Causal analysis on the use of dental services among old-age pensioners in Norway

The purpose of the present paper was to establish and evaluate a causal model on the use of dental services in Norwegian old-age pensioners living at home. Data were derived from the Health Survey of 1975. The independent variables sex, age, education and dental status were dichotomized and arranged in this assumed temporal sequence. The dependent variable was also dichotomized into use versus non-use of dental services last year. The analysis started with the bivariate percentage table for sex and use, and then proceeded by adding one by one of the independent variables in the order of their time sequence. In this way the statistical relationship between an independent and the dependent variable was decomposed into direct, indirect and spurious effects. Dental status was the most influential determinant of use of services, having the greatest direct effect (0.40) and mediating indirect effect of the prior variables. Education came second with a direct effect of 0.18 and an indirect effect (0.11) by affecting dental status. There were only negligible differences in use rates between the age groups (less than 75 / greater than or equal to 75 years of age) when the other variables were held constant (0.05, P greater than 0.05). The greater part of the direct effect of sex (0.11) was limited to dentate persons aged 65-74 with low education. Both age and sex affected use of services indirectly through dental status (0.09 and --0.04, respectively).