John A Yiamouyiannis - Water Fluoridation and Tooth Decay
John A Yiamouyiannis - Water Fluoridation and Tooth Decay
John A Yiamouyiannis - Water Fluoridation and Tooth Decay
Safe \Vater Foundation, 6439 Taggart Road, Delaware, Ohio 43015, USA,
55
Volume 23, No. 2
56 Yiamouyiannis IVater Fluoridation and Tooth Decay 57
groups, average d f t and DMFT rates per child were determined for each of
t h e 84 areas. Age-adjusted DMFT rates for 5- to' 17-year-olds were calculated Figure 1
by addlng t h e DMFT rates for each of the 13 oge groups and divlding b! Tooth decay in fluoridated (F), partially fluoridated (PF), and non-
13 (IO). fluoridated (NF) areas: Permanent Teeth.
We obtained data regardlng the fluoridation status of the areas surveyed
from Natural Fluoride Content of Community Water Supplies, Fluoridatlon 6.00 T n
Census 1969, Fluoridation Census 1975, and Fluorldation Census 1985, all
published by the U.S. Public Health Service. In some cases, local authorities 5.00
were also contacted to determine the fluoridation status of an area.
DECAY ED,
Average DMFT (and d f t ) rates lor the F, NF, and PF groups were cal- MISSING, 4.00
culated for each age. Average-age-adjusted DMFT (and d f t ) rates for the AND FILLED
F, NF, and PF groups were calculated by taking the average of the age- PERMANENT 3.00
adjusted rates for t h e respective groups (IO). TEETH
The percentage of "caries-free" children was calculated for each age-group (DMFT) PER 2.00
for each area. Age-adjusted "caries-free" rates were also calculated. A student CHILD
was considered to b e "caries-free" so long as they had no DMFT or dft. For
example, a child who had lost all their teeth and no longer had any left 1.oo
to be decayed or filled would not be recorded as a "carles-free" student.
0.00
Through the United States Freedom of Information Act, we also obtained
resldence data lor east1 o l the above schoolchildren which allowed us to cal- 5 6 7 8 9 1011 12 131415 1 6 1 7
culate tooth decay rates for those in F, NF, and PF areas who had Ii\.ed
at the same resldence for their entire life. AGE
The two-tailed t-test was used to determine 95% confidence Intervals
and to determine statistlcal significance (at t h e 959/0 confidence level). A Table 2
two-sided Wilcoxon rank s u m test ( 1 1 ) was used to determine whether there Average-age-adjusted DMFT rates for 39,207 U.S. schoolchildren and 17,336
was a statistically slgnlflcant difference (at the 95% confidence level) in life-long resident schoolchildren in 84 areas throughout the United States,
the rank order of t h e DMFT rates of F and NF areas. Standard deviations are given in parentheses.
Results
Total Life-long
Table I presents the number of students examlned and the age-adjusted
DMFT r a t e for each of t h e 84 areas In t h e order of increasing tooth decay No. of No. of No. of
Areas Students DMFT Students DMFT
rate. There is no statlstlcally signlllcant dllference In the rank order of the
age-adlusted DMFT rates of F and NF areas. As can be seen by examination
of column I , there Is no clusterlng of fluoridated areas at the top of the Fluor Ida ted 27 12,747 1.96 1.97
(0.4 15) 6'272 (0.465)
table. In the quartile with the lowest age-adjusted DMFT rates, 9 are non-
fluoridated, 3 are partially fluoridated, and 9 are fluoridated. In the quartile Partially 2.18 2.25
wlth the hlghest DMFT rates, 5 a r e nonfluoridated, IO are partially fluoridated, Fluoridated 27 '2'578 (0.465) 5'642 (0.470)
end 6 a r e fluorldated. Table 1 also Indicates that there Is no biased geo-
graphical dlstrlbution of F and NF areas that Is hldlng some potential decay- Nonfluoridated 30 1.99 2.05
13'882 (0.408) 51422 (0.517)
preventive effect of water fluorldation.
There is no statistically significant difference between the average DhfFT adjusted DhlFT rates In F and NF areas are 1.96 and 1.99, respectively. The
rates for the F and NF groups a t any age (Flgure I). The average DhfFT 95% confldence Interval for the DMFT r a t e i n F areas minus the DMFT r a t e
rates of the PF groups a r e higher than those of the F and NF groups a t i n NF areas is (-0.19, 0.25); thus we can rule o u t , w l t h a certainty o f 95%,
every age w i t h the exception of 14-year-olds. the posslblllty that t h e DMFT r a t e In F areas is more than one-fourth o l
a tooth less t h a n In the NF areas. We can also rule out, w i t h a certainty
There is no statistically slgnlficant difference i n the average-age-adjusted of 95%, the posslbllty that t h e DMFT r a t e in NF areas is more than one-fifth
DMFT rates among the F, PF, and NF groups (Table 2). The average-age- of a tooth less than In the F areas.
there are no statistically slgniflcant dlfferences i n tooth decay rates between Figure 2C
permanent residents of F and N F areas at any age (Flgure 2A). I f water Tooth decay rates of white females.
fluoridation were t o have reduced tooth decay as measured by DhlFT, tooth
decay rates for life-long residents llvlng i n fluoridated areas should be l o u e r 6
than tesldents who had not spent thelr e n t i r e lives In these areas. This was
not found to be the case. Flgures 2B and 2C show that among white males
5
Figure 2A Decayed,
Tooth decay in residents of fluoridated (F), nonfluoridated (NF), and missing, 4
partially fluoridated (PF) areas who l i v e d their entire l i f e in the and filled
same household.
permanent 3
7.00 - teeth
(DMFT) per 2
child
DECAYED, 5.00 1
MISSING,
AND FILLED 4.00
PERMANENT 0
TEETH 3.00 5 6 7 8 9 1011 1 2 1 3 1 4 1 5 1 6 1 7
(DMFT) PER
CHILD 2.00 Age
arid u i t e females (which make up about 70% of all the children tudied),
there i s no signlficant dlfference in DMFT rates i n the F and NF areas a t
any age group.
5 6 7 8 9 1011 1 2 1 3 1 4 1 5 1 6 1 7 In contrast, notably lower tooth decay rates were observed in the decl-
AGE Figure 3
Tooth decay i n fluoridated (F), Partially fluoridated (PF), and non-
Figure 28 fluoridated (NF) areas: Deciduous Teeth.
Tooth decay rates of white males.
2.50 T
2.00
Decayed,
missing, 4 DECAYED
and filled AND FILLED 1.50
DECIDUOUS
permanent 3
TEETH (dtt) 1 .oo
teeth
PER CHILD
(DMFT)per 2
child 0.50
1
I
0 0.00
5 6 7 8 9 1011121314151617 5 6 7 8 9 1011121314151617
Age AGE
duous t e e t h of young children living In F areas. The 5-, 6-, and 7-year-olds Figure 4
in the F group have dft r a t e s 22Y0, 9 % and 6% lower than those of t h e NF "Caries-free" rates in nonfluoridated (NF), partially fluoridated (PF),
group, rspectively (Figure 3). Although the average-age-adjusted dft rates i and fluoridated (F) areas.
f o r F, NF, and PF groups were not significantly different statistically, the!
w e r e higher for t h e NF groups (0.96 k0.25) for t h e P F Groups (0.93 ? 0 . 2 4 ) ,
whlch in turn is slightly higher than t h e F group (0.89 kO.19).
To focus in on dft r a t e s among children 5-8, t h e eight areas which corn-
T
menced w a t e r fluoridation between 1970 and 1978 were removed from the
PF group and added t o t h e F group. The 5-, 6-, and 7-year-olds in the ne\\
F (F*) group have dft r a t e s 24%, IO%, and i0Y0 lower than those of the 50%
N F group, respectively, and t h e dft r a t e of 5-year-olds in the F* group is
significantly lower ( p 0.05) than that of the N F group. PERCENT
OF 4 0%
Moreover among 5-, 6-, and 7-year-old life-long residents i n the F* group, "CARIES-
d f t rates were 42%, 18% and I I % lower than those of the NF group, FREE" 30%
respectively, and t h e dft r a t e of 5-year-olds in the F* group was significantly
lower ( p e 0.002) than t h a t of t h e N F group (Table 3). I f water fluoridation CHILDREN
were t o have reduced tooth decay a s measured by dft among 5-year-olds, 20% 1-
iL+m-m+m
tooth decay rates for life-long 5-year-old residents livlng in fluoridated areas
should have been lower t h a n those o f residents who had not spent their entire
lives In these areas. This was found t o be t h e case. From Table 3, i t can 10%
also be seen that this large and significant reduction disappears a l t e r a couple
of years. 0%
Fluoride may have caused a reduction in d f t by delaying deciduous tooth 5 6 7 8 9 1011 121314151617
eruptlon. This Is consistent w i t h t h e fact that the dft rate in tile F and F*
groups reaches a maximum later than In the N F group. Fluoride-induced delays AGE
in tooth eruptlon have been reviewed elsewhere ( I 2,13) with contradictor>,
conclusions, but more recent studies examining 5-year-olds have indica1cd
delayed eruption that could account for such a difference in tooth decay Discussion
r a t e s (14).
7he data presented here a r e consistent with data reported elsewhere
The percentage of decay-free children in F, PF, and N F areas is 34.5%, in large U.S. surveys. In 1977, the Rand Corporation examined the tooth decay
31.9%, and 35.1Y0 respectively. There Is no statisticaly slgnificant difference r a t e of 25.000 chlldren In ( 5 F and 5 NF) nonrandomly selected areas (15).
between t h e average "caries-free" r a t e s for t h e F and NF groups at any age In t h e three areas In thelr study that were Included in t h e present study,
a e compared the tooth decay rates of 12-year-olds. There was good agreement
(Flgure 4).
between this study and theirs wlth regard to tooth decay rate, after converting
Table 3 DXlFS (decayed, missing and filled permanent tooth surfaces) t o DMFT (16)
and considering t h e acknowledged 36Oh decrease i n DMFS from 1979-1980
Percentage change in d f t r a t e s In all residents a n d life-long residents of to 1986-1987 (17).
F and F* areas in comparison t o NF areas.
In 1983-1984, Hlldebolt @. ( 4 ) examined t h e tooth decay r a t e s of over
Total L i fe-long
6500 \lissouri rural schoolchildren from grades 2 (average age 7.5) and 6
Age (NF-F)/NF (NF-F*)/NF (NF-F)/NF (NF-F*)/NF laberage age 11.5). Among 6th graders living in t h e most intensively studied
reqions, t h e average DhlFTtdft r a t e was 2.07 for those drinking nonfluorldated
u a t e r and 2.17 for those drlnking fluoridated water, compared t o t h e
5 22% 24% 36% 4 2%
DfITT+dft rate of 2.00 reported for 1 I-year-olds living in fjolcomb, Missouri
(p < 0.05) (p < 0.02) (p 8 0.002)
in our study.
6 9% 10% 14% 18%
7 6% 10% 5% 11%
In 1986, Kumar al, examined 1446 schoolchildren aged 7-14 from
Neahurgh, New York (fluoridated In 1945) and cohorts from nonfluoridated
8 -4% 1% -5% 1% Kingston, New York (18). The sample selection was nonrandom and had a
response r a t e of only 50-65%. Nonetheless, t h e age-adjusted DMFT rates ob-
served (1.5 for fluoridated Newburgh and 2.0 for nonlluoridated Kingston) :\nother recent study by Bruneile and Carlos (311, which also uses the
were in ilne with t h e corresponding values obtained in this study for commun- Same database that we used, reports a 17.7"/0 lower DMFS r a t e in t h e F areas.
ities in the area (1.5 for nonfluorldated New Paltz, New York and 1.7 for This study has a number of maJor deflciencles which render the study of
fluoridated New York City). little or no value.
Conclusions I. I t contains extremely serious errors. For example, by a cursory inspection,
we found two values that a r e off by 100% o r more. In their Table 9,
Does water fluoridation reduce tooth decay? i ] This study and other recent t h e DhlFS figure for life-long F exposure residents of Region VI1 should
studies (3-8) show that t h e r e is currently no significant difference in tooth be about 3, not 1.46 as reported. From their Table 3, t h e percent of
decay r a t e s in F and NF areas and that decreases in tooth decay rates over 5-year-olds who have caries is 1.0%, not the 2.7% that c a n be calculated
t h e last 25 years have been comparable regardless of fluoridation status; from t h e table (100%-97.3%). When I pointed out this error to Dr. Carlos,
I f this is true, t h e r e was no significant difference In the tooth decay r a t e s he admitted that only 19 out of the 1851 5-year-olds had caries: 19/1851
between these areas 25 years ago. i i ] From 1970 to t h e present, total fluoride = I%,but refused to make t h e correction (32).
Intake studies indicate an average intake o f 1-2 mg per day i n nonfiuoridated
areas and 3-5 mg per day in fluoridated areas (19,ZO); thus, i t is difficult 2. It fails t o report t h e tooth decay rates for each of t h e 84 geographical
t o claim that the reason tooth decay differentials between fluoridated and areas surveyed. This covers up t h e fact that there is n o difference In
nonfluorldated areas have disappeared is because t h e fluoride intakes in these t h e tooth decay rates of t h e fluoridated and nonfluoridated areas surveyed.
areas a r e now similar. Furthermore, t h e substantially higher incidence o l The Brunelle/Carlos study even fails t o list t h e areas studied. As a result,
dental fluorosis In fluoridated areas confirms that residents in these areas they produce misleading illustratlons; for example, their Figure 3 implies
a r e consuming substantially higher levels of fluoride than those living in non- that Arizona and New Mexico have the lowest tooth d e c a y rates, when, '
fluoridated areas (2 1-23), lii] Dramatic reductions in tooth decay have occurred in fact, not a single a r e a was surveyed in elther of t h e t w o states.
i n developing countries where there is no water fluoridation and there is
llttle reason to suspect that there would be elevated levels of fluoride in 3. It falls to control for geographlcal differences In tooth decay r a t e s by
the food chain (7,9,24,25). iv] In addition to recent studies, a number of earl). indiscriminantly and disproportionately bunching children from all parts
studies have also shown no signlflcant reduction in tooth decay a s a result o f the country into 2 groups, F and NF.
of water fluoridation (7,26-28). v] Serious questions have been raised regarding
the reliability of earlier studies claiming that fluoridation causes a reduction 4. It fails to do the statistical analysis (or even provide t h e data, i.e. the
in tooth decay (29). standard deviation and sample number) necessary t o determine whether
t h e values found for F and NF areas a r e significantly different. Our
Acknowledgements calculations show that even i f thelr d a t a were accurate, t h e 17.7% figure
does not reflect a statistically significant difference between t h e F and
1 thank Kimberly Close-Hittle, Jerry Putnam, Margot Yiamouyiannis, and NF groups.
Opal Kuhn for thelr help In t h e calculation and verification of summary data
as well a s Jill Pitts and Chris Hiatt for their lightning fast speed in entering 5. It fails to report t h e data for t h e approxlmately 23,000 schoolchildren
d a t a into our computer. Without the generousity of Dr. Leo Roy, Dr. Reuben who were not IIfe-time resldents of either t h e F or N F areas ( t h e PF
Benner, Dr. H. Charles Kaplan, Dr. Gerard Judd, Richard Barmakian, John group). I f fluoridation reduced tooth decay, the DMFS r a t e of t h e P F
C. justice, Len Greenall, Mr. and Mrs. Andrew Yimoyines, Wini Silko, A l h l group should have been greater than that of the F group and less than
International, Inc., and o t h e r patrons of t h e Center for Health Action and that o f the NF group. Our data indlcate that the P F group would have
t h e S a f e Water Foundation, t h e preparatlon and publication of this article had a DhlFS r a t e higher (although not slgniflcantly higher) than either
would not have been possible. Finally, I thank Ray Fahey for correcting an t h e F o r NF groups.
error we had made in assignlng t h e fluoridation status of E. Orange, NJ.
6. It fails to report t h e d a t a for t h e percentages of decay-free children
Addendum in f and NF areas. Our d a t a indicate that had these calculations been
done by Bruneile and Carlos, t h e results may have actually indicated
Recently Bruneiie (301, using the s a m e database that w e used, reported better (although not significantly better) dental health in t h e NF areas.
26% fewer dfs (decayed and filled deciduous tooth surfaces) in children \rho
had always resided in F communities than those who never lived in F commun- Drunelle and Carlos, as w e l l as their employer, t h e NIDR, have recently
ities. This finding agrees reasonably well with the d a t a outlined in our Table come under attack for presenting erroneous data and designing poor experi-
3, which shows a statistlcally significantly lower d f t r a t e in life-long 5-year-old ments which promoted t h e fluoride mouthrinse program (33). The apparent
resldents of fluorldated areas. However, by omission of age-specific data, poor quality of their research regardlng t h e 1986-1987 survey (30,31) is not
t h e Brunelle study covers up t h e f a c t that this difference in tooth decay an isolated case.
is no longer significant In 6-year-olds and disappears entirely among 8-year-
olds.
References and Notes Demonstration Program. R-2862-RWJ. Rand Corporation, Santa Monlca,
CA, 1982.
1. Green, J.C., Louie, R. and Wycoff, S.J.: Preventive Dentistry I. Dental 16. Jarvinen, S.: Epldemlologlc Characterlstlcs of Dental Caries: Relation
Caries. J. Amer. Med. Assn., 262:3456-3463, 1989. of DhlFS to DMFT. Cornmunlty Dent. Oral Epldemlol., I1:363-366, 1983.
2. Szpunar, S.M. and Burt, B.A.: Dental Carles, Fluorosis and Fluoride E x p ~ 17. Johnson, S. HHS News (U.S. Department of Health and Human Servlces:
sure i n Michlgan Schoolchildren. J. Dent. Res., 67:802, 1988. National Institutes of Health) J u n e 21, 1988.
3. Colquhoun, J.: Influence of Social Class and Fluoridation on Child Dental 18. Kumar, J.V., Green, E.L., Wallace, W. and Carnahan, T.: Trends in Dental
Health. Community Dent. Oral Epidemlol.. 13:37-41, 1985. Fluorosis and Dental Caries Prevalences In Newburgh and Kingston, NY.
4. Colquhoun, J.: Child Dental Health Differences in New Zealand. Community Amer. J. Pub. Health, 79:565-569, 1989.
Health Studies, 11:85-90, 1987. 19. Rose, D. and Marler, J.R.: Environmental Fluoride, 1977. NRCC No. 16081.
5. Gray, AS.: Fluoridation: Time for a New Baseline? J. Canadian Dent. National Research Council of Canada, Ottawa, Ontario, 1977, pp. 75-83.
ASSOC., 53:763-765, 1987. 20. Featherstone, J.D.B. and Shields, C.P.: A Study of Fluoride Intake In
6. Hlldebolt, C.F., Elvln-Lewis, M., Molnar, S., McKee, J.K., Perkins, h1.D. New York S t a t e Resldents. 01 14Uc1288-I, Eastrnan Dental Center,
and Young, K.L.: Caries Prevalences Among Geochemical Regions of Rochester, N Y , 1988.
Missourl. Amer. J. Physical Anthropol., 78:79-92, 1989. 21. Segreto, A S . , Collins, E.M., Carnann, D. and Smith, C.T.: A Current
7. Dlesendorf, M.: The Mystery of Deciinlng Tooth Decay. Nature, 322: 125- S t u d y of Mottled Enamel In Texas, J. Amer. Dent. Assoc., 108:56-59,
129, 1986. 1984.
8. Johnston, D.W., Gralnger, R.M. and Ryan, R.K.: The Decline of Dental 22. Leveret, D.: Prevalence o f Dental Fluorosls in Fluoridated and Nonfluorl-
Carles In Ontario School Children. J. Canadlan Dent. Assoc., 5 2 4 I 1-4 17, dated Comrnunltles - A Prellmlnary Investlgatlon. J. Pub. Health Dent.,
1986. 46:184- 187. 1986.
9. Luorna, A-R. and Ronnberg, K.: Twelve-Year Follow-up or Caries Preval- 23. Colquhoun, J.: Disfiguring Dental Fluorosis In Auckland, New Zealand.
e n c e and Incidence In Children and Young Adults i n Espoo, Finland. Fluoride, I7:234-242, 1984.
Community Dent. Oral Epidemlol., 15:29-32, 1987. 24. Poulsen, S., Amaratunge, A. and Risager, J.: Changes In the Epidemlc-
IO. Hill, A.B.: Medical Statistics. Hodder and Stoughton, London, 1977, p. logical Pattern of Dental Caries In a Danish Rural Cornmunlty over a
183. While the numerous age-specific comparisons of the dental health IO-Year Period. Community Dent. Oral Epldemlol., 10:345-351, 1982.
of chlldren at different ages provides t h e best evidence, i t is occasionally 25. Backman, B., Crossner, C-G. and Holm, A-K.: Reductlon of Caries In
desirable t o have a summary r a t e t o enable an overall comparison of 8-Year-Old Swedish Chlldren between 1967 and 1979. Cornmunlty Dent.
dlfferent populations. For thls purpose, w e have used the age-standardized Oral Epldemlol., 10:178-181, 1983.
or age-adjusted rates, In order t o avoid giving dlsproportionate weighting 26. Scrivener, C.: Unfavorable Report from Kansas Community Uslng Artlflcal
t o larger numbers of chlldren from one particular age-group that would Fluorldation of Clty Water Supply for Three-Year Period. J. Dent. Res.,
tend to distort t h e summary figure. In uslng these rates, a standard popu- 30:465. 1951.
latlon must be chosen. The one most commonly used I s the hypothetical 27. Galagan, D.J.: Climate and Controlled Fluorldatlon. J. Amer. Dent. Assoc.,
population w l t h equal populations at each age group, which merely results 4 7 : 159- 170, 1953.
from taking an arithmetic mean of t h e age-specific tooth decay rates 28. Schroeder, P.: Dental Health i n Children In Rural Regions without School
measured. I n t h e above reference, Austln Bradford Hill addresses this Clinics. J. Dent. Res., SO(Supp1ernent Part l):l231, 1971.
method In a dlscusslon of the handllng of mortality rates under a section 29. Yiamouyiannis. J.: Fluoride, t h e Aging Factor. Health Actlon Press,
tltled "The Equlvalent Average Death-Rate." Analogously, equal weights Delaware, Ohlo, 1986, pp. 94-110.
were glven to each of the 84 geographical areas t o prevent a distortion
w h l c h rnlght be Induced by the varlatlon of the area sample sizes, since References for Addendum
certaln geographical areas have characterlstlcally hlgher (or lower) tooth
decay rates than others. 30. Brunelle, J.A.: Caries Attack In t h e Primary Dentition of U.S. Children.
I I. Wllcoxon, F., Kattl, S.K. and Wllcox, R.A.: Critical Values and Probablllty J . Dent. R e s , 69(Speclal Issue):180 [Abstr. No. 5753, 1990.
Levels for the Wllcoxon Rank Sum Test and the Wilcoxon Signed Rank 31. Brunelle, J.A. and Carlos, J.P.: Recent Trends In Dental Caries In U.S.
Test. Selected Tables In Mathematfcal Statistics, Markham Publishing Children and the Effect of Water Fluorldatlon. J. Dent. Res., 69(Speclal
Co., Chlcago,, 1:197, 201, 1970. lssue):723-728, 1990.
12. Waldbott, G.L., Burgstahler, A.W. and McKinney, H.L.: Fluoridation, the 32. Carlos, J.P.: Personal communicatlon, 1989,
G r e a t Dilemma. Coronado Press, Lawrence, Kansas, 1978, 423 pp. 3 3 . Disney, J.A., Dohannan, H.M., Kleln, S.P. and Bell, R.M.: A Case Study
13. El-Badrawy, H.E.: Dental Development In Optlmal and Suboptlmal Fluorlde i n Contesting the Conventional Wisdom: School-Based Fluoride Mouthrinse
Communitles. J. Canadlan Dent. Assoc., 50:761-764, 1984. Programs in t h e USA. Community Dent. Oral Epidemlol., 18:46-56, 1990.
14. Krylov, S.S. and Pemrolyd, K.: Declduous Tooth Eruption and Fluorosis
In the Case of Increased Fluorine Content In t h e Drlnklng Water. **********
Stomatologlie (Mosk), 6 1:75-77, 1982.
15. Bell, R.M., Klein, S.P., Bohannan, H.M., Graves, R.C. and Dlsney, J.A.:
Results of Basellne Dental Exams In t h e National Preventlve Dentlstry
.
Volume 23, No. 2
I. ...LA
Fluoride