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

-------