A random effects model for some epidemiological features of dental caries.

We describe a random effects model for caries lesion development and progression based on considering the effects of the pH fluctuations over time in microbial dental plaque as a Wiener process with a single absorptive barrier. The model predicts that the period of greatest risk to developing caries occurs shortly after eruption, but thereafter the longer a surface survives without developing a lesion, the less likely will it be that a lesion will subsequently develop. The model is able to anticipate why the effect of water fluoridation on caries prevalence is most pronounced when caries is diagnosed at cavity level. This model offers one way in which the variability which characterizes the complex ecosystem associated with dental caries may be considered a subject of interest for enhancing our understanding of its pathogenesis and epidemiology.