IJE Advance Access originally published online on June 17, 2005
International Journal of Epidemiology 2005 34(4):750-754; doi:10.1093/ije/dyi125
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Article |
Cohort Profile: The Boyd Orr lifegrid sub-sample—medical sociology study of life course influences on early old age
Department of Primary Care and Social Medicine, Imperial College London, St Dunstan's Road, London W6 8RP, UK. E-mail: dblane{at}imperial.ac.uk
 |
How did the study come about? |
|---|
Top
How did the study...
What does the study...The study had its origins in Sir John Boyd Orr's 1937–39 study of the diet and health of British children. In the mid-1990s, the surviving study participants were traced to their present-day locations by means of the National Health Service Central Register (for details, see linked article).1 This follow-up after 60 years was organised by Stephen Frankel, George Davey Smith, and David Gunnell, in pursuit of their interest in life course influences on adult health—an interest that had developed in the context of the then current work on birthweight and later health by David Barker and his colleagues.
In 1996, the UK Economic and Social Research Council (ESRC) funded intensive study of a stratified random sample of the surviving members of the full Boyd Orr cohort. This funding was a part of the ESRC Research Programme Social Variations in Health. The study was lead by David Blane and George Davey Smith, who had worked together on the West of Scotland Collaborative Study, where they had developed ideas about the impact on health of accumulation across the life course of socio-economic disadvantages. The study of this stratified random sample of the Boyd Orr cohort, called the lifegrid sub-sample here, was designed to identify life course and contemporary influences on health in early old age (in this profile, the term early old age refers to the phase of the life course between labour market exit and the onset of physical dependency; it does not refer to a fixed chronology, although its core can be seen as the years 60/65–75, with a periphery stretching 55–85 years). Fieldwork for the ESRC Social Variations in Health project was conducted during 1997–98.
Members of the lifegrid sub-sample were re-visited by postal questionnaire during the year 2000, in a study funded by the ESRC Research Programme Growing Older. The study was lead by David Blane, Paul Higgs and Richard Wiggins and was designed to identify life course and contemporary influences on quality of life in early old age. At the same time, individuals drawn from the two extreme one-thirds of the distribution of Healthy Diet Index scores (for details, see linked article) were interviewed in-depth. This study was funded by the LINK Research Programme Eating, Food and Health, which involved ESRC, the Biotechnology and Biological Science Research Council and the Ministry of Agriculture, Fisheries and Food, with matched funding from Research into Ageing. The study was designed to identify life course and contemporary influences on diet in early old age and was lead by David Blane, David Gunnell, Andrew Ness, and Maria Maynard.
All members of the lifegrid sub-sample are being visited for a third time (or fourth time, if the original 1937–39 survey is taken into account) during 2004–05. The study is funded by the ESRC Priority Research Network Capability and Resilience across the life course. The study is designed to identify protective factors, both cross-sectional and longitudinal, which make it possible for individuals to flourish despite adversity. The study is lead by David Blane, Scott Montgomery, and Richard Wiggins.
|
What does the study cover—and how has this changed? |
|---|
The focus of the 1997–98 study was social inequality across the life course and its influence on health in early old age. Besides measuring social class at each stage of the life course, the study was concerned with measuring exposure to the material and psychosocial hazards that are the biologically plausible links between social class and health. A person's social class position was seen as rooted in both the sphere of production and the sphere of consumption. Exposure to occupational fumes and dust, physically arduous work and low job control were seen as the main biologically plausible hazards belonging to the sphere of production; and exposure to atmospheric pollution in the area of residence, residential damp and inadequate nutrition are the main biologically plausible hazards belonging to the sphere of consumption (exposure to tobacco smoke was included as a control variable). The level of exposure to any one of these hazards was seen as varying with the specific nature of a person's occupation and geographical locality (agricultural labourer versus assembly line worker, for example; or fishing village versus industrial conurbation). Beyond these specifics, however, a person's social class was seen as the main determinant of, and as being indexed by, their total exposure to all the hazards combined. These exposures could be summed cross-sectionally or accumulated across the life course. Health in early old age was measured as self-reported long-standing illness, self-reported presence of serious disease, researcher-observed prescribed medication and researcher-measured anthropometric and physiological status.
Follow-up studies of the lifegrid sub-sample shifted their focus somewhat from a primary concern with health. The age of the study members happened to coincide with growing scientific and policy interest in a range of issues associated with the ageing population. Social gerontology, as a result of the contemporary increase in healthy life expectancy and the spread of second pensions, had widened its focus to include such issues as how to postpone the onset of physical dependency and how to maximise quality of life during the so-called Third Age (synonymous with early old age, in the above described sense). This coincidence between study characteristics and socio-demographic change was reinforced by a further coincidence between the longitudinal nature of the study and growing scientific acceptance of the life course approach, which had changed from being innovative to near-orthodoxy. The result was a series of follow-up studies, the foci of which were determined by a succession of UK Research Council programmes, although each study was centred around social inequalities across the life course.
The first of these follow-up studies concerned quality of life in early old age, for which we developed a new measure. The new measure was designed to: (i) include the positive aspects of life in early old age; (ii) be based in contemporary social theory about modern society and the place within it of early old age; and (iii) be independent of the factors that might influence it, such as health, friendship, and financial circumstances. At the same time a parallel follow-up study was concerned with diet and nutrition in early old age, for which we developed a new measure based on the level of agreement between a person's diet and the detailed food group recommendations of the UK Department of Health's Committee on the Medical Aspects of Nutrition. Those whose diets were best and worst according to this measure were interviewed in-depth about any changes in their diet over the years and the factors that may have influenced these changes. The most recent, and presently on-going, follow-up study is concerned with changes in health and quality of life during early old age. The study is guided by social psychology's concept of resilience (flourishing despite adversity) and Amartya Sen's concept of capabilities (freedom to flourish). Based on disparate intellectual sources (Berlin Ageing Study; Ray Pahl; Richard Wilkinson; Amartya Sen), new measures such as styles of coping with adversity, types of social networks, experience of equality in relationships, and capabilities across the life course have been developed.
|
Who is in the sample? |
|---|
The lifegrid sub-sample is a stratified random sample of surviving, traced members of the Boyd Orr cohort, who were aged between 5 and 14 years at the time of the original survey and who were physically examined. Among the information collected in 1937–39 were details of household per capita food expenditure. We used this as our measure of childhood socio-economic circumstances, because the father's occupation was often recorded as ‘unemployed’, which prevented allocation to any one of the Registrar General's social classes. The cohort was stratified into three groups of equal size on the basis of household per capita food expenditure during the subject's childhood. Power calculations had indicated that the research questions of the ESRC Social Variation in Health study required 100 interviews from each stratum. The names of subjects who comprised the random sample were listed in random order. From this list sample, subjects were invited consecutively for interview up to the point of completion of 100 interviews in each stratum. Non-responders to the first invitation were sent a reminder letter.
At the first mailing, 257 subjects volunteered for interview and 37 at the second mailing giving a total of 294 interviews. A further 126 failed to respond to mailing and 259 subjects refused an interview. Ninety-nine subjects were lost between drawing the sample and contacting for interview, of whom 59 could not be traced by the health authority, 24 had died, 13 declined an interview because they were seriously ill, and 3 were ineligible because of mistaken identity at tracing by the NHS Central Registry. We learned subsequently that around three-quarters of these ‘failed contacts’ arose because of minor discrepancies between the research database and the health authority records in the NHS number and the spelling of the subject's surname. The remaining one-quarter of ‘failed contacts’ comprised subjects who recently had moved away from the health authority's jurisdiction, emigrated or who were no longer registered with a General Practitioner. The 294 subjects who were interviewed represented a response rate of 39% if the dead and ineligible were excluded and 43% if, in addition, the ‘not traced’ and ‘seriously ill’ were excluded. The interviewed subjects, when compared with those who refused an interview or failed to reply to the request for interview, were less disadvantaged during childhood in terms of affluence, social class and health. The sample, in consequence, was biased conservatively.
Those interviewed proved reasonably representative, in several key respects, of their age-peers within the British population. The social class distribution of the fathers of those interviewed in 1937–39 was similar to the social class distribution of all males aged 20–44 years at the 1931 decennial census (the main differences were a slight over-representation of Social Class II and a slight under-representation of Social Class III). The socio-demographic characteristics of those interviewed at the follow-up in 1997–98 were similar to the socio-demographic characteristics of those aged 65–74 years at the 1991 decennial census (the main differences were a higher proportion living in rural areas and a lower proportion lacking access to a motor car). Finally, the self-reported and measured health of those interviewed in 1997–98 was broadly similar to those aged 65–74 years in the 1995 Health Survey for England (interviewees had somewhat higher blood pressure and slightly higher lung function).
In summary, the initial sample had a low response rate and is biased conservatively, probably as a result of the stratified sampling. Nevertheless the sample is broadly representative of its age-peers within the British population. In this sense, it perhaps gives a unique insight into the generation that precedes the first of the prospective British birth cohorts, born in 1946.1
We returned to the sample, by postal survey, in the year 2000. Of the 294 people interviewed in 1997–98, 12 had died or were seriously ill or untraceable. The remaining 282 were mailed a self-completion questionnaire with, where appropriate, a second mailing after 14 days. Those who did not respond to either mailing were interviewed, if willing, by telephone. In total, 264 cohort members completed the questionnaire (182, first mailing; 74, second mailing; 8, telephone), representing a response rate of 90% of those interviewed in 1997–98 and 93% of those asked to complete the questionnaire. Those lost to study (died; seriously ill; untraceable; non-responders to postal and telephone survey) did not differ from the survey respondents in terms of their gender and social class composition, although their rate of limiting long-standing illness was somewhat higher.
In parallel with the postal survey, in-depth interviews were conducted for the LINK Programme Eating, Food and Health. Purposive sampling was used to select potential interviewees. Interviewees were selected on the basis of age, gender, current living arrangements (with spouse or family vs alone), income, geographical location and health—all of which previous research had shown to be associated with variations in diet. Interviewees were selected also on the basis of their current diet, with half of those selected having a poor diet and half a good diet, as indicated by their Healthy Diet Index score. Potential interviewees were selected from the upper and lower one-thirds of the distribution of Healthy Diet Index scores to ensure that a wide range of eating habits was represented. Thirty-one people were interviewed: 16 with poor diets (9 men; 7 women) and 15 with good diets (8 men; 7 women), with the interviewer blind to each interviewee's Healthy Diet Index status.
The present Cohort Profile is being written midway through the fieldwork for the 2004–05 survey, so we are unable to report on the numbers and characteristics of those in the latest sample.
|
How often have they been followed up? |
|---|
The dataset contains information from the following surveys: (i) Original Boyd Orr survey 1937–39 (data collected by interview); (ii) Social Variations in Health survey 1997–98 (data collected by interview); (iii) Healthy Diet Index scores from food frequency questionnaire survey 1997–98 (data collected by postal questionnaire); (iv) Growing Older survey Year 2000 (data collected by postal questionnaire); (v) Eating, Food and Health study 1998 (data collected by in-depth interview of purposively selected sub-sample); and (vi) Capability and resilience survey 2004–05 (on-going data collection by interview).
|
What has been measured? |
|---|
|
|
What is attrition like? |
|---|
For details of attrition between the original Boyd Orr survey in 1937–39 and the re-establishment of contact in 1996, See linked article. Of the 294 people interviewed in 1997–98, 12 had died or were too ill to participate further or untraceable by the postal survey in the year 2000. Information on rate of attrition at the home interview survey in 2004–05 will not be available until late 2005.
|
What has it found—key findings and publications? |
|---|
Accumulation of disadvantage
Adult social class acts as an indicator of accumulated lifetime exposure to a range of biologically plausible hazards located in the spheres of production and consumption.2 Disadvantage during childhood, whether expressed in terms of health3 or socio-economic circumstances,4 predicts forwards to the level of exposure to biologically plausible hazards that will be accumulated across the life course. Disadvantage during early old age, whether expressed in terms of housing tenure, receipt of welfare benefits or social class of working life occupation,5 predicts backwards to the level of exposure to biologically plausible hazards that have been accumulated across the life course.
Health
There are at least three ways in which life course influences health in early old age. In some cases, such as lung function, the influences may accumulate across the whole life course, involving factors in childhood, adulthood and early old age. Alternatively, the later life relationship may have been determined, as in the case of adult height, at a much earlier stage of life. Finally, as in the case of blood pressure, the relationship may be conditional in that factors from different stages in the life course have to occur sequentially before the later life effect is produced.6 The case of blood pressure in early old age shows that later health can be influenced by life course processes which are complex. Slow pre-pubertal growth during childhood predicts systolic blood pressure and pulse pressure during early old age.7 Child height predicts the level of lifetime exposure to low job control.3 Slow pre-pubertal growth interacts with low job control during adulthood to further raise systolic blood pressure and pulse pressure during early old age.8 These findings can be interpreted in terms of life course processes as a critical period effect (pre-pubertal growth associated with the formation of mechanisms controlling blood pressure in later life), the accumulation of disadvantage (childhood disadvantage, as indicated by child height, predicting later exposure to low job control) and interaction between a critical period effect and an accumulation effect (raised blood pressure associated with interaction between slow child growth and high adult exposure to low job control.4 The ability to explain variation in the health of women in early old age is increased, but only modestly, when lifetime accumulated domestic labour is added to the hazard exposures accumulated via formal, paid employment.9
Nutrition
People in early old age often describe long-term influences on their current nutrition; for example, those who had eaten a lot of fruits and vegetables as children, usually from rural areas, continued to do so throughout their life and were high consumers of fruit and vegetables in early old age. Such relationships were not fixed; some with urban childhood acquired in early old age a taste for fruit and vegetables, perhaps from a fitness club joined to maintain a youthful appearance.10 People in early old age tend to describe their diet as good, irrespective of its objective nutritional content. Medical advice about nutrition is valued by this age group, but little was available despite numerous occasions (onset of chronic disease, bereavement, shift of residence) when contact with General Practitioners was frequent and such advice would have been welcome.11
Quality of life
No single factor has a decisive influence on the positive quality of life in early old age, as measured by CASP-19;12 instead, quality of life is best seen as a summary response to the whole range of factors that constitute existence. Some of these derive from earlier stages of the life course (level of accumulated disadvantage; forced early retirement), but current circumstances in early old age are of greater importance.13 Key predictors of quality of life in early old age are current financial circumstances, health status, the quality and emotional closeness of social contacts and the nature of the neighbourhood environment.14
Methodology
Resuscitating a long-forgotten study15 and using a lifegrid to collect retrospective information from the missing decades16,17 are useful methodological innovations.
|
What are the main strengths and weaknesses? |
|---|
The main strengths are: (i) the study contains full life course information on the generation before the first of the British birth cohort studies who were born in 1946; (ii) currently the cohort is completing the transition through early old age—an important age group for current policy concerns.
The main weaknesses are: (i) small number of subjects, so low statistical power; (ii) selective loss to the study of the more disadvantaged individuals, through death and so forth, although the resulting bias is conservative; (iii) failure to collect information on the use of boarding schools and the employment of domestic servants, with consequent social class bias in the calculation of domestic labour.
|
Can I get hold of the data—where can I find out more? |
|---|
The datasets from the Social Variation in Health study and the Growing Older study are deposited at the UK Data Archive.
Additional information can be obtained from David Blane at d.blane{at}imperial.ac.uk.
|
Complete list of publications |
|---|
Wiggins R, Higgs P, Hyde M, Blane D. Quality of life in the third age: key predictors of the CASP-19 measure. Ageing Soc 2004;24:693–708.[CrossRef][Web of Science]
Gunnell D, Berney L, Holland P, Maynard M, Blane D, Davey Smith G, Frankel S. Does the misreporting of adult body size depend upon an individual's height and weight? Methodological debate. Int J Epidemiol 2004;33:1–2.
Blane D, Higgs P, Hyde M, Wiggins R. Life course influences on quality of life in early old age. Soc Sci Med 2004; 58:2171–79.[CrossRef][Medline]
Blane D, Abraham L, Gunnell D, Maynard M, Ness A. Background influences on dietary choices in early old age. J R Soc Health 2003;123:204–09.[Medline]
Berney L, Blane D. The life grid method of collecting retrospective information from people at older ages. Res Policy Planning 2003;21:13–22.
Higgs P, Hyde M, Wiggins R, Blane D. Researching quality of life in early old age: the importance of the sociological dimension. Social Policy Administration 2003;37:239–52.
Hyde M, Wiggins R, Higgs P, Blane D. A measure of quality of life in early old age: the theory, development and properties of a needs satisfaction model (CASP-19). Aging Ment Health 2003;7:186–94.[CrossRef][Web of Science][Medline]
Abraham L, Blane D, Gunnell D, Maynard M, Ness A. A qualitative study of healthy and unhealthy eaters in early old age. Proc Br Psychol Soc 2003;11:114.
Abraham L, Blane D, Gunnell D, Maynard M, Ness A. Influences on diet in early old age. Proc Br Psychol Soc 2003;11:48.
Hyde M, Blane D, Higgs P, Wiggins R. Influences on quality of life in early old age. Findings Series, ESRC Growing Older Programme 2003.
Blane D, Berney L, Montgomery S. Domestic labour, formal employment and women's health. Soc Sci Med 2001; 52:959–65.[CrossRef][Medline]
Hyde M, Blane D, Higgs P, Wiggins R. Changing parameters of Ageing in the UK: evidence for the existence of the Third Age. Gerontology 2001;47(suppl 1):20.
Higgs P, Blane D, Wiggins R, Hyde M. Quality of life in early old age: development of a needs satisfaction model. Gerontology 2001;47:677.
Hyde M, Blane D, Higgs P, Wiggins R. Development of a sociological measure of quality of life in early old age. ESRC Growing Older Programme Newsletter 2001;4:5.
Gunnell D, Berney L, Holland P, Maynard M, Blane D, Frankel S, Davey Smith G. How accurately are height, weight and leg length reported by the elderly and how closely are they related to measurements recorded in childhood? Int J Epidemiol 2000;29:456–64.
Berney L, Blane D, Davey Smith G, Gunnell D, Holland P, Montgomery S. Socioeconomic measures in early old age as indicators of previous lifetime exposure to environmental health hazards. Sociol Health Illn 2000;22:415–30.[CrossRef]
Montgomery S, Berney L, Blane D. Pre-pubertal growth and blood pressure in early old age. Arch Dis Child 2000; 82:358–63.
Holland P, Berney L, Blane D, Davey Smith G, Gunnell D, Montgomery S. Life course accumulation of disadvantage. Soc Sci Med 2000;50:1285–95.[CrossRef][Web of Science][Medline]
Berney L, Blane D, Davey Smith G, Holland P. Life course influences on health in early old age. In: Graham H (ed). Understanding Health Inequalities. Buckingham: Open University Press, 2000, pp. 79–95.
Holland P, Berney L, Blane D, Davey Smith G. Life course influences on health in early old age. ESRC Health Variations Programme Newsletter 6, September 2000.
Blane D, Berney L, Davey Smith G, Gunnell D, Holland P. Reconstructing the life course: a 60 year follow-up study based on the Boyd Orr cohort. Public Health 1999; 113:117–24.[CrossRef][Web of Science][Medline]
Holland P, Berney L, Blane D, Davey Smith G. The lifegrid method in health inequalities research. ESRC Health Variations Programme Newsletter 1999, pp. 8–9.
Blane D, Montgomery S, Berney L. Social class differences in lifetime exposure to environmental hazards. Sociol Health Illn 1998;20:532–36.[CrossRef]
Berney L, Blane D. Collecting retrospective data: accuracy of recall after 50 years judged against historical records. Soc Sci Med 1997;45:1519–25.[CrossRef][Web of Science][Medline]
Blane D. Collecting retrospective data: development of a reliable method and a pilot study of its use. Soc Sci Med 1996;42:751–57.[CrossRef][Medline]
|
Acknowledgments |
|---|
I am most grateful to the study participants for their continuing support of our research and to my colleagues who have worked on its various components—Lucy Abraham, Lee Berney, George Davey Smith, David Gunnell, Paul Higgs, Zoe Hildon, Paula Holland, Martin Hyde, Maria Maynard, Scott Montgomery, Andrew Ness, and Richard Wiggins.
|
References |
|---|
1 Martin RM, Gunnell D, Pemberton J, Frankel S, Davey Smith G. Cohort Profile: The Boyd Orr cohort—an historical cohort study based on the 65 year follow-up of the Carnegie Survey of Diet and Health (1937–39). Int J Epidemiol 2005;34:742–9.
2 Wadsworth MEJ. The Imprint of Time: Childhood, History and Adult Life. Oxford: Clarendon Press, 1991.
3 Blane D, Montgomery S, Berney L. Social class differences in lifetime exposure to environmental hazards. Sociol Health Illn 1998; 20:532–36.
4 Holland P, Berney L, Blane D, Davey Smith G, Gunnell D, Montgomery S. Life course accumulation of disadvantage: childhood health and hazard exposure during adulthood. Soc Sci Med 2000; 50:1285–95.[CrossRef][Web of Science][Medline]
5 Blane D. The life course, the social gradient and health. In: Marmot M, Wilkinson R (eds). Social Determinants of Health. 2nd edn. Oxford: Oxford University Press, 2005.
6 Berney L, Blane D, Davey Smith G, Gunnell D, Holland P, Montgomery S. Socioeconomic measures in early old age as indicators of previous lifetime exposure to environmental health hazards. Sociol Health Illn 2000;22:415–30.
7 Holland P, Berney L, Blane D, Davey Smith G. Life course influences on health in early old age. ESRC Health Variations Programme Newsletter 6, September 2000.
8 Montgomery S, Berney L, Blane D. Pre-pubertal growth and blood pressure in early old age. Arch Dis Child 2000;82:358–63.
9 Berney L, Blane D, Davey Smith G, Holland P. Life course influences on health in early old age. In: Graham H (ed). Understanding Health Inequalities. Buckingham: Open University Press, 2000, pp. 79–95.
10 Blane D, Berney L, Montgomery S. Domestic labour, formal employment and women's health. Soc Sci Med 2001;52:959–65.[CrossRef][Medline]
11 Blane D, Abraham L, Gunnell D, Maynard M, Ness A. Background influences on dietary choices in early old age. J R Soc Health 2003;123:204–09.[Medline]
12 Abraham L, Blane D, Gunnell D, Maynard M, Ness A. Older people's accounts of their diet. J Health Psychol (in press).
13 Hyde M, Wiggins R, Higgs P, Blane D. A measure of quality of life in early old age: the theory, development and properties of a needs satisfaction model (CASP-19). Aging Ment Health 2003;7:186–94.[CrossRef][Web of Science][Medline]
14 Blane D, Higgs P, Hyde M, Wiggins R. Life course influences on quality of life in early old age. Soc Sci Med 2004;58:2171–79.[CrossRef][Medline]
15 Wiggins R, Higgs P, Hyde M, Blane D. Quality of life in the third age: key predictors of the CASP-19 measure. Ageing Soc 2004; 24:693–708.
16 Blane D, Berney L, Davey Smith G, Gunnell D, Holland P. Reconstructing the life course: a 60 year follow-up study based on the Boyd Orr cohort. Public Health 1999;113:117–24.[CrossRef][Web of Science][Medline]
17 Blane D. Collecting retrospective data: development of a reliable method and a pilot study of its use. Soc Sci Med 1996;42:751–57.[CrossRef][Medline]
18 Berney L, Blane D. Collecting retrospective data: accuracy of recall after 50 years judged against historical records. Soc Sci Med 1997;45:1519–25.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  Z. Hildon, S. M. Montgomery, D. Blane, R. D. Wiggins, and G. Netuveli Examining Resilience of Quality of Life in the Face of Health-Related and Psychosocial Adversity at Older Ages: What is "Right" About the Way We Age? Gerontologist, June 23, 2009; (2009) gnp067v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. M Martin, D. Gunnell, J. Pemberton, S. Frankel, and G. Davey Smith Cohort Profile: The Boyd Orr cohort--an historical cohort study based on the 65 year follow-up of the Carnegie Survey of Diet and Health (1937-39) Int. J. Epidemiol., August 1, 2005; 34(4): 742 - 749. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||