Evidence from cross-sectional studies
Numerous cross-sectional observational studies in children have used dietary interview and questionnaire methods to study the relationship between caries prevalence and consumption of sugar and sweets. The results are somewhat conflicting (Rugg-Gunn, 1989): A significant, but not very strong, correlation between caries and the total quantity of sugar consumed has been found in some studies but not in others. A closer relationship has been demonstrated between caries and the quantities of sweets and confectionery consumed, probably because these products are consumed in ways that enhance cariogenicity¾between meals and over long periods¾whereas consumption of even large quantities of sugar at meals seems to do little harm.
There may be several reasons for these findings. Most of the studies have shown a weaker correlation between sugar intake and caries than might, in theory, be expected. A general methodologic weakness is that dietary data obtained by questionnaires, 24- hour recall, or diet history interviews cover a very limited period, from only 1 day to some months, while caries data express total caries experience over the years (Birkhed, 1990). Furthermore, some people, such as the obese, are known to underreport their intake of sugar.
Interstudy comparison is difficult, because the studies have been carried out with a large range of variables: with different age groups of subjects, at different times, in different countries, and in specific populations. For example, dietary information has been collected in a variety of ways: Some reports subdivided confectionery into types of sweets, only some of which were significantly related to caries experience. The term sugary food was seldom defined, which frequently made interpretation of the correlation between caries and frequency of sugar intake difficult. Some studies were limited to only one aspect of sugar consumption, such as bedtime eating habits. In most of the studies, children were not selected for inclusion on the basis of their level of caries experience, but some studies compared only the eating habits of children at the two extremes of the range of caries experience.
Absolute figures for caries experience were not reported: In many studies only correlation coefficients were reported. In some studies, although significant correlations were found, the absolute differences in caries prevalence were small. In other cases large differences in sugar consumption habits were observed but insufficient data were presented to allow their inclusion.
Several studies have investigated the effect of infant-feeding practices on caries, particularly “rampant caries” (or labial incisor caries) in the very young. Five British studies (Goose, 1967; Goose and Gittus, 1968; James et al, 1957; Winter et al, 1966; Winter et al, 1971) have all shown a strong relationship between labial incisor caries and sugared infant pacifiers, especially nursing bottles. One study which did not show such a relationship was reported by Richardson et al (1981a) in South Africa. The worldwide use of comforters and their effect on oral health has been reviewed by Winter (1980).
Two studies by Granath et al (1976, 1978) are of particular interest: Not only was the
level of consumption of sugary foods compared with caries severity, but also two
other important confounding factors, fluoride supplementation and oral hygiene
practices, were taken into account. The first study, involving 6 year olds, was small
(179 children) and the higher level of caries found in the children who consumed
larger amounts of sugary foods between meals was not statistically significant.
However, the second study, involving 4 year olds, was larger (515 children) and
differences between the dietary groups were highly significant. When the effects of
oral hygiene and fluoride were kept constant, the children with low between-meal
sugar intake had 86% fewer buccal and lingual carious lesions and 68% fewer
approximal carious lesions than did children with high between-meal sugar intake.
Hausen et al (1981), in a study involving more than 2,000 Finnish children, aged 7 to
16 years, reported that water fluoride level, toothbrushing frequency, and sugar
exposure were all important determinants of caries prevalence, but least important
was sugar exposure. Similarly, in another study in Finland, involving 543 children in
three age groups (5, 9, and 13 years), Kleemola-Kujala and Rasanen (1982) found a
stronger relationship between poor oral hygiene and caries than between high sugar
consumption and caries, although both relationships were important. However, the
combination of poor oral hygiene and poor dietary habits seemed to be synergistic.
Very similar results were reported among 159 12- to 16-year-old French Canadians
(Lachapelle-Harvey and Sevigny, 1985).
Holund et al (1985) reported more frequent consumption of sugary drinks in cariesactive
than caries-inactive 14-year-old Danes. Continuing the work begun by Granath
in the 1970s, Schroeder and Granath (1983) found that poor dietary habits and poor
oral hygiene were both good predictors of caries in 3-year-old Swedish children. A
few years later, Schroeder and Edwardsson (1987) reported that the predictive
potential of diet and oral hygiene can be enhanced by the addition of salivary
Lactobacillus and Streptococcus mutans counts.
In another group of Swedish 13-year-old schoolchildren, positive salivary S mutans
values were found to be a significant but weak risk indicator for caries, but evaluation
of the intake of sticky sugar products according to an estimated point scale disclosed
no correlation with caries prevalence (Kristoffersson et al, 1986).
Stecksen-Blicks et al (1985) conducted a large survey of the relationship between
dietary and toothbrushing habits and caries prevalence in children of three age groups
(4, 8, and 13 years) living in two northern communities and one southern community
in Sweden. Children from the south had considerably more carious lesions in both
primary and permanent teeth. This was attributed to differences in toothbrushing
frequency and the age at which dental care started. The lack of observed differences
in diet between north and south indicated that diet was not an important factor.
A large cross-sectional study in the US specifically investigated the relationship
between consumption of soft drinks and caries prevalence (Ismail et al, 1984).
Analyses of data from 3,194 Americans, aged 9 to 29 years, revealed significant
positive associations between frequency of between-meal consumption of soft drinks
and high decayed, missing, or filled teeth scores. These associations remained even
after the researchers allowed for the reported concurrent consumption of other sugary
foods and other confounding variables.
In some studies, caries experience has been correlated with the subjects’ dietary habits
some years previously. Persson et al (1985) reported a positive correlation between
the consumption of sucrose-rich foods at 12 months of age and the presence of caries
at 3 years of age in 275 Swedish children. Both factors were linked to the educational
status of the mother. The importance of social factors as determinants of eating habits
and caries experience of young children has been highlighted in a number of studies;
for example, Blinkhorn (1982) reported caries and sugar consumption in Edinburgh,
Scotland, to be much higher in children from socially deprived backgrounds. The role
of socioeconomic factors will be discussed later in this chapter.
In Hertford, England, Silver (1987) collected data on infant feeding and caries status
in children aged 3 years and compared the data to dietary habits and caries status
when the subjects were aged 8 to 10 years. “Poor infant feeding” (including the use of
sugared foods and drinks) was positively correlated with the subjects’ caries
experience at 3 years and at 8 to 10 years. Children who in infancy had been bottlefed
with sweet drinks were more likely to be consuming sugar-containing snacks at
the age of 8 to 10 years, supporting the concept that a taste for sweet food, acquired in
infancy, persists in later childhood.
The importance of establishing good oral health habits as early as possible and
postponing bad habits for as long as possible has recently been highlighted in a 2-year
prospective study by Wendt (1995). Almost 700 infants were examined at the age of 1
year and reexamined after 1 and 2 years. At the baseline examination, the amount of
plaque, gingival conditions, caries prevalence (decayed or filled surfaces), and
salivary S mutans levels were recorded. At the annual examinations, the
accompanying parent was interviewed about the child’s oral hygiene and dietary
habits. The percentage of caries-free children decreased from 99.5% to 71.7% at the
age of 3 years. Among children (n = 61) of immigrant parents, only 35% were
cariesfree (Wendt et al, 1992).
Children who were regularly bottle-fed with sweet drinks at night or breast-fed for
more than 12 months (mostly at night, when salivary function is at resting level)
developed significantly more new carious lesions than did children with more
disciplined dietary habits. Bottle-feeding with sweet drinks was common among
children of immigrants (Wendt and Birkhed, 1995). Because this was a prospective
study, it confirmed that regular bottle-feeding with sweet drinks, and prolonged
breast-feeding at night, should be regarded as risk factors for caries development in
infants and toddlers.
Children who were caries free at 3 years of age had had their teeth brushed more
regularly and frequently: At 1 and 2 years of age, these children already had less
visible plaque than did children with caries. Immigrant children had had their teeth
brushed less frequently, used fluoride toothpaste less frequently, and, at 1 year of age,
already had a higher prevalence of visible plaque than did nonimmigrant children
(Wendt et al, 1994).
If dietary risk behavior was already apparent at 1 year of age, the chance of remaining
caries free until 3 years of age was highest if good oral hygiene habits were
established by the age of 2 years. Caries-related behavioral patterns established during
infancy, such as oral hygiene and dietary habits, persisted throughout early childhood
(Wendt et al, 1996).
In this context, it is of interest to note that in the county of Varmland, Sweden, largescale
preventive programs at maternal and child welfare centers emphasize early
establishment of good oral hygiene and dietary habits: As a result, from 1973 to 1993,
the percentage of caries-free 3 year olds increased from 35% to 97%.
Although a few studies (eg, those investigating sugar intake in infant feeding) have
attempted to assess lifelong habits of sugar consumption, nearly all cross-sectional
studies have attempted to relate current caries prevalence to current consumption of
sugar or sweets or, at the most, consumption over the previous 3 to 7 days. As
discussed earlier, this approach may be valid in young children, whose teeth have
erupted and developed caries within the preceding few years and whose sugar
consumption habits may have remained relatively constant since the time of tooth
eruption; for older groups, its validity is questionable. In a child of 12 years, caries
experience is typically confined mainly to the permanent first molars, which erupted 6
years previously and may have developed caries quite early. It cannot be assumed that
there has been no change in sugar consumption habits between the ages of 6 and 12
years.
It is therefore not valid to relate the dietary habits at one point in time (eg, at 12 years)
to caries experience over a very much longer period (eg, 6 to 12 years). However,
most cross-sectional studies have attempted to do just this. A typical example is the
study by Mansbridge (1960), reporting that caries prevalence was 13% greater in 12
to 14 year olds who admitted consuming more than 8 oz (227 g) of sweets per week
than it was in those claiming to consume less. The difference, although statistically
significant, was modest. First molar caries prevalence was similar in the two groups,
but the difference was pronounced for premolar and second molar caries. The first
molars had erupted about 6 to 8 years before the sweet-eating habits were assessed,
compared to fewer than 4 years for the premolars and second molars.
The many cross-sectional studies conducted several decades ago showed that, at the
time, frequent intake of sugar-containing products was often a risk indicator for caries
in very young individuals with relatively high caries prevalence. However, recent
studies of populations older than 12 years, with good oral hygiene, including regular
daily use of fluoride toothpaste, generally show very weak or no correlation between
intake of sugar-containing products and caries prevalence. However, a combination of
poor oral hygiene and a high frequency of sugar intake seems to have a synergistic
cariogenic effect.
