Figure. Cross-level interaction of tertiles of neighborhood socio-economic position and individuals’ household income on number of retained teeth, adjusted for age, sex, and education. Datapoints are adjusted means and standard errors from a multi-level linear regression model where the dependent variable is the reported number of remaining teeth. Explanatory variables in the model include 2 ‘dummy’ variables for 3 levels of neighborhood socio-economic position, 3 dummy variables for 4 levels of annual household income, and 6 dummy variables for the interaction between neighborhood socio-economic position and household income. The estimates are adjusted for age, sex, and individual’s education. The model estimates a statistically significantly difference in number of teeth between the lowest- and highest-income groups in neighborhoods with poor socio-economic position. In contrast, the model estimates no statistically significant differences between income groups in neighborhoods with high socio-economic position.