental Protection Agency.
Washington, DC.
Cook-Mozaffari P, Bulusu L, Doll R. 1981. Fluorldatlon of water supplies and
cancer mortality. I: a search for an effect In the UK on rlpV of death fror.
cancer. J. Epldeir.lol. Com. Hlth. 35:277-232.
Cook-Mozaffarl P, Doll R. 1981. Fluorldation of water supplies and cancer
mortality. II: mortality trends ;after fluoridation. J. Epidetr.iol. Conan.
Hlth. 35":233-238.
Dambacher MA, Lauffenburgcr TH, Lammle B, Haais HC. 1978. Long-tern effects
of sodium fluoride in osteoporosis. In: Courvoisler B. Donath A, Baud CA.
eds. Fluoride and bone. Second symposium, Centre d1Etude des Maladies
Osr"o-artlculaires de Geneve. Bern: Han Huber Publishers, pp. 238-241.
Dean HT. 1933. Distribution of mottled teeth in the United States. Public
Health Rep. 48:703-734.
Dean HT. 1934. Classification of mottled enamel diagnosis. J. Ac. Dent.
Assoc. 21:1621-1426.
Dean HT. 1942. The investigation of physiological effects by the epideaio-
logical method. In: Moulton FR, ed. Fluorine and dental health. Washington,
DC: American Association for the Advancement of Science, Publication No. 19,
pp. 23-jl.
Dean HT, Elvove E. 1935. Studies on the minimal threshold of the dental sign
of chronic endemic fluorosis (mottled enamel^. Public Health Rep.
50:1719-1729.
Dean HT, Elvove E. 1937. Further studies on the minimal threshold of chronic
endemic dental fluorosis. Public Health Rep. 52:1249-1264.
Dinmar. BD, Elder MJ, Bonney TB, Bovard PG, Colwell MO. 1976. ( A 15-year
retrospective study of fluoride excretion and bony radiopacity among aluminum
smelter workers - Pt. 4. J. Occup. Med. 18:21-25.
Drlscoll WS, Horowitz HS, Meyers RJ, Heifetz SB, Kirgcan A, Zimmerman ER.
19"3. Prevalence of dental caries and dental fluorosis in areas with optimal
and above optimal water fluoride concentrations. JADA 107:42-47.
Drury. JS, Ensminger JT, Hammons AS et al. 1980. Reviews of the environmental
effects of pollutants: IX. Fluoride. Washington, DC: (Oak Ridge National
Laboratory, ORKL/EIS-85) U.S. Environmental Protection Agency. EPA-600/1-78-050.
Eager JM. 1901. Denti di Chiale teeth (teeth). Pub. Health Rep. 16:2576.
» • •%
Ekstrand J, Alven G, Boreus LO, Norlin A. 1977. Pharmacokinetics of fluoride
in man after single and multiple oral doses. Europ. J. Clin. Phannacol.
12:311-317.
IX-2
-------
Ekstrand J, Koch G, Petersson LG. 1983. Plasma fluoride concentrations in
pre-school children after ir.gestion of fluoride tablets and toothpaste.
Caries F.es. 17:379-36-.
Environnental Health Directorate. 1977. Fluoridation and cancer: an
analysis of Canadian drinking water fluoridation and cancer mortality data.
Canada: Ministry of Health and. Welfare. 77-EHD-18.
Environmental Health Directorate. 1982. Tap water consumption in ranada.
Canada: Ministry of Health and Welfare. 82-EKD-f-O.
Ericsson Y. 1958. The state of fluorine in miik and its absorption and
retention when administered in milk. Acta Odontol. Scand. 16:51-77.
Evans FG, Wood JL. 1976. Mechanical properties and density of bone in a case
of severe endemic fluorosis. Acta Orthop. Scand. £7:489-495.
Faccini JM, Teotia SPS. 19^4. Histopathological essesstaent of endemic
skeletal fluorcsis. Calcif. Tissue Res. 16:45-57.
Farley JR, Vergedal JE, Baylink DJ. 1983. Fluoride directly stimulates
proliferation and alkaline phosphatase activity of bone forming cells.
Science 222:330-332.
Feldman I, Morken D, Hodge HC. 1957. The state of fluoride in drinking
water. J. Dental Res. 36:192-202.
Franke J, Rath F. Runge H, Fengler F, Auermann E, Lenart G. 1975. Industrial
fluorosis. Fluoride 8:61-85.
Fritz H. 1958. Roentgenologische und pathologische-anatomische betrachtungen
zuc fluoroseproblem. Med. Habilitation-sschr. (In German).
Gabovich RD. 1952. In: Gabovich RD, Ovrutsky GD, eds. 1969. p. 612.
Gabovich RD, Ovrutsky GD. 1969. Fluorine la stomatology and hygiene.
Bethesda, MD: U.S. Department of Health, Education and Welfare, Public Health
Service. DHEW Pub. No. (NIH) 78-785. Translated and published by U.S. DHEV
ir. 1977.
Galagan DJ, La-nson GG. 1953. Climate and endemic d.ntal fluorosis. Public .
Health Rep. 68:497-508.
Gall JF. 1966. Kirk-Othaner enclyclopedia of chemical technology. New York:
John Wiley & Sons, Inc., pp. 610-624.
Geever EF, Leone NC, Geiser ?, Liebenaan JE. 1958. Patholog-ic studies in man
after prolonged ingestior. of fluoride in drinking water. 1. Necropsy findings
in a community with a water level of 2.5 ppm. J. Am. Dent. Assoc. 56:499-507.
Grinbergen GV. 1974. A double blind test for determination of intolerance to
fluoridated water (preliminary report). Fluoride 71:146-152.
IX-3
-------
Hagan TL, Pasternack M, Scholz GC. 1954. Waterborne fluorides and mortality.
Public Health Rep. 69:450-454.
Hanhijarvi H, Anttonen VM, Pekkarinen A, Pentitila A. 1972. The effect of
artificially fluoridated drinking water on the plasna of ionized fluoride
content in certain clinical diseases and in normal individuals. Acta Phartaacol.
Toxicol. 31:104.
Hansen K. 1978. Effects of fluoridated water in tissues of CSE female mice.
Bios 19':51-55.
Harris NO, Hayes RL. 1955. A tracer study of the effect of acute and chronic
exposure to sodlun fluoride on the thyroid iodine metabolism of rats. J.
Dent. Res. 34:470-477.
Heasman MA, Martin AE. 1962. Mortality in areas containing natural fluoride
in their water supplies. Mon. Bull., Min. Health Public Health Lab.
Ser. 21:150-160.
Hem JW, Bonner JF, Brudevold F, Snith FA, Hodge HC. 1956. Distribution in
the soft tissue of the rat of radioactive fluoride administered as sodium
fluoride. Nature 178:1295-1296.
Hobbs CS, Merriman CM. 1962. Fluorosis in beef cattle. Tennessee Agri-
cultural Experiment Station Bulletin 351. Knoxville: University of
Tennessee.
Hobbs CS, -Moorman RP Jr, Griffith JM et al. 1954. Fluorosis in cattle and
sheep. Tennessee Agricultural Experiment Station Bulletin 235. Knoxville:
University of Tennessee.
Hodge HC, Smith FA. 1965. Biological effects of inorganic fluorides. Ir:
Simons JH, ed. Fluorine chemistry, Vol. IV. New York: Academic Press, pp.
3-7, 10-13, 29-33, 89-113, 518.
Hodge HC, Smith FA. 1970. Air quality criteria for the effects of fluorides
on man. J. Air Pollut. Control Assoc. 20:226-232.
Hodge HC, Smith FA. 1981. Chapt. 12. Fluoride. In: Bonner F, Colburn J*.
eds. Disorders of mineral metabolism. Vol. 1. Trace minerals. New York:
Academic Press, pp. 439-483.
Hodge HC, Smith FA, Gedalia I. 1970. Excretion routes other than urinary.
In: Fluorides and human health. Geneva: World Health Organization, pp.
142-144.
Hodges PC, Farced OJ, Ruggy G, Chudnoff JS. 1941. Skeletal sclerosis in
chronic sodium fluoride poisoning. JAMA 117:1938.
Hoover RN,-McKay FW, Frauiner.i JF Jr. 1976. Fluoridated drinking water and
the occurrence of cancer. J. Natl. Cancer Inst. 57:757-768.
IX-4
-------
Illinois Department of Public Health, Bureau of Statistics. 195?- Mortality
In fluoride and non-fluoride areas. Health Statistics Bulletin No. 20.
Springfield. 1L: Illinois Department of Public Health.
1ARC. 198?. International Agency for Res*arch on Cancer, WHO. Inorganic
fluorides used In drinking-water and dental reparations. In: IARC
monographs on the evaluation of the carcinogenic risk of chealcals to humans.
Vol. 27. pp. 237-303.
Jackson D, Veldmann SM. 1958. Fluorine in human bone related to age and the
water supply of different regions. J. Pathol. Bacteriol. 76:451-459.
Jackson D, Weldmann SK. 1959. The relationship between aged and Che fluorine
content of human dentine and enamel: a regional survey. Brit. Dental J.
107:303-306.
Jagltllo C, Lin J-S. 1974. Sodium fluoride at potential outagrn in mammalian
cggi, Arch. Environ. Health 29i230-235,
Jowsey J, Rlggs BL, Kelly PJ, Hoffman DL. 1972. Effect of combined therapy
with sodium fluoride, vitamin D and calcium in osteoporosis. Ain. J. Hed.
53:43-49.
Juncos LI, Donadio JV. 1972. Renal failure and fluorosis. JAKA 222:783-785.
Kay MI, Young RA, -Posner AS. 1964. Crystal structure of hydroxyapatite.
Nature 204:1050-1052.
Kempf GA, McKay FS. 1930. Mottled enamel ic a segregated population. Public
Health Rep. 45*:2923-2940.
Klnlen L, Doll R. 1981. Fluoridatlon of water supplies and cancer mortality.
Ill: a re-examination of mortality in cities In the USA. J. Epidemlol.
Community Health. 35:239-244. •
Kleck RE. 1984. Report to the Office of Drinking Water. U.S. Environmental
Protection Agency, Washington, D.C. from Che review panel psychological/behavioral
effects of dental fluorosis. November 17.
Knizhnikov VA. 1958. Effect of the natural fluorine content of water on
certain health ind'ces in an adult population. Gig. Sanit. 23:18-23. (In
Russian; translated by Kettering Laboratory,.Cincinnati, Ohio).
Koop CE. 1982. Letter Co John W. Hernandez Jr. July 30.
Koop CE. 1964. Letter to William D. Ruckelshaus. January 23.
Kran D, Schneider EL, Singer L, Martin GR. 1978. The effects of high arid low
fluoride diets on the frequencies of sister chromatid exchanges. Mutat. Res.
57:51-55.
IX-5
-------
Kramer L, Osls D, Wl*trowskl E, Spencer H. 1974. Dietary fluoride lr
different areas of the United States. Am. J. Clln. Nutrit. 77:590-594.
Largent EJ. 1960. The metabolism of fluorides In man. Am. Ked. Assoc. Arch.
Ind. Health 21:318-323.
Leone NC, Ceever EF, Moran NC. 1956. Acute and subacute toxlcity studies of
sodiuc fluoride In animals. Public Health Rep. 71:459-667.
Leone NC, Shimkin KB, Arnold FA, et al. 1954. Medical aspects of excessive
fluoride in a water supply. Public Health Rep. 69:925-936.
Leone NC, Stevenson CA, Besse B, Haves LE, Davber TR. 1960. The effects of
the aosorptlon of fluoride. II. A radiological investigation of five hundred
forty-six human residents of an area in which the drinking water contained
only a minimum trace of fluoride. Am. Med. Assoc. Arch. Ind. Health
21:326-327,
Leone NC, Stevenson CA, Hllblsh TF, Sosman MC. 1955. A roentgenologlc study
of a human population exposed to high-fluoride domestic water (a 10-year
study). Am. J. Roentg. RadiumTher- Nucl. Med. 74:874-885.
Letklewicz FD. 1984. Fluoride occurrence in drinking water, food and air
Prepared by JRB Associates, McLean, VA, for Office of Drinking Water, U.S.
Environmental Protection Agency. Washington, DC. EPA Contract 68-01-6388.
Lewis F, Faine RC. A fluorosis study in selected South Carolina coastal
communities.
Lldbeck VL, Hill IB, Beeman JA. 1943. Acute sodium fluoride poisoning. JAMA
121:826-827.
Luoma H, Helminen SKJ, Ranta H, Rytomaa I, Meurman JH. 1973. Relationships
between the fluoride and magnesium concentrations in drinking water and some
components in serum related to cardiovascular diseases in men from four rural
districts in Finland. Scand. J. Clin. Lab. Invert. 32:217-224.
Kachle W, Largent EJ. 1943. The absorption and excretion of fluoride. J.
Ind. Hyg. 25:112-123.
Ma-ians J, Kleiiana'n JC, Coruoni-Iiuntley J. 1983. The relationship bt.-veen hip
fracture and water fluoridation: an analysis of national data. Ac. J. Pub.
Health 73:296-298.
Martin GR, Brown KS, Matheson DW, Lebowitz H, Singer L, Ophaug R. 1979. Lack
of cytogenetic effects in mice or mutations in Salmonella receiving sodiuc
fluoride. Mutat. Res. 66:159-167.
Maynard EA, Downs WL, LeSher MT. 1951. Toxicology of some fluoride compounds.
Univ. Rochester At Energy Proj. Quart. Tech. Rep. UR-164, pp. 73-77.
McCann HG, Bullock FA. 1957. The effect of fluoride ingestion on the composi-
tion and solubility of mineralized tissues of the rat. J. Dental Res. 36:391-398.
IX-6
-------
McCauley HB, McClure F.i. 1954. Effecr of fluoride In drinking water on the
osseous development of the hand and wrist In children. Public Health Rep.
69:671-683.
McClure FJ. 1944. Fluoride domestic waters and fysteroic effects. I.
relation to bone fracture experience, height anri weight of high school boys
and young selectees of the armed forces of the I'nlted States. Public Health
Rep. 59:1543-1558.
McClure FJ, Likins RC. 1951. J. Dental Res. 30:172-176.
McClure FJ, Mitchell HH, Hamilton TS, Kinser CA. 1945. Balances of fluoride.
ingested from various sources in food and water by five young men. J. Ind .
Hyg. Toxlcol.' 27:159-170.
McCollum EV, Siomonds N, Becker JE, Bunting RW. 1925. The effect of additions of
fluorine to the diet of the rat on the quality of the teeth. J. Biol. Chen.
63:553-562.
Mohamed AH, Chandler MEW. 1976. Cytological efrects of sodium fluoride on
oitotic and meiotic chromosomes of m*» " . Aaer- Chem. Soc. Abst. Paper. 172
fFluo):No. 24.
NAS. 1971. National Academy of Sciences. Committee on Biologic Effects of
Atmospheric Pollutants. Fluorides: biologic effects of atmospheric
pollutants. Washington, DC: National Academy of Sciences, pp. 175-176,
227-232.
NAS. 1977. National Academy of Sciences. Safe Drinking Water Committee.
Drinking water and health. Part 1, Chapters 1-5. Washington, DC: National
Academy of Sciences, pp. V-184-V-348.
WAS. 1963. tfatiaffAl Ae*4«»y at Sciences, Prinking v«t*r *o4 n««ti.t), Vet . 3.
Washington, DC: National Academy Press, pp. 279-283.
NCI. 1975. National Cancer Institute. Statement by the National Cancer
Institute on fluoridatlon studies by the Rational Health Federation. Bethesda,
KD: U.S. Department of Health, Education ard Welfare, Division of Dentistry,
FL-75. March.
NDWAC. 1982. National Drinking Water Advisory Cou.cil. Minutes of meeting,
October 26, 19S2. U.S. Environmental Protection Agency, Office of Drinking
Water, Washington, DC. October.
NDWAC. 1984. National Drinking Water Advisory Council. Minutes of meeting,
August 2 and 3, 1984. U.S. Environmental Protection Agency, Office of
Drinking Water, Washington, DC. August;
NDWAC. 1985. National Drinking Water Advisory Council. Minutes of meeting,
December 6 and 7, 1984. U.S. Environmental Protection Agency. Office of '
Drinking Water, Washington, DC. August.
TX-7
-------
Needlenar HI., Pueschel SM, Rothman KJ. 1974. Fluoridatlor and the occurrence
of Down's syndrome. N. Engl. J. Med. 291:821-823.
Neumar. WF, Neuroan MV, Main ER, O'Leary J, Snith FA. 1950. The surface cherlstry
of bone. II. Fluoride deposition. J. Biol. Chen. 187:655-661.
NIflSH. 1975. National Institute for Occupational Safety and Health.
Criteria for A Recommended Standard . . . Occupational Exposure to Inorganic
Fluorides. U.S. Dept. of Health, Education and Welfare, Public Health
Service, Center for Disease Control. 191. pp.
NIOSH. 1976. National Institute for Occupational Safety and Health. Criteria
for a Recommended Standard . . . Occupational Exposure to Inorganic Fluorides.
U.S. Dept. of Health, Education and Welfare, Public Health Service, Center for
Disease Control. 181. pp.
Office of Science and Technology, State of Michigan. 1979. Report of the
Governor's Task Force on Fluorides.
Okushl I. 1954. Changes of the heart muscle due to chronic fluorosls. Part
I. Electrocardiogram and heart X-ray picture made in inhabitants of
high-fluorine zone. Shikoku Acta Med< 5:159-165.
Oldh'am PD, Newell DJ. 1977. Fluoridation of water supplies and cancer - a
possible association? Appl. Stet. 26:125-135.
Oreopoulos DE, Taves DR, Rabinovich S et al. 1976. Fluoride and dialysis
osteodystrophy: results of double-blind study. Trans. An. Soc. Artif. Int.
Organs 20:203-208.
Osls D, Kramer L, Wiatrowski E, Spencer H. 1974. Diet fluoride intake in
man. J. Nutrit. 106:313. Cited in. Singer et al. 1980.
Peirce AW. 1959. Studies on fluorosis of sheep. III. The toxicity of
water-borne fluoride for the grazing sheep throughout its life. Aust. J. Agr.
Res. 10:186-199.
Perkinson JD, Whitney IB; Monroe RA, Lotz WE, Comar CL. 1955. Metabolism of
fluorine-18 in domestic animals. Am. J. Physiol. 182:383-389.
Perry RH, Cnilton CK, eds. 1973. Chemical engineers' handbook. 5th ed. New
York: McGraw-Hill Book Company.
Rapaport I. 1959. Nouvelles recherches sur le mongolisme. A propos du role
pathogenlque du fluor. Bull. Acad. Nat. Med. 143:367-377.
Richards LF, Westmoreland WW, Toshiro M, McKay CH, Morrison JT. 1967.
Determining optimum fluoride levels for community water supplies in relation
to temperature. J. Am. Dent. Assoc. .74:389-397.
• •»
Richmond JB. 1980. Letter to Dr. I. Lawrence Kerr. April 4.
IX-8
-------
Riggs BL, Seeoan E, HodgPor SF, Taves DR, O'Fallon WK. 1982. Effect of t>>e
fluoride/calcium regimen on vertebral fracture occurrence In postroenopausa]
osteoporosis. N. Eng]. J. Med. 306:446-450.
Rogot E, Sharre.tt AR, Feinleib M, Fabsltz RR. 1978. Trends in urban
mortality in relation to fluoridation status. Am. J. Epldemiol. 107:104-112.
Roholm K. 1937. Fluorine intoxication: a clinical hygienic study, with a
review of the literature and some experimental investigations. London: Lewis
ard Co. , pp. 140-147.
Roman RJ, Carter JR, North WC, Kauker ML. 1977. Renal tubular site of action
of fluoride in Fischer 344 rats. Anesthesiol. 46:260-264.
Royal College of Physicians. 1976. Fluoride, teeth and health. London:
Pitman Medical and Scientific Publishing Co., Ltd., pp. 51-56.
Rush GF, Willis LR. 1982. Renal tubular effects of sodium fluoride. J.
Phannacol. Exptl. Therap. 223:275-279.
Schiffl HH, Binswanger U. 1980. Human urinary fluoride excretion as
influenced by renal functional impairment. Nephron. 26:69-72.
Schlegel H. 1974. Industrielle skelettfluorose vorlanfiger bericht uber 61
falle ans alurcinumhutten. Soz. Praventivraed. 19:269-274.
Schlesinger ER, Overton DE, Chase KC. 1956a. Study of children drinking
fluoridated and nonfluoridated water. Quantitative urinary excretion of
albumin and formed elements. J. Am. Med. Assoc. 160:21-24.
Schlesinger ER, Overton DE, Chase HC, Cantwell KT. 1956b. Newburgh-Kingston
caries-fluorine study XIII. Pedlatric findings after ter. years. J. Am. Dent.
Assoc. 52:296-306.
Schour I, Poncher KG. 1937. Rate of opposition of enamel and dentin, measured
by the effect of acute fluorosis. Amer. J. Dls. Child. 54:757-796.
Schour I, Smith MC. 1934. The histologic changes in the enamel and dentin
of the rat incisor in acute and chronic experimental fluorosis. Arizona Agri-
cultural Experiment Stations Technical Bulletin 52. Tucson, AZ: University of
Arizona, pp. 69-9!.
Schultz JA, Lamb R. 1925. The effect of fluorine as sodium fluoride on the
growth and reproduction of albino rats. Science 61:93-94.
Segreto VA, Collins EK, Camarm D. Smith CT. 1984. A current study of mottled
enimel in Texas. J. Am. Dent. Assoc. 108 56-59.
Shambaugh GE, Causse J. 1974. Ten years experience vith fluoride in
otosclerotic (otospongiotlc) patients. Ann. Otol. 83.635-642.
IX-9
-------
Shapiro JR. 19P3. Report tc the Surgeon Generfl!: by the ad hoc committee on
the non-dental health effects of fluoride in drinking water- September 2fi.
Based upon discussion and recommendations made during'a meeting held in
Bethesde, Maryland on April 18-19, 1983.
Shortt HE, McRobert ER, Barnard TV, Nayar ASM. 1937«. Endemic fluorosls in
the Madras presidency. Indian J. Med. Res. 25:553-568.
Shortt HE, Pandit CG, Raghavachari TNS. 1937b. Endemic fluorosls in Nellore
district of South India. Indian Med. Gaz. 72:396-398.
Shupe JL. 1982. Proceedings of the International Fluoride Symposium. Logan
Utah: Utah State University. May.
Shupe JL, Miner ML, Greenwood DA, Harris LE, Stoddard GE. 1963. The effect
of fluorine on dairy cattle. II. Clinical and pathologic effects. Amer. J.
Vet. Res. 26:966-984.
Singer L, Armstrong WD. 1965. Determination of fluoride. Anal. Biochen.
10:495. Cited in Kramer et al. 1974.
Singer L, Orphaug RH, Harland BF. 1980. Fluoride intake of young male adults
In the United States. An. J. Clin. Nutrit. 33:328-332.
Singh A, Dass R, Hayreh SS, Jolly SS. 1962. Skeletal changes in endemic
fluorosis. J. Bone Joint. Surg. 448:806-815.
Singh A, Jolly SS. 1970. Chronic toxic effects on the skeletal system. In:
Fluorides and human health. Geneva: World Health Organization, pp. 238-249.
Singh A, Jolly SS, Bansal BC, Mathur CC. 1963. Endemic fluorosis. Epideci-
ological, clinical and biochemical study .of chronic fluorine intoxlcat.ion in
Panjab (India). Med. 42:229-246.
Small JS. 1983. Information specialis.t: fluorides and health. National
Institute of Dental Research, National Institutes of Health, Bethesda, MD '
20205. Personal Communication.
Smith FA. 1983a. Dose-response effects of fluoride in mammalian species.
In: Proceedings of the 17th Annual Conference on Trace Substances in
F-.vironaental Health, University of Missouri, Columbia, June 13-16, .983.
Smith FA. 1983b. Personal communication.
Smith FA, Gardner DE, Hodge HC. 1953. Age Increase in fluoride content in
hunan bone. Federation Proceedings 12:368.
Smith FA, Gardner DE, Leone NC, Hodge HE. 1960. The effects of the
absorption of fluoride. A.M.A. Arch, of Ind. Health 21:330-332.
IX-10
-------
Smith FA, Hodpe HC. 1959. Chapt. 1. Fluoride toxlcity. In: Muhlcr JC,
Hlne MX, eds. Fluorine and dentn] health. The pharmacology and toxicology of
fluorine. Bloomlnpton, IN: University of Indiana Press, pp. 19-20.
Smith FA, Hodge HC. 1979. Airborne fluorides and mar: part I. CRC Crlt.
Rev. Environ. Control 8:293-371.
Scith MC, Lantz EM, Smith HV. 1931. Univ. Ariz. Agr. Exprl. Sta. Bull. 32.
Stevenson CA, Vacson AR. 1957. Fluoride osteosclerosls. Am. J. Roentg.
Radium Ther- Nucl. Med. 78:13-18.
Still CN, Kelley P. 1980. On the incidence of primary degenerative dementia
versus vater fluoride content in South Carolina.
Stoddard GE, Harris LE, Bateman GQ, Shupe JL, Greenwood DA. 1963. Effects of
fluorine on dairy cattle. I. Growth and feed consumption. J. Dairy Sci.
46:1094-1102.
Strassburg MA, Greenland 1979. Methodologlc problems ir evaluating the
carcinogenic risk of environmental agents. J. Environ. Health 41:214-217.
Strubelt 0, Iven H, Younes K. 1982. The pathophyslologlcal profile of the
acute cardiovascular toxicity of sodium fluoride. Toxicol. 24:313-323.
Suttie JV, Gesteland R, Phillips PH. 1961. Efferts of dietary sodium
fluoride on dairy cows. VI. In young heifers. J. Dairy Sci. 4£:2250-2258.
Suttie JV, Miller RF, Phillips PH. 1957a. Studies of the effects of dietary
sodium fluoride on dairy cows. ' I. The physiological effects and the
development .symptoms of fluorosis. J. Nutr. 63:211-224.
Suttie JW, Miller RF, Phillips PH. 1957b. Studies of the effects of dietary
sodium fluoride on dairy cows. II. Effects on milk production. J. Dairy Sci.
40:1485-1491.
Suttie JV, Phillips PH, Miller RF. 1958. Studies of the effects of dietary
sodium fluoride on dairy cows. III. Skeletal and soft tissue fluorine
deposition and fluorine toxicosis. J. Nutr. 65:293-304.
Suttie JV, Phillips PH. 1959. The effect of age c.. the rate of fluorine
deposition in the femur of the rat. Archives Biochea. Biophys. 83:355-359.
Takaoori T. 1955. Recent studies on fluorosis. Tokushima J. Exptl. Med.
2:25-44.
taves, DR. 1978. Fluoridation and mortality due to heart disease. Nature
272:361-362.
Taves DR. 1979. Chapt. 13. Claims of harm from fluoridatlon. In:
Continuing evaluation of the use of fluorides. Vashington, DC; Am. Assoc.
Adv, Sci., pp. 295-321.
IX-11
-------
Taylor JK. Scott JX, Maynard EM. Smith FA, Kodge HC. 1961. Toxic effects of
fluoride on the rat kidney. I. Acute injury from single large doses.
Toxlcol. Appl. Phannacol. 3:278-289.
Teotia M, Teotia SPS, Kunwar DK. 1971. Endemic skeletal fluorosis. Arch.
Dis. Child. 46:686-691.
Teotia M, Teotia SPS, Singh RK. 1979. Skeletal fluoride toxicity in
children. Indian J. Fed. 46:389-396.
Underwood EJ. 1973. Toxicants naturally'occurring in foods. National
Academy of Sciences. Committee on Food Protection.
USEPA. 1976. U.S. Environmental Protection Agency. National interim primary
drinking water regulations. Washington, DC: U.S. Environmental Protection
Agency. EPA-570/9-76-003.
USEPA. 1979. U.S. Envlronoental Protection Agency, Office of Drinking Water-
Guidance for the issuance of variances and exemptions. Washington, DC: U.S,
Environmental Protection Agency.
Van Rensburg SWJ, De Vos WH. 1966. The influence of excess fluorine Intake
in the drinking water on reproductive efficiency in bovines. Onderstepoort J.
Vet. Res. 33:185-194.
Wagner MJ, Stookey GK, Muhler JC. 1958. Deposition of fluoride in soft
tissues following skeletal saturation. Proc. Soc. Exp. Biol. Med. 99:102-105.
Waldbott EL. l',62. Int. Arch. Allergy 20 (Suppl. I): 1-60.
Weast RC, ed. 1980. CRC handbook of chemistry and physics, 60th ed. Boca
Raton, FL: CRC Press, Inc., pp. B-l, B-9, B-51, B-79, B-83, B-121, B-126.
Wenzel A, ThyIstup A. 1982. Dental fluorosis and localized enamel opacities
in fluoride and nonfluoride Danish communities. Caries Res. 16:340-348.
Wenzel A, Thylstrup A, Melson B. 1982a. Skeletal development and dental
fluorosis in 12-14-year-old Danish girls from a fluoride and a nonfluoride
community. Scand. J. Dent. Res. 90:83-88.
Wenzel A, Thylstrup A, Melsen B, Fsjerskov 0. 1982b. The relationship
between water-borne fluoride, dental fluorosis and .skeletal development in
11-15 year old Tanzanian girls. Archs. Oral Biol. 27:1007-1011.
WHO. 1970. World Health Organization. Fluorides and human health. Geneva:
World Health Organization, p. 117.
WHO. 1982. World Health Organization. 1ARC Monographs on the evaluation of
•the carcinogenic risk of chemicals to humans. Chemicals, industry processes
and industries associated with cancer in humans. International Agency for
Research on Cancer Monographs Vol. 1 to 29, Supplement 4. Geneva: World
Health Organization..
IX-12
-------
WHO. 19?i. World Health Orjzaniratlon. Guidelines for ririrklng water
quality. Volure 7 - Recoasuncarions. Geneva: World Health Orgpnj z?.t lor .
?. 68.
Vhitford G. laves DR. 197!. Flurride-induced diuresis: plasma coricentratlp
in the rat. Prcc. Soc . Ex-tl. Biol. Hed. ;3T:
. .
Windholz M, ed. 1976. The Merck index. 9th cd. Rahway NJ : Kerck & Co.
Yiacouyiannis J, Burk 0. 1977. Fluoridatien and cancer age - d«?peno'4:ice cf
cer morca'.iry related to artif ici«i] f luoridation. Fluoride 10:102-125.
Zipkin I, Lee WA, Leone NC. 1957. Rate of urinary fluoride output in norrvsl
adults, ,-us. J. Public Healch- 47:8^8-851 .
2ipkin I, Likin? -RC. 1957. Absorption of various fluorine compounds frc^
the gastrointestir.al tract of the rat. Am. J. Physiol. 39!:549-550.
Zipkin I, Likins SC, McClurc FJ , Steere AG. 1956. Urinary fluoride levels
associated .with use of fluoridated waters. Public Health Rep. 71 :7f.7-7/r:.
Zipkin I. hcCIure FJ, Leone NC, Lee WA. 1953. Fluoride deposition in hu=ar
bones after prolonged ir.gescior. of fluoride in drinking water. Public Kealf
Kep. 73:732-7AO.
IX-13
------- |