Microbial risk markers for childhood caries in pediatricians' offices

Abstract

Dental caries in pre-school children has significant public health and health disparity implications. To determine microbial risk markers for this infection, this study aimed to compare the microbiota of children with early childhood caries with that of caries-free children. Plaque samples from incisors, molars, and the tongue from 195 children attending pediatricians' offices were assayed by 74 DNA probes and by PCR to Streptococcus mutans. Caries-associated factors included visible plaque, child age, race, and snacking habits. Species were detected more frequently from tooth than tongue samples. Lactobacillus gasseri (p < 0.01), Lactobacillus fermentum, Lactobacillus vaginalis, and S. mutans with Streptococcus sobrinus (all p < 0.05) were positively associated with caries. By multifactorial analysis, the probiotic Lactobacillus acidophilus was negatively associated with caries. Prevotella nigrescens was the only species (p < 0.05) significantly associated with caries by the 'false discovery' rate. Analysis of the data suggests that selected Lactobacillus species, in addition to mutans streptococci, are risk markers for early childhood caries.

Figure 1.

Microbiota of caries-affected and caries-free children. Bacterial species withz > 5% difference in detection between caries-affected and caries-free children are in the same order as in Table 2. Most bacterial species assayed were detected more frequently (*p ≤ 0.05 or **p ≤ 0.01) from children with caries than from caries-free children. P. nigrescens association remained significant after adjustment for multiple comparisons by the false-discovery rate (***p ≤ 0.05 FDR). A. actinomycet. is Aggregatibacter actinomycetemcomitans, and F.n. polymorphum is Fusobacterium nucleatum subspecies polymorphum.

Figure 2.

Scatter plot of Partial Least Squares (PLS) weights w*c for bacterial data modeled on caries prevalence (dft) in children with visible plaque (n = 102). For a given PLS model, one vector of X-weights w*_a and one vector of Y-weights c_a are obtained for each model component (a). The X- and Y-weights are plotted together for the first two components (w*c [1] and w*c [2]). Variable importance in projection (VIP) indicates how influential the outcome (microbial) variables are in explaining the variable of interest (decayed filled teeth-dft). VIP > 1.0 is influential, and VIP ≥ 1.5 is highly influential. In this plot, species located close to the outcome variable dft were positively associated with dft, whereas those in the opposite direction represent species negatively associated with dft. Encircled ▲ indicate positions of highly influential species (VIP > 1.5), and encircled △ positions of influential (VIP 1.0-1.4) or borderline influential species (VIP 0.7-0.9). Non-labeled, not encircled △ indicate positions of non-influential species. S. mutans and S. sobrinus (when detected by checkerboard) were positively associated with caries, but were non-influential in this model.