A survey of over 8,000 people aged 75 and over in 23 general practices in Great Britain was undertaken as part of a trial of the assessment and management of the health of older people in the community. Analyses were undertaken of their quality of life in relation to their socio-economic position.
Most people aged 75 and over do not report many problems with home management, mobility, body care and movement, social interaction, or morale. Chances of being in the worst fifth of quality of life increase with age for all problems except morale. Women have higher chances of poor outcomes than men except for social interaction, the gender difference being greatest for morale. People who are living with someone other than a spouse (categorised as 'dependent') have worse scores on home management, mobility, and body care and movement than people considered 'independent', defined as living alone or with their spouse and not in sheltered housing or a residential home.
Among independent people:
- Having taken account of gender, age, and marital status, independent people in rented homes have 50-70 per cent excess odds of having any one of the five poor quality of life outcomes compared to those in owner-occupation.
- Infrequent social contact with people outside the household correlates with greater chances of poor quality of life but there is little difference in frequency of contact between owner occupiers and renters.
- Contrary to expectation, receipt of formal or informal help does not account for any tenure differences in morale.
- Health and lifestyle (smoking and alcohol consumption) jointly explain a substantial part of the tenure differential for each quality of life dimension.
- Among dependent people the major differences in quality of life refer to social interaction and morale. Adjusting for the health, lifestyle or social support factors barely alters the differences by tenure.
Social class and housing tenure for most of one's adult life are relevant to quality of life in addition to housing tenure in old age. Moving 'down' from owner occupation during most of adult life to social sector housing in old age carries significantly greater chances of three out of five poor outcomes. An 'upward' move in the opposite direction carries reduced chances of two outcomes compared with staying in the social sector. Among the independent groups, people who were in social housing at both times had double the odds of all poor outcomes compared with those who were in owner occupation both times.
Housing tenure continues to be associated with all five poor outcomes when area deprivation and population density are taken into account. Having adjusted for housing tenure and other area factors, the chances of poor morale are higher in the more densely populated areas, chances of poor mobility lower in the least deprived areas, and chances of poor home management or body care higher the more deprived the surrounding enumeration districts (using average score).
Context
The Labour Governments of 1997 and 2001 declared an interest in reducing inequalities in educational and working opportunities, in income, and also in the health differences by socio-economic group. This project linked these two issues together by analysing differences in quality of life according to socio-economic status of people aged 75 years and over in the community.
Socio-economic status in old age
It is already known that being in owner occupation carries an advantage over being in rented accommodation in terms of mortality, self-reported limiting long-term illness, and later moves into institutional care. Because ill health can influence where people live - in particular if it leads to living with children or moving to sheltered accommodation - separate comparisons are made between people in rented and in owner-occupied tenures for i) people who live alone or with spouse (independent) and ii) people who live with others (dependent). Those who were in sheltered or residential homes were separated out into a fifth group (supported housing). The largest group (53 per cent) comprised people in owner-occupation and considered independent, the second largest group comprised their counterparts in rented homes (21 per cent). The dependent groups were much smaller (7 per cent in owner-occupation and 3 per cent in rented homes); finally, 13 per cent lived in sheltered housing and 3 per cent in residential homes.
Variation in quality of life by gender, age and marital status
Most people in the study had good quality of life shown by low (favourable) scores on these measures. Women tend to have higher scores than men and older people than younger ones. The difference between men and women is most pronounced for the morale component and negligible for social interaction. The age differential is least for morale. Single men are at a disadvantage in terms of poor social interaction and, if aged under 80 years, of home management. On the other hand single women are at an advantage with regard to morale. Older widowed people have greater chances of poor home management, and younger ones of poor morale, than others of their age group and gender.
Variation in chances of poor quality of life by housing tenure and dependency
Having taken account of gender and age differences between tenure groups, 'independent' people in rented homes have 50-70 per cent excess odds of a poor quality of life outcome compared to those in owner-occupied ones. Among the dependent groups, there is no clear tenure differential for home management and body care and movement but otherwise the differentials are similar to those in independent groups. People in the dependent groups are more likely to have each of the three physical SIP outcomes than their independent counterparts (not shown); this is not true of poor social interaction or morale. People in supported homes are no more likely than those in the dependent rented groups to have poor physical quality of life or poor social interaction outcomes but are a little more likely to have poor morale (not shown). People in supported housing are not considered further here.
Personal factors associated both with poor quality of life and with housing tenure
An association between housing tenure and quality of life could be attributable to a third factor which is both more common among the more socio-economically disadvantaged and a cause of poor quality of life. For a sub-sample of 73 per cent of the quality of life interviewees, there was information on health, lifestyle and social contact. For five out of six health problems there are higher percentages of people in rented homes than in owner-occupied ones among the independent groups. For the dependent groups this was true for three health problems. People with the health problem have higher prevalence of poor quality of life scores than those without. After taking account of gender, age and marital status, the six problems jointly account for some of the differences in the chances of poor quality of life by housing tenure. Having a poor outcome on one of the three physical sickness impact profile dimensions is in itself a risk factor for poor morale in addition to most of these six health problems and this accounts for a small part of the difference in chances of low morale in the two sectors (both for the independent and the dependent groups).
With respect to lifestyle, there is an inverse association between alcohol consumption and prevalence of all the measures of poor quality of life. Non-smokers are less likely to have poor social interaction or morale than smokers. These two factors account for a small part of the tenure differentials among the independent people but have negligible impact on the tenure differentials among dependent people - there is some evidence that variation in quality of life by smoking or alcohol consumption is smaller among people in the 'dependent' circumstances. Self-perceived physical activity is only considered in relation to social interaction and morale, as many of the components of the physical SIP instruments could be considered as physical activity. It is strongly negatively associated with poor social interaction and morale; the physically inactive have about seven times the odds of poor scores as the very active. It noticeably attenuates the tenure differentials particularly for the dependents and has a lesser impact on the differential among the independents. Greater frequency of contact, for example with relatives, friends and neighbours, is accompanied by lower prevalence of poor morale. Frequency of external contact is similar for owner-occupiers and renters, and therefore does not contribute to tenure differentials. Information was available on
informal help received for practical tasks and on use of formal services. In practice, those who receive help are more likely to have poor morale than those who do not, even after allowing for the presence of poor physical quality of life. People in rented homes are more likely to receive informal help than those in owner-occupied homes (among independents 37 per cent of renters and 30 per cent of owner-occupiers received help; among the dependent groups the percentages were 65 per cent and 59 per cent respectively). However, instead of helping to offset the tenure differentials in morale, it makes no difference to them.
Characteristics of the areas in which people live
The personal scores could be linked to the Carstairs score and the population density (persons per square kilometre) for the enumeration district (ED) in which the participant lived, and to the mean and standard deviation of the Carstairs scores for adjacent EDs. The Carstairs score is an unweighted combination of four standardised variables: percentages of unemployed, of unskilled social class, in overcrowded housing, and without a car.
The clearest patterns that emerged were:
- Continuing differentials between renters and owneroccupiers among independent groups with odds ratios of the order of magnitude of 1.5-1.7 for all five outcomes;
- Continuing differentials between renters and owneroccupiers among dependent groups for social interaction and morale (the only ones which have statistically significant differentials before adjustment for area) - the odds ratios are 1.6;
- People in the least deprived area have a lower chance of poor mobility than people in other areas; Higher average level of deprivation in surrounding EDs being associated with higher chances of poor home management or poor body care and movement;
- Greater population density being associated with higher chances of poor morale.
Socio-economic status in mid-life and later life
People in class IIIM or IV/V are more likely to have poor scores than those in classes I/II. For every outcome, having been in local authority accommodation most of their lives increased risk of poor outcomes, even after adjusting for social class. However, ever having owned a car does not add to these two socio-economic measures. Knowing social class and housing tenure during earlier adult life reduces, but does not remove, residual differences in the chances of poor quality of life by housing tenure in old age.
Changes in housing tenure
Quality of life was analysed by change in tenure between 'most of adult life' and time of interview for those who were independent at the latter time (numbers were too small in the dependent groups). Nearly three-quarters of the participants were still in 'independent' tenure groups in old age. The largest group (43 per cent of the whole sample or 59 per cent of the independent group) reported owner-occupation for most of their adult life and at the time of interview and were the least likely to have any of the five outcomes. People who were in social sector housing at both times have about twice the chances of any one of the poor quality of life outcomes as people who were in owner-occupation both times. People who moved 'down' from owner-occupation to social sector housing were more likely to have poor body care and movement score, social interaction score, or morale than those who remained in owner-occupation such that they were not significantly better off in these respects than people who had stayed in social sector housing. On the other hand people who moved 'up' from social sector housing to owner-occupation were less likely to have poor home management or social interaction score than those who stayed in social sector, although, except for social interaction, they were still worse off with respect to quality of life than people who had stayed in owner-occupation. Being in a manual social class is associated with greater chances of poor quality of life than being in a non-manual class for people who are independent and in owner-occupation in old age. However, there are not such clear hierarchical distinctions by class among people in independent social housing in old age. Taking a different perspective, within social classes I-IIIM chances of poor quality of life appear greater among those in independent social housing in old age than among those in owner-occupation but not within social class IV/V.
Conclusions
Among the independent older people, adjusting for health factors leads to the greatest reduction in estimated chances of poor quality of life comparing renters with owner-occupiers, followed by the combined effects of smoking and alcohol consumption. Measures that reduce differences in perceived health between tenure groups - e.g. reducing or coping with shortness of breath, might reduce the differences in prevalence of poor quality of life. Social contact is not responsible for the differentials although if renters had experienced less frequent contact with friends and neighbours the differences between them and owner-occupiers could have been greater.
Among the smaller groups of dependent people, there is some suggestion that renters are worse off than owner-occupiers with respect to poor physical sickness impact profile outcomes but the differences are not statistically significant. Health differences are smaller in these groups and differences in smoking and alcohol consumption appear to have weaker associations with the physical SIP dimensions. However, the dependent people in rented accommodation have twice the chances of poor social interaction or morale compared to people in owner-occupation. Self-perceived physical inactivity explains some of the differentials but may itself be a manifestation of poor morale. Quality of life was measured at the same time as the health problems and other factors so that the chronological sequence of events is unknown. However, the results can be used to generate ideas about options that might be worth pursuing. It is hypothesised that there is a cumulative process whereby, for example, health problems lead to functional limitations that lead to further health problems and a continuing feedback occurs.
There is mixed evidence as to whether support might reduce chances of poor morale or not. Despite greater restrictions in home management, mobility and body care and movement among the dependent than the independent group, the dependent group were no more likely to report low morale than their counterparts in the same tenure group. This might reflect support they received but, in contrast to this, receiving informal practical help or formal services carried increased odds of low morale. We do not know whether morale would have been even worse without this help because the survey was cross-sectional.
Greater deprivation of the area the person lived in is associated with greater chances of poor quality of life when personal housing tenure and other area factors are not taken into account. There are some area effects in addition to those of personal housing tenure for all but social interaction. The deprivation level of the immediate area is only a significant factor for mobility, the least deprived areas performing better than the rest. Deprivation in a wider area than one's own Enumeration District is an independent factor for home management and body care and movement whereas population density is positively associated with chances of poor morale once other factors are taken into account. It can only be surmised why these associations arise: the 'user-friendliness' of the immediate environment might affect mobility, services available in a wider area might affect home management and body care and movement, and the general alienation of more densely populated areas might affect morale.
Socio-economic status in middle age as well as old age matters for quality of life in old age. For all outcomes, social class, housing tenure most of adult life, and housing tenure in old age if independent, have cumulative effects on quality of life. People in social housing at both stages of their life are most consistently prone to the various aspects of poor quality of life relative to people in owner-occupation. A 'downward' transition from owner-occupation to social housing is clearly accompanied by greater chances of three outcomes and odds not significantly any better than those for people who have been in social housing during working life. Even for the two outcomes where the 'downward' transition does not significantly raise the overall odds of poor scores, the reason for the poor scores differs. Health reasons are more likely to be given by those who have changed tenure than by those who have stayed in owner-occupation (not shown). People who moved 'up' from social housing to owner occupation and remained non-dependent are at a clear advantage over those who had not changed 'up' for social interaction and home management. Although the sequence of events cannot be disentangled in this study, the results show sufficient evidence of differences in chances of poor quality of life by socio-economic status to warrant concern.
About the study
The interviewees came from 23 General Practices in Britain that were taking part in a large-scale Trial of the Assessment and Management of Elderly People in the Community, funded by the Medical Research Council (the MRC Trial) and set within general practice. Anyone aged 75 years and over at the time of the assessment was eligible provided that they were still under the care of the General Practitioner, were neither in long-stay nursing care nor terminally ill. Quality of life, which is one of the trial outcomes, was assessed at baseline, and during a 36-month follow-up period. Trained interviewers administered a questionnaire that contained standard quality of life instruments and questions about socio-economic attributes and about use of informal and formal services. The General Practice nurse then invited eligible people to complete a brief assessment designed to cover the seven areas recommended for the annual screening that General Practitioners are contracted to undertake for this age group, namely sensory function, mobility, mental condition, physical condition, use of medicines, and social environment. In total, 9573 people on the age-sex registers of the 23 practices were eligible for the MRC Trial and quality of life questionnaires are available for 8734 (91 per cent). Their median age was 80.3 years. There are brief assessments for 6405 people, 73 per cent of those with quality of life information. Quality of life was measured by four sets of questions from the Sickness Impact Profile (SIP) and the Philadelphia Geriatric Morale Scale (PGMS). Three of the SIP dimensions refer to physical aspects of quality of life: home management, mobility, and body care and movement and the fourth one is social interaction.
These findings concern the relative chances of being in the worst fifth of scores for these five indicators (four SIP dimensions and PGMS) according to demographic and socio-economic position. The groups with the worst 20 per cent of scores for each quality of life indicator were substantially more likely than others to report any one of the problems which were included in that indicator and clearly had a poorer quality of life.
Further details
A full version (with tables and references) is available at the GO Programme website: http://www.shef.ac.uk/uni/projects/gop/gop8.html/.
Investigators on this GO project were Elizabeth Breeze, Chris Grundy, Professor Astrid Fletcher, Dr Paul Wilkinson at the London School of Hygiene and Tropical Medicine. Dr Dee Jones and Amina Latif from Cardiff organised and ran the Quality of Life component of the study. Professor Christopher Bulpitt at Imperial College is an investigator on the MRC Trial.
Key words
Older age, quality of life, inequality
View all other award details