Ethnicity, socio-economic status and health

Objectives: To disentangle statistically the distinct roles of ethnicity and Socio-economic status (SES) as influences on health inequalities.

Methods: We use data from the UK Household Longitudinal Survey (UKHLS) 2009-19 and the Health Survey for England (HSE) 2010-14. General health is measured through self-reported assessments of health, quality of life and disability, and specific conditions through self-reports of diagnoses. SES measures cover household income and educational attainment. Ethnicity is self-selected from a list of standard categories and grouped into nine. The UKHLS also distinguishes those born abroad from UK-born. Random effects (UKHLS) and pooled data (HSE) models are estimated, with a variance ratio measure used to summarise the relative predictive importance of ethnicity and SES.

Results: For most general health measures, ethnicity and SES have highly statistically significant effects, with SES capturing most of the observed inter-ethnic differences. The pattern of ethnic advantage and disadvantage in general health is not uniform across minority ethnic groups, nor across health measures. For diagnosed conditions, there is also a complex pattern across ethnic groups and conditions and some conflict between HSE and UKHLS results, The strongest evidence is for those born outside the UK, revealing health disadvantage for South Asian groups with respect to diabetes in men and women and also cardiovascular disease among Bangladeshis. But we find SES-adjusted health advantage with respect to respiratory disease and cancer for members of the same groups.

Conclusions: The role of ethnicity relative to SES in health inequalities is complex. Different minority ethnic groups have different patterns of health advantage and disadvantage, which also vary with the health measure used. There is some conflict between the detailed findings from analyses based on different datasets. When designing policies to reduce health inequalities, it is important to avoid overly simple views of the nature of inequalities.