摘要：Objectives. We examined racial/ethnic disparities in dental caries among kindergarten students in North Carolina and the cross-level effects between students’ race/ethnicity and school poverty status. Methods. We adjusted the analysis of oral health surveillance information (2009–2010) for individual-, school-, and county-level variables. We included a cross-level interaction of student’s race/ethnicity (White, Black, Hispanic) and school National School Lunch Program (NSLP) participation (< 75% vs ≥ 75% of students), which we used as a compositional school-level variable measuring poverty among families of enrolled students. Results. Among 70 089 students in 1067 schools in 95 counties, the prevalence of dental caries was 30.4% for White, 39.0% for Black, and 51.7% for Hispanic students. The adjusted difference in caries experience between Black and White students was significantly greater in schools with NSLP participation of less than 75%. Conclusions. Racial/ethnic oral health disparities exist among kindergarten students in North Carolina as a whole and regardless of school’s poverty status. Furthermore, disparities between White and Black students are larger in nonpoor schools than in poor schools. Further studies are needed to explore causal pathways that might lead to these disparities. Disparities in access to dental health services and oral health status exist among population subgroups in the United States. 1–4 Among the most pronounced and persistent disparities in pediatric oral health are those defined by race and ethnicity. 2,3,5–8 According to the 2011–2012 National Health and Nutrition Examination Survey, 45.7% of Hispanic children aged 2 to 8 years had experienced dental caries in their primary dentitions, compared with 43.6% of non-Hispanic Black children and 30.5% of non-Hispanic White children. 5 Although disparities in oral health are well described and have been recognized as important, the causes of racial/ethnic disparities have not been fully explained in previous research. Individual-level studies generally identify residual differences in racial/ethnic groups after control for socioeconomic status (SES) and other confounders. 4,6–9 One promising direction in understanding and resolving these disparities, particularly at the conceptual level, is the consideration of risk factors occurring at multiple levels, such as the individual child, family, community, or society as a whole. 10–13 This approach might help explain reported residual racial/ethnic disparities, and such knowledge could lead to the design of interventions that target risk factors at the appropriate levels. Several studies have used multilevel analysis to explore the role of community-level factors in child oral health. 14–17 However, to our knowledge, only 1 study has considered determinants of racial/ethnic disparities in oral health in a pediatric population in the United States. Fisher-Owens et al. 3 used information from 2 levels (children and states) to test a conceptual model with 4 levels (child, family, neighborhood, and state). They found that state-level factors had almost no impact on racial/ethnic disparities in a global measure of self-reported oral health status. This negative finding likely resulted because their definition of community (i.e., the state) was too large a geographic area to be causally related to the outcome. Other multilevel studies of children have also found that determinants beyond the individual level tend to be weakly associated with oral health. 15,16 Yet given the small number of multilevel dental studies of racial/ethnic disparities in children, their potential limitations in defining the higher levels, and the repeated finding of residual racial/ethnic disparities in oral health after controlling for a large number of individual factors, an exploration of higher-level determinants might lead to a better understanding of the mechanism through which racial/ethnic disparities affect oral health. Further justification for a multilevel study of oral health disparities is provided by research on other health conditions demonstrating that determinants of disease at one level can be modified by determinants at other levels. 18–20 Only 1 study has explored cross-level interactions for predictors of child oral health. Martins et al. 16 examined the interaction between children’s household income and type of school attended (public or private) on dental caries experience in primary teeth; they found no statistically significant cross-level effect. To our knowledge, no study has explored the effect of a cross-level interaction involving individual race/ethnicity and pediatric oral health outcomes in a US population. In this study, we examine racial/ethnic differences in dental caries experience among kindergarten students in North Carolina using a multilevel analysis with 2 levels, examining individual students within schools. Our particular interest is in the variation in dental caries experience by race/ethnicity at the individual level and its cross-level interaction with a compositional school-level variable measuring poverty among families of enrolled students.
