as well as by oral microflora and host factors (e.g., genetic susceptibility and the composition and flow of saliva). However, as McDonald (1985a) points out, even after 23 centuries we know only a little more than Aristotle about the relative cariogenicity of foods.
Among the major carbohydrates in the dietcomplex carbohydrates (e.g., starches) and simple sugars (e.g., sucrose and lactose)sucrose appears to have the greatest cariogenic potential. Using data collected by the World Health Organization, Sreebny (1982a) reported a correlation coefficient of .72 between sugar availability in grams per person per day and DMFT in 12-year-old children in 47 countries. However, the correlation coefficient for dmft in 6-year-old children in 23 countries was only .31. Takeuchi (1961) provided time-trend data supportive of Sreebny's cross-sectional findings and reported that the prevalence of dental caries in Japanese children decreased precipitously during the late 1940s in association with the severe reduction in sugar supplies during World War II. Similar observations were made in Europe (Sognnaes, 1948; Toverud, 1957). In the United States, correlation analyses of time-trend data on per-capita sugar consumption and caries incidence suggest (1) that caries incidence increases when per-capita sugar consumption exceeds 40 g/day in the absence of fluoride use and 50 g/day when fluoride is used and (2) that the increase in caries incidence reaches a plateau when the per-capita consumption reaches approximately 130 g/day (Lehner, 1980; Newbrun, 1982; Sheiham, 1983, 1984; Sreebny, 1982b). As Glinsmann et al. (1986) noted, these data suggest that the current mean intake of added and total sugars in the United States (53 and 95 g/day, respectively) contributes substantially to overall caries risk and that reduction in sugar intake could be expected to reduce that risk. However, findings of other studies indicate that the correlation between sugar intake and caries occurrence is not entirely consistent. For example, caries incidence in Great Britain did not change appreciably from 1940 to 1977, despite an apparent doubling of sugar intake (Jackson, 1979). Likewise, the 32% decline in caries prevalence in the United States in the 1970s appears to have occurred despite a continued high intake of sugars. A similar observation was made in a study by DePaola et al. (1982), who noted that caries incidence in Massachusetts schoolchildren dropped markedly during a period in which total sugar consumption increased and then leveled out. Although total sugar consumption increased during the study period, however, the amount of sucrose consumed actually decreased.
Sucrose in solution stimulates the formation of plaque (Geddes et al., 1978)a substance comprising microbial colonies embedded in a matrix of salivary proteins and extracellular polymers, which may serve as a medium for growth of caries-promoting bacteria. It also increases the mineral content of plaque and saliva, suggesting increased mineral resorption from the teeth (Tenovuo et al., 1984). Frequent rinsing with a sucrose solution over 2 months produced changes characteristic of early demineralization of tooth surfaces (Geddes et al., 1978). Small slabs of bovine enamel attached to human teeth for short periods also underwent demineralization when frequently exposed to sucrose (Pearce and Gallagher, 1979; Tehrani et al., 1983).
Sucrose in foods is also cariogenic. In a clinical trial at the University of Turku in Finland (Scheinin et al., 1975a,b), three groups consuming diets containing sucrose, fructose, and xylitol, respectively, were followed for 2 years. By the study's end, the average number of decayed, missing, or filled tooth surfaces (DMFS) was twice as high in the group consuming sucrose than in the fructose group. The xylitol group had virtually no DMFS. The lower cariogenicity of fructose relative to sucrose may explain in part the inability of some studies to demonstrate a cariogenic potential of presweetened foods such as cereals (Finn and Jamison, 1980; Glass and Fleisch, 1974), which differ considerably in their content of specific sugars (Glinsmann et al., 1986). The decline in caries prevalence in the United States since the 1970s, despite a continued high consumption of total sugars, may be partially due to the nation's increasing consumption of corn-derived sweeteners such as fructose and the declining use of sucrose (Glinsmann et al., 1986).
The composition of dietary carbohydrate also appears to influence cariogenicity. Early enamel erosion, a risk factor for caries, was noted in 12 children ages 9 to 15 years who had consumed large quantities of soft drinks (Asher and Read, 1987). The authors concluded that a major contributing factor was the high citric acid content and resulting low pH of the drinks. Consumption of canned pears and apples has also been noted to lower plaque pHa factor believed to promote