Ethnic density effects on physical health
Kate Pickett, Richard Shaw and Christo Albor, University of York
[Project number 30123]
Our proposed analyses using the Office for National Statistics Longitudindal Study (ONS-LS) form part of a wider programme of research examining the effects of ethnic density on physical health in UK and US ethnic minority population. This wider project is funded by the MRC
We plan to study the impact of ethnic density on mortality in the UK ONS-LS, and in the US Compressed Mortality File, as well as limiting long-term illness in the ONS-LS. To study mortality in the ONS-LS will require longitudinal event data on study members deaths from 1991-2001, whilst the analyses of limiting long-term illness will be a cross-sectional analysis, using data from the 2001 census. Different samples and variables will be needed for the analyses, although the choice of potential confounding and moderating variables will be essentially the same. In both cases, data on individuals will be linked to area-level data from the Census.
The sample for the mortality analyses will include all Indian, Black Caribbean and Pakistani adults enumerated at the 1991 census and aged 25 or more years at the time of the 1991 census. The sample for the analyses of limiting long term illness will include all people aged 25 or more at the time of the 1991 census of Indian, Pakistani, White Irish, Black Caribbean, Black African, Bangladeshi and Chinese origin enumerated at the 1991.
Depending on results, we may also investigate White British adults meeting the same criteria to test if ethnic density effects are specific to minority groups or if the presence of a particular ethnic group influences the wider population.
We propose two sets of analyses on the impact of ward-level % same ethnicity
on ethnic minority health. First, ward and individual-level data from
the1991 census will be linked to mortality data from 1991-2001. Second,
ward and individual-level data from the 2001 census will be linked to
investigate limiting illness.