Latent Class Analysis Identifies Preschool-age Children’s Oral Health-related Behavior Profiles

Objectives: Early childhood caries (ECC) is the most common chronic disease in preschool-age children. Its incidence is strongly associated with oral health-related behaviors and practices (OHB) which are almost entirely under the control of parents and guardians. We hypothesized that patterns of modifiable OHBs cluster within families. In this study, we sought to identify OHB “profiles” (i.e., clusters of participants with distinct favorable and unfavorable OHB patterns) among a large community-based sample of parents/caregivers and children.
Methods: We used parent survey data (n=7,856; 8% in Spanish) collected in the context of an ongoing, community-based epidemiologic study of early childhood oral health among 3-5-year-old children in NC. Responses to 6 questionnaire items pertaining to modifiable OHB domains (i.e., diet, oral health practices and dental attendance) were carried forward to a latent class analysis (LCA) framework. The optimal number of latent classes (i.e., “profiles”) was identified via the Bayesian Information Criterion (BIC). Subsequently, we contrasted profile membership with children’s clinical and parent-reported oral health status using bivariate testing methods and a p<0.05 criterion. Analyses were done using Mplus v.8.1 and Stata 15.1.
Results: We identified two OHB profiles (A and B) comprising 75% and 25% of the analytical sample, respectively. Class A participants had more favorable OHBs in terms of tooth brushing frequency, use of fluoridated toothpaste, regular dental check-ups, consumption of sugar-containing snacks and beverages. Children in Class A had 27% (95 confidence interval=27%-36%) lower prevalence of ECC and significantly better parent-reported oral health status (“fair-poor”—Class A: 12% vs B: 27%, p<0.05).
Conclusions: Our results suggest that favorable and unfavorable modifiable OHBs cluster in groups of individuals. Identifying and working with these community groups rather than focusing on individual behaviors may be a promising alternative for promoting early childhood oral health at the population level.