Census based self reported health measures: an examination of changes between the 1991 and 2001 censuses

Dermot O'Reilly and Michael Rosato, Queen's University Belfast

[Project number 30044]

Census based self reported health (SRH) measures are important indicators of morbidity both in individuals and at area level, and are used routinely in identifying social and health 'need' and in subsequent resource allocation. The 1991 Census for England and Wales contained a single question on limiting long-term illness (LLTI) - the extent of '.. health problems or handicap ..'. By 2001 the scope had extended to two questions: one on general health in the previous twelve months; and a modified LLTI question (on 'disability' rather than 'handicap').

Levels of self-reported ill-health are increasing. The Northern Ireland census shows an increase in the proportion of reported LLTI from 11.1% to 20.8% between 1991-2001; and, according to the GHS (GB) LLTI prevalence has increased from 21% to 35% (1971-2000).
This is set within a secular trend of rising life expectancy and changes in perception of what 'normal' health or well-being is at any point in the lifecycle. Age, gender and pre-existing chronic illness all influence SRH levels (1,2) and previous LS work has shown the importance of ethnicity (3). There is also evidence of UK-wide regional variation (4), and it has been suggested that changes to localised economic conditions can influence employment chances and the flow of marginally ill people out of economic activity into ill-health ('hidden unemployment') (5,6). Another question raised by the 2001 questions relates to differences in scope: general health asks about persisting problems, while LLTI is about limitations on daily living. Research shows LLTI relates more to physical limitation and less on mental and social wellbeing (these latter better reflected in the general health question). Therefore, though they correlate highly at both area and individual level, they may differ in relation to cause-specific mortality.

We will examine implications of changes to the SRH questions over two broad areas:

  • Comparing LLTI at 2001 with its 1991 version - implications of the changed context and how this can be examined in relation to changing trends. We examine the degree of continuity in LLTI between censuses; transitions in associated demographic and socio-economic factors in relation to changing SRH status; and compare LLTI and short-term mortality after each census (1991-94, 2001-04).
  • The relationship between LLTI and general health at the 2001 census and subsequent mortality (2001 to 2004) - using both all-cause and cause-specific mortality.

References:
1. Carr A, Gibson B, Robinson (2001) Measuring quality of life: Is quality of life determined by expectations or experience? British Medical Journal 322; 1240-1243.
2. Daltroy L, Larson M, Eaton H, Phillips C, Liang M. Discrepancies between self reported and observed physical function in the elderly: the influence of response shift and other factors. Social Science and Medicine (1999); 48:549-1561.
3. Harding, S, Balarajan R. Limiting long-term illness among black Caribbeans, black Africans, Indians, Pakistanis, Bangladeshis and Chinese born in the UK Ethnicity & Health 2000; 5: 41-46.
4. O'Reilly D, Rosato M, Patterson C Self-reported health and mortality: an analysis across the UK
(re)submitted to the British Medical Journal: 16/04/2005
5. Bartley M, Sacker A, Clarke P. Employment status, employment conditions, and limiting illness: prospective evidence from the British household panel survey 1991-2001 Journal of Epidemiology and Public Health (2004); 58; 501-506 doi: 10.1136/jech.2003.009878
6. Beatty C, Fothergill S, MacMillan R. A theory of employment, unemployment and sickness. Regional Studies (2000). 34:617- 630.

The longitudinal study (LS), based on a continuing, representative, 1% sample of the population of England and Wales, is a high quality multi-cohort research database with ongoing linkage of routinely collected vital statistics data with the census returns from 1971, 1981, 1991 and 2001. It therefore provides a rare opportunity to track transitions in socio-demographic and socio-economic circumstances between census points, and examine how these relate to health outcomes. This project links with two others which use the LS to examine different aspects of self-reported health in 1991 and 2001, and which have already been accepted by the research board (Harding and Rosato: Social trajectories and long term illness among foreign-born and UK-born ethnic minority groups). It is anticipated these will form part of the analysis for a PhD thesis - the details of which are still to be worked out, but which should begin in September 2005.

Top of page