Abstract
INTRODUCTION: Measures of socioeconomic position (SEP) are widely used in health research.
AIM: To provide future researchers with empirically based guidance about the relative utility of five measures of SEP in predicting health outcomes.
METHODS: Data from 12 488 adults were obtained from the 2006 New Zealand Health Survey.Seven health-related outcome measures with expected variations by SEP are modelled using five measures of SEP: a census-based small-area index of relative socioeconomic deprivation,NZDep2006; a questionnaire-based individual-level index of socioeconomic deprivation,NZiDep; an index of living standards,ELSI; education,measured by highest qualification; and equivalised household income.
RESULTS: After including the individual measure of deprivation,the area-based measure of deprivation adds useful explanatory power,and,separately,the broader spectrum provided by the living standards index adds only a small amount of extra explanatory power.The education and household income variables add little extra explanatory power.
DISCUSSION: Both NZiDep and ELSI are useful health-outcome predictors.NZiDep is the cheapest data to obtain and less prone to missing data.The area index,NZDep,is a useful addition to the arsenal of individual SEP indicators,and is a reasonable alternative to them where the use of individual measures is impracticable.Education and household income,using commonly used measurement tools,may be of limited use in research if more proximal indicators of SEP are available.NZDep and NZiDep are cost-effective measures of SEP in health research.Other or additional measures may be useful if costs allow and/ or for topic-related hypothesis testing.