status (p<0.0001) and prevalence of diabetes (p<0.0001) for both countries. Level
of income for Korea showed (P=0.010).
2) Results of performing chi-square analysis with prevalence of edentulism as
the dependent variable showed that edentulism was a statistically significant factor
for age group (p<0.0001), education attainment level (p<0.0001), level of income (p<0.0001), insurance status (p<0.0001) and prevalence of diabetes (p<0.0001). For
NHANES gender (p=0.453), smoking history for KNHANES (p=0.059) and
NHANES (p=0.168) did not show statistically significant results.
3) Results performing t-test and ANOVA analysis with number of tooth loss as
the dependent variable showed that increased number of tooth loss with increased
in age group for both KNHANES and NHANES (p<0.0001) were shown. Gender
(p<0.0001), education attainment level (p<0.0001), level of income (p<0.0001),
insurance status (p<0.0001) prevalence of diabetes (p<0.0001) and smoking history
(p<0.0001) were all statistically significant factor for number of tooth loss for both
KNHANES and NHANES.
4) Results of performing simple logistic regression analysis with edentulism as
the dependent variable showed that smoking history factor for KNHANES (odds
ratio 1.16) and NHANES (odds ratio 0.88) was a statistically not significant factor
(p=0.06) and (p=0.16) respectively.
5) Results of performing multiple logistic regression analysis with edentulism as
the dependent variable with demographic factors and prevalence of diabetes
showed that smoking history (p<0.0001), age group (p<0.0001) and education
attainment level (p<0.0001) were statistically significant factors. Also other
statistically significant factors were level of income (upper vs. low) for KNHANES
and NHANES (p<0.0001).
From the above results, smoking was the risk factor for tooth loss and