Original Paper

Caries Res 1999;33:8187

Received: March 9, 1998

Accepted after revision: May 6, 1998

Enamel Erosion by Some Soft Drinks and

Orange Juices Relative to Their pH, Buffering
Effect and Contents of Calcium Phosphate
M.J. Larsen B. Nyvad
Department of Operative Dentistry and Endodontics, Royal Dental College, Faculty of Health Sciences,
Aarhus C, Denmark

Key Words
Buffering effect Enamel dissolution Erosion Orange
juice pH Soft drinks

Abstract
The capability of a soft drink or a juice to erode dental
enamel depends not only on the pH of the drink, but also on its buffering effect. As the latter is the ability of the
drink to resist a change of pH it may add to the effects of
the actual pH. The aim of the present study was to compare the pH and the buffering effect of various soft
drinks with their erosive effects and the solubility of apatite. In 18 soft drinks, mineral waters and juices available on the Danish market, pH and the concentrations of
calcium, phosphate and fluoride were determined. The
buffering effect was determined by titration with NaOH.
Human teeth (n = 54) covered with nail varnish except
for 3E 4-mm windows were exposed to 1.5 liters of the
drink for either 7 days or 24 h under constant agitation.
The depth of the erosions was assessed in longitudinal
sections. The depth was found to vary greatly from 3
mm eroded by the most acidic drinks and fresh orange
juice to only slightly affected surfaces by most of the
mineral waters. The dissolution of enamel increased
logarithmically inversely with the pH of the drink and
parallel with the solubility of enamel apatite. Orange
juice, pH 4.0, supplemented with 40 mmol/l calcium and
30 mmol/l phosphate did not erode the enamel as the

1999 S.Karger AG, Basel

00086568/99/03310081 $15.00/0
Fax +41 61 306 12 34
E-Mail [email protected]
www.karger.com

Accessible online at:

http://BioMedNet.com/karger

calcium and phosphate saturated the drink with respect

to apatite. Generally, the lower the pH the more NaOH
was necessary to bring the pH to neutrality. In particular
the buffering effect of the juice was high. For all drinks,
no effect of their low fluoride concentrations was observed.

Over the last decades, the prevalence of erosion seems to

have increased [Nunn, 1996], presumably due to an increasing consumption of soft drinks and fruit juices. The factors
known to cause erosion include all kinds of acidic foods
with a low concentration of calcium and phosphate [Meurman et al., 1987; Lussi et al., 1993; Zero, 1996]. The erosive
agents are characterized as being highly undersaturated
with respect to both hydroxyapatite and fluorapatite
[Larsen, 1975; Meurman and ten Cate, 1996] so that enamel apatite dissolves without any mutual formation of surface
fluorhydroxyapatite [Larsen, 1973].
When enamel dissolves in vivo the dissolution causes either of two distinctly different types of lesions, a caries lesion or erosion. By definition, a caries lesion is caused by
acids formed by bacterial degradation of sugars, whilst erosion is defined as chemical dissolution caused by acids of
any other origin [Imfeld, 1996]. Due to the nature and origin of the acid, the caries lesion is localized under plaque
whilst erosion appears widespread on exposed surfaces. The
two lesions are histologically clearly different, the caries le-

M. Joost Larsen
Department of Operative Dentistry and Endodontics
Royal Dental College, Faculty of Health Sciences
Vennelyst Boulevard, DK8000 Aarhus C (Denmark)
Tel. +4589 424 093, Fax +4586 202 202, E-Mail [email protected]

Fig. 1. Apatite solubility as a function of pH in the range 74 as calculated by a previously published computer program [Larsen, 1986].
The conditions were: a Ca/P ratio of 5/3, a pKHAp of 58.5 [McDowell
et al., 1977] and that enamel contains 38% calcium. For comparison,
pH in plaque after a sucrose rinse considered to be cariogenic usually
ranges above pH 4.5. The dotted horizontal line that intersects the
curve around the critical pH of 5.5 indicates calcium and phosphate
levels in saliva.

Fig. 2. Apatite solubility as a function of pH in the range from 7 to 2.4

sion being a gradual demineralization of subsurface enamel

and dentin with the outer 20- to 50-m-thick surface layer
preserved more or less intact, in contrast to the erosion in
which the dental mineral is dissolved layer by layer leaving
the subsurface enamel almost unaltered. A softening of the
eroded enamel surface indicative of a slight demineralization of the very surface enamel was observed by Meurman
et al. [1990].
In erosion the amount of mineral dissolved from enamel
depends on a number of conditions: pH, the buffering effect
or the concentration of acids and the length of the exposure
time. Finally, the presence of suitable concentrations of calcium, phosphate and fluoride in the drink may counteract
the dissolution.
Figures 1 and 2 demonstrate that the solubility of apatite
(defined as the amount of apatite that can exist dissolved at
equilibrium at the actual pH) increases logarithmically with
a drop in pH, the increase in solubility being largest at low
pH. Although the solubility of apatite is considerable
around pH 4 it seems reasonable to counteract enamel dis-

solution by addition of calcium phosphate to a drink with a

pH as low as 4. The solubility of apatite increases from
75 g/l at pH 3 to 400 g/l at pH 2.5. In comparison, the entire
permanent dentition weighs less than 50 g [Bolk, 1925].
However, this does not mean that a dentition can be dissolved in 1 liter of a pH 3 juice because during the dissolution of dental apatite, pH shifts to higher values as a result
of the dissolution of the apatite [Larsen et al., 1993]. Further, the above calculation is based on equilibrium conditions which do never occur in the oral cavity: most of the
drinks wash quickly through the mouth without even being
in contact with the teeth.
Apart from its pH, the ability of an acidic solution to dissolve enamel apatite depends on its ability to keep pH unaffected by the dissolution of apatite and dilution with saliva,
i.e. its buffering effect. The higher the buffering effect or
concentration of acids in the drink, the more apatite will be
dissolved before neutral pH is approached and the dissolution brought to an end. Therefore, it seems likely that an
acidic solution with a high buffering effect in the low pH

82

Caries Res 1999;33:8187

as calculated by a previously published program [Larsen, 1986] under

conditions as those in figure 1. A tremendous increase in apatite solubility in the lower pH range can be seen. The horizontal line indicating calcium phosphate levels in saliva coincides with the abscissa and
can therefore not be distinguished.

Larsen/Nyvad

Table 1. Concentrations of some inorganic ions and pH in soft drinks, mineral waters and orange juices

pH

Schweppes Indian Tonic

Schweppes Dry Grape
Fanta Orange
Sprite Light
Coca Cola
Pepsi Cola
7 Up
Orange Juice MD
Orange Juice MD/calcium
Carlsberg Kurvand/citrus-lime
Carlsberg Kurvand
Tuborg Blue/citrus-lime
Tuborg Blue Special
Aqua Minerale/citrus-lime
Aqua Minerale
Apollinaris/citrus-lime
Maarum
Apollinaris
Distilled water
1

2.48
2.76
2.86
2.98
2.40
2.53
3.20
3.84
4.03
4.02
4.59
5.42
5.44
5.01
5.11
4.44
5.38
4.88
5.60

Buffer effect1
mmol/l
dpH 5.5

dpH 7.0

47
47
32
16
9
8
15
78
78
5
2
0
0
0
0
0
0
0

68
66
51
31
25
18
33
107
109
17
17
11
21
11
11
5
20
10

Calcium
concentration
mmol/l

Phosphate
concentration
mmol/l

Fluoride
concentration
ppm

0.43
0.58
0.38
0.36
0.26
0.09
0.19
2.97
42.90
1.18
1.15
0.35
0.38
0.34
0.33
0.42
0.90
3.03

0
0.18
0.13
0
5.47
5.46
0
3.55
31.20
0
0
0
0
0
0
0.01
0
0.01

0.41
0.33
0.20
0.58
0.20
0.45
0.58
0.80
0.09
0.06
0.09
0.20
0.35
0.52
0.55
0.17
1.78
0.41

Buffer effect defined as the amount of alkali needed to bring pH to 5.5 or 7.0 (in mmol/l).

range may be more harmful to enamel compared to a solution with the same buffering effect, however, in the higher
pH range. To which extent the value of pH or the buffering
effect is the most important is unclear.
An additional factor of significance in erosion is the concentration of calcium, phosphate and fluoride in the acidic
drink. Generally, the concentrations of these ions are low
and far from sufficient for saturation of the drink, but modification of the drink by supplementation has been shown
to significantly counteract dental erosion [for review see
Grenby, 1996a]. Because the solubility of enamel apatite increases so steeply (fig. 1, 2) only drinks with pH above 3.5
may realistically benefit from calcium phosphate supplementation in counteracting enamel dissolution.
The aim of the present study was to examine the pH of
various soft drinks and juices available on the Danish market, their buffering effect with emphasis on the pH range in
which the buffer is efficient, and their contents of calcium,
phosphate and fluoride, for a comparison with their ability
to dissolve dental enamel in vitro.

Enamel Erosion by Soft Drinks and Orange

Juice

Materials and Methods

Eighteen commonly available soft drinks and juices on the Danish
market were chosen for the study (table 1). Each drink was by chemical analysis characterized by its concentration of ions of dental interest by the following methods: calcium by AA spectrophotometry
[Willis, 1961], phosphate by color spectroscopy [Chen et al., 1956],
and fluoride by electrode (Orion); pH was determined by glass electrode immediately after opening the bottle and subsequently in a
poured 10-ml volume of the drink. The buffering effect in 3 ml of
each drink was determined by monitoring pH after serial additions of
0.025 ml 0.2 mol/l NaOH. For comparative purposes, the amount of
NaOH necessary to change the pH of 1 liter of the actual drink to
pH 4.0, 5.0, 5.5 and to 7.0 was calculated.
The erosive potential of the drinks and distilled water was studied
in a total of 54 freshly extracted human premolars and molars which
prior to experimentation were covered with nail varnish except for
3E4-mm windows left for exposure. During the exposure to the
drinks the teeth were hung onto closed 2-ml, test tubes serving as
floating buoys. Bottles containing 1.5 liters of soft drink were cautiously opened, a few milliliters of the content removed for analysis,
and a tooth hanging under its test tube was set to float with the tooth
submersed in the drink before the bottles were closed again. Care was
taken to avoid loss of carbon dioxide during opening and the subsequent exposure period. All bottles were continuously gently agitated
on a table for 24 h or 1 week before discontinuation of the exposure.
After the exposure, the teeth were sectioned longitudinally and
ground to a thickness of approximately 100 m. The depth of the ero-

Caries Res 1999;33:8187

83

Fig. 3. Depths of artificial erosions in enamel after exposure to soft

drinks (J), mineral waters (L), orange juices with and without calcium phosphate supplementation (*) and distilled water (K) for 1 week
as a function of pH of the drink. Note that the calcium phosphate-supplemented orange juice did not erode the enamel. If the orange juices
are disregarded, the series of points tends to indicate a curved line of
a shape similar to those in figures 1 and 2.

Fig. 4. Depths of artificial erosions in enamel after exposure to

lemonade-like carbonated soft drinks (J), orange juice without calcium phosphate supplementation (*) and distilled water (K) for 24 h as
a function of pH of the drink.

sions was assessed using a Quantimet 550+ (Leica). For each lesion
the greatest distance from the eroded surface perpendicular to the redrawn original anatomical enamel surface was recorded as the depth
of the erosion. The mineral distribution of the eroded teeth was examined in microradiographs produced by a Machlett X-ray tube operated at 20 mA and 20 mV.
All experiments and analyses were duplicated in two separate
runs. For calculation of enamel solubilities, a previously published
computer program [Larsen, 1986] was used. The solubility product
for hydroxyapatite was 1058.5 [McDowell et al., 1977] and that for
fluorapatite 1059.6 [McCann, 1968]. A regression analysis was made
according to Armitage [1977]. The use of human teeth for this experiment was approved by an ethical committee.

like soft drinks had a low pH and at least some of them a

considerable buffering effect. The pH of the orange juices
ranged around 4 and exhibited a high buffering effect,
whilst the pH of the mineral water-type drinks varied
around 5 and above and their buffering was generally weak.
The buffering effect of the two types of juice (orange
juice with or without calcium phosphate complementation,
respectively) did not differ significantly, despite the fact
that the addition of phosphate in concentrations of
28 mmol/l represents a buffering effect of approximately
1015 mmol/l in the pH range 67. The calcium and phosphate concentrations in the orange juice supplemented with
minerals ranged around 43 and 31 mmol/l, which by calculation corresponds to the solubility of apatite in distilled water at pH 4.
In contrast to the juices, when the carbonated drinks
were poured some carbon dioxide escaped and pH rose
somewhat. The pH rise was considerable in the mineral waters and less in the lemonade-type carbonated drinks. The
pH of the lemonade-type of soft drinks with sugar was lower than that of the juices, but the buffering effect was con-

Results

Chemical Analysis
The soft drinks and juices were found to fall into three
different categories (table 1): (a) carbonated lemonade-like
soft drinks with carbohydrates (Schweppes, Coca-Cola,
Pepsi), (b) orange juices, and (c) mineral water-like drinks
without sugar. For an overall comparison, the lemonade-

84

Caries Res 1999;33:8187

Larsen/Nyvad

siderably smaller, with Schweppes drinks at the upper end

and the colas at the lower end of the range. The colas contained around 5.7 mmol/l phosphate in contrast to the other
drinks. The pH of the mineral water-type of drinks was
comparatively high and the buffering effect of this kind of
drink was low, irrespective of the addition of citrus and
lime. In this category pH rose almost half a unit upon pouring.
The fluoride concentration was low in all drinks, except
for the Maarum mineral water in which 1.8 ppm fluoride
was found.
Histologic Examination of Exposed Teeth
From figures 1 and 2 it appears that 1.5 liters of distilled
water at pH 7 can dissolve 0.008 g of enamel, at pH 5
around 0.5 g, at pH 4 around 5 g, and at pH 3 approximately 85 g. In the actual experiments, only small amounts of
enamel and dentin were dissolved: An erosive effect of the
mineral water-type of drinks could be difficult to distinguish directly. However, the lesions were clearly identified
and measured in the microradiograms. In contrast to the
mineral waters, the lemonade-type of soft drinks caused a
considerable erosion of the enamel with some of the lesions
extending to the dentin. As the organic material of the
dentin appeared to protect the underlying mineral and thus
to slow down further dissolution, lesions deeper than 2 mm
should be evaluated cautiously as they might have been
deeper had the enamel been thicker. A plot of the depths of
the lesions induced by exposure for 7 days to the drinks in
seen in figure 3: The lemonade-like carbonated soft drinks
with pH below pH 3.3 induced the deeper lesions, whilst the
mineral-like waters affected the enamel much less.
When the lesion reached the dentin the rate of demineralization was reduced and the depth determination would
therefore be inaccurate. Therefore, a second experiment on
the most erosive drinks, the carbonated soft drinks and the
orange juice, was made by inducing erosions for a shorter
period of 24 h. The examination of the teeth exposed for only 24 h to the drink showed that, in principle, the lower pH
the deeper was the lesion (fig. 4).
The orange juice that did not contain added calcium and
phosphate caused a deep lesion in enamel in contrast to the
juice that contained the two ions. Calculation showed that
given a pH of 4.03 and concentrations of calcium 43 mmol/l,
phosphate 31 mmol/l and fluoride 0.09 ppm, then the ion
products in distilled water would be 1053.9 for fluorapatite
and that of hydroxyapatite 1058.4, i.e. a supersaturation with
respect to fluorapatite and saturation with respect to hydroxyapatite. Thus, the addition of calcium and phosphate
in amounts corresponding to the actual enamel solubility

Enamel Erosion by Soft Drinks and Orange

Juice

Table 2. Regression analysis of logarithms of depth of erosion as a

function of pH of the soft drinks and the mineral waters (for details,
see text)

SSDregression
7.5190
SSDaround
1.0647
SSDtotal
8.5837
Equation Y = 1.8130.633X
Correlation coefficient r = 0.936

DF 1
DF 13
DF 14

F 91.8103

was close to establish saturation with respect to hydroxyapatite and, as observed by experimentation, prevented the
enamel from dissolving.
Comparisons of Chemical Composition and in vitro
Enamel Dissolution
Overall, figure 3 demonstrates that the erosive effect of
drinks above pH 4.2 is limited. At pH 4 and below, the depth
of the erosion was found to increase steeply and parallel
with the increase in apatite solubility in the low pH range
(fig. 1, 2). Apart from the juices, no particular measurable
effect on lesion depth could be attributed neither to the
buffering effect of the drinks nor to the content of calcium,
phosphate or fluoride. Yet, the Maarum mineral water with
1.8 ppm fluoride did not cause any observable erosion, in
contrast to the two Tuborg Blue waters at the same pH and
with only 0.20.3 ppm fluoride.
Statistical Results
From figure 3 it is seen that the depths of the erosions as
induced by the drinks decrease logarithmically with pH.
Therefore, a regression of the logarithms of the depths on
pH was made (table 2). Due to their special character the
juices, with or without added calcium phosphate, were not
included in the regression. The Maarum mineral water
could not enter the regression as it did not cause any observable erosion. The regression confirmed the close relation between the depth of erosion and the pH of a drink,
r = 0.936.

Discussion

No generally accepted golden standard for in vitro development of erosions has been developed [Grenby,
1996b]. In the present study erosions were induced in 1.5
liters of acidic drinks under continuous agitation during 1
week. It must be realized that this exposure was extremely
intensive and far from that of real life. The high intensity of

Caries Res 1999;33:8187

85

the exposure was deliberately chosen because buffer effects

are more likely to show under prolonged exposure. Furthermore, also drinks with less erosive capabilities may give
evidence of erosion under such conditions.
It should be noted that the curves in figures 1 and 2 are
derived from an apatite solubility product of 1058.5 [McDowell et al., 1977] and is as such theoretical. Below pH 4
it is not possible to verify the curve experimentally because
a competing precipitation of brushite masks the results. The
reason for that is that the solubility of apatite increases more
than that of brushite below pH 4 [Larsen, 1986; Larsen and
Jensen, 1989]. Therefore, at e.g. pH 3, as apatite dissolves,
a competing precipitation of brushite will prevent saturation
of the solution and thus maintain a continuous dissolution
of enamel apatite for almost as long as pH is that low. In
conclusion, the theoretical curves cannot be verified experimentally and secondly, if brought to the end the dissolution
of apatite might be higher than the curves indicate.
Comparisons of the curves in figures 1 and 2 and the
plots in figures 3 and 4 showed that apart from the juices
the depth of the lesions was a direct function of the enamel apatite solubility at the pH of the drink, a result comparable to that of Rytmaa et al. [1988]. A characteristic of the
curves is their logarithmic nature. Therefore, when regressions are made it seems sensible to use logarithmic plots
in order to obtain a good fit for straight line regression
(table 2).
The 0.1 mmol/l calcium and 5.5 mmol/l phosphate of the
colas (table 1) could be supposed to prevent some of the dissolution of enamel. However, at their pH around 2.5 the solubility of apatite in distilled water ranges around 500 g apatite per liter (fig. 1), corresponding to 1,000 mmol/l apatite
or 5,000 mmol/l calcium and 3,000 mmol/l phosphate per
liter. In this context, the 5.5 mmol/l phosphate in the colas
cannot be expected to have a noticeable inhibitory effect on
erosion. Further, we observed that the colas dissolved neither more nor less apatite than their pH suggested.
The calcium and phosphate added to the orange juice
had a strong preventive effect on erosion. This particular
juice has recently been launched as a functional food intended to counteract osteoporosis. The concentrations of
calcium and phosphate in the juice are similar to those in
cows milk products. In the present study the added calcium
and phosphate appear to be able to prevent the erosion of
the juice itself.
The rather low concentrations of fluoride in the drinks
did not affect the depth of the lesions. In the literature, the
effect of fluoride has been studied under various conditions
with conflicting results [Holloway et al., 1958; Meurman
and ten Cate, 1996; Sorvari et al., 1994; Spencer and Ellis,

86

Caries Res 1999;33:8187

1950]. Because fluoride is supposed to act in concert with

the other ions of the apatite lattice (calcium and phosphate)
and in concentrations that saturate the solution with respect
to fluorapatite [Larsen et al., 1976], it does not seem likely
that fluoride alone has the ability to suppress erosion. However, more work is needed to clarify this problem. For the
present study it is concluded that none of the inorganic ions
in the drinks had any observable effect on the erosive capabilities, except for the high amounts of calcium phosphate
added to the orange juice.
The drinks could be subdivided into three groups according to their buffering effect with the orange juices
showing high buffer effects, with the lemonade-like soft
drinks as a medium group, and the mineral waters almost
without any buffers. It seems likely that the high buffering
effect of the juice caused the juice to create a deep erosion
in the enamel corresponding to what a drink of one pH unit
lower would have caused. Part of the erosive potential of
the juice could probably also be ascribed to its content of organic acids forming organo-calcium complexes which will
increase the solubility of enamel apatite beyond what the
curves in figures 1 and 2 suggest. For the lemonade-like
drinks the buffering effect varied considerably with
Schweppes tonic waters in the upper range. A closer examination of the data in the plot figures 3 and 4 shows that the
high buffering effect of these tonics did not lead to a deeper
erosion than judged from their pH. On the other hand, a
comparison of the buffering effect of the tonics with that of
juice, as seen in table 1, suggests that an effect might exist.
In conclusion, the present work has shown that mineral
waters erode enamel only to a limited extent in contrast to
lemonade-like soft drinks and orange juice. Except for the
orange juice, the erosive capability of the drinks increased
logarithmically inversely with the pH of the drink and parallel with the solubility of apatite. Further, apart from the orange juice, no effect was observed, neither of calcium,
phosphate, and fluoride in the drink nor of their buffering
effect. For the orange juice, the very high buffering effect
and a very high content of calcium and phosphate in the
modified juice affected the erosion significantly.

Larsen/Nyvad

References
Armitage P: Statistical Methods in Medical Research. Oxford, Blackwell, 1977.
Bolk L: Das Gewicht der Zhne. Anat Anz 1925;
59:572574.
Chen PS, Toribara TY, Warner H: Microdetermination of phosphorus. Anal Chem 1956;28:1756
1758.
Grenby TH: Lessening dental erosive potential by
product modification. Eur J Oral Sci 1996a;
104:221228.
Grenby TH: Methods of assessing erosion and erosive potential. Eur J Oral Sci 1996b;104:207
214.
Holloway PJ, Mellanby M, Stewart RJC: Fruit
drinks and tooth erosion. Br Dent J 1958;104:
305309.
Imfeld T: Dental erosion: Definition, classification, and links. Eur J Oral Sci 1996;104:151
155.
Larsen MJ: Dissolution of enamel. Scand J Dent
Res 1973;81:518522.
Larsen MJ: Degrees of saturation with respect to
apatites in fruit juices and acidic drinks. Scand
J Dent Res 1975;83:1317.
Larsen MJ: An investigation of the theoretical
background for the stability of the calcium
phosphate salts and their mutual conversion in
aqueous solutions at pH 510. Arch Oral Biol
1986;31:757761.

Enamel Erosion by Soft Drinks and Orange

Juice

Larsen MJ, V d Fehr FR, Birkeland JM: Effect of

fluoride on the saturation of an acetate buffer
with respect to hydroxyapatite. Arch Oral Biol
1976;21:723728.
Larsen MJ, Jensen SJ: Stability and mutual conversion of enamel apatite and brushite at 20C as
a function of pH of the aqueous phase. Arch
Oral Biol 1989;34:963968.
Larsen MJ, Pearce EIF, Jensen SJ: Notes on the
dissolution of human dental enamel in dilute
acid solutions at high solid/solution ratio.
Caries Res 1993;27:8795.
Lussi A, Jggi T, Schrer S: The influence of different factors on in vitro enamel erosion.
Caries Res 1993;7:387393.
McCann HG: The solubility of fluorapatite and its
relationship to that of calcium fluoride. Arch
Oral Biol 1968;13:9871001.
McDowell H, Gregory TM, Brown WE: Solubility
of Ca5((PO)4)3OH in the system Ca(OH)2H3PO4-H2O at 5, 15, 25, and 35.5 C. J Res
Natn Bur Stand 1977;81A:773782.
Meurman JH, ten Cate JM: Pathogenesis and modifying factors of dental erosion. Eur J Oral Sci
1996;104:199206.
Meurman JH, Hrknen M, Nveri H, Koskinen J,
Torkko H, Rytmaa I, Jrvinen V, Turunen R:
Experimental sports drinks with minimal dental erosion effect. Scand J Dent Res 1990;98:
120128.

Caries Res 1999;33:8187

Meurman JH, Rytmaa I, Kari K, Mutomaa H:

Salivary pH and glucose after consuming various beverages, including sugar-containing
drinks. Caries Res 1987;21:353359.
Nunn JH: Prevalence of dental erosion and the implications for oral health. Eur J Oral Sci
1996;104:156161.
Rytmaa I, Meurman JH, Koskinen J, Laakso T,
Gharazi L, Turunen R: In vitro erosion of
bovine enamel caused by acidic drinks and
other foodstuffs. Scand J Dent Res 1988;96:
324333.
Sorvari R, Meurman JH, Alakuijala P, Frank RM:
Effect of fluoride varnish and solution on
enamel erosion in vitro. Caries Res 1994;28:
227232.
Spencer AJ, Ellis LM: The effect of fluoride and
grapefruit on the etching of teeth. J Nutr 1950;
43:107115.
Willis JB: Determination of calcium and magnesium in urine by atomic absorption spectroscopy. Anal Chem 1961;33:556559.
Zero DT: Etiology of dental erosion extrinsic
factors. Eur J Oral Sci 1996;104:162177.

87