Methods: All Japanese municipalities (n=2,522) had been stratified into 9 regions with 3 caries levels and 44 municipalities were randomly selected. Community health service workers collected information on sociodemographic characteristics, oral health behavior and dental condition for 3-year-old children during community dental health check-ups. Municipality average income was used as a neighborhood-level socioeconomic status and occupation of household as an individual-level socioeconomic status. Multilevel logistic regression models selected oral health behavior variables related to significant neighborhood variation. Continuous analyses were then conducted to estimate the contribution of individual- and neighborhood-level socioeconomic status to oral health behaviors.
Results: A total of 3,301 parents (79.9%) in 39 municipalities participated in our survey. After excluding missing data, 3,005 data were used. Five oral health behavior variables were significantly varied between municipalities: weekly frequency of intake of sweet drinks, dairy frequency of intake of sweet foods and drinks, frequency of toothbrushing, initiation of fluoride toothpaste usage and initiation of toothbrushing by parents. There was a significant association between neighborhood-level socioeconomic status and frequency of intake of sweet foods and drinks per day with considering individual-level socioeconomic status. After adjusted for individual- and neighborhood-level socioeconomic status, two models, which outcome were initiation of fluoride toothpaste usage and initiation of toothbrushing by parents respectively, still showed significant neighborhood variation.
Conclusion: Neighborhood-level socioeconomic status is independently associated with individual eating behavior. Community wide interventions are needed to improve population oral health.