Do Standard Measures of Deprivation Reflect Health Inequalities in Older People?
By Alwan, Nisreen Wilkinson, Margaret; Birks, Dorothy; Wright, John
ABSTRACT The objective of the study is to examine the relationship between different deprivation indicators and both self- rated health and emergency admission rates of older people to determine which indicators best predict the health of people in this age group. The method employed an ecological study design using data from all 100 neighbourhoods in Sheffield in 2004 and analysing relationships in three age groups 50-64, 65-74 and over 75 years. Analysis was performed using Pearson correlation coefficient. For people aged 50-64 years, receipt of income support was the best predictor of poor self-reported health (R = 0.85). For people aged 64-75 years, lack of formal educational qualifications showed the strongest relationship with poor health (R = 0.88), although there was still a significantly strong relationship between poor self- rated health and both non-property ownership (R = 0.8) and receipt of income support (R = 0.7) in this age group. For people aged 75 years and over, lack of formal qualifications showed the strongest relationship (R = 0.6, P < 0.001). This was reinforced by a strong relationship between this indicator and emergency admission rates. In conclusion, caution should be used when using conventional deprivation/poverty measures to select older populations to be targeted for services. Our analysis has shown that the deprivation indicator that correlates best with the subjective health rating of people aged 75+ is educational qualification.
Journal of Public Health Policy (2007) 28, 356-362.
doi:10.1057/palgrave.jphp.3200137
Keywords: health, older adults, deprivation
INTRODUCTION
Deprivation indicators are typically used to target health interventions at populations. There is substantial evidence examining the relationship between different poverty and deprivation indicators with poor health. The vast majority comes from research focusing on working age populations. Health and social care initiatives have been historically targeted at older populations using the same indicators of disadvantage as younger populations.
For example, in Sheffield, the percentage of households in a neighbourhood receiving income support benefits is used as a deprivation measure to target programmes aimed at reducing health inequalities.
Despite the fact that older people suffer much more ill health, there has been little research on the suitability of standard measures of poverty and deprivation to predict the health of older people. There is little to suggest that such social and economic indicators predict ill health at older age as well as they do in people of working age.
We have examined the relationship between different area measures of deprivation and the health of older people in the same areas, using both self-rated health and emergency admissions rates, to find out which deprivation measure correlates best with older people’s health.
DATA AND METHODS
This is an ecological study examining the relationship between standard measures of deprivation often used to assist targeting health initiatives and the health of all people aged 60 years and over in the city of Sheffield.
The study was performed in Sheffield, England. The population’s age and ethnicity structure in Sheffield is fairly representative of England’s population (Table 1). According to the 2001 UK census data, the total number of people aged 60 years and over in Sheffield was 114,634.
The data used to characterise the degree of affluence or deprivation of the 100 neighbourhoods in Sheffield came from two sources:
(1) The 2001 census, which provided the single indicators of housing tenure and educational attainment for people aged 65 years and over.
(2) Social security benefits data were used to derive two area measures related to receipt of income support (a social security benefit paid as an income supplement to those with less money coming in than the law says they need to live on, and is calculated on a case-by-case basis (1)):
(i) The percentage of households claiming income support in a neighbourhood (poverty postcodes are postcodes where 30% or more of households claim income support).
(ii) The percentage of people aged 60 years and over living in households claiming income support benefits in each neighbourhood.
(3) The data used to describe the health of older people came from two sources:
(i) The 2001 census that provided a self-rated health measure (How would you describe your general health?). This is a new question added to the 2001 census separate from the long-term limiting illness question. The response of “not good health” was used as an indicator of poor health.
(ii) Hospital Episode Statistics that provided emergency admission rates for people aged 65 years and above in each neighbourhood in Sheffield for the year 2004.
The strength of association between variables was measures using Pearson correlation coefficients. Analysis was done in three age bands: 50-64 years, 65-74 years and 75 years and over using the following deprivation measures:
(1) Poverty postcodes (a poverty postcode is defined as more than 30% of its households claiming income support benefits).
(2) The percentage of people over 60 years living in households claiming income support benefits in a neighbourhood.
(3) The percentage of people aged 65 + in a neighbourhood with no or unknown qualifications (Figure 1 and Table 2).
(4) The percentage of people aged 65 + in neighbourhood who do not own property.
RESULTS
The current deprivation measure that is used in Sheffield by the City council and the Health Informatics Service to target health initiatives, the poverty postcode, showed a strong relationship with “not good health” in the age band 50-64 (R = 0.85). However, the relationship was less strong with the older age bands (Table 3).
Regarding the older age bands (65 + years), the deprivation measure that correlated best with poor self-rated health was the absence of formal qualifications (R = 0.88 for 65-74 years and 0.6 for 75+ years). Non-property ownership showed a strong relationship with poor self-rated health in the 65-74 years age band but the correlation was less strong for the over 75 years (Table 3).
In the 75 + age band, no formal qualification was found to be the best measure to predict poor health (R = 0.6, P The strength of the relationship between educational attainment and an objective proxy for the health status of people aged 65 years and over – emergency admission rates – was also examined. There was still a relatively strong positive relationship between the two variables (R = 0.57, i.e. 33 % of variation in emergency admission rates in people aged 65 years and over is explained by educational attainment). DISCUSSION We have found that educational attainment was the deprivation indicator that showed the strongest relationship with the self-rated health of people aged 75 years and over among those examined. The current indicator used in Sheffield, the poverty postcode, correlated better with the self-rated health of the younger age bands (50-74 years). There is evidence that the link between mortality and social inequalities persists in old age (2) and that income is the best predictor of mortality in German and North American elderly population (3). However, in the UK, housing tenure and living in a household with access to car are the indicators that have been shown to be the best indicators associated with reduced mortality in older people (4). There is also evidence that income support receipt for those over 75 years is strongly correlated with mortality (5). In all these studies, morbidity or health status were not examined. Using self-reported general health, a combination of individual measures (either education or occupational social class) and Townsend deprivation index was found to provide the best indicator of deprivation; however, the study only used data for people aged 55- 69 years (6). The analysis from our study extends to include the older age groups as well. We used self-rated health as it represents a more holistic indicator of health in the sense of well-being. However, it is a subjective measure of health. When educational attainment, expressed as the lack of formal qualifications, was examined against an objective measure of health, emergency admission rates in people aged 65 years and over, there was still a relatively strong relationship. Using education as a deprivation measure has the benefit of being a relatively stable measure. It also avoids the problems of reverse causation (bad health leading to deprivation rather than the opposite) that accompany the use of other measures such as occupation (e.g. occupational social class), as education is normally fixed early in life, and it is available in the UK from routine census data. However, there is limited differentiation between groups when using this indicator, perhaps allowing the most advantaged to be differentiated from the rest of the population, although this pattern may change in the future with more people gaining qualifications than previously. Another limitation is that the evidence that supports a strong correlation with health comes mainly from ecological studies, including this one, with the limitations that this type of research carries. Based on the findings from this analysis, we suggest that educational attainment should be considered for use as a deprivation indicator when targeting health interventions at people aged 75 years and over. Competing interests: none. REFERENCES 1. Public Health Electronic Library. www.phel.gov.uk (accessed 07 November 2006). 2. Huisman M, Kunst AE, Anderson O, Bopp M, Borgan JK, Borrell C, et al. Socioeconomic inequalities in mortality among elderly people in 11 European populations. J Epidemiol Community Health. 2.004;58: 468-75. 3. Von Dem Knesebeck 0, Luschen G, Cockerham WC, Siegrist J. Socioecenomic status and health among the aged in the United States and Germany: a comparative cross-sectional study. Soc Sci Med. 2.003;57:1643-52. 4. Breeze E, Slogett A, Fletcher A. Socioeconomic and demographic predictors of mortality and institutional residence among middle aged and older people: results from the longitudinal study. J Epidemiol Community Health. 1999;53:765-74. 5. O’Reilly D. Standard indicators of deprivation: do they disadvantage older people? Age and Aging. 2002;31:197-202. 6. Grundy E, Holt G. The socioeconomic status of older adults: how should we measure it in studies of health inequalities? J Epidemiol Community Health. 2001;55:895-904. NISREEN ALWAN*, MARGARET WILKINSON, DOROTHY BIRKS and JOHN WRIGHT * Address for Correspondence: Leeds Primary Care Trust, North West House, Ring Road, West Park, Leeds LS16 6QG, UK. E-mail: nisreen.alwan@nhs.net NlSREEN ALWAN, M.B. ChB, MPH, MRCP is at Leeds PCT, Public Health, North West House, Ring Road West Park, Leeds, LS 16 6QG, UK. nisreen.alwan@nhs.net DOROTHY BirkS is Director of Public Health, Sheffield South West Primary Care Trust, Sheffield, South Yorkshire, UK. dorothy.birks@sheffieldsw-pct.nhs.uk MARGARET WILKINSON is Senior Information Specialist, Sheffield Health Informatics Service, Sheffield West Primary Care Trust, Sheffield, UK. margaret.wilkinson@ewr-pct.nhs.uk JOHN WRIGHT is Consultant in Epidemiology and Public Health Medicine, Bradford Royal Infirmary, Bradford, UK. john.wright@bradfordhospitals.nhs.uk Copyright Palgrave Macmillan Limited 2007 (c) 2007 Journal of Public Health Policy. Provided by ProQuest Information and Learning. All rights Reserved.