Risk indicators for tooth decay: results from the multivariate analysisa.

aHurdle model, adjusted for all the variables in the table plus examiner to account for inter-examiner variability.bModels the probability of the women having tooth decay.cModels the number of decayed teeth among the women having at least one.dAdjusted odds ratio.e95% Confidence Interval.fexp(β) can be interpreted as follows: while holding all other variables constant in the model, among women having tooth decay, a woman in a given class has on average exp(β) more decayed teeth than a woman in the reference class. For example, among women having tooth decay, a woman aged 18–24 years has on average 1.37 more decayed teeth (or 37% more decayed teeth) than a woman aged 30 or more (reference class for age).*Significant at α = 5%.