Community Dent Oral Epidemiol 1998; 26: 31623 Copyright C Munksgaard 1998

Printed in Denmark . All rights reserved

ISSN 0301-5661

Urinary fluoride levels and Siegward D. Heintze,

Jose Roberto de Magalhaes Bastos and
Roosevelt Bastos

prevalence of dental fluorosis in Bauru Dental School, University of

Sao Paulo, Brazil

three Brazilian cities with different

fluoride concentrations in the
drinking water
Heintze SD, Bastos JRM, Bastos R: Urinary fluoride levels and prevalence of
dental fluorosis in three Brazilian cities with different fluoride concentrations in
the drinking water. Community Dent Oral Epidemiol 1998; 26: 31623. C Munks-
gaard, 1998

Abstract Urine samples from three daytime periods were collected from 545
550-year-old residents of three different Brazilian cities: Garca had fluoridated
drinking water since 1973, Bauru since 1975 and Itapolis was not fluoridated. Den-
tal fluorosis was examined in 985 524-year-olds using the Thylstrup-Fejerskov
index (TF). The subjects were asked to estimate their daily intake of liquids and
frequency of beverage consumption. The analysis of 94 water samples showed
high variations in the fluoride content of the drinking water. The mean fluoride
concentration of the water samples in Garca was 0.9 mg/L (range 0.751.2), in
Bauru 0.64 mg/L (range 0.011.3), and in Itapolis 0.02 mg/L. Mean urinary fluo-
ride concentration was 1.31 mg/L (s 0.61) in Garca, 0.88 mg/L (s 0.49) in Bauru,
and 0.39 mg/L (s 0.21) in Itapolis. Self-reported daily liquid intake was not related
to urinary fluoride concentration. The mean prevalence of fluorosis was 13.3% in
Garca, 6.8% in Bauru, and 1.7% in Itapolis, with mainly categories TF 1 and TF 2
being recorded. Subjects with dental fluorosis tended to show a higher mean
urinary fluoride concentration but the difference was not statistically significant. Key words: Brazil; fluorosis; urinary fluoride
The study showed that fluoride exposure measured by urinary fluoride excretion concentration; water fluoridation
was within the range expected for the level of fluoride concentration in the drink-
Jose Roberto de Magalhaes Bastos,
ing water. However, enamel fluorosis tended to be markedly lower than ex- Department of Social Dentistry, Bauru
pected. This study revealed that fluoride levels in the two cities with fluoridated Dental School, University of Sao Paulo,
drinking water were variable. To optimise anticaries benefits and minimise the Al. Dr. Octavio P. Brisolla, 9-75,
risk of fluorosis greater control of the fluoride dosing of the drinking water is 17.043-101 Bauru SP Brazil
required. Accepted for publication 2 March 1998

When assessing the safety of various levels of
fluoride ingestion it is important to consider the
fluoride intake from all potential sources. These
sources might include drinking water, fluoride oral
care products and the environment as well as food
and beverages. Estimates of fluoride intake using
dietary histories may not accurately reflect fluoride
ingestion from sources such as toothpaste, mouth-
rinses and fluoride air pollution. As urine is the
main excretion route for ingested fluoride, analysis
of the fluoride concentration in urine is a useful
way to estimate the overall fluoride intake of a
population (1). In communities with a fluoride con-
tent of 0.51.5 mg/L in drinking water, early stud-
ies from the 1950s and 1960s showed that urinary
fluoride concentration was almost identical with

316

the fluoride concentration in drinking water (2, 3).
Nowadays, with the widespread use of toothpastes
containing fluoride and other topically applied
fluorides, some of which are partly retained in the
mouth or swallowed, especially by young children
(4), urinary fluoride concentration is usually higher
in both fluoridated and nonfluoridated communi-
ties (5, 6). People living in nonfluoridated commu-
nities consume food and beverages prepared with
fluoridated water and may, therefore, have higher
urinary fluoride concentrations than expected (7).
Fluoridation of the drinking water supply is
increasingly being questioned as caries levels have
fallen in many countries and the prevalence of den-
tal fluorosis may be increasing (8).
In Brazil, the situation is somewhat different
from that in European countries or the United
States. Toothpastes containing fluoride have only
been available since the end of the 1980s and are
only rarely used by people with a low income (9).
Fluoride concentration in commercial beverages is
generally low. In a recent study, however, one-
fourth of the analysed soft drinks, which are
increasingly consumed by children and adoles-
cents, showed a fluoride concentration between 0.5
mg/L and 0.8 mg/L (10). Furthermore, the speci-
fied fluoride concentration in the drinking water
does not necessarily reflect fluoride exposure, since
the Brazilian waterworks are reported not to main-
tain a constant fluoride level in the drinking water
(11) because of lack of control systems, personnel
and money (12). This may result in excess fluoride,
as well as suboptimum fluoride levels. As 40% of
the whole Brazilian population and 90% of the
population living in the state of Sao Paulo are said
to be provided with fluoridated water (12), it
seemed necessary to determine the fluoride expo-
sure of different age groups exposed to different
fluoride levels in the drinking water. Within the
context of a large risk/benefit study of water fluo-
ridation, urine samples from 550-year-old subjects
of three cities in the state of Sao Paulo with dif-
ferent fluoride levels in the drinking water were
collected and analysed for fluoride concentration.
Furthermore, the prevalence of fluorosis was as-
sessed since high prevalence is a sign of an overex-
posure to fluoride.
Material and methods
Three cities in the state of Sao Paulo with similar
socio-economic and socio-demographic conditions
were chosen for the study:
Urinary fluoride and dental fluorosis
Bauru: water fluoridation since 1975; 260 382 in-
habitants in 1992; 683 m above sea-level; mean
annual temperature 24.2C
Garca: water fluoridation since 1973; 41 351 in-
habitants in 1992; 526 m above sea-level; mean
annual temperature 22C
Itapolis: low fluoride content in the drinking
water; 30 111 inhabitants in 1992; 491 m above
sea-level; mean annual temperature 23C.
Between June and August 1995 and in May 1996, a
total of 1328 subjects aged 5 to 50 years were exam-
ined in health centres, schools and factories. The
population was selected to include subjects from
all social strata. For subjects aged 524 years, den-
tal fluorosis was recorded using the Thylstrup-Fej-
erskov index (TF) by a trained and calibrated
examiner (13). All permanent teeth were examined
immediately after drying with cotton rolls. The
subjects were also asked to estimate their intake of
liquid derived from tap water (including coffee
and tea) on the day before (number of glasses of
200 ml or cups of 100 ml). Other questions ad-
dressed the type of water source, the use of water
purifiers, as well as measures of oral hygiene in-
cluding the use of toothpaste or mouthrinses con-
taining fluoride.
After the oral examination and interview, the
subjects were given three 60-ml plastic bottles and
asked to collect urine at three different times dur-
ing the next day:
Between 8 am and 10 am
Between 2 pm and 4 pm
Between 7 pm and 9 pm.
The urine samples were returned the next day or
collected by a member of the examination team
and frozen to a temperature of 20C immediately
after receipt. The frozen plastic bottles were subse-
quently brought to the Dental School in Bauru in
polystyrene cooling boxes and stored in freezers
until further analysis. A defined number of bottles
were taken out of the freezer and allowed to de-
frost the evening before the analysis.
Tap water samples were collected from random-
ly selected homes and stored for fluoride analysis.
In Bauru, water samples were taken before and af-
ter filtration to determine the influence of the filter
on the fluoride concentration. In Brazil, water puri-
fiers are commonly used to improve the water
quality (to eliminate chloride, bacteria, etc.).
For fluoride analysis, the fluoride-specific
electrode Orion 9409 and the reference electrode
Orion 9001 (Orion Research) were used. The exact
method is specified in other reports (14).
317
Heintze et al.

Only results from individuals who reported

using tap water and took urine samples from all
three daytime periods were included in the final
analysis.
Statistical analysis was performed by the SPSS
program. To evaluate differences in urinary fluo-
ride concentration among pooled age groups, the
ANOVA procedure was used. To estimate whether
the self-reported daily liquid intake could explain
fluoride concentration in the urine, regression anal-
ysis was used. The significance level for all opera-
tions was set to 0.01.
Results
The fluoride concentration of the drinking water
from the two cities with water fluoridation was
variable, especially in the city of Bauru. Additional
samples were taken in Bauru because of unex-
pectedly low fluoride concentrations in some of the
samples (Table 1). The mean fluoride concentration
in Bauru was 0.64 (s 0.30) with a range from 0.01
to 1.3 mg/L. In Garca the concentration was 0.90
mg/L (s 0.18) with a range from 0.75 to 1.2 mg/L.
In Bauru, water samples taken in 7 of the 25 dis-

Table 1. Mean fluoride concentration (mean F) in drinking water samples of the three cities

Mean F Confidence
City (mg/L) s Range interval n

Bauru (not filtered) 0.64 0.30 0.011.30 0.560.71 67

Bauru (filtered) 0.61 0.27 0.011.30 0.520.70 40
Garca 0.90 0.18 0.751.20 0.791.01 10
Itapolis 0.02 0.02 0.010.08 0.010.04 17

Table 2. Mean urinary fluoride concentration (mean F), according to age and for all ages (total)

Mean F Median F Confidence

Age group/city (mg/L) (mg/L) s Range interval n

59 years
Bauru 1.04 1.03 0.33 0.551.51 0.851.23 12
Garca 0.96 0.86 0.37 0.461.83 0.821.10 29
Itapolis 0.40 0.32 0.19 0.200.94 0.320.48 24
1014 years
Bauru 1.06 1.02 0.45 0.382.13 0.911.21 34
Garca 1.36 1.20 0.78 0.223.90 1.111.60 39
Itapolis 0.40 0.39 0.20 0.060.92 0.340.45 49
1519 years
Bauru 0.93 0.86 0.58 0.212.57 0.621.24 14
Garca 1.11 1.03 0.47 0.402.73 0.961.26 37
Itapolis 0.30 0.27 0.17 0.020.79 0.250.35 42
2024 years
Bauru 0.79 0.66 0.46 0.271.73 0.740.97 25
Garca 1.45 1.31 0.70 0.374.07 1.241.66 43
Itapolis 0.39 0.39 0.19 0.100.86 0.290.48 16
2534 years
Bauru 0.90 0.74 0.57 0.293.37 0.731.07 46
Garca 1.41 1.43 0.37 0.752.30 1.241.53 36
Itapolis 0.51 0.47 0.26 0.100.98 0.310.71 19
34 years
Bauru 0.68 0.58 0.38 0.131.77 0.550.81 36
Garca 1.61 1.50 0.65 0.713.50 1.331.89 22
Itapolis 0.45 0.39 0.22 0.180.86 0.100.80 22
Total
Bauru 0.88 0.75 0.49 0.133.37 0.800.96 167
Garca 1.31 1.19 0.61 0.224.07 1.231.40 206
Itapolis 0.39 0.36 0.21 0.020.98 0.360.42 172

318

Table 3. Mean urinary fluoride concentration (mean F) for pooled ages
519 years
City Mean F (mg/L)
Bauru 1.02
Garca 1.16
Itapolis 0.36
tricts examined had fluoride concentrations below
0.6 mg/L; one district showed more than 1 mg/L
of fluoride. In Itapolis, where the drinking water is
not fluoridated, the mean fluoride concentration
was 0.02 mg/L (s 0.02) with a range from 0.01 to
0.08 mg/L. Filtration did not significantly influence
the fluoride concentration in the drinking water
(Table 1), although there was a tendency towards
slightly lower fluoride concentration in filtered
water (0.61 vs 0.64). In the interview, 60.6% of the
subjects declared that they used some sort of water
filter, mostly ceramic (78.4%) or activated carbon
filters (9.9%). Of those using filters, 73% used it to
purify the water before drinking, 20% for cooking
meals and 7% gave no answer.
In all, 1635 urine samples from 545 individuals
were available for fluoride analysis (Table 2). This
represents 41% of the 1328 subjects included in the
study. The mean urinary fluoride concentration
(mean fluoride concentration of the three samples)
of the subjects from Garca was 1.31 mg/L, 0.92
mg/L higher than that of the subjects from Itapolis
Fig. 1. Scatterplot mean urinary fluo-
ride concentrations (mg/L) by self-re-
ported liquid intake (ml) of the sub-
jects of all ages in Garca and Bauru.
Regression lines for Bauru (R20.012;
P0.21) and Garca (R20.003; P0.45).
Urinary fluoride and dental fluorosis
20 years
Mean F (mg/L) F-statistic P
0.80 8.29 0.01
1.47 13.75 0.001
0.46 8.11 0.01
and 0.43 mg/L higher than that of the subjects
from Bauru. For all age groups combined, the
average difference between Bauru and Itapolis was
found to be 0.49 mg/L (Table 2). These differences
were statistically significant (P0.001). In the cities
with water fluoridation, a large individual range
was observed (between 0.13 and 4.07 mg/L).
The association between urinary fluoride concen-
tration and age differed in the three cities. In Garca
the urinary fluoride concentration appeared to
increase with age and in Bauru it decreased. When
the pooled age group 5 to 19 years (bone growth)
was compared with the age pool 20 to 50 years (no
more bone growth, only remodelling), statistically
significant differences were found in all three cities.
However, the differences were most pronounced in
Garca (Table 3).
For the combined age groups, a correlation be-
tween self-reported liquid intake (in millilitres) and
urinary fluoride concentration was observed in the
two cities with water fluoridation but it was not
statistically significant (Fig. 1). However, in the age
319
Heintze et al.
Fig. 2. Prevalence of fluorosis (percentage of subjects) in Gar-
ca (G), Bauru (B) and Itapolis (I), according to age group and
maximum TF score for each individual.
group 1014 years in the city of Bauru, regression
analysis revealed a significant correlation between
self-reported liquid intake (in millilitres) and fluo-
ride concentration (R20.263, P0.009).
Ninety-eight percent of the individuals inter-
viewed said that they regularly used toothpaste
containing fluoride and 16.5% used a fluoride
mouthrinse daily or weekly. No association, how-
ever, was observed between the use of a fluoride
mouthrinse and urinary fluoride concentration.
For the 524-year-old subjects dental fluorosis
was recorded using the TF index. A total of 985
subjects were examined and of these 207 were from
Bauru, 430 from Garca and 348 from Itapolis. In
Garca, 18.9% of 59-year-old children, 13.9% of the
1014-year-olds, and only 6.7% of the 2024-year-
olds suffered from dental fluorosis (Fig. 2). For the
combined age groups, the mean prevalence of flu-
orosis was 13.3% in Garca, 6.8% in Bauru, and 1.7%
in Itapolis. The disease predominantly affected the
central and lateral incisors, as well as the first mo-
lars. The majority of the fluorosis was recorded for
the categories TF 1 and TF 2. However, in the 10
14 and 1519 age groups, nine cases of TF 3 and
one case of TF 4 were recorded in Garca and Bauru.
In Bauru, the highest prevalence was found in the
age group 1519 years (12.8%).
Subjects with signs of fluorosis (TF0) tended to
have a higher mean urinary fluoride concentration.
However, this was statistically significant only for
the 1014-year-olds (Fig. 3).
320
Fig. 3. Mean urinary fluoride concentration according to den-
tal fluorosis (TF) and age group (P0.01) in Bauru and
Garca.
Discussion
Fluctuations in the fluoride concentration of
monthly water samples from both artificially and
naturally fluoridated drinking water have been re-
ported from various regions in the world (15). A
retrospective study from the city of Porto Alegre
revealed that fluoride was not added to the drink-
ing water for almost a quarter of the observed
period (13 years). Even when fluoride was added,
fluctuations (between 0.6 mg/L and 2.0 mg/L)
were observed (11). In the present study, the fluo-
ride concentration in Bauru was between 0.01
mg/L and 1.3 mg/L. The problem in Bauru is that
the public drinking water comes from 26 different
natural wells, and fluoride has to be added to each
of them. A recent report on water fluoridation pro-
grammes in the state of Sao Paulo points out the
problems some Brazilian cities are facing with re-
gard to maintaining a constant fluoride level (12):
lack of trained personnel, control, automatic dos-
age devices, and money to purchase sufficient
quantities of fluoride. In the city of Garca, fluoride
concentration was found to be more stable. Unlike
Bauru, Garca obtains its drinking water from sur-
face water. Therefore, only one plant processes the
water and adds fluoride. This task is carried out
manually four times a day by a chemical engineer
who monitors the fluoride concentration by the vi-
sual colorimetric method. The variability of the
fluoride level in the drinking water and the inter-

ruption of the supply reduces the anticaries effec-
tiveness (16). However, fluoride levels that are too
high may increase the risk of dental fluorosis (17).
To improve the situation, Brazilian waterworks
should be equipped with automatic dosage devices
and fluoride specific electrodes. The mean fluoride
concentration in the drinking water in the city of
Bauru was 0.16 mg/L lower and that of Garca 0.1
mg/L higher than the value recommended by
WHO in 1986 as the optimum fluoride concentra-
tion for regions with an annual temperature be-
tween 21.5C and 26.2C (18). These deviations
from the optimal concentration are minor if the
considerable variations in the absolute concentra-
tions are considered. Another factor that may com-
plicate the estimation of fluoride intake from drink-
ing water is that most Brazilians use some sort of
water purifier. However, the difference between
filtered and non-filtered water in the water samples
of Bauru was small. This suggests that the filters
most frequently used (ceramic filters) do not re-
move fluoride. It is also known that more sophisti-
cated filters, such as those of the reverse-osmosis
type, do remove fluoride (19).
Urinary fluoride concentration showed a dif-
ferent pattern in the two cities with water fluorida-
tion. This may be explained by the fluctuation of
fluoride concentration found in the drinking water.
In the city of Garca, urinary fluoride concentration
increased with age, while a decrease was observed
in Bauru. Urinary fluoride levels of adults may be
substantially influenced by high or low fluoride in-
take 2 or more years earlier (20). This may not be
the case for children, in whom urinary fluoride ad-
justs more rapidly to the fluoride intake (19). How-
ever, it is unclear why adults (20 years) in Bauru
should have a lower level of urinary fluoride than
children (20 years) whereas in Garca the reverse
was found. It is possible that fluctuations in the
fluoride concentration of the drinking water or dif-
ferential use of other sources of fluoride such as
fluoride toothpaste or beverages might account for
the anomaly. In the city of Garca, the mean preva-
lence of fluorosis for 524-year-olds was 13.3%. The
highest prevalence of 18.9% was found in the age
group 59 years, suggesting that the level of fluo-
rosis is increasing in this city. However, alterations
of enamel in adolescents and adults owing to wear
and discoloration may mask dental fluorosis and
render its diagnosis difficult (21). In Bauru, the
highest percentage was found in the age group 15
19 (12.9%). In Itapolis, this phenomenon was prac-
tically non-existent. Again, suboptimum fluoride
Urinary fluoride and dental fluorosis
levels in the drinking water and/or interruption of
the fluoride supply to the drinking water may ex-
plain the lower prevalence of fluorosis in Bauru.
Estimating fluoride excretion by means of 24-h
samples is more reliable than spot or single sam-
ples (22). This sampling procedure has been exten-
sively used to determine optimum levels of fluo-
ride in salt (2325). As fluoride is rapidly excreted
after intake with the peak within 2 h (26), spot or
single samples may not reflect the average concen-
tration and even less the total fluoride intake.
Furthermore, the consumption of large amounts of
liquids (as can be expected in a tropical country
like Brazil) dilutes urine and lowers fluoride con-
centration. The fluoride concentration of a single
sample may vary considerably from that of a 24-h
sample. However, the average concentration from
three urine samples taken in this study may rea-
sonably well reflect the average fluoride concentra-
tion in a 24-h sample as found by Zipkin et al. (2).
Early studies indicate that urinary fluoride levels
increase with age between 1 and 12 years of age
(3). A study comparing urine samples from English
and Sri Lankan subjects reported a mean fluoride
level of 1.19 mg/L already in 4-year-old children
(6). An Australian study even reported a mean
fluoride concentration of 1.72 in 34-year-olds in a
community with a water fluoride level of 1 mg/L
(5). In the present study, large individual variations
were observed, which may be the result of both
different eating and drinking habits and the fluc-
tuation of the fluoride levels in the drinking water,
especially in Bauru. Other factors that may account
for inter-individual differences are urine pH, glo-
merular filtration rate, urine flow rate and kidney
status (19). In the present study, urinary fluoride
concentration was poorly correlated with self-re-
ported liquid intake from drinking water and may
not serve as an accurate guide for fluoride intake.
The significant correlation found in one age group
in one city may be interpreted as a random event.
Drinking water still plays an important role for
Brazilian consumers. Data collected in the context
of this study showed that about 80% of the subjects
interviewed drink tap water at least once a day and
30% even three times a day between meals. Fifty
percent of the subjects drink water with main
meals. These results were similar for all age
groups. The trend observed in European countries
and the United States to replace drinking water
with industrialized beverages like soft drinks and
bottled or mineral waters cannot be confirmed for
Brazil. In our survey, bottled waters played only a
321
Heintze et al.
minor role. Soft drinks, although increasingly pop-
ular, are still used less than tap water. Neverthe-
less, 50% of 618-year-olds prefer soft drinks with
main meals and about 40% drink them between
meals. Tea-drinking is not a widespread custom
among Brazilians and does not contribute to the
daily intake of fluoride. Occasionally, mate-tea is
consumed, but this type of tea contains less than
0.1 mg/L fluoride (10).
In some studies, urinary fluoride concentration
has been used as an indicator for fluorosis risk. In
a Belgian study with 3022 521-year-olds, urinary
fluoride concentration was closely related to the
presence of fluoride opacities (27). A similar
trend was found in the present study but differ-
ences were only statistically significant in one of
four age groups. A precondition for using urinary
fluoride samples as an indicator of fluorosis risk
would be the assumption that the fluoride inges-
tion of a population group is more or less constant
during life, which of course is speculative.
Data on the prevalence of fluorosis in Brazil are
scarce. A study in 1995 examining 2397 512-year-
old children in the hot coastal city of Santos, Sao
Paulo (number of inhabitants: 430 000; mean annu-
al temperature: 26.5C), which has had its water
fluoridated since 1983 (0.60.8 mg/L fluoride),
showed a fluorosis prevalence of 30.2%, of which
19.3% were in the category very mild and 8.6%
in the category mild according to the Dean index
(28). As there are no publications of epidemiologi-
cal data of fluorosis collected 10 or 20 years ago in
these cities or in Brazil in general, no estimations
of changes in the level of fluorosis can be made.
Comparing the results of the present study with
the latest publications on fluorosis risk in the
United States, it can be said that fluorosis is cur-
rently not a major concern in Brazil. However,
there may be a tendency of increasing prevalence.
In the United States, the prevalence of fluorosis has
risen from 6% to 18% in regions without water
fluoridation and from 22% to 43% in regions with
optimum fluoride concentration in the drinking
water during the last decade (8). Not only has the
percentage of affected children risen in the United
States, but also its severity. The lower fluorosis
prevalence among Brazilian children may have
four reasons: (1) toothpastes containing fluoride
were not available before the end of the 1980s; (2)
not all Brazilians regularly use toothpastes contain-
ing fluoride, particularly those with a low income;
(3) fluoride supplements are rarely prescribed for
children; and (4) maintaining an optimum fluoride
322
level in the drinking water cannot be guaranteed,
which often results in suboptimum fluoride levels.
Further research on the prevalence of dental fluo-
rosis in Brazilian children and adolescents is neces-
sary to monitor the development of this side-effect
of fluoride over an extended period of time.
References
1. World Health Organization. Fluorides and oral health.
Technical Report Series No. 846. Geneva: WHO; 1994.
2. Zipkin I, Likins RC, McClure FJ, Steere AC. Urinary
fluoride levels associated with uses of fluoridated water.
Public Health Rep 1956;71:76772.
3. Gedalia I. Urinary fluoride levels of children and adults.
J Dent Res 1958;37:6014.
4. Ripa LW. A critique of topical fluoride methods (denti-
frices, mouthrinses, operator-, and self-applied gels) in
an era of decreased caries and increased fluorosis preva-
lence. J Public Health Dent 1991;51:2341.
5. Dooland MB, Carr SM. Urinary fluoride levels in South
Australian pre-school children in summer and winter.
Aust Dent J 1985;30:4103.
6. Rugg-Gunn AJ, Nunn JH, Ekanayake L, Saparamadu
KDG, Wright WG. Urinary fluoride excretion in 4-year-
old children in Sri Lanka and England. Caries Res
1993;27:47883.
7. Ripa LW. A half-century of community water fluorida-
tion in the United States: review and commentary. J Pub-
lic Health Dent 1993;53:1744.
8. Clark DC. Trends in prevalence of dental fluorosis in
North America. Community Dent Oral Epidemiol
1994;22:14852.
9. Witt MC. Pattern of caries experience in a 12-year-old
Brazilian population related to socioeconomic back-
ground. Acta Odontol Scand 1992;50:2530.
10. Heintze SD, Bastos JRM, Tomita N. Fluoride content and
pH of beverages found on the Brazilian market. J Dent
Res 1996;75:192.
11. Barros ERC, Scapini MFTC. Analise crtica da fluoretacao
de aguas no municpio de Porto Alegre, RS. Rev Gaucho
Odont 1990;38:24754.
12. Ferreira RA. Driblando a carie. Revista da APCD
1996;50:819.
13. Thylstrup A, Fejerskov O. Clinical appearance of dental
fluorosis in permanent teeth in relation to histological
changes. Community Dent Oral Epidemiol 1978;6:315
28.
14. Obry-Musset AM, Bettembourg D, Cahen PM, Voegel JC,
Frank RM. Urinary fluoride excretion in children using
potassium fluoride containing salt or sodium fluoride
supplements. Caries Res 1992;26:36770.
15. Larsen MJ, Fejerskov O, Bojen O, Senderowitz F. Fluctua-
tion of fluoride concentrations in drinking waters: a col-
laborative study. Int Dent J 1989;39:1406.
16. Kunzel W. Effect of an interruption in water fluoridation
on the caries prevalence of the primary and secondary
dentition. Caries Res 1980;14:30410.
17. Mller IJ. Fluorides and dental fluorosis. Int Dent J
1982;32:13547.
18. World Health Organization. Appropriate use of fluorides
for human health. Geneva: WHO; 1986.

19. Whitford GM. Intake and metabolism of fluoride. Adv
Dent Res 1994;8:514.
20. Likins RC, McClure FJ, Steere AC. Urinary excretion of
fluoride following defluoridation of a water supply. Pub-
lic Health Rep 1956;71:21720.
21. Fejerskov O, Manji F, Baelum V, Mller IJ. Dental fluoro-
sis a handbook for health workers. Copenhagen:
Munksgaard; 1988.
22. Hodge HC, Smith FA, Gedalia I. Excretion of fluorides.
In: Fluorides and human health. Geneva: WHO; 1970. p.
14161.
23. Ericsson Y. Urinary estimation of optimal fluoride dos-
age with domestic salt. Acta Odontol Scand 1971;29:43
51.
24. Menghini G, de Crousaz P, Steiner M, Helfenstein U, Sen-
er B. Excretion urinaire de fluorures chez des ecoliers de
Urinary fluoride and dental fluorosis
Geneve et Lausanne, en relation avec la fluoration du sel.
Schweiz Monatsschr Zahnmed 1989;99:29298.
25. Marthaler TM, Steiner M, Menghini G, de Crousaz P. Uri-
nary fluoride excretion in children with low fluoride in-
take or consuming fluoridated salt. Caries Res
1995;29:2634.
26. Zipkin I, Leone NC. Rate of urinary fluoride outcome in
normal adults. Am J Public Health 1957;47:84851.
27. van Nieuwenhuysen J-P, DHoore W. Caries dentaires,
fluorures urinaires et opacites de lemail. Schweiz Mo-
natsschr Zahnmed 1992;102:27985.
28. Delage Silva DRA, Gomes Silva ELL, Covello L, Sartori
LC. Levantamento epidemiologico em saude bucal: San-
tos, SO, 1995. University of Sao Paulo, Faculty of Public
Health, 1995.
323