Measuring socio-economic inequalities in edentate Scottish adults--cross-sectional analyses using Scottish Health...

To investigate the appropriateness of different measures of socio-economic inequalities, in relation to adult oral health in Scotland, utilizing data from a series of large, representative population surveys.

The Scottish Health Surveys (SHeS) (1995; 1998; 2003; 2008/09) are cross-sectional national population-based surveys used to monitor health status in those living in private households. The age groups included in this study are as follows: 45-54; 55-64 years: all survey years; 65-74: 1998 onwards; 75+: 2003 onwards. Primary outcome was no natural teeth (edentulism). Three measures of socio-economic position: Occupational social class, Education, Carstairs deprivation score (2001) were used. Simple (absolute/relative differences) and complex measures (Slope Index, Relative Index, Concentration Index and c-index) of inequality were produced for each age group across all four surveys.

Simple and complex (absolute) measures of inequality have both demonstrated narrowing disparities in edentulism over time in the 45- to 64-year-old group, a levelling off in those aged 65 and above, and a rise in those aged 75+. Complex relative measures (RII, Concentration Index and c-index), however, show an increasing trend in inequalities over time for all age groups, suggesting that rates of improvement in edentulism rates are not uniform across all social groups.

Simple absolute inequality provides a quick and easy indication of the extent of disparities between extreme groups, whereas complex measures (absolute and relative) consider the gradient in health across all social groups. We have demonstrated that both are useful measures of inequality and should be considered complementary to one another. The appropriate choice of complex measure of inequality will depend on the audience to whom the results are to be communicated. This methodological approach is not confined to oral health but is applicable to other health outcomes that are socially patterned.