Original Paper

Caries Res 2008;42:98–104 Received: January 12, 2007

Accepted after revision: December 13, 2007
DOI: 10.1159/000116118
Published online: February 15, 2008

Influence of Beverage Composition on the

Results of Erosive Potential Measurement by
Different Measurement Techniques
D.H.J. Jager A.M. Vieira J.L. Ruben M.C.D.N.J.M. Huysmans
Academic Center for Oral Health, University Medical Center Groningen, Groningen, The Netherlands

Key Words the erosive potential with chemical analyses. Drink com-
Acidic beverages ⴢ Calcium analysis ⴢ Erosion ⴢ Inorganic position also influenced the effect of small versus large ex-
phosphorus analysis ⴢ Mineral composition ⴢ Optical posure volumes, indicating the need for standardization of
profilometry exposure parameters. Copyright © 2008 S. Karger AG, Basel

Abstract
The influence of beverage composition on the measure- Dental erosion is defined as an irreversible loss of den-
ment of erosive potential is unclear. The aim of this study was tal hard tissue due to a chemical process without involve-
to evaluate whether beverage composition influences the ment of microorganisms [Imfeld, 1996]. Dental erosion
measurement of erosive potential and to evaluate the influ- may be caused by either extrinsic or intrinsic factors. One
ence of exposure in small and large volumes. Eleven bever- of the extrinsic causes of dental erosion is excessive con-
ages were included: water (control), 3 alcopops, 2 beers and sumption of acidic beverages [Dugmore and Rock, 2004].
5 soft drinks. For each beverage 15 bovine enamel samples Different techniques are available to assess the erosive
were used: 5 for chemical and 10 for profilometric analysis. potential of acidic beverages. Frequently used techniques
After exposure to the beverages (63 min) the resulting solu- include profilometry and chemical analysis.
tions were analyzed for Ca and inorganic phosphorus (Pi) Calcium determination and inorganic phosphorus
content. The samples for optical profilometry were sub- (Pi) determination are used to measure the loss of miner-
mersed sequentially in 500 ml or in 1 ml of the drinks for 3, als from the enamel and are well established and accurate
6, 9, 15 and 30 min (total 63 min). For some of the beverages [Barbour and Rees, 2004]. Small concentrations of ions
high baseline concentrations of Ca (energy drink) or Pi (cola released from the enamel can be measured, so it is pos-
drink, cola lemon drink, beer, beer lemon) were found. Some sible to observe the initial stages of erosion and to use
of the beverages showed a good correlation between the small volumes of the examined solutions. It also allows
chemical methods. Profilometry (both for 1 and 500 ml) the use of natural tooth surfaces since polishing is not
showed generally lower enamel losses than the chemical needed [Barbour and Rees, 2004]. However, chemical
methods. Lower enamel losses were found for the profilom- methods provide only quantitative and not morphologi-
etry 1 ml compared to the profilometry 500 ml only for the cal or mechanical data [Grenby, 1996].
cola drinks. It can be concluded that the composition of the For measuring the loss of surface layers the method of
beverages had a significant effect on the determination of choice is optical or contact profilometry. In optical pro-

© 2008 S. Karger AG, Basel Derk-Jan Jager

0008–6568/08/0422–0098$24.50/0 Academic Center for Oral Health, University Medical Center Groningen
Fax +41 61 306 12 34 Antonius Deusinglaan 1
E-Mail karger@[Link] Accessible online at: NL–9713 AV Groningen (The Netherlands)
[Link] [Link]/cre Tel. +31 50 3633 116, Fax +31 50 3632 696, E-Mail [Link]@[Link]
Table 1. Beverages used in this study, with details of their composition

Beverage Producer pH Ca, mmol/l Pi, mmol/l DS (HA)

Cola drink Coca Cola Coca-Cola Enterprises Nederland BV, 2.47 0.8780.04 4.7680.15 0.005
Dongen, The Netherlands
Cola lemon drink Coca Cola light Coca-Cola Enterprises Nederland BV, 2.73 0.7380.01 4.9080.06 0.008
lemon Dongen, The Netherlands
Orange drink Fanta orange Coca-Cola Enterprises Nederland BV, 3.03 0.0680.01 0.1980.01 0.001
Dongen, The Netherlands
Fruit drink Dubbel friss orange/ Riedel Beverages, Ede, The Netherlands 3.35 1.3080.03 0.5180.03 0.018
pink grapefruit
Vodka alcopop Smirnoff Ice Diageo, London, UK 3.43 0.1580.005 0.00480.001 <0.001
Energy drink Red Bull Red Bull, de Bilt, The Netherlands 3.43 2.4080.21 0.0180.02 <0.001
Ice tea Lipton Ice tea Unilever, Rotterdam, The Netherlands 3.80 0.1280.01 0.2580.004 0.009
Beer lemon Grolsch beer lemon Inbev Nederland, Breda, The Netherlands 3.83 0.9680.02 3.5180.08 0.068
Rum lime alcopop Breezer Lime Bacardi Martini NV, Gouda, The Netherl. 3.87 0.1780.01 0.0280.001 <0.001
Beer Bavaria beer Bavaria NV, Lieshout, The Netherlands 4.20 0.7280.02 5.3080.14 0.125
Bottled water Sourcy bottled water Vrumona BV, Bunnik, The Netherlands 8.09 1.2080.03 n.m. –

DS = Degree of saturation; HA = hydroxyapatite; n.m. = not measured.

filometry there is no physical contact between the probe zerland) and cut into blocks of approximately 5 ! 3 mm using
and the surface, so no damage will occur by scratching of a vertical sawing machine with a diamond saw blade (11-4243,
Buehler, Düsseldorf, Germany). The blocks were embedded in
the softened surface [Barbour and Rees, 2004]. A draw- acrylic resin (Autoplast polymer, Candulor AG, Wangen, Switzer-
back of profilometric techniques is that enamel losses be- land) leaving the enamel surface uncovered and subsequently the
low 2 ␮m are difficult to measure. samples were polished flat (800–1,200 grit grinding paper) and
Little is known about the correlation between different thoroughly rinsed with tap water. The samples were randomly
methods and the lack of a ‘gold standard’ is a shortcom- divided into three groups of 55 samples each: one group for chem-
ical analysis and two groups for profilometric analysis.
ing in the field of erosion research. Moreover, the influ- Before inclusion in the experiment the area of exposure of each
ence of the composition of the beverages on the measure- of the 55 samples used for chemical analyses was measured with
ments is unclear. In earlier studies comparing methods a stereomicroscope equipped with a measuring grid (Leitz Duri-
for measuring the erosive potential of beverages, stan- met, Wetzlar, Germany) fitted out with a digital xy table (Sony
dard solutions such as citric acid or lactate buffer were Magnescale LY101, Tokyo, Japan).
The 110 samples used for the profilometric analysis were part-
used [Ganss et al., 2005; Zero et al., 1990]. As a result, the ly covered with PVC tape exposing an area of approximately 3 !
influence of the composition of beverages on the results 3 mm in the center of the enamel sample.
obtained by the different methods could not be deter-
mined. We hypothesized that the chemical composition Beverages
Eleven beverages, all available in the Netherlands, were in-
of soft drinks and the volume used influence the deter-
cluded in this study (table 1). Immediately after opening the bot-
mination of erosive potential. tles, and also after degassing (the drinks were placed on a shaking
The aim of this study was to evaluate whether bever- table set at 200 rpm until no bubbles were visible), the pH was
age composition influences the measurement of erosive measured 5 times using a calibrated glass pH electrode (Radiom-
potential and to evaluate the influence of exposure in eter, PHM 84 Research Meter, G202C, Copenhagen, Denmark) in
100 ml of the beverages. Room temperature in the laboratory was
small and large volumes.
21 ° C with a possible variation of 82 ° C. Standard buffers, pH 7.01
and 4.00 (20 ° C), were used (measurement uncertainty for both
80.015 units) (Merck KGaA, Darmstadt, Germany). Calibration
Materials and Methods was performed with these buffers every day.

Preparation of Samples Demineralization Procedures

A total of 165 buccal surfaces of extracted bovine incisors, Before starting the demineralization procedure the samples
stored in water, were ground flat with water-cooled silicon car- for chemical analysis were submersed for 3 min in 3 ml of a stan-
bide 220 grit grinding discs (SIA siawat P220, Frauenfeld, Swit- dard solution of 50 mM citric acid, 0.4 mM KH2PO4, 0.4 mM CaCl2

Influence of Beverage Composition on Caries Res 2008;42:98–104 99

Erosive Potential Measurement
and 1 mM NaN3 (pH = 3) to remove the smear layer from the pol- equipment’s software. Erosion depth (␮m) was calculated by di-
ished surfaces and subsequently rinsed with tap water. All the viding the volume loss by the exposed enamel area of the scanned
beverages were decarbonated. surface. The erosion depths of the 5 samples were averaged. The
For the chemical analyses each of the 5 enamel samples was profilometry resulted in two further depth parameters: d(prof1)
submersed in 1 ml of each beverage in a test tube for 3 min under and d(prof500).
constant agitation on a shaking table at 100 rpm. After 3 min the
samples were lifted from the beverages and the enamel surface pH Changes and Degree of Saturation
was rinsed with 3 ml of demineralized water which was collected The pH of the solutions after each exposure in the profilom-
in the test tube. From the resulting mixtures 1 ml was used for Ca etry (1-ml) group was measured. The baseline degree of satura-
analysis and 1 ml for Pi analysis. The same procedure was repeat- tion of the beverage with regard to hydroxyapatite and dicalcium
ed for exposures of 6, 9, 15 and 30 min (total 63 min). The expo- phosphate dihydrate (DCPD) was calculated by means of a com-
sures for the different times were made sequentially on the same puter program [Shellis, 1988], using the baseline Ca and Pi con-
specimens. centrations of the beverages, together with the pH measured after
For the profilometric analysis each of the 5 enamel samples degassing. To determine the possible influence of saturation of
was submersed in 1 ml of each beverage in a test tube for 3, 6, 9, the beverages on the measurement results during the erosion pro-
15 and 30 min under constant agitation on a shaking table (100 cess, the Ca and Pi concentrations and pH after the 30-min incu-
rpm). The pH of these solutions was measured after each expo- bation were used to calculate the change in degree of saturation
sure. Another set of 5 samples was submersed in 500 ml of each with regard to hydroxyapatite and DCPD after the erosion re-
beverage for 63 min under constant agitation on a shaking table gime.
(100 rpm) in beakers with a diameter of 9.5 cm. All experiments
were performed at room temperature (21 8 2 ° C). Statistical Analysis
For investigation of the relationship between the change in Ca
Chemical Analysis and Pi concentrations, linear least squares regression was per-
Pi concentration in the beverages was determined using a formed. The Pi concentration was the independent (x) variable. A
phosphomolybdate spectrophotometric method [Chen et al., one-way ANOVA followed by a Bonferroni post-hoc test in SPSS
1956]. The concentration of Ca in the beverages was determined 12.01 (SPSS , Chicago, Ill., USA) was used to test differences be-
by atomic absorption spectroscopy (PerkinElmer Analytical In- tween the cumulative erosive depths at 63 min obtained by the
struments, Shelton, Conn., USA) [Vieira et al., 2005]. This was chemical methods [average of d(Ca) and d(P): d(CaP), d(prof1)
performed in the presence of lanthanum (0.326%) in order to sup- and the d(prof500)]. The significance level for all tests was set at
press phosphate interference. An air/C2H2 flame and a wave- 0.05.
length of 422.7 nm were used.
For the chemical analyses all the beverages had to be diluted
with demineralized water. For Pi analysis most of the beverages
were diluted 16 times in total. The beverages with high Pi concen- Results
trations (the colas and the beers) had to be diluted 80 times. For
Ca analysis all the beverages were diluted 18.4 times. The pH of the beverages ranged from 2.4 (cola) to 8.1
The Ca and Pi losses from the enamel samples were deter- (bottled water) (table 1). Table 1 also shows the baseline
mined by subtracting the Ca or Pi content of the beverages before Ca and Pi concentrations and table 2 shows the changes
the enamel exposure (average of 10 measurements) from the total
Ca or Pi content of the solution after exposure. In addition, the in Ca and Pi concentrations for all erosion times and all
ratio of the Ca dissolved to the Pi dissolved (⌬Ca/⌬P) was calcu- drinks. Pi concentration ranged from not detectable (bot-
lated for each exposure time. tled water) to 5.3 mmol/l (beer). Baseline Ca concentra-
Lesion depth was calculated from the Ca and Pi loss using the tion ranged from 0.06 mmol/l (orange soft drink) to 1.3
average Ca and Pi content per unit volume for bovine enamel and mmol/l (fruit drink). For most of the drinks the ⌬Ca/⌬P
the exposed enamel area [Dijkman et al., 1983]. A Ca concentra-
tion in enamel of 25.1%, a P concentration in enamel of 17.61% ratio did not differ significantly from 1.6 except for some
were assumed and an average enamel density of 2.93 g/cm3 was of the low exposure times (3 and 6 min), and for the cola
assumed. This resulted in two depth parameters: d(Ca) and d(P), drink, orange soft drink, and the ice tea. In table 3, the
lesion depth estimated from Ca loss or Pi loss, respectively. The parameters for the linear least squares regression analysis
estimated erosion depth (␮m) of the 5 samples was averaged. of the Ca and the Pi concentrations for all beverages are
Profilometric Analysis presented. In most cases a high linear correlation (r2 1
Erosion depths were measured using an optical profilometer 0.8) was found, except for the beers (r2 = 0.07 and r2 =
(Proscan 2000, Scantron, Taunton, England). Before inclusion of 0.19), cola drink (r2 = 0.76), energy drink (r2 = 0.63) and
the enamel samples in the experiment, baseline measurements cola lemon drink (r2 = 0.53). For this reason and because
were performed on each sample in order to confirm the flatness of the problems measuring the Pi concentration in drinks
of the polished enamel surfaces.
After the demineralization procedure the PVC tape was re- with a high baseline Pi concentration, d(Ca) of the beers,
moved. The samples were scanned over the reference and eroded cola drink and cola lemon was used for the comparison
surfaces. The volume lost due to erosion was calculated with the with profilometry. In table 4 and in figure 1, the cumula-

100 Caries Res 2008;42:98–104 Jager/Vieira/Ruben/Huysmans

Table 2. Changes in calcium (Ca) and inorganic phosphorus (Pi) concentrations (mmol/l) of drinks after each exposure, together with
the ratio of the changes in Ca and Pi (⌬Ca/⌬Pi)

3 min 6 min 9 min 15 min 30 min DS

(HA)
Ca Pi Ca/Pi Ca Pi Ca/Pi Ca Pi Ca/Pi Ca Pi Ca/Pi Ca Pi Ca/Pi

Cola drink 0.15 0.15 1.16 0.26 0.26 1.22 0.32 0.33 1.89 0.42 0.61 0.70 0.61 0.87 0.71 0.009
(0.02) (0.05) (0.36) (0.02) (0.09) (0.84) (0.05) (0.26) (1.49) (0.05) (0.12) (0.13) (0.05) (0.06) (0.05)
Cola lemon 0.17 0.26 0.71 0.25 0.25 1.81 0.33 0.20 1.97 0.47 0.17 3.16 0.85 0.74 1.21 0.001
drink (0.03) (0.09) (0.32) (0.02) (0.15) (1.85) (0.03) (0.10) (0.91) (0.02) (0.07) (1.17) (0.02) (0.18) (0.31)
Orange 0.13 0.05 2.47 0.14 0.10 1.37 0.19 0.15 1.30 0.26 0.26 1.03 0.48 0.53 0.90 0.007
soft drink (0.02) (0.01) (0.56) (0.03) (0.02) (0.15) (0.05) (0.04) (0.12) (0.05) (0.06) (0.03) (0.10) (0.12) (0.04)
Fruit drink 0.17 0.07 2.67 0.29 0.15 2.06 0.37 0.17 2.17 0.46 0.26 1.76 0.94 0.50 1.90 0.029
(0.04) (0.03) (0.59) (0.05) (0.03) (0.66) (0.04) (0.03) (0.28) (0.07) (0.04) (0.12) (0.15) (0.09) (0.32)
Vodka 0.13 0.06 2.14 0.21 0.11 1.97 0.28 0.15 1.84 0.36 0.22 1.66 0.71 0.46 1.54 0.017
alcopop (0.02) (0.02) (0.33) (0.02) (0.01) (0.17) (0.04) (0.02) (0.07) (0.04) (0.03) (0.06) (0.14) (0.09) (0.03)
Energy –0.06 0.05 –1.18 0.00 0.02 –0.07 0.17 0.15 1.06 0.47 0.22 2.15 0.83 0.46 1.83 0.032
drink (0.05) (0.01) (0.97) (0.10) (0.10) (1.25) (0.20) (0.04) (1.38) (0.15) (0.05) (0.69) (0.11) (0.09) (0.19)
Ice tea 0.11 0.04 2.48 0.15 0.06 2.34 0.20 0.09 2.19 0.21 0.12 1.74 0.37 0.28 1.32 0.024
(0.01) (0.001) (0.23) (0.01) (0.01) (0.29) (0.02) (0.01) (0.09) (0.03) (0.02) (0.09) (0.12) (0.02) (0.40)
Beer 0.09 –0.02 1.75 0.07 –0.14 2.48 0.11 –0.02 –0.82 0.12 –0.17 0.28 0.21 0.03 2.84 0.077
lemon (0.04) (0.10) (4.12) (0.04) (0.22) (4.23) (0.04) (0.10) (3.23) (0.04) (0.17) (2.54) (0.04) (0.19) (4.08)
Rum lime 0.13 0.05 2.78 0.17 0.07 2.53 0.21 0.09 2.33 0.28 0.15 1.81 0.48 0.28 1.72 0.028
alcopop (0.02) (0.01) (0.32) (0.02) (0.02) (0.41) (0.06) (0.03) (0.33) (0.07) (0.03) (0.14) (0.10) (0.07) (0.10)
Beer 0.10 –0.09 –4.12 0.09 –0.08 –1.26 0.10 –0.02 1.80 0.05 –0.13 –0.17 0.11 0.27 0.51 0.146
(0.02) (0.07) (7.07) (0.01) (0.13) (2.42) (0.02) (0.10) (3.29) (0.06) (0.07) (0.88) (0.03) (0.13) (0.42)
Bottled –0.02 n.m. n.m. –0.01 n.m. n.m. –0.03 n.m. n.m. –0.01 n.m. n.m. –0.02 n.m. n.m. –
water (0.01) (0.02) (0.02) (0.04) (0.03)

Means with SD in parentheses. The degree of saturation (DS) with respect to hydroxyapatite (HA) after the final 30-min exposure is also given.
n.m. = Not measured.

tive results of Ca, Pi and the profilometric analyses are Table 3. Parameters for the linear least squares regressions of Ca
presented. The highest enamel loss was found for cola concentration on Pi concentration for all beverages
lemon drink in the d(prof500) group (13.54 ␮m). The
Slope y intercept r2
drinks concentrating in the middle part of the graph
(dashed lines) showed lower erosive potential for pro- Cola drink 0.47 0.14 0.75
filometry compared to the chemical analysis. The two Cola drink lemon 0.75 0.17 0.53
colas showed lower erosive potential in the d(prof1) group Orange soft drink 0.76 0.07 0.98
compared to the d(prof500) group and higher erosive po- Fruit drink 1.69 0.06 0.91
Vodka alcopop 1.45 0.05 0.99
tential for the d(Pi) compared to the d(Ca). In figure 1,
Energy drink 2.17 0.14 0.86
also the rank order in which the different methods placed Ice tea 1.07 0.08 0.80
the drinks can be assessed. For the beverages the influ- Beer lemon 0.10 0.44 0.19
ence of the measurement method on its rank order in Rum lime alcopop 1.49 0.06 0.97
erosiveness (1 is lowest, 11 is highest erosion) is marked, Beer 0.05 0.09 0.07
Bottled water 6.94 0.09 0.18
e.g., the orange soft drink is the 7th most erosive drink in
d(Ca) but the 4th most erosive with d(Pi); similarly, ice The Pi concentration was the independent (x) variable.
tea is 8th with d(Ca) and 4th with d(prof500). One-way
ANOVA showed a significant effect of measuring tech-
nique (p ! 0.05) for all beverages except ice tea and the
fruit drink. d(CaP) showed an enamel loss that was sig- the d(prof500) the d(CaP) showed a significantly lower
nificantly higher (p ! 0.05) than that of d(prof1) for the enamel loss for the cola lemon drink (p = 0.004), and a
rum lime alcopop (p ! 0.0001), energy drink (p = 0.007), significantly higher enamel loss was found for rum lime
vodka alcopop (p = 0.034), beer [p ! 0.0001, d(Ca) only] alcopop (p ! 0.0001), energy drink (p = 0.022), vodka al-
and orange soft drink (p ! 0.0001). When compared to copop (p = 0.034), beer (p ! 0.0001), beer lemon (p =

Influence of Beverage Composition on Caries Res 2008;42:98–104 101

Erosive Potential Measurement
 Table 4. Cumulative loss of enamel after 63-min total exposure to the beverages

d(Ca) d(Pi) d(prof1) d(prof500)

Cola drink 4.4480.22 9.2281.25 2.0880.58 8.0483.62

Cola lemon drink 6.7280.36 8.9781.75 6.4281.15 13.5484.31
Orange soft drink 3.6480.22 5.5080.38 2.2980.88 2.3780.51
Fruit drink 6.5580.53 5.6780.72 4.2482.53 3.2781.17
Vodka alcopop 5.0080.85 4.9480.88 2.9881.06 2.6981.35
Energy drink 4.2580.96 4.8480.18 2.3480.85 2.6980.96
Ice tea 3.2380.48 3.1180.33 1.8081.34 3.0880.63
Beer lemon 1.9980.32 –1.7982.02 1.1280.98 0.00
Rum lime alcopop 4.0980.31 3.4380.29 0.8480.70 1.4780.72
Beer 1.3080.23 –0.3881.16 0.00 0.00
Bottled water –0.3380.92 0.0180.01 0.00 0.00

Means with SD.

0.001) and orange soft drink (p = 0.001). The d(prof1) The Pi analysis of the beers and the Ca analysis of the
showed a significantly lower enamel loss than the energy drink yielded negative values. For these beverages
d(prof500) only for the cola drink (p = 0.002) and the the differences between the Ca analysis and the Pi analy-
cola lemon drink (p = 0.003). sis and the negative values may possibly be explained by
The results obtained for the pH measurements after their chemical composition. Because of the high baseline
each exposure in the 1-ml profilometry showed very little concentrations of Ca and Pi it was sometimes necessary
change in pH (–0.02 to +0.03) after the erosion process to use high dilutions which may have increased the mea-
for most beverages. Only the cola and the cola lemon surement error. A high concentration of Pi in the colas
showed a small rise in pH (0.1) after 30 min. None of the and the beers could also have interfered with the Ca mea-
beverages was supersaturated with respect to hydroxy- surements by calcium binding, but this was prevented by
apatite or DCPD after a 30-min erosive exposure in 1 ml adding a high concentration of lanthanum. However, the
of the samples (table 2). The highest degree of saturation amount of Ca and Pi released from the enamel into solu-
for hydroxyapatite was found for the beers. The highest tion by erosion was for some beverages relatively small
rise in the degree of saturation after 30 min was found for compared to the baseline concentration of Ca or Pi in the
the energy drink. beverages, especially for the short exposure times. This
resulted in small changes in mineral concentration, which
did not always exceed the measurement error and made
Discussion it difficult to obtain reliable measurements.
Previous studies, using standard erosive solutions
In this study bovine enamel was used. Most in vitro such as citric acid or lactate buffer reported almost per-
studies use bovine enamel since it is considered a suitable fect agreement between the Ca and Pi analyses [Ganss et
substitute for human enamel [Zero, 1996]. Although al., 2005; Zero et al., 1990]. Because of the use of standard
Meurman and Frank [1991] did not observe any differ- solutions in these studies the influence of the composi-
ence in the progression of erosion or the surface ultra- tion of the erosive solution on the results could not be
structure of erosive lesions between bovine and human determined. In the present study, for three of the drinks
prismatic enamel, another study showed that morpholog- a Ca/P ratio (table 2) with significant deviations from 1.6
ical differences such as a higher porosity exist when com- (the Ca/P ratio of bovine enamel) was found. This sug-
pared to human enamel, which results in higher rates of gests that reprecipitation occurred during the erosive
artificial caries lesion formation [Featherstone and Mell- exposure. However, in none of the drinks could a super-
berg, 1981]. It should be considered that in this study a saturation with respect to other calcium phosphates be
comparison between the methods was performed and not calculated. No explanation could be found for the phe-
an extrapolation of the results to the clinical situation. nomenon which, especially for the orange drink, does

102 Caries Res 2008;42:98–104 Jager/Vieira/Ruben/Huysmans

 seem to be systematic, looking at the decline of the Ca/P
ratio with increasing exposure time. The short exposures
and the drinks with low erosive potential were found to
be irrelevant for the calculation of the Ca/P ratio because
of the low enamel losses.
Profilometric analysis showed a trend for lower enam-
el loss compared to Ca and Pi analysis. These results are
in agreement with the findings of a previous study where
Ca/Pi analysis and contact profilometry were compared
[Ganss et al., 2005]. The difference found between the
chemical methods and the profilometry may be explained
by the fact that the erosion process does not only remove
enamel layers but also causes a ‘softening’. Profilometry
does not account for the subsurface loss of the softened
layer. The depth of the softened layer is unknown but may
be more than 10 ␮m [Eisenburger et al., 2004].
Removal of reaction products and the supply of fresh
acid are important for the continued formation of erosion
lesions [Eisenburger and Addy, 2003]. In a study investi-
gating the relationship between enamel erosion and liq-
uid flow rate, it was concluded that the rate of erosion is
dependent on liquid velocity, exposure time and the total
volume of the acidic solution [Shellis et al., 2005]. In this
study agitation of 1 ml probably results in a different re-
placement of liquid at the enamel surface compared to
agitation of a 500-ml reservoir, thus influencing the ero-
sion rate. However, regarding the exposure volume of the
beverage, our study found significant differences be-
tween the measured erosive potential only for the two
cola drinks.
Some authors have observed that a small change in the
degree of saturation resulted in a difference in the disso-
lution of enamel and that it is an important parameter
that defines the ability of a solution to demineralize
enamel [Barbour et al., 2003; Finke et al., 2000; Margolis
et al., 1999; Tanaka and Kadoma, 2000]. Only in a small
volume of the beverages would a significant rise in satu-
ration be expected. To ascertain whether this was the case
in the present study, the saturation with respect to hy-
droxyapatite was determined for all 1-ml volumes before
Fig. 1. Cumulative loss of enamel as measured by the four tech-
niques, showing both the quantitative loss and the rank order for
and after exposure. It was found that the degree of satu-
each beverage. To facilitate comparison between the techniques ration rose with increasing exposure time. However, this
for each beverage, the points have been connected. The plot area was seen in all beverages and the highest rise was found
is divided into three areas: little or no erosion (–2 to 2 ␮m), mod- for the energy drink. This did not explain our findings as
erate erosion (2–6 ␮m), and severe erosion (6–14 ␮m). only the cola drinks showed a lower erosion in the 1-ml
exposure.
Although pH is a parameter in the calculation of the
degree of saturation, it has been reported as a separate
factor in erosive potential [Larsen and Nyvad, 1999; Mar-
golis et al., 1999]. A rise in pH would result in a slowing

Influence of Beverage Composition on Caries Res 2008;42:98–104 103

Erosive Potential Measurement
down of the erosion process. Only for the cola drink and appropriate measurement method. Optical profilometry
the cola lemon drink a measurable, if low (0.1), rise in pH is suggested as a beverage-independent alternative. Bev-
was found. As this corresponds to the observed reduced erage composition also influences the effect of small ver-
erosion for the two cola drinks in the 1-ml exposure mod- sus large exposure volumes, indicating the need for stan-
el, we assume that pH was a determining factor. dardization of exposure parameters such as exposure
In conclusion, the present study has shown that the times, volumes and flow rate of the drinks during expo-
composition of the beverages had a significant effect on sure to prevent differences in erosion rate due to differ-
the determination of the erosive potential with chemical ences in liquid velocities.
analyses. This should be considered when choosing an

References

Barbour ME, Parker DM, Jandt KD: Enamel dis- Featherstone JD, Mellberg JR: Relative rates of Shellis RP: A microcomputer program to evalu-
solution as a function of solution degree of progress of artificial carious lesions in bo- ate the saturation of complex solutions with
saturation with respect to hydroxyapatite: a vine, ovine and human enamel. Caries Res respect to biominerals. Comput Appl Biosci
nanoindentation study. J Colloid Interface 1981;15:109–114. 1988;4:373–379.
Sci 2003;265:9–14. Finke M, Jandt KD, Parker DM: The early stages Shellis RP, Finke M, Eisenburger M, Parker DM,
Barbour ME, Rees JS: The laboratory assessment of native enamel dissolution studied with Addy M: Relationship between enamel ero-
of enamel erosion: a review. J Dent 2004; 32: atomic force microscopy. J Colloid Interface sion and liquid flow rate. Eur J Oral Sci 2005;
591–602. Sci 2000;232:156–164. 113:232–238.
Chen PS, Toribara TY, Warner H: Microdeter- Ganss C, Lussi A, Klimek J: Comparison of cal- Tanaka M, Kadoma Y: Comparative reduction
mination of phosphorus. Anal Chem 1956; cium/phosphorus analysis, longitudinal mi- of enamel demineralization by calcium and
28:1756–1758. croradiography and profilometry for the phosphate in vitro. Caries Res 2000; 34:241–
Dijkman AG, Tak J, Jongebloed WL, Arends J: quantitative assessment of erosive deminer- 245.
Influence of HClO4 strength and etching alisation. Caries Res 2005;39:178–184. Vieira A, Ruben JL, Huysmans MC: Effect of ti-
time on rate of etching and surface rough- Grenby TH: Methods of assessing erosion and tanium tetrafluoride, amine fluoride and
ness of human enamel. Caries Res 1983; 17: erosive potential. Eur J Oral Sci 1996; 104: fluoride varnish on enamel erosion in vitro.
14–22. 207–214. Caries Res 2005;39:371–379.
Dugmore CR, Rock WP: A multifactorial analy- Imfeld T: Dental erosion: definition, classifica- Zero DT: Etiology of dental erosion – extrinsic
sis of factors associated with dental erosion. tion and links. Eur J Oral Sci 1996; 104: 151– factors. Eur J Oral Sci 1996;104:162–177.
Br Dent J 2004;196:283–286. 155. Zero DT, Rahbek I, Fu J, Proskin HM, Feather-
Eisenburger M, Addy M: Influence of liquid tem- Larsen MJ, Nyvad B: Enamel erosion by some stone JD: Comparison of the iodide perme-
perature and flow rate on enamel erosion soft drinks and orange juices relative to their ability test, the surface microhardness test,
and surface softening. J Oral Rehabil 2003; pH, buffering effect and contents of calcium and mineral dissolution of bovine enamel
30:1076–1080. phosphate. Caries Res 1999; 33:81–87. following acid challenge. Caries Res 1990;24:
Eisenburger M, Shellis RP, Addy M: Scanning Margolis HC, Zhang YP, Lee CY, Kent RL Jr, 181–188.
electron microscopy of softened enamel. Moreno EC: Kinetics of enamel demineral-
Caries Res 2004;38:67–74. ization in vitro. J Dent Res 1999; 78: 1326–
1335.
Meurman JH, Frank RM: Progression and sur-
face ultrastructure of in vitro caused erosive
lesions in human and bovine enamel. Caries
Res 1991;25:81–87.

104 Caries Res 2008;42:98–104 Jager/Vieira/Ruben/Huysmans