Article Text


Cutting edge methodology
P1-22 Which measures of socio-economic position perform most consistently across ethnic groups? Retrospective cohort study using census data linkage
  1. C Fischbacher1,
  2. G Brin2,
  3. N Bansal2,
  4. J Pearce2,
  5. R Bhopal2
  1. 1Information Services Division, Edinburgh, UK
  2. 2University of Edinburgh, Edinburgh, UK


Introduction Ethnic health inequalities are substantial but incompletely understood. Some studies have attributed ethnic inequalities to socio-economic status (SES) and adjusted ethnic comparisons SES measures. In the absence of individual level data, area level deprivation measures are routinely used for this purpose. However it is not clear whether these or other available SES measures perform similarly across ethnic groups. We examined the association between educational, occupational and economic indicators and incident cardiovascular disease (CVD) in ethnic groups in Scotland.

Methods We obtained educational, occupational and economic indicators for those aged over 30 years and resident in Scotland in April 2001 (n=2.97 million) using the Scottish Census. Data on first CVD events up to April 2008 obtained from hospital discharge and mortality databases were combined with census sources using probability linkage methods. In each ethnic group we examined associations between CVD rates and SES indicators.

Results There were marked SES differences between ethnic groups but the SES measures were differently associated with health. In men individual educational qualification and highest household occupational group were most consistently associated across ethnic groups with CVD. Among women individual educational qualification and individual occupational group were the most consistent measures. Areal deprivation, car ownership, household tenure and employment status performed much less consistently.

Conclusions Studies of ethnic inequalities should take account of SES. Area measures of material deprivation do not predict cardiovascular outcomes consistently across ethnic groups. Educational level may provide a better means of taking account of such differences.

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.