Environ Geochem Health (2007) 29:51–58
DOI 10.1007/s10653-006-9062-0
ORIGINAL PAPER
Dental fluorosis associated with drinking water from hot
springs in Choma district in southern province, Zambia
V. Shitumbanuma Æ F. Tembo Æ J. M. Tembo Æ
S. Chilala Æ E. Van Ranst
Received: 31 March 2006 / Accepted: 2 October 2006 / Published online: 25 November 2006
Ó Springer Science+Business Media B.V. 2006
Abstract This study was conducted to investi-
gate the high incidence of mottled teeth among
residents of an area with hot springs in the Choma
District of the Southern Province of Zambia. A
survey involving 128 pupils was conducted at a
Basic School to collect data on pupil’s back-
grounds and their main sources of drinking water
between birth and age 7. A dental specialist
examined the pupils’ teeth and samples of drink-
ing water were collected from locations where the
majority of the pupils lived. It was analysed for
fluorides and other drinking water quality param-
V. Shitumbanuma (&)
Department of Soil Science, School of Agricultural
Sciences, University of Zambia, 32379, Lusaka,
Zambia
e-mail: shitumbanuma@zamtel.zm
F. Tembo
Department of Geology, School of Mines, University
of Zambia, 32379, Lusaka, Zambia
J. M. Tembo
Department of Civil and Environmental Engineering,
School of Engineering, University of Zambia, 32379,
Lusaka, Zambia
S. Chilala
Choma General Hospital, 630063, Choma, Zambia
E. Van Ranst
Laboratory of Soil Science, Geological Institute,
Ghent University, Krijgslaan 281 (S8), 9000, Gent,
Belgium
eters. Results of the survey showed a highly
significant (P < 0.001) association between pu-
pils’ main sources of drinking water between birth
and age 7 and the incidence of discoloured teeth.
All (100%) pupils who drank water from hot
springs before age 7 had moderate to severe
fluorosis, while the majority (96.7%) of the pupils
who drank water from other sources had no
dental fluorosis. Fluoride concentrations ranged
from 5.95 to 10.09 mg/l in water from hot springs,
and from 0.03 to 0.6 mg/l in water from other
sources. Fluoride levels in water from hot spring
water samples exceeded the 1.5 mg/l WHO
guideline value for drinking water, while those
in water from other sources were significantly
(P < 0.05) lower than this. We conclude that the
high prevalence of mottled teeth among residents
of the study area is a case of endemic dental
fluorosis associated with drinking water from hot
springs containing high concentrations of fluoride.
Keywords Dental fluorosis
Drinking water

Faults
Fluoride
Granite
Hot springs
Zambia
Introduction
Dental fluorosis, or mottled enamel, is a common
disorder of teeth associated with drinking water
containing high concentrations of fluoride. The
critical period of exposure to fluorides that leads
123
52
to manifestation of enamel or dental fluorosis is
during the formation of permanent teeth, gener-
ally from birth to the age of 7 or 8 (Billings,
Berkowit, & Watson, 2004). Dental fluorosis has
been reported in communities living in areas
where drinking water contains fluorides at con-
centrations as low as 1.0 mg/l (Smith, 1939;
WHO, 1996; Heller, Eklund, & Burt, 1997). It is
particularly prevalent in the Rift Valley regions of
East Africa, especially in Kenya and Ethiopia,
where volcanic sources of fluoride are important,
and in India and West Africa, where salts and
sedimentary fluorine-bearing minerals are the
primary sources of fluorine (Appelo & Postma,
1994).
In Zambia, dental fluorosis is not very common
and is generally not well documented. However,
an area with hot springs in the Choma District of
the Southern Province of Zambia has a high
incidence of people with discoloured and mottled
teeth. Residents of the area have long suspected
this problem to be associated with the drinking of
water from hot springs. Prior to this study, no
scientific investigation had been conducted to
determine whether there was any link between
the occurrence of discoloured and mottled teeth
among residents of the area in Choma with hot
springs and consumption of water from the
springs. This study was conducted to establish
whether there was an association between the
incidence of discoloured teeth among residents of
this area and the sources of drinking water in the
area, and whether the incidence of discoloured
teeth was associated with levels of fluoride in
drinking water.
Materials and methods
Location and description of study area
The study was conducted in the Choma District of
the Southern Province of Zambia (Fig. 1). The
Choma District is located on the Southern Plateau
of Zambia, which is part of the Central African
Plateau, a generally flat landscape at an elevation
of 1,200 m above sea level (Dalal-Clayton,
English, Williams, & Spaargaren, 1985). Much of
the Choma District is underlain by granite rocks
123
Environ Geochem Health (2007) 29:51–58
that are part of the large granitoid mass known as
the Choma–Kalomo batholith, which extends for
several tens of kilometres between the Choma and
Kalomo Districts (Newton, 1963).
The Choma District is in the dryer part of
Zambia, locally referred to as the medium rainfall
region, which in the past two decades has expe-
rienced frequent dry spells and droughts. Avail-
ability of water for domestic and agricultural use
is often a problem in this region, particularly in
the hot dry months of August–October, when
most seasonal streams dry up. During such
periods, people depend on water from deep wells
or have to travel long distances to get it. In
locations with hot springs, residents have used
water from the springs for their domestic require-
ments, livestock and irrigating crops in the dry
season. The majority of the residents in our study
area are workers on commercial tobacco farms
and ranches, while others are small-scale farmers,
engaged in growing maize or rearing cattle. Our
target group in the study was the children of this
low income group, who are the main victims of
the problem of mottled and discoloured teeth.
Methods employed in the study
A survey involving 128 pupils at a Basic School in
the study area was conducted, using a structured
questionnaire. Data captured during this survey
included pupils’ places of birth, their main source
of drinking water between birth and age 7, their
current residence, age and sex, among other
parameters. A physical examination of all the
pupils’ teeth was also carried out by a dentist to
check for clinical signs of dental fluorosis using
Deans’ Index. Most of the children with clinical
signs of dental fluorosis had moderate to severe
dental fluorosis according to Deans’ Index.
Results of the dental examination, indicating
whether or not pupils had clinical manifestation
of dental fluorosis and the degree of severity of
the symptoms were captured on the same sheet
that had other data on the background. After the
survey water samples were collected from main
drinking water sources at locations where the
majority of the pupils interviewed had lived
between birth and age 7. The water samples were
collected in triplicate at each location in 1-l
Environ Geochem Health (2007) 29:51–58
Fig. 1 Map of Zambia
showing location of the
study in the Choma
District
polyethylene bottles and transported to the lab.
The water temperature, pH and electrical con-
ductivity were measured on site using a portable
pH, EC Temperature, multiparameter analyser,
model Eijkelkamp 12.28. The following day, the
water samples were analysed for fluoride concen-
trations using an Ion selective electrode and for
other drinking water quality parameters follow-
ing standard methods outlined in Clesceri,
Greenberg, and Eaton (1998). Furthermore, sam-
ples were collected from rock outcrops, at hot
spring sites, at locations where outcrops were
present at or near the hot springs. Thin sections of
rock samples were made and studied using a
Reichert Neovar-Pol petrographic microscope, in
order to determine the mineralogy of the rocks.
To test whether there was an association
between the incidence of discoloured teeth among
pupils and main sources of drinking water between
birth and the age of 7 a Chi-squared test was used to
analyse the data obtained from the survey. The
sources of drinking water were grouped into two
classes, namely hot springs and non-hot spring
sources, hereafter referred to as ‘‘other sources’’.
To establish whether there was a significant
difference between the levels of fluorides in water
from hot springs and that from other drinking
53
water sources, a t test was used. The statistical
analyses were carried out using the software
packages SPSS version 11.0 and SAS version 6.12.
Results
Relation between occurrence of mottled teeth
and drinking water source
The incidence levels of discoloured teeth and
other symptoms associated with dental fluorosis
among pupils who drank water from hot springs
and other sources of drinking water are presented
in Table 1. All the 37 pupils (100%) who drank
water from hot springs during the period of
formation of their permanent teeth had symptoms
associated with dental fluorosis, while only 3
(3.3%) of the 91 pupils who drank water from
other sources had these symptoms. The overall
incidence of moderate to severe dental fluorosis
at the school was 31.25%, while that of pupils at
the school who had consumed water from hot
springs was 100%, and that of pupils who drank
water from other sources was 3.3%.
Results of the Chi-squared tests of indepen-
dence between the source of drinking water and
123
54
Table 1 Occurrence of moderate to severe dental fluoro-
sis among pupils in relation to their main source of
drinking water between birth and age 7 in the study area
Water source Clinical symptoms of Total
dental fluorosis
Absent (%) Present (%)
Other (non-hot spring) 88 (96.7) 3 (3.3) 91
Hot spring 0 (0) 37 (100) 37
Total 88 (68.75) 40 (31.25) 128
incidence of occurrence of dental fluorosis indi-
cated a highly significant (P < 0.001) association
between pupil’s sources of drinking water be-
tween birth and age 7 and the incidence of
discoloured teeth. Furthermore, the measures of
association between the occurrence of dental
fluorosis among pupils and their main source of
drinking water also indicate a highly significant
(P < 0.001) strong (0.946–1.00) association. More
specifically, the statistics indicated that the occur-
rence of clinical symptoms of dental fluorosis was
associated with drinking of water from hot
springs, while the absence of symptoms of dental
fluorosis was associated with drinking water from
other (non-hot springs) sources. The lambda
values for the Chi-squared test indicated that a
knowledge of a pupil’s source of drinking water
between birth and age 7 reduced the error in
predicting whether a pupil has dental fluorosis or
not by more than 90%.
Fluoride concentrations in water
from drinking water sources
Selected results of the analyses of the water
samples collected from the study area are pre-
sented in Table 2. The temperature of water from
hot springs measured in the field ranged from 28.2
to 70.6°C, while that of samples collected from
other sources (bore holes and shallow wells)
ranged from 24 to 26.6°C. The concentration of
fluoride in water samples from hot springs was
significantly (P < 0.0001) higher than that of
samples from other sources, with mean values of
6.56 mg/l for hot springs and 0.15 mg/l for other
water sources. In general, water with higher
temperatures contained higher fluoride concen-
123
Environ Geochem Health (2007) 29:51–58
trations than that with lower temperatures. There
was, however, no statistically significant relation-
ship between water temperature and the concen-
tration of fluoride in the water.
The range of fluoride concentrations of water
samples from hot springs and from other water
sources is shown in Fig. 2. Concentrations of
fluoride in water from hot springs are significantly
higher (P = 0.001) than those in water from other
sources and also significantly greater than the
WHO (1996) guideline value for fluoride in
drinking water of 1.5 mg/l. In contrast, fluoride
concentrations in water from other sources are
generally far below the WHO guideline, except
for samples collected from a shallow well at
Miller Farm, which is less than a kilometre away
from a hot spring.
Geology of hot spring areas
The hot springs of the study area emanate from
within granitic rocks that make up the large
granitoid mass called the Choma–Kalomo batho-
lith. The granites are poorly exposed at the sites
where most of the hot springs are located, except
at Chibimbi, where there are more extensive
outcrops.
Two varieties of granite were observed in the
field. The most common variety is a pink medium
to coarse grained porphyritic granite that varies in
composition from leucocratic K-feldspar granite
to mesocratic biotite granite. In places, aplite
dykes of up to 10 cm thick are observed, whose
dominant minerals are K-feldspar and quartz.
The second variety is a grey, porphyritic, coarse
to medium-grained biotite-hornblende granite.
Both rock types are weakly to strongly foliated
and in the more strongly deformed parts, the
preferred orientation of the minerals, particularly
biotite, give the rock a gneissic appearance.
The surfaces of rock outcrops close to the hot
springs appear whitish due to weathering of the
feldspars. The granite is a coarse grained rock
that consists of potassium and sodium feldspars,
biotite, hornblende and quartz with accessory
apatite and iron ore. All the hot springs are
located along fault structures in the granite.
Faulting is indicated by sheared granite where
the rocks are exposed at the surface.
Environ Geochem Health (2007) 29:51–58 55

Table 2 Temperature
Location Water source Temperature (°C) F– (mg/l)
and mean fluoride
concentrations (with B Miller Farm Hot spring 70.6 7.48 ± 0.01
standard errors) of water Reservoir * 21.5 10.10 ± 0.10
samples from different Shallow well 25.9 2.28 ± 0.01
sources in the study area Ross Farm Hot spring 28.2 6.82 ± 0.02
Chibimbi Hot spring 45.8 6.40 ± 0.66
Hot spring 32.8 8.27 ± 0.03
Semahwa river Hot spring 50.7 5.96 ± 0.01
ND measurement was not Moomba Farm Bore hole 24.8 0.60 ± 0.01
determined Nchenga Farm Bore hole 24.3 0.04 ± 0.01
Kachenje Shallow well ND 0.04 ± 0.00
* Originally derived from Choma Lodge River 26.6 0.03 ± 0.00
hot spring

10.0 crystals associated with biotite. The granite from

Chibimbi hot springs is pink in colour and in the
Fluoride concentration (mg/L)

8.0
main fault zone has been transformed into
6.0
mylonite, resembling quartzite. In the thin sec-
4.0 tion, the rock consists of microcline, sodic plagio-
2.5
clase and quartz, all of which show a high degree
2.0

1.5
WHO (1996) Drinking water guideline for F- of recrystallisation and granulation.
1.0
The grey granite consists of sodic plagioclase
0.5 (50%), which occurs as both porphyroclasts and
0.0 in the matrix, and hornblende (10%). The other
Hot Spring Others minerals in the rock are microcline (10%), quartz
Water Source
(20%) and biotite (10%). Accessory minerals
Fig. 2 Fluoride concentrations in water from hot springs include iron oxide, epidote and apatite. The grey
and other drinking water sources in the study area (Choma granite also displays varying degrees of alteration
District) of the feldspars to sericite and kaolinite. The
quartz in both rocks occurs as anhedral grains and
shows undulose extinction indicative of the defor-
Petrography and mineralogy mation suffered by the rocks.

Results of the petrographic analyses showed that Fault structures and location of the hot springs
the pink granite has a porphyritic texture in which
1–2 cm of pink microcline porphyroclasts are All the hot springs in the study area are located at
embedded in a medium-grained matrix of micro- edges of gently sloping topographic highs that
cline (50%), plagioclase (10%), quartz (25%), separate shallow, broad valleys that characterize
biotite (10%) and hornblende (1%), with iron the drainage pattern in the area. In all the areas
oxides, apatite and epidote as accessory minerals. investigated, the hot springs were observed to be
The microcline is subhedral to anhedral and located along faults that have been identified
shows moderate to strong alteration to sericite. using both satellite imagery and field observa-
Plagioclase occurs as a ground mass mineral with tions. In the field, the presence of faults is
anhedral to subhedral grain shapes showing indicated by the presence of fault breccia,
deformation twinning and moderate to strong silicification and shearing of the granite. Figure 3
kaolinisation and epidotisation. The biotite is shows the location of hot springs at Miller
green and occurs as unevenly distributed clots and and Ross Farms, and at Semahwa River, in
flakes scattered in the rock. Hornblende occurs as relation to the geology and occurrence of faults
isolated euhedral to subhedral green pleochroic in the area.

123
56
Fig. 3 Geology and
location of three of the
hot spring sites in the
study area
Discussion and conclusion
It is important to mention that the study area is in
a rural district with no industrial activity that
could possibly serve as an anthropogenic source
of fluoride. Additionally, the study area and its
surrounding environment are very sparsely pop-
ulated. With the exception of the few locations
with human dwellings, most of the area is under
the natural Miombo woodland that characterises
much of the plateau region of Zambia. Conse-
quently, the mostly likely major sources of fluo-
ride for the residents of this area are the water
they drink and the food they eat. Considering that
the majority of children included in the study are
the offspring of farm workers and peasant farm-
ers, much of the food they eat is locally produced.
123
Environ Geochem Health (2007) 29:51–58
Furthermore, because the majority of the children
in this study came from poor families who often
cannot afford toothpaste and therefore do not
usually use it, fluoridated dental products are not
likely to be significant sources of fluoride in this
population. There is no fluoridation of water in
Zambia, so the consumption of municipally fluo-
ridated water cannot be considered a possible
source of fluoride either.
Results of this study have shown that the high
incidence of discoloured teeth among the resi-
dents of the area of the Choma District with its
hot springs is strongly associated with the drink-
ing of water from hot springs that contain high
levels of fluoride. Mamuse (2003) reports a
similar observation in the Chimanimani District
in Eastern Zimbabwe in an area with hot springs
Environ Geochem Health (2007) 29:51–58
where the incidence of dental fluorosis among
school children is 61.9% and the fluoride levels in
the water are up to 5 mg/l. A number of
researchers (Smith, 1939, Alarcon-Herrera,
Martin-Dominguez, Trejo-Vasqez, & Rodriguez-
Dozal, 2001; Louw, Grobler, & van Kotze, 2002;
Heller, Eklund, & Burt, 1997) have reported that
drinking water containing 1.0 mg/l fluoride or
lower causes dental fluorosis.
Recent research on dental fluorosis indicates
that the effects of fluoride on enamel in man are
cumulative, and that rather than requiring a
specific threshold dose, it depends on the total
intake from all sources, and on the duration of the
intake (Aoba & Fejerskov, 2002). Our estimate of
daily fluoride intakes for children in the study area
drinking hot spring water with an average
concentration of 6.56 mg F/l, assuming children’s
body weight to range from 15 kg to 30 kg, and
water consumption levels of 500 ml to 1,000 ml
per day, was 0.11–0.21 mg/kg of body weight.
These fluoride intake values are much higher than
the 0.03 mg/kg of body weight reported by Aoba
and Fejerskov (2002) to be certain to result in
dental fluorosis, and the range of 0.03 to 0.1 mg F/
kg of body weight reported by Billings, Berkowitz,
and Watson (2004) to be the fluoride intake levels
from birth to age 8 in populations where dental
fluorosis is a problem. On the other hand, children
drinking water from other sources with an average
F concentration of 0.15 mg/l, with the same range
of body weights and water consumption levels,
would have intakes of 0.003–0.005 mg F/kg of
body weight, which are far below the above
reported values.
In view of the above discussions and of the fact
that fluoride concentrations in water samples from
hot springs are much higher than the suggested
maximum limit for drinking water of 1.5 mg/L
allowed by WHO (1996), while those of water
samples from other sources are much lower than
this limit, we are compelled to conclude that the
high prevalence of discoloured teeth among res-
idents of the study area is a case of endemic dental
fluorosis caused by drinking water from
hot springs that have high concentrations of
fluoride.
Unlike the Rift Valley areas of East Africa
where fluoride in water is derived from volcanic
57
activity (Appelo & Postma, 1994), we think that
the sources of fluoride in the hot spring water in
Choma are fluoride-bearing minerals in granitic
rocks. Minerals such as micas (biotite, muscovite
and phlogopite), amphiboles such as horn-
blende, and apatite are reported to be among
the sources of fluorine in rocks (Wedepohl,
1974; Goldschmidt & Muir, 1970). Biotite,
hornblende and apatite are all present in granite
samples obtained from sites where the hot
springs are located. The high fluoride levels in
water from hot springs are most probably due to
the chemical weathering of fluoride-bearing
minerals, which happens upon contact with the
hot water. Dissanayake (2005) has also attrib-
uted the high levels of fluoride in fluorosis-
affected regions of Sri Lanka to the presence of
high-grade metamorphic rocks containing fluo-
rine-bearing minerals such as micas, hornblende
and fluorite.
Our studies have revealed evidence of chem-
ical weathering of granitic rock at sites with hot
springs. Results of petrographic analyses of thin
sections of rocks show that hydrothermal alter-
ation has occurred in rocks at hot springs, as
evidenced by the transformation of K- feldspars
to sericite and the alteration of plagioclase to
epidote and kaolinite. In addition, the fluorine-
bearing minerals biotite and hornblende have also
been altered to various degrees and have com-
pletely disappeared in the fault rocks. We think
that the chemical alteration of the rocks has been
facilitated by fault structures that serve as easy
channels for water movement over a larger
surface area of exposed minerals.
Acknowledgements We are grateful to the VLIR Inter-
University Cooperation for providing the research grant to
conduct this work. We thank farmers in the study area,
particularly Mr. Miller and Mr. Ross for allowing us access
to their farms and teachers at Kachenje Basic School for
their help during the course of our field work. We further
acknowledge the assistance of the Choma District Health
Management Board for allowing their member of staff to
assist with the dental studies and for allowing us to conduct
this work in their district. We are indebted to Mr. Paul
Zimba for the cartographic work and to Dr. Habatwa
Mweene for greatly helping to improve the clarity of the
presentation of the work in this paper. Lastly, our
appreciation goes to the reviewers of this paper whose
suggestions further enhanced the quality of its
presentation.
123
58
References
Alarcon-Herrera, T. M., Martin-Dominguez, I. R., Trejo-
Vasqez, R., & Rodriguez-Dozal, S. (2001). Well water
fluoride, dental fluorosis and bone fractures in Gua-
diana Valley of Mexico. Fluoride, 34(2), 139–149.
Aoba, T., & Fejerskov, O. (2002). Dental fluorosis:
chemistry and biology. Critical Reviews in Oral
Biology and Medicine, 13(2), 155–170.
Appelo, C. A. J, & Postma, D. (1994). Geochemistry,
Groundwater and Pollution. Rotterdam: Balkema, p.
536.
Billings, R. J., Berkowitz, R. J., & Watson., G. (2004).
Teeth. Pediatrics, 113(4), 1120–1127.
Clesceri, L. S., Greenberg, A. E., & Eaton, A. D. (Eds.).
(1998) Standard Methods for Examination of Water
and Waste Water (20th ed.). Washington, DC:
American Public Health Association (APHA), Amer-
ican Water Works Association (AWWA) and Water
Environment Federation (WEF).
Dalal-Clayton, D. B., English, C., Williams, G. J., &
Spaargaren, O. (1985). A Geomorphic Legend for
Zambia. Technical Guide No. 15. Soil Survey Unit.
Research Branch. Department of Agriculture. Mt
Makulu. Ministry of Agriculture and Water Develop-
ment. Republic of Zambia: Government Printers.
Dissanayake, C. C. (2005). Of stones and health: medical
geology in Sri Lanka. Science, 305, 883–885.
Goldschmidt, V. M., Muir, A. (1970). Geochemistry.
Oxford: Oxford University Press, pp. 568–583.
123
Environ Geochem Health (2007) 29:51–58
Heller, K. E., Eklund, S. A., & Burt, B. A. (1997). Dental
caries and dental fluorosis at varying fluoride concen-
trations. J Public Health Dent, 57(3), 136–143.
Louw, A. J., Grobler, S. R., & van W. Kotze, T. J. (2002).
Degree of fluorosis in areas of South Africa with
different levels of fluoride in drinking water. General
Dentistry, July–August 50(4), 352–356.
Mamuse, A. (2003). Fluoride Contaminated water in
Gokwe District (NW Zimbabwe): Spatial Distribution,
Lithostratigraphic Controls and Implications for Hu-
man Health. Msc Thesis, Curtin University of Tech-
nology, Australia.
Newton, A. R. (1963). The Geology of the Country
Between Choma and Gwembe. Explanation of the
Degree Sheet 1627, SW and SE Quarter. Northern
Rhodesian Ministry of Labour and Mines. Report of
the geological survey No. 8. Lusaka, Zambia: The
Government Printer.
Smith, M. C. (1939). Iodine and fluorine. In: Food and
Life. Year Book of Agriculture. United States
Department of Agriculture, Washington, DC, pp.
211–213.
Wedepohl, K. H. (1974). Handbook of Geochemistry.Vol
II/4. Fluorine. 9 (A1–A6). Berlin, Heidelberg, New
York: Springer-Verlag.
World Health Organization. (1996). Guidelines for Drink-
ing Water Quality, (Vol. 2). Health Criteria and other
supporting information (2nd ed.). Geneva: WHO, pp.
231–233.