IJE Advance Access originally published online on November 7, 2006
International Journal of Epidemiology 2007 36(1):58-65; doi:10.1093/ije/dyl241
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Self-rated life expectancy and lifetime socio-economic position: cross-sectional analysis of the British household panel survey
Research Unit in Health, Behaviour and Change, University of Edinburgh Medical School, Teviot Place, Edinburgh EH8 9AG, UK.
* Corresponding author: E-mail: f.popham{at}ed.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionConclusionReferences |
|---|
Background The association between mortality risk and socio-economic position (SEP) across the lifecourse is established. This study investigates whether people's own ratings of their life expectancy are also associated with lifetime SEP. Health behaviour messages, which often emphasize the long-term benefits of behavioural change, may be received differently depending on people's perceptions of their life chances.
Methods Cross-sectional analysis of 4780 adults aged 25–64 interviewed in the British Household Panel Survey in 2001.
Results Just under a quarter of respondents did not think it likely they would live to 75 or older. People in lower SEPs were more likely to be pessimistic about their life expectancy. This applied across a number of socio-economic measures (father's social class, educational achievement, own social class, and household income). Eight socio-economic lifecourse pathways were compared. In comparison to those following the most advantaged pathway, those experiencing sustained socio-economic disadvantage were most likely to be pessimistic about their longevity, but those experiencing sustained upward mobility did not differ. Comparisons with measures of self-rated general health and limiting illness suggest that self-rated life expectancy is at least partially independent of current health status.
Conclusions This study shows that people's own perceptions of their life expectancy are associated with lifetime SEP. Self-rated life expectancy, in part, appears to reflect something over and above current health status and smoking behaviour. Given its ease of collection, it would be informative to include self-rated life expectancy in future studies.
Keywords Self-rated life expectancy, socio-economic position, lifecourse, BHPS
Accepted 5 October 2006
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
While life expectancy has increased year on year in the UK,1 social inequalities have continued to widen.2 The association between premature mortality risk and socio-economic position (SEP), measured at various points across the lifecourse, has been established.3 Recently, studies have reported people's own ratings of their life expectancy to be associated with their SEP.4,5 In the UK, more pessimistic self-ratings of life expectancy have been shown to be associated with lower social class.4 In a US study, pessimistic ratings were associated with low education and economic hardship.5 However, to date there has been little work assessing the association between lifetime SEP and self-rated life expectancy.
Self-rated life expectancy has often been used to investigate whether people's perceptions of their mortality risk reflect their actuarial predicted life expectancy, given their health behaviours. The aim of this type of work is to assess whether individuals understand the mortality risk which may result from their behaviour.6–8 However, anthropological research suggests that people have a sophisticated ‘lay’ epidemiological understanding of mortality risk, informed not only by simple health promotion messages, but also by the health experiences of family, friends, and wider societal experiences.9,10 Health promotion messages are widely recognized by their target audiences, but behavioural change amongst particular groups has been difficult to achieve. One clear example of this is the relative failure of smoking cessation efforts amongst lower socio-economic groups.11 It has been suggested that this failure to heed ‘the message’ should be understood as a rational response of these groups given their more limited health and life chances.11 Without improvements in SEP and associated improvements in health and life chances, it is argued, behavioural change is less likely.11
Furthermore, it is not clear whether self-rated life expectancy simply reflects experiences and perceptions of morbidity. In previous studies, an association between self-rated life expectancy and social class was largely explained by adjustment for self-rated health,4 while associations remained with education and economic hardship after adjustment.5 Moreover, self-rated life expectancy has been shown to be predictive of mortality after adjustment for general health.12,13
Risk of morbidity and mortality are, in general, strongly associated with exposure to disadvantage over the lifecourse and three principle mechanisms by which this occurs have been identified.14–23 The first is ‘critical period’, in which exposure to disadvantage at one life stage may have a particular impact on health. For example, low childhood SEP has been shown to have an association independent of adult SEP with mortality from some diseases.14 The second is ‘accumulated risk’, in which each exposure to disadvantage across the lifecourse cumulatively impacts on health. For example, mortality risk has been shown to increase with each exposure to disadvantage.15 The third is ‘social mobility’, in which SEP trajectories across the lifecourse are studied. This allows assessment of the impact of upward or downward social mobility on health. Evidence shows that both upward and downward mobility in adult SEP reduce and increase, respectively, mortality risk; however, mobile groups do not reach the risk level of their destination SEP.16 These three mechanisms are not necessarily mutually exclusive and separating their particular impact is not straightforward.17
In this study we adopt a pathway model similar to that used by Nicholson et al.18 to assess the impact of lifecourse SEP (childhood through education to present adult SEP) on self-rated life expectancy. We also report results for analyses using the ‘critical period’ and ‘accumulated’ exposure approaches to offer comparative perspectives.
The main aims of the study were thus:
- To assess the association between SEP across the lifecourse and self-rated life expectancy.
- To compare self-rated life expectancy with two established self-rated health measures to assess whether it captures something other than current health condition.
- To assess how different socio-economic pathways across the lifecourse are associated with self-ratings of life expectancy and health.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
The sample
The British Household Panel Survey (BHPS) began in 1991 with a representative sample of 10 264 adults aged 16 or over in 5511 households recruited through a stratified random sample with households clustered in selected postcode sectors. These original sample members have been resurveyed every year since. Children of sample members are added to the main sample on reaching 16 as are adults joining the households of sample members. A detailed overview of the BHPS's methodology is given elsewhere.24
Self-report measures
In the 2001/02 wave only, respondents aged under 65 were asked whether they thought it likely that they would live to be 75 or over (very likely, likely, unlikely, very unlikely, and do not know). For the present analysis we dichotomized responses as likely (very likely, and likely) and not likely (all other categories including do not know).
General health was assessed by the following question. ‘Please think back over the last 12 months about how your health has been. Compared to people of your own age, would you say that your health has on the whole been: excellent, good, fair, poor, or very poor?’ For the present analysis answers were dichotomized as excellent to good and fair to very poor.
Finally, respondents were asked ‘does your health in any way limit your daily activities compared to most people of your age? Yes or no’.
Socio-economic position
In health inequalities research, lifetime SEP is usually either assessed by repeating the same measure during a person's life15 or by collating different measures that reflect SEP at various time points in the lifecourse.18 Given the data available we chose the latter approach for this study. SEP was captured at three sequential points through the lifecourse: childhood, education, and current adult SEP.
SEP in childhood was based on the occupation of the respondent's father (or mother if the father was out of work or the father's class was unknown) when the respondent was aged 14. Education level was based on the individual's highest educational qualification to date and included both academic and vocational qualifications. It was initially coded into the following groups:
- Degree level.
- Other higher educational qualifications or their equivalent.
- A-level (a post-compulsory school qualification usually taken at age 17 or 18) or its equivalent.
- O-level (compulsory level qualification usually taken at age 16 now superseded by GCSE) or its equivalent and any qualifications below this level.
- None of these educational qualifications.
For current adult SEP we test two measures: occupational social class (based on respondent's most recent job) and household income.
Gross annual household income in the previous year was derived by the BHPS team from the detailed information collected. Some missing values were imputed.24 Income was adjusted to account for differences in household composition and size, using the McClements scale.25 Households were split into income quartiles (for all households including eligible people). Occupational social class was coded using the Registrar General's classification.
Health behaviours
Unfortunately, the BHPS does not measure alcohol consumption, dietary patterns, or, in the year of interest, physical activity. However, to assess the impact of present health behaviours on self-rated life expectancy and its relationship with SEP we included present smoking behaviour, measured by the person's average daily cigarette consumption.
Negative affect
To assess whether self-rated life expectancy simply reflects the negative affect associated with depression, we included a measure of potential psychiatric morbidity, the General Health Questionnaire 12 point version (GHQ12), coded using a standard cut-off with a 'case' being any person scoring 4 or more.26
Analysis
Sample attrition and additions since the original BHPS sample was recruited in 1991 have affected its representativeness. All analyses used the appropriate cross-sectional correctional weight for 2001, provided by the BHPS team. We used three-level multilevel logistic regression models to account for the non-independence of households resident in the same postcode sector and adults resident in the same households. To compare the relative association of the measures of SEP from the different periods of the lifecourse with self-rated life expectancy we used the relative index of inequality (RII), this can be taken as a measure of risk between the most advantaged and disadvantaged groups.27 For the lifecourse models we used only one of the measures of current SEP and chose occupational social class as this was the scale used to measure childhood SEP. To limit the number of possible categories in the accumulated and pathway lifecourse models we dichotomized our measures of SEP as follows.
- Occupational social class: non-manual/manual.
- Education: highest qualification post-compulsory school level and above (an A-level or vocational equivalent and above)/compulsory school level or below (an O-level, GCSE or vocational equivalent and below).
Stata version 9 was used for all analysis.28 Analysis was limited to those aged 25–64 who were not in full-time education at the time of the survey in 2001. As people could have gained their highest educational qualification after they had last worked we excluded those whose date of leaving full-time education was after they had left their most recent job. Data were missing from 17% of possible cases mainly (12%) due to parents' occupation not being recorded. People with missing information were more pessimistic about their life expectancy and rated their health worse. They were also significantly more likely to be in manual class occupations and live in low income households. Complete data on all variables of interest were available for 4780 individuals living in 3194 households in 1493 postcode sectors.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
Twenty-four per cent of individuals did not think it likely they would live to be 75, 28% rated their general health as fair to very poor, and 14% said they had a limiting illness. The self-reported measures of health and life expectancy were correlated with each other. The strongest association was between general health and limiting illness (0.50, P < 0.001), with weaker associations between self-rated life expectancy and general health (0.17, P < 0.001), and self-rated life expectancy and limiting long-term illness (0.15, P < 0.001). Figure 1 gives the joint distribution of all three self-rated measures and shows that just over half the sample expected to live to 75 or over, had excellent or good general health and did not have a limiting illness. It also shows that the majority of those who did not expect to live to 75 had neither a limiting illness nor poor general health.
|
Table 1 reports the distribution of the individual measures of SEP, smoking behaviour, and GHQ, and gives the associated odds ratios of reporting poor health, limiting illness, and belief that the respondent will not live to 75. Men were less likely to report poor health but more likely to be pessimistic about their life expectancy than women. Self-rated life expectancy was not significantly associated with age but both limiting illness and general health were, with highest rates in these measures amongst those aged 55–64. For all measures, lower SEP was associated with lower expectations of life and poorer health. Smoking was associated with poorer self-rated health and low self-rated life expectancy, with 49% of the heaviest smokers being pessimistic about their longevity. As expected, all measures of SEP were associated with smoking behaviour, with low SEP associated with increased smoking prevalence (results not shown). GHQ was significantly associated with all three self-rated measures with higher rates of possible psychiatric morbidity amongst people reporting low life expectancy and poorer health.
|
Lifetime socio-economic position
Table 2 shows results for the ‘critical period’ analysis. It presents the associations between the measure of SEP from each period of the lifecourse and self-rated life expectancy, adjusted for the other measures of SEP. After adjustment, education was still significantly associated, and also had the largest RII. Occupational social class was also significantly associated, but father's social class at age 14 was not.
|
Table 3 shows results for the ‘accumulated disadvantage’ analysis. Using the dichotomized versions of the measures of SEP, disadvantage was scored one and advantage zero and then summed across the three measures to produce an accumulated measure of SEP (range 0–3). Table 3 shows the odds ratios for each cumulative exposure to disadvantage compared with no disadvantage (0). One period of disadvantage was associated with lower self-rated life expectancy but not significantly but two and three periods of disadvantage were significantly associated.
|
Table 4 shows the results of the pathway analyses. Eight socio-economic pathways from childhood to adulthood are shown, labelled from A (the most advantaged) through H (the most disadvantaged). Pathway A is equivalent to scoring zero in the accumulated model while pathway H is equivalent to scoring three. However, the pathway model additionally allows us to explore whether there is an impact of downward (for example pathway D) and upward (for example pathway E) mobility on the self-reported measures. Age- and sex -adjusted odds ratios associated with each pathway are shown for each self-report measure. Pathway A, in which the respondent had a higher SEP in childhood, had educational qualifications from beyond compulsory schooling and attained a higher SEP in adulthood, is the reference category.
|
To assess the overall association between self-rated life expectancy and lifetime SEP we conducted a likelihood ratio test comparing the age- and sex-adjusted model with the model additionally including lifetime SEP. This showed a significant improvement in fit suggesting an overall association between lifetime SEP and self-rated life expectancy (
2 = 49.58, df = 7, P < 0.001). For self-rated life expectancy, five of the seven pathways (B, D, F, G, and H) showed significantly higher odds after adjustment for age and sex. Pathways B, D, and F feature some degree of downward mobility. In pathway B, for example, father's social class was non-manual and the respondent attained a post-compulsory level qualification, but their own social class was manual. Pathways G and H both feature people from manual backgrounds who did not attain a post-compulsory level qualification. However, pathway E, in which those from a manual background experienced sustained upward mobility, showed no significant difference from the most advantaged pathway.
The associations between self-rated life expectancy and SEP were reduced by adjusting for smoking behaviour, with only pathways G and H remaining significantly associated. Additional adjustment for GHQ had little impact on the results. Finally, adjustment for the two measures of self-rated health reduced the odds ratios further, although pathways G and H remained significantly associated.
The results differed for the two self-reported health measures. Firstly, in comparison to pathway A, all other pathways were associated with higher odds of reporting poor general health. The risk of reporting poor general health for those from a manual background was in particular heightened by being in the manual social class in adulthood, whatever the educational route (pathways F and H). For limiting illness there was a particular association between higher likelihood of poor health and low education attainment and adult manual social class. This occurred even when respondents came from an advantaged background (pathway D).
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
Principal findings
Using data from a large-scale population survey we have shown that self-rating of life expectancy was associated with a number of individual measures of SEP (confirming results from previous studies4,5), and with lifetime SEP. The pathway model suggests that, for those born into a manual class, subsequent educational attainment may be important for self-ratings of life expectancy. For this group, low educational attainment led to a higher likelihood of pessimistic ratings, while, interestingly, respondents from a manual background who were upwardly mobile both in education and adult social class were just as likely to view their life expectancy positively as the most advantaged.
Some of the variation in self-rated life expectancy was explained by adjustment for smoking. Smokers were far more likely to be pessimistic about their life expectancy. Additionally, some variation was explained by the higher rates of poor health in the most disadvantaged groups. However, the fact that associations remained for the most disadvantaged in childhood and education after adjustment for smoking and the measures of morbidity suggests that ratings of life expectancy may capture something other than simply current health condition.
There is an ongoing debate about the extent to which social disadvantage features in people's health discourse.29,30 Evidence from this study suggests that, whether consciously or not, SEP across the lifecourse may influence people's judgements about their life expectancy. Further work is needed to understand how people make judgements about their life expectancy and how these judgements are related to SEP.
Smoking contributes significantly to inequalities in health31; however, this study adds weight to the notion that different groups may view their future chances differently, and it has been suggested that it may impact on how health promotion is received.4 Tackling SEP differences in life chances may be a crucial step in developing an environment for health promotion to succeed11; however, no evidence for whether or not such a strategy would be successful can be provided by this study.
Single question self-report measures of health are established in inequalities research as a simple and effective way of assessing health.32 Regularly including a question on self-rated life expectancy in health surveys may add to our understanding of health inequalities and inform public health debates.
Strengths and limitations
Our data came from a respected social survey with a large sample size that contains excellent measures of SEP. Father's social class was collected retrospectively rather than prospectively. Research comparing father's social class measured prospectively and retrospectively found moderate levels of agreement, with disagreement biased towards more favourable SEP in retrospective recall.33 This would have the impact of underestimating rather than inflating any effects of childhood SEP on self-ratings of life expectancy in this study. Missing data on father's social class was quite common. It was more likely to be missing among those presently in low SEP, this again may mean that childhood SEP's impact is underestimated in this study.
For measuring health we were dependent on self-report measures as the BHPS does not collect more objective measures of health. Self-rated general health has been shown to be a valid measure of mental and physical health status34 and to be associated with mortality risk independently of more objective health measures.35 Limiting illness is also predictive of mortality risk.36
Smoking behaviour was also based on self-report. It is possible that people under-report their smoking and so its impact on the associations presented here may not have been fully captured. Generally, self-reports of smoking have been found to be accurate compared with more objective measures37 and rates of discrepancy are not related to SEP.38
One potentially important confounder not captured by the BHPS was whether people's parents were still alive and, if not, the age of their death. It is likely that those from a lower SEP background would be more to likely to have a parent who had died at a young age. Parental age of death may be an important factor in how people judge their own longevity.
Unfortunately, to date, respondents have only been asked once in the BHPS about their self-rated life expectancy, so we could not assess how ratings changed prospectively with SEP and health across the years of the survey. Further longitudinal work is needed.
It is debatable whether education itself should be regarded as a measure of SEP.39 It is certainly an important mechanism for maintaining socio-economic advantage from generation to generation.40 As an alternative design we could have summarized prospective measures of adult SEP (for example income and social class) since the BHPS started in 1991 for each individual. However, because of missing data in the years up to 2001 this would have reduced sample size quite considerably and while methods are available for correcting for this, either through longitudinal weighting or imputation,41 we preferred using education as a simple proxy for adult SEP up to current adult SEP measured in 2001. In particular, prospective measures of adult SEP such as income and social class would have been left censored in 1991.
|
Conclusion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
Self-rated life expectancy was associated with lifecourse SEP, some of the association was explained by self-reports of current health condition and smoking behaviour. Those experiencing disadvantage in both childhood and education had the highest odds of pessimistic ratings with significantly higher odds even after adjustment for smoking and self-rated morbidity. In contrast, those experiencing sustained upward mobility were no more likely to be pessimistic about their life expectancy than the most advantaged. Further work is needed to understand how people rate their life expectancy and how ratings change over time; however, that certain SEP groups are more pessimistic may have implications for how public health advice is received.
|
Acknowledgments |
|---|
The data used in this article were made available through the UK Data Archive. The data were originally collected by the ESRC Research Centre on Micro-social Change at the University of Essex, now incorporated within the Institute for Social and Economic Research. Neither the original collectors of the data nor the Archive bear any responsibility for the analyses or interpretations presented here. FP and RM are funded by the Chief Scientist Office of the Scottish Executive Health Department. These are the opinions of the authors, not the funders.
Author contributions FP and RM conceived the study. FP conducted the analysis and both authors contributed to the writing of the article. FP is the guarantor.
Conflict of interest None declared.
KEY MESSAGES
|
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
1 Donkin A, Goldblatt P, Lynch K. Inequalities in life expectancy by social class 1972–1999. Health Stat Q (2002) 15:5–15.
2 Davey Smith G, Dorling D, Mitchell R, Shaw M. Health inequalities in Britain: continuing increases up to the end of the 20th century. J Epidemiol Community Health (2002) 56:434–35.
3 Davey Smith G. Health Inequalities. Lifecourse Approaches. (2003) Bristol: Policy Press.
4 Wardle J, Steptoe A. Socioeconomic differences in attitudes and beliefs about healthy lifestyles. J Epidemiol Community Health (2003) 57:440–43.
5 Mirowsky J, Ross CE. Socioeconomic status and subjective life expectancy. Soc Psychol Q (2000) 63:133–51.[CrossRef][Web of Science]
6 Falba TA, Busch SH. Survival expectations of the obese: is excess mortality reflected in perceptions? Obes Res (2005) 13:754–61.[Web of Science][Medline]
7 Hamermesh DS, Hamermesh FW. Does perception of life expectancy reflect health knowledge? Am J Public Health (1983) 73:911–14.
8 Schoenbaum M. Do smokers understand the mortality effects of smoking? Evidence from the health and retirement survey. Am J Public Health (1997) 87:755–59.
9 Davison C, Davey Smith G, Frankel S. Lay epidemiology and the prevention paradox: the implications of coronary candidacy for health education. Sociol Health Illn (1991) 13:1–19.[Medline]
10 Frankel S, Davison C, Davey Smith G. Lay epidemiology and the rationality of responses to health education. Br J Gen Pract (1991) 41:428–30.[Web of Science][Medline]
11 Lawlor DA, Frankel S, Shaw M, Ebrahim S, Davey Smith G. Smoking and ill health: does lay epidemiology explain the failure of smoking cessation programs among deprived populations? Am J Public Health (2003) 93:266–70.
12 Siegel M, Bradley EH, Kasl SV. Self-rated life expectancy as a predictor of mortality: evidence from the HRS and AHEAD surveys. Gerontology (2003) 49:265–71.[CrossRef][Web of Science][Medline]
13 van Doorn C, Kasl SV. Can parental longevity and self-rated life expectancy predict mortality among older persons? Results from an Australian cohort. J Gerontol B Psychol Sci Soc Sci (1998) 53:S28–S34.[Abstract]
14 Galobardes B, Lynch JW, Davey Smith G. Childhood socioeconomic circumstances and cause-specific mortality in adulthood: systematic review and interpretation. Epidemiol Rev (2004) 26:7–21.
15 Davey Smith G, Hart C, Blane D, Gillis C, Hawthorne V. Lifetime socioeconomic position and mortality: prospective observational study. BMJ (1997) 314:547–52.
16 Blane D, Harding S, Rosato M. Does social mobility affect the size of the socioeconomic mortality differential?: evidence from the Office for National Statistics Longitudinal Study. J R Stat Soc Ser A Stat Soc (1999) 162:59–70.[CrossRef][Medline]
17 Hallqvist J, Lynch J, Bartley M, Lang T, Blane D. Can we disentangle life course processes of accumulation, critical period and social mobility? An analysis of disadvantaged socio-economic positions and myocardial infarction in the Stockholm Heart Epidemiology Program. Soc Sci Med (2004) 58:1555–62.[CrossRef][Web of Science][Medline]
18 Nicholson A, Bobak M, Murphy M, Rose R, Marmot M. Socio-economic influences on self-rated health in Russian men and women—a life course approach. Soc Sci Med (2005) 61:2345–54.[CrossRef][Web of Science][Medline]
19 Bartley M, Plewis I. Accumulated labour market disadvantage and limiting long-term illness: data from the 1971–1991 Office for National Statistics' Longitudinal Study. Int J Epidemiol (2002) 31:336–41.
20 Lahelma E, Martikainen P, Laaksonen M, Aittomaki A. Pathways between socioeconomic determinants of health. J Epidemiol Community Health (2004) 58:327–32.
21 Power C, Matthews S, Manor O. Inequalities in self rated health in the 1958 birth cohort: lifetime social circumstances or social mobility? BMJ (1996) 313:449–53.
22 Power C, Matthews S, Manor O. Inequalities in self-rated health: explanations from different stages of life. Lancet (1998) 351:1009–13.[CrossRef][Web of Science][Medline]
23 Bartley M, Plewis I. Does health-selective mobility account for socioeconomic differences in health? Evidence from England and Wales, 1971–1991. J Health Soc Behav (1997) 38:376–86.[CrossRef][Web of Science][Medline]
24 Taylor MF, Brice J, Buck N, Prentice-Lane E. British Household Panel Survey User Manual Volume A: Introduction, Technical Report and Appendicies. (2004) Colchester: University of Essex.
25 Department for Work and Pensions. Households Below Average Income 1994/95—2003/04. (2004) London: Department for Work and Pensions.
26 Goldberg D, Williams P. A User's Guide to the General Health Questionnaire. (1988) Windsor: NFER-Nelson.
27 Mackenbach JP, Kunst AE. Measuring the magnitude of socio-economic inequalities in health: an overview of available measures illustrated with two examples from Europe. Soc Sci Med (1997) 44:757–71.[CrossRef][Web of Science][Medline]
28 StataCorp. Stata Statistical Software: Release 9. (2005) College Station, Texas: StataCorp LP.
29 Blaxter M. Whose fault is it? People's own conceptions of the reasons for health inequalities. Soc Sci Med (1997) 44:747–56.[CrossRef][Web of Science][Medline]
30 Davidson R, Kitzinger J, Hunt K. The wealthy get healthy, the poor get poorly? Lay perceptions of health inequalities. Soc Sci Med (2006) 62:2171–82.[CrossRef][Web of Science][Medline]
31 Acheson D. Independent Inquiry into Inequalities in Health Report. (1998) London: TSO.
32 Bowling A. Just one question: if one question works, why ask several? J Epidemiol Community Health (2005) 59:342–5.
33 Batty GD, Lawlor DA, Macintyre S, Clark H, Leon DA. Accuracy of adults' recall of childhood social class: findings from the Aberdeen children of the 1950s study. J Epidemiol Community Health (2005) 59:898–903.
34 Singh-Manoux A, Martikainen P, Ferrie J, Zins M, Marmot M, Goldberg M. What does self rated health measure? Results from the British Whitehall II and French Gazel cohort studies. J Epidemiol Community Health (2006) 60:364–72.
35 Idler EL, Benyamini Y. Self-rated health and mortality: a review of 27 community studies. J Health Soc Behav (1997) 38:21–37.[CrossRef][Web of Science][Medline]
36 Rosato M, O'Reilly D. The relationship between limiting long term illness and mortality in England and Wales: a 12-year follow-up study of the ONS Longitudinal Study. J Epidemiol Community Health (2005) 59(Suppl 1):A5. A6.
37 Patrick DL, Cheadle A, Thompson DC, Diehr P, Koepsell T, Kinne S. The validity of self-reported smoking: a review and meta-analysis. Am J Public Health (1994) 84:1086–93.
38 Boreham R, Erens B, Falaschetti E, Hirani V, Primateston P. Risk factors for cardiovascular disease. In: Health Survey for England 1998: Cardiovascular Disease—Erens B, Primateston P, eds. (1999) London: TSO.
39 Blane D. Commentary: The place in life course research of validated measures of socioeconomic position. Int J Epidemiol (2006) 35:139–40.
40 Bynner J, Joshi H. Equality and opportunity in education: evidence from the 1958 and 1970 birth cohort studies. Oxford Rev Educ (2002) 28:405–26.[CrossRef]
41 Wiggins RD, Schofield P, Sacker A, Head J, Bartley M. Social position and minor psychiatric morbidity over time in the British Household Panel Survey 1991–1998. J Epidemiol Community Health (2004) 58:779–87.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||