Dental Erosion, Oral Hygiene, and Nutrition in

Eating Disorders

Alex Milosevic
David A. Brodie
Peter D. Slade
(Accepted 19 January 1996)

Objective: To determine the influence of oral hygiene practices and additional fluoride on
erosive tooth wear in eating disorders. The proportional dietary intake of carbohydrates, fats,
and proteins was also investigated. Method: Tooth wear was measured with the use of the
tooth wear index (TWI). All subjects completed a questionnaire on past dental history as well
as a 1-week diet sheet. A total of 33 subjects participated in the study, 20 of whom were
follow-ups, allowing the progress of dental erosion to be made. All subjects were referred
from the Department of Clinical Psychology. Results: Oral hygiene practices between sub-
jects with and without severe erosion were not significantly different. Only 8 bulimics spent
more time brushing after vomiting than at other times. The pH of vomitus from 6 subjects
ranged between 2.9 and 5.0, with a mean of 3.8, well below the critical pH for enamel
demineralization to occur. Of the 20 follow-up subjects, 12 (60%) exhibited worsening tooth
wear. The mean values for daily carbohydrate, protein, and fat intake were not significantly
different at baseline and at recall, and the proportional dietary intake was similar to recom-
mended energy provision at 47%, 40%, and 13%, respectively. Discussion: The contribution
by toothbrush abrasion to the overall wear in the eroded dentition of bulimics is not signifi-
cant. Therefore, immediate post-vomiting oral hygiene practices can be recommended. The
proportional nutritional intake values of carbohydrates, fats, and proteins in this group of
bulimics are acceptable. © 1997 by John Wiley & Sons, Inc. Int J Eat Disord 21: 195–199,
1997.

Dental erosion is a frequent sequela to self-induced vomiting in bulimia nervosa (Roberts

& Li, 1987; Milosevic & Slade, 1989; Robb, Smith, & Geidrys-Leeper, 1995). However, the
occurrence of erosion is not invariable since it was not linearly associated with the fre-
quency and duration of self-induced vomiting or the total number of vomiting episodes

Alex Milosevic, B.D.S., Ph.D., F.D.S., is Lecturer in the Academic Unit of Restorative Dentistry within the Depart-
ment of Clinical Dental Sciences, The University of Liverpool, where David A. Brodie, B.Ed., M.Sc., Ph.D., is
Professor of Movement Science and Physical Education and head of the department, and Peter D. Slade, B.A., Ph.D.,
M.Phil., former Professor of Clinical Psychology and head of the department. Address reprint requests to Dr. A.
Milosevic at The School of Dentistry, The University of Liverpool, Liverpool L69 3BX, United Kingdom.

International Journal of Eating Disorders, Vol. 21, No. 2, 195–199 (1997)

© 1997 by John Wiley & Sons, Inc. CCC 0276–3478/97/020195–05

Prod. #1162
 1098108x, 1997, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/(SICI)1098-108X(199703)21:2<195::AID-EAT11>3.0.CO;2-1 by <Shibboleth>-member@dundee.ac.uk, Wiley Online Library on [12/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
196 Milosevic, Brodie, and Slade

(Milosevic & Slade, 1989). Other factors or cofactors must be involved in the erosive
process. The influence of oral hygiene practices may be important since toothbrushing
immediately after an acid attack abraded away the softened, demineralized tooth surface
in vitro (Davis & Winter, 1980). Fluoride in toothpaste strengthens teeth against plaque
acids that cause dental decay as well as in vitro dental erosion (Bartlett, Smith, & Wilson,
1994). The use of fluoride supplements in childhood or having lived in an area with an
optimally fluoridated water supply would also be protective. The contribution of tooth-
brush and dentrifice abrasion to the eroded dentition in the eating disorders as well as
other aspects of oral hygiene have not been previously investigated. Furthermore, the
influence of the type of food that was ingested and subsequently vomited may have a
bearing on the acidity (pH) of regurgitated gastric contents, and therefore erosive poten-
tial, since the critical pH for enamel demineralization occurrence is 5.5 (Stephan, 1940).
It may be expected that the proportion of carbohydrates, fats, and proteins in the diet
of bulimics could differ from normal, healthy individuals. This study aimed to assess
dental history and oral hygiene practices as well as progress of dental erosion in a group
of vomiting bulimics. The nutritional intake was also investigated.

METHOD
Dental Assessment
Our previous work (Milosevic & Slade, 1989) assessed the dental status of 33 vomiting
bulimics, 7 nonvomiting bulimics, and 18 anorexics (N = 58) compared with 50 healthy
controls. Twenty of the original 58 eating disorders subjects agreed to take part in this
follow-up study, plus a further 13 new cases, of whom all were bulimics engaged in
self-induced vomiting. The subjects were categorized according to criteria outlined in the
3rd rev. ed. of the Diagnostic and statistical manual of mental disorders (DSM-III R; American
Psychiatric Association, 1987), as in the original study. All participants underwent a tooth
wear charting for erosion using the tooth wear index (TWI; Smith & Knight, 1984). For the
20 follow-up cases, the second erosion chart could identify progress of the erosion. A
two-section questionnaire on oral hygiene practices was designed. In the first section, the
questions asked about area of birth, use of fluoride supplements, and cleaning habits, and
the second section was specifically aimed at vomiting bulimics with questions asking
about oral hygiene practices with respect to the time of vomiting. The 20 follow-up
subjects were asked whether they would be prepared to bring a sample of their vomit for
pH analysis. Those subjects who agreed were instructed to note the time interval between
last food intake and sample collection. The sample was then stored in the home freezer.
The subjects brought the samples to a subsequent appointment when the pH of the
thawed vomitus was measured in an Orion pH meter (Orion Research, Inc., Boston, MA).
A 1-week food intake record from each subject allowed a nutritional analysis with the
aid of the ‘‘Microdiet’’ software (Microdiet System, Version 6.4, Salford University, UK).
Another dietary record was obtained at a recall visit, 1 month after the first visit.
Statistical significance was assessed by using chi-square analysis for categorical data
and Student’s paired t test for interval data.

RESULTS
A total of 33 subjects answered the questionnaire. Twenty-six were vomiting bulimics
(mean age 28.3 years), 2 were nonvomiting bulimics (mean age 23 years), and 5 were
 1098108x, 1997, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/(SICI)1098-108X(199703)21:2<195::AID-EAT11>3.0.CO;2-1 by <Shibboleth>-member@dundee.ac.uk, Wiley Online Library on [12/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Dental Erosion 197

anorexics (mean age 22.6 years); age differences between the groups were not significant.
Twenty-seven subjects were born in the northwest of England, which has a nonfluori-
dated water supply, of whom only 3 received fluoride supplements. These three subjects
did not have dental erosion. Twenty-two cases brushed twice daily with a 50/50 split
between the erosion absent and present groups. Most subjects brushed their teeth for a
time period between 1 and 3 min (67%), whereas 5 subjects brushed their teeth for longer
than 3 min. Nineteen bulimics brushed their teeth after vomiting, and 5 did not. However,
a further breakdown of this group into those with erosion present and with erosion absent
found no significant differences between the groups. Of the 15 subjects without dental
erosion, 12 claimed to have brushed their teeth immediately after vomiting.
Nine individuals brushed within 1 min of vomiting and 8 brushed up to 5 min after-
ward. Only 8 bulimics (33%) spent more time brushing after vomiting than at other times.
Seven subjects used a mouthrinse after vomiting with no significant differences between
those with and without tooth wear. Again, there were no differences between the tooth
wear groups in their response to the length of time between vomiting and brushing and
the length of time spent brushing.
Six subjects returned sample of frozen vomitus. None of these individuals exhibited
tooth wear. The mean pH was 3.8 (SD 0.89) with a range of 2.9–5.0 (Table 1). One subject
vomited 4 hr after eating, but the others ranged from 10 to 20 min.
Fifteen subjects took part in the nutritional analysis. The mean age of this group was
25.7 years (range 18–43), with a mean height of 161.3 cm (range 150–175) and a mean mass
of 55.4 kg (range 27–80). The mean values for the nutritional analyses at baseline and
recall are presented in Table 2. Student’s t test on each pair of values found nonsignificant
differences. Two individuals declined to attend for repeat assessment. When the average
efficiency of digestion is taken into account, the net kilocalorie values for carbohydrates,
fats, and proteins become 4.0, 9.0, and 4.0 kcal, respectively (Katch & McArdle, 1983, p.
81). To construct Table 2, the total gram intake of carbohydrate, fat, and protein at baseline
and recall was multiplied by the respective kilocalories value and the proportion of each
expressed as a percentage of energy load.

DISCUSSION

The analysis of oral hygiene practices does not shed light as to why some of the
vomiting bulimics have severe erosion and others do not. The upper palatal surfaces are
not readily brushable, but neither the pattern nor the time spent brushing nor mouthrinse

Table 1. The time interval between last food intake and

production of vomit with the respective pH values

Time Between Last Food Intake

Subject and Vomiting (min) pH of Vomitus

1 Not known 3.3

2 240 3.5
3 10 2.9
4 15 3.4
5 20 5.0
6 10 4.9

Note: N = 6, mean time = 59 min, mean pH = 3.8 (SD 0.89).

 1098108x, 1997, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/(SICI)1098-108X(199703)21:2<195::AID-EAT11>3.0.CO;2-1 by <Shibboleth>-member@dundee.ac.uk, Wiley Online Library on [12/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
198 Milosevic, Brodie, and Slade

Table 2. Proportional daily food intake with mean values for each food type at baseline
and recall

Baseline Recall

Food Type Na g kcal % Na g kcal % t p Value

Carbohydrates 15 208.5 834 47 13 278.0 1,112 48 1.4 NS

Fat 15 76.8 691 39 13 101.4 913 40 1.3 NS
Protein 15 61.3 245 14 13 70.0 280 12 .9 NS
Total N = 15 346.6 1,770 100 N = 13 449.4 2,305 100
a
N = number of subjects. Two subjects withdrew from the recall.

use was any different between those vomiting bulimics with pathological erosion and
those without. Most of the subjects (87.5%) used a soft or medium toothbrush, replacing
it between 3 and 6 months in 60% of cases. This is acceptable toothbrush usage.
The total individual pathological TWI score was obtained by adding the scores for each
of the 16 sets of tooth surfaces (Smith & Knight, 1984). The total pathological TWI score
increased in 5 (38.5%) vomiting bulimics, in the 2 nonvomiting bulimic cases, and in 4 of
the 5 anorexics (80%). Four vomiting bulimics exhibited a decrease in pathological TWI
score because having crossed into an older age group, the acceptable threshold TWI levels
changed, rendering fewer surfaces pathological.
Seven vomiting bulimics had more surfaces with exposed dentine at recall as did both
nonvomiting bulimics and 3 anorexics. One subject in the vomiting bulimic group exhib-
ited a reduction in pathological TWI score from 5.29 to 1.84 with 16 fewer surfaces with
exposed dentine because she had virtually all her upper teeth crowned as a result of
erosion. Restored surfaces are not scored. The erosion did progress in 12 of the 20 follow-
up subjects (60%). This increase in erosion occurred mainly on the upper and lower incisal
edges with loss of the enamel and consequent exposure of the underlying dentine. How-
ever, this was not considered to be pathological in any of the subjects according to the age
criteria of the TWI.
The six vomitus pH values were all below the critical pH for enamel demineralization.
However, none of the 6 subjects exhibited pathological erosion which is supportive of
previous results reporting poor correlation between vomiting variables and erosion.
On both baseline and recall occasions it can be seen from Table 2 that the percentage of
calories attributed to each major foodstuff is 47% for carbohydrates, 40% for fats, and 13%
for proteins. This can be compared with values reported by Gregory, Foster, Tyler, and
Wiseman (1990) which quote 43%, 38%, and 19% respectively, showing a slightly lower
value for both carbohydrates and fats and a commensurate higher value for proteins.
Interestingly, the dietary reference value recommended by the Department of Health
(1991) is a reduction for energy source from fat to 33%, leaving protein levels at about the
same as those in subjects in this study. The World Health Organization (1990) has even
more extreme population goals, suggesting that the carbohydrates should make up be-
tween 50 and 70% of dietary energy and fats should make up between 15 and 30% of the
energy load. It would appear, in summary, that this sample does not differ markedly from
normal populations in terms of the proportions of fat, carbohydrate, and protein in their
diet.

REFERENCES
American Psychiatric Association. (1987). Diagnostic and statistical manual of mental disorders (3rd Rev. ed.).
Washington, DC: Author.
 1098108x, 1997, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/(SICI)1098-108X(199703)21:2<195::AID-EAT11>3.0.CO;2-1 by <Shibboleth>-member@dundee.ac.uk, Wiley Online Library on [12/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Dental Erosion 199

Bartlett, D. W., Smith, B. G. N., & Wilson, R. F. (1994). Comparison of the effect of fluoride and non-fluoride
toothpaste on tooth wear in vitro and the influence of enamel fluoride concentration and hardness of enamel.
British Dental Journal, 176, 346–348.
Davis, W. B., & Winter, P. J. (1980). The effect of abrasion on enamel and dentine after exposure to dietary acid.
British Dental Journal, 148, 253–256.
Department of Health. (1991). The health of the nation. London: HMSO.
Gregory, J., Foster, K., Tyler, H., & Wiseman, M. (1990). The dietary and nutrition survey of adults. London: HMSO.
Katch, F. C., & McArdle, W. D. (1993). Nutrition, weight control and exercise (2nd ed.). Philadelphia: Lea & Febiger.
Milosevic, A., & Slade, P. D. (1989). The orodental status of anorexics and bulimics. British Dental Journal, 167,
66–70.
Robb, N. D., Smith, B. G. N., & Geidrys-Leeper, E. (1995). The distribution of erosion in the dentitions of patients
with eating disorders. British Dental Journal, 178, 171–175.
Roberts, M. W., & Li, S.-H. (1987). Oral findings in anorexia nervosa and bulimia nervosa: A study of 47 cases.
Journal of the American Dental Association, 115, 407–410.
Smith, B. G. N., & Knight, J. K. (1984). An index for measuring the wear of teeth. British Dental Journal, 156,
435–438.
Stephan, R. M. (1940). Changes in the hydrogen ion concentration on tooth surfaces and in carious lesions.
Journal of the American Dental Association, 27, 718–723.
World Health Organization. (1990). Diet, nutrition and the prevention of chronic diseases. (WHO Tech. Rep. No. 797).
Geneva: Author.