This article was originally published in and reproduced here with the permission of, Dental Digest, which is provided as a service to health professionals through an educational grant from the Sugar Bureau.
Sugars, Starches and Dental caries : frequency versus quantity
Over the past hundred years, there has been a continued debate over the relationship of sugars, and to a lesser extent starches, to dental caries and their degree of importance. Is the quantity of sugar used each day more important than frequency of use?
Miller, at the end of the 19th century, in his seminal work that produced the acidogenic theory of dental caries, demonstrated that enamel demineralisation was related to acids produced by the incubation of certain types of bacteria with various dietary substrates. Miller focused his attention on potato starches and sugar was not prominent in his research. Nevertheless, since the time of Miller, dental clinicians, researchers and public health officials have focused entirely on sugar. With few exceptions, the mantra has been to reduce sugar quantity. Those who think there is a strong positive correlation of quantity sugar consumed to dental caries claim strong evidence from a number of studies carried out in the 1940-60s. For example, the classic work by Gustafsson et al (1) is cited as evidence that quantity of sugar was significantly related to dental caries. Yet a close examination of the results of that study shows that the frequency of use of the sugar or sugary foods was more significantly related to caries incidence. Similarly, other studies often cited as proving a direct relationship of sugars to caries actually show a complex relationship with many other factors than sugar(s) involved.
As a result of the debate on sugars and caries, some studies were carried out to relate dental caries prevalence on a national basis to reported quantities of sugars used. The main study was that of Sreebny et al (2) whose conclusion was that there was a linear relationship of sugar used to dental caries. However, this study was based on very crude caries data, much of it derived from weak, limited epidemiology. No allowance was made for the presence of fluoride, which had become more widely used, particularly in more advanced societies, or for the confounding correlation of frequency of consumption with the amount consumed.
Over the past 100 years, there have been many publications on caries and sugars. Studies have variously attempted to identify any relationship between the quantity of sugar used or the frequency of use of sugars and dental caries. Such studies have been based on an incidence of caries over time, usually, a minimum of two years, related to sugar use and consumption, usually assessed by diet diaries. An analysis of these studies was first carried out by Burt et al (3) who reported that any relationship for either quantity or frequency was weak. A later study by Anderson (4) showed that there were 94 such reports in the literature since 1966 that met some of a set of assessment criteria. This study did review all published reports going back for over a hundred years but all prior to 1966 were deficient in one way or another and could not be considered. However, when the analysis was completed it was found that only one study met all the necessary criteria, while another 32 met nearly all criteria. These 32 papers were then tabulated for whether significant relationships were found for dental caries with frequency or quantity of sugar use. Of 22 papers reporting on quantity, only 6 found a significant relationship. By contrast 19 of the 32 papers reported a significant relationship between frequency of use of sugar.
These analyses (3,4) have, therefore, shown that the evidence for any relationship of sugars to caries was absent for quantity and present but weak for frequency. Interestingly the selection of studies in both of these analyses was very similar, although there were some differences and the Anderson study was several years later, so included some more up to date reports.
The conclusion of these two studies is therefore that there is a relationship, but that it is weak. However, it should be noted that the data that these studies were very weak as only a few of the relevant variables were examined.
The best scientific evidence is always experimentally based. Accordingly, by the 1960- 70s, several researchers set out to test the question of frequency versus quantity by controlled experiments. The first approach for such studies used the animal model, in which caries- susceptible animals, mainly rats, were fed various diets, but also sugars at different frequencies. The seminal work of Bowen and co-workers (5) showed a significant relationship of caries related to intervals between meals and hence, the frequency of use of sugars. Related studies, looking at a variety of snack foods, including sugars and starches, showed that the relationship was complex, but there were significant relationships between caries incidence with the frequency of the use of snack foods, that were combinations of sugars with other food components such as starches(6).The greatest cariogenicity was found to be snack foods that combined sugar(s) and cooked starches.
Sugar is after all rarely if ever, eaten as sugar, but usually in combination with other dietary components. For example, cakes and biscuits are made essentially of starch and sugar. These days the starch, in particular, is white flour, which is high in gluten. The combination of glutinous cooked starch with sugars, of various types, makes a very cariogenic food. When consumed by rats less than seven times a day the cariogenic potential of the food (its ability to promote dental caries) is lower than if used more than 11 times a day. The maximum production of caries occurs with a frequency of 17 times a day. It should be borne in mind that these frequencies are higher in the rat model than would be in humans because of the nature of the oral conditions pertaining to the rat.
A second approach has used an in situ caries model, in which slabs of dental enamel are positioned on mandibular flanges of removable prosthetic appliances, with Dacron gauze to facilitate plaque accumulation. The flanges of the appliance are then, at various frequencies, immersed in test solutions over a sufficient number of days to produce early enamel demineralisation. The latter can be measured by various techniques such as indentation, microradiography or image analysis. Tests by Duggal et al (7) using this approach, showed that enamel demineralisation, as an indicator of dental caries, increased when the frequency of challenge exceeded 3 times per day. Volunteers continued with their usual oral hygiene practices, but a non-fluoride toothpaste was used throughout the study.
However, in modern advanced societies the use of preventive measures, such as fluoride, has to be taken into account in studies on caries. Accordingly, the later experiment of Duggal et al (7) included a second leg where the use of fluoride was allowed for. The volunteers used fluoride toothpaste as a routine measure twice daily. When the enamel mineralisation lesions were assessed, with fluoride use it was found that there was a significant reduction in lesion depth between test and control enamel slabs. Use of carbohydrate, in the form of a sugar solution, was not statistically different to the control for 1, 3 and 5 times a day challenges. An increase in lesion depth did not occur until 7 or 10 challenges per day were performed. Thus, this study showed not only that frequency of use was significantly related to dental caries but also that the use of fluoride offset the effects of frequency.
These findings, that it is frequency of use of sugars (often in combination with starches), that are characteristic of a cariogenic diet, and that oral hygiene with fluoride toothpaste is an essential preventative measure, provide the evidence for developing effective prevention programmes. Based on this premise studies have been carried out in which the frequency of use of cariogenic foods and oral hygiene have been targeted in at-risk groups. A recent such study (8) focused on the mothers of very young infants, aged 8 months, whose primary teeth had mostly not yet erupted. Through home visits every three to twelve months, the mothers were advised about preventive measures such as fluoridated children’s toothpaste, as well as the use of foods at key times during the day. At the end of the three-year period of study, only two children had developed early childhood caries. Further analysis of this study showed that it was highly cost effective.
A hundred years of literature and innumerable studies have failed to show a statistically significant relationship between sugar consumption, as quantity, to dental caries. Studies on the frequency of sugar use, however, have indicated that there is a correlation, albeit not a very strong one. The experimental evidence is much stronger and shows that frequent use of refined carbohydrates (more than 3 times a day), including sugars, but in the absence of fluoride, is related to the development of dental caries. However, the frequency of challenge before enamel demineralisation occurs increases to more than 7 times a day, when fluoride toothpaste is used regularly twice daily. A useful oral health message is that frequency of eating, of any foods, should be five times a day or less coupled with the use of fluoride toothpaste twice a day.
Gustafsson et at. The Vipeholm dental caries study. Acta Odont Scand (1954) 11:232-364. 2. Sreebny LM Sugar availability, sugar consumption and dental caries. Community Dent Oral Epidemiol (1982) 10:1-7. 3. Burt BA, Pai S, Satischandra P. Sugar consumption and caries risk: a systematic review. J Dent Educ (2001) 65:1017-1023. 4. Anderson C. Sucrose and dental caries: A systematic review of the literature. MDentSci Thesis, University of Leeds, 2003. 5. Bowen WH et al. Effects of varying intervals between meals on dental caries in rats. Caries Res (1983) 17:466-471. 6. Mundorff SA. et al. Cariogenic potential of foods. I. Caries in the rat model. Caries Res (1990) 24:344-355. 7. Duggal MS, et al. Enamel demineralisation in situ with varying frequency of carbohydrate consumption with and without fluoride toothpaste J Dent Res (2001) 80: 1721-1724. 8. Kowash T, Effectiveness on oral health of a long term health education programme for mothers with young children. Brit Dent J (2000) 188;201-205.
Martin Curzon, Emeritus Professor of Child Dental Health, Leeds Dental Institute, University of Leeds.
33 Weymouth Street, London, W1G 7BY