You are here

Article Text

Article menu

Download PDFPDF

Research report
What determines Self-Rated Health (SRH)? A cross-sectional study of SF-36 health domains in the EPIC-Norfolk cohort
  1. Nahal Mavaddat1,
  2. Ann Louise Kinmonth1,
  3. Simon Sanderson1,
  4. Paul Surtees2,
  5. Sheila Bingham3,
  6. Kay Tee Khaw4
  1. 1General Practice and Primary Care Research Unit, Department of Public Health and Primary Care, Institute of Public Health, University of Cambridge, Cambridge, UK
  2. 2Strangeways Research Laboratory, Institute of Public Health, University of Cambridge, Cambridge, UK
  3. 3MRC Centre for Nutritional Epidemiology Cancer Prevention and Survival, Department of Public Health and Primary Care, Institute of Public Health University of Cambridge, Cambridge, UK
  4. 4Clinical Gerontology Unit, Department of Public Health and Primary Care, Institute of Public Health, University of Cambridge
  1. Correspondence to Dr Nahal Mavaddat, General Practice and Primary Care Research Unit, Department of Public Health and Primary Care, Institute of Public Health, University of Cambridge, Forvie Site, Robinson Way, Cambridge CB2 0SR, UK; nm212{at}medschl.cam.ac.uk

Abstract

Background Self-Rated Health (SRH) as assessed by a single-item measure is an independent predictor of health outcomes. However, it remains uncertain which elements of the subjective health experience it most strongly captures. In view of its ability to predict outcomes, elucidation of what determines SRH is potentially important in the provision of services. This study aimed to determine the extent to which dimensions of physical, mental and social functioning are associated with SRH.

Methods We studied 20 853 men and women aged 39–79 years from a population-based cohort study (European Prospective Investigation of Cancer study) who had completed an SRH (Short Form (SF)-1) measure and SF-36 questionnaire. SF-36 subscales were used to quantify dimensions of health best predicting poor or fair SRH within a logistic regression model.

Results In multivariate models adjusting for age, gender, social class, medical conditions and depression, all subscales of the SF-36 were independently associated with SRH, with the Physical Functioning subscale more strongly associated with poor or fair compared with excellent, very good or good health (OR 3.7 (95% CI 3.3 to 4.1)) than Mental Health (OR 1.4 (95% CI 1.2 to 1.5)) or Social Functioning subscales (OR 1.8 (95% CI 1.6 to 2.0)) for those below and above the median.

Conclusion This study confirms that physical functioning is more strongly associated with SRH than mental health and social functioning, even where the relative associations between each dimension and SRH may be expected to differ, such as in those with depression. It suggests that the way people take account of physical, mental and social dimensions of function when rating their health may be relatively stable across groups.

  • self-rated health
  • health status
  • subjective health
  • SF-36
  • functional status

https://doi.org/10.1136/jech.2009.090845

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Self-Rated Health (SRH) assessed by a simple single-item measure has been demonstrated to be a robust predictor of health outcomes independent of many objective measurable physical and biological factors.1 2 A number of large studies across varied populations have found SRH to be a reliable predictor of all-cause mortality, and morbidity and mortality in a range of conditions including cardiovascular disease and cancer, despite adjustments for physical and biological factors and for comorbid illness.1–8 It nevertheless remains uncertain what considerations individuals draw on to rate their health and which elements of the health experience are most strongly captured by the single subjective measure. If SRH is to be used to guide risk-reduction strategies for health, it is important to study what lies beneath the measure in order to identify potentially modifiable factors influencing SRH and in turn health outcomes. Studying basic determinants of SRH in the population is a first step towards this goal.

    The single measure of subjective health, SRH, has been proposed as a global assessment, an ‘effective summary’ of multiple measures and dimensions of health, incorporating a wide range of medical and social predictors.2 9–11 Nevertheless, models of subjective health experience purport predominantly a psychological or social, or a mainly physical or biological determinant of SRH. In the psychological models, self-perception of health coloured by dispositional attitudes such as optimism, resilience or sense of coherence, or by emotional factors such as depression or psychological distress, and in the social model, factors such as social class and situation and social coping resources, are believed to contribute to an individual's well-being or demise.12–21 The pathophysiological models of SRH, on the other hand, propose how one rates one's health to be largely a reflection of biological inputs, physical functioning and symptoms and even of subclinical disease or ‘unmeasured biological process’ yet to be discovered.2 11 12 22–25 Which of these models predominantly predicts SRH is potentially important in the provision of services. Should SRH largely reflect poor mental health or social deprivation, efforts should be made to improve the mental well-being and social circumstances of those identified with poor SRH. If it is a preclinical indicator of poor physical health on the other hand, early screening for diseases may be the most pertinent approach. Understanding influences on perception of SRH will inform theories of optimal subjective health that can be used to guide and evaluate strategies for improving the health of individuals and populations.

    Further elucidation of factors contributing to how individuals rate their health requires ongoing research, both qualitative and quantitative. This study examines how individuals rate their health using data from a large, well-characterised population cohort derived from the European Prospective Investigation of Cancer study (EPIC-Norfolk) study. Measures include scales of the Medical Outcomes Study 36-Item Short Form (SF-36), a widely validated self-assessment tool comprising eight multi-item independent dimensions addressing physical, mental and other health dimensions including social functioning and vitality.26 The study aims to quantify and determine using these scales which dimensions of health are most strongly associated with the SRH measure. For our proposed model of potential factors contributing to the subjective assessment of health, see figure 1.

    Figure 1
    Figure 1

    Model of influences on self-rated health (SRH).

    Methods

    Study population

    The EPIC study population comprises men and women aged 39 to 79 years, resident in Norfolk, UK and recruited from general practice age sex registers as part of EPIC-Norfolk.27 Detailed descriptions of study methodology have been reported previously.28 Approval for the study was obtained from the Norfolk Local Research Ethics Committee.

    A total of 25 639 men and women attended a health examination between 1993 and 1997, gave informed signed consent and completed a detailed health questionnaire. During the years 1996–2000, surviving participants were approached again by mail and asked to complete further questionnaires including the UK version of the SF-36.29 The aim was for participants to complete a further questionnaire approximately 18 months after study entry, but for those who entered the study in 1993 and early 1994, this was a longer period. For the purposes of analysis, all data were, however, considered as cross-sectional. For a full description of study methods, see Surtees et al.30 A total of 20 921 participants (73.2% of the total eligible EPIC sample) responded. Of these, 20 853 completed the SRH (SF-1) question of the SF-36 and were included in the study.

    Measures

    SRH was measured by SF-1, the first question of the SF-36. Participants were asked to assess their general health with the question ‘In general, would you say your health is?’ Response options were ‘excellent, very good, good, fair or poor.’

    The eight scales of the SF-36 include Physical Functioning (capacity to carry out activities during a typical day), Role-Physical (need to cut down or limit activities as a result of poor physical function), Vitality (amount of energy, feelings of being worn out or tired), Mental health (emotional well-being such as being nervous or down in the dumps), Role-Emotional (limitation of activities due to poor emotional well-being), Bodily Pain (amount of bodily pain and how much this interferes with activities), Social Functioning (degree to which physical health or emotional problems have interfered with social activities) and General Health. Scores for each dimension or subscale were obtained by summing scores from individual relevant responses in that dimension. For each subscale, raw scores were transformed into a scale from 0 to 100 using a scoring system described by Ware et al (1994), with 0 representing the worst and 100 best possible health.26

    Prevalent major medical conditions were determined by a positive response in the health questionnaire to the question: ‘Has a doctor ever told you that you have any of the following?’ followed by a list of options including cancer, stroke, heart attack, diabetes, respiratory disease (asthma or bronchitis) and arthritis. Any positive answer was recorded as the presence or past history of a major chronic medical condition. Depression was determined in the same way or as a positive response to a question determining the use of antidepressant medication.

    Participants were also asked about current or most recent occupation. Social class was classified according to the Registrar General's occupation-based classification scheme with the detailed classification described elsewhere.31 32 For the purpose of this study, social class was grouped into non-manual (I—professionals, II—managerial and technical, and III—non-manual skilled workers) and manual classes (III—manual skilled, IV—partly skilled and V—unskilled manual workers).

    Statistical analyses

    The Statistical Package for Social Sciences for Windows V 14.0 (SPSS, Chicago) was used for all analyses.

    Descriptive statistics (means and percentages) were used to show the characteristics of the study sample. The assumption of equal variances was verified. Two-sample t tests were used to compare differences in mean values between men and women. Differences in proportions were compared using χ2 tests. Analysis of variance was used to obtain mean values of each descriptive variable for each category of SRH (SF-1) and adjusted for age. Differences in mean values across groups were evaluated using F tests. All p values reported were for two-sided significance tests with a p value of <0.05 regarded as statistically significant.

    Correlation coefficient tables were produced for SF-1 and subscales of the SF-36 to determine how closely measures were related to each other. Partial correlations were reported as r values.

    The relationship between SRH (SF-1) and SF-36 dimensions was examined using logistic regression models, with a poor or fair (=‘poor’) rating compared with good or very good or excellent (=‘good’) rating as the outcome. Using scores for all participants created quartiles for SF-36 subscale scores. Univariate models were used to calculate ORs for ‘poor’ health for each quartile of SF-36 subscale scores separately for men and women after adjusting for age, and for those below and above 65 at study entry. A multivariate model was further created with ‘poor’ SRH (SF-1) as the outcome, but with SF-36 scores dichotomised above and below the median. The General Health (GH) subscale was not used in the models, since SRH is an item incorporated within this subscale. Six models were tested. All controlled for age, gender and social class, as these have been previously related to SRH.17 33–37 Some models also controlled variably for prevalent major medical conditions and depression or alternatively excluded all those with, or included only those with, medical conditions and depression (see table 4).

    Results

    A total of 20 853 men and women aged 39–79 years at study entry completed the SF-36 questionnaire including the SRH (SF-1) question. There were slightly more women, those in non-manual social classes and those without medical conditions among respondents. This may have impacted upon mean levels of functioning on SRH and SF-36 measures in the study. Demographic information and distribution of SRH and mean SF-36 subscale scores in men and women are shown in table 1. Table 2 shows the distribution of SRH by age, social class, medical conditions and depression, as well as mean SF-36 subscale scores by SRH. Functional status and SRH in the population were generally high. Women had slightly poorer scores on the SF-36 than men. However, similar proportions of women and men reported ‘poor’ SRH (18.1% vs 17.9%). A greater proportion of older participants aged over 65 compared with those below 50 (24.6% vs 13.6%), more in manual than non-manual classes (22.6% vs 14.7%), with medical conditions than without (28.5% vs 10.9%), and with depression than without (24.6% vs 16.6%), reported ‘poor’ SRH.

    View this table:
    Table 1

    Demographics, medical conditions, Self-Rated Health (SF-1) and SF-36 subscale scores combined and by gender (percentage count and mean SD)

    View this table:
    Table 2

    Distribution of Self-Rated Health (SF-1) by SF-36 subscale scores and other covariates: demographics and medical conditions (percentage count and mean sd)

    An association was seen between SRH (SF-1) and all SF-36 subscales (all p<0.0001), such that those in poorer SRH categories had lower SF-36 scores. This was particularly strong in gradient for Physical Functioning, Role Physical, General Health and Vitality subscales. Partial correlation coefficients for the relationship between SRH and each SF-36 subscale showed these to be highly related to each other and to the SRH score (all correlations p<0.0001). These results are shown in an additional table.

    The OR of ‘poor’ health compared with ‘good’ health for each quartile of scores on the SF-36 subscales is shown separately for men and women and by age for those below and over 65 at study entry in table 3.

    View this table:
    Table 3

    SF-36 Subscale Scores association with Poor/Fair Self-Rated Health (SF-1) by gender and by age (OR and 95% CI)

    Odds of ‘poor’ health were consistently greater for each SF-36 subscale in the lowest quartiles compared with the highest quartiles of each scale, suggesting that all subscales of the SF-36 are highly related to SRH. The magnitude of the OR for ‘poor’ health for the lowest quartiles was, however, greater in Physical Functioning and Vitality than that for Mental health and Social Functioning, as was Role Physical compared with Role Emotional for all groups. Physical functioning was, however, less strongly associated with SRH at older than younger ages in univariate models, but nevertheless remained a stronger association than mental health or social functioning in older patients.

    Table 4 shows the results of multiple regression models for ‘poor’ SRH (SF-1) and SF-36 dichotomised subscale scores in the full sample. In multivariate models adjusting for age, gender and social class and excluding the General Health scale, all remaining seven subscales of the SF-36 were independently associated with SRH. However, Physical Functioning showed a more than twofold stronger association with ‘poor’ health (OR 3.7 (95% CI 3.3 to 4.1)) than Mental Health (OR 1.4(95% CI 1.2 to 1.5)), as did Vitality (OR 3.6 (95% CI 3.2 to 4.1)). While age group, manual social class, prevalent major medical conditions and depression were all independently associated with ‘poor’ SRH, adjusting for these, excluding these from analyses or including only those with medical conditions and depression in the analyses did not alter results. In addition, when multivariate analyses were carried out separately in two age groups (less than and over 65 at study entry) there was little difference in results (see additional table).

    View this table:
    Table 4

    Multiple regression models for SF-36 subscale scores showing the relationship with poor/fair Self-Rated Heath (SF-1) (ORs and 95% CI)

    Discussion

    This study adds to our understanding of the meaning behind the SRH measure. It demonstrates the importance of underpinning perceptions of physical ability and function, as well as vitality in the self-rating of health across a range of groups. In a sample of over 20 000 middle-aged and older men and women from the population-based EPIC-Norfolk cohort, all dimensions of the SF-36 were independently associated with the single measure of SRH, supporting the notion that SRH reflects a global assessment of health and incorporates perceptions of a range of physical, mental and social factors. However, the strength of association between SRH and physical functioning in our population was more than twice that of mental health and almost twice that of social functioning. When covariates such as age, gender, social class and chronic medical conditions or depression were analysed separately or controlled for, the relative association of each dimension with SRH remained unchanged, suggesting that in general, poor physical function has a stronger association with poor subjective health than does mental health and social functioning. We also found vitality, defined as a positive feeling of aliveness or energy related to both levels of mental and physical well-being, but distinguishable from measures of mental ill health,38–40 to have a significant independent association with SRH, similar in strength to that of physical functioning.

    A body of literature, including that in patients with physical disabilities, suggests strong associations between the subjective assessment of overall health and the ability to perform physical functions.11 12 22 23 25 41 42 For example, in a previous prospective study of 7505 participating in the National Population Health Survey in Canada in which psychological and physical factors were both measured, SRH was most closely related to physical factors and the experience of symptoms.12 43 Similarly, in a large cross-sectional analysis of the British Whitehall and French Gazel studies involving around 28 000 participants, physical mobility was one of the most significant contributors to SRH.11 Further, in a cross-sectional study of 9332 in a younger UK population aged 18–64, also using the SF-36, SRH was found to be more strongly related to Physical Functioning than to other dimensions.23 Our study, however, has provided important additional quantification of the strength of relationships between SRH and physical, mental and social function in a larger population and with data stratified for age and sex, and controlled for the effects of social class and the presence of major chronic diseases. It has also demonstrated the consistency of these relationships between groups.

    It has been suggested that the subjective assessment of health made through the single SRH question is a highly individualised process where individuals ‘decide for themselves how to combine the various dimensions’ and adopt different strategies in their response.41 44 45 Others have suggested that various dimensions may impact differently on the assessment of SRH in different groups. For example, some literature suggests that the experience of physical functioning may be less likely to impact upon SRH in some, such as those with chronic disabilities or older people.22 25 41 Our study posits that there is, nevertheless, likely an underlying core consistency in how people perceive their health. For example, in our study, the relative associations of physical, mental and social dimensions with SRH were similar in those below and over 65 at study entry once adjustments were made for confounders (although our population did not include extremes of age). This was also true in men and women. We also tested whether patients with depression may be more likely to draw on mental domains when assessing their health (or alternatively that in those with depression, poor SRH might affect mental health functioning disproportionately). We found no greater association between SRH and mental health function among those with reported depression than in the rest of the study population. Finally, physical functioning was just as strongly associated with SRH in those with and without major medical conditions.

    These data cannot infer a causative relationship between poor physical function and SRH. However, we propose that SRH is likely a composite response in the individual to a range of physical, emotional and social inputs, but that the experience of physical function is the strongest contributor from among these to the subjective health experience in both the presence and absence of disease. When considering disease states, poor physical function may be seen as a final common pathway in symptomatology in a range of conditions and hence associated with both objective and in turn potentially subjective poor health. Further research, however, is required to elucidate more fully the relationship between physical functioning and subjective health in individual conditions and to determine whether attempting to improve aspects of physical functioning in patients with chronic diseases may impact upon their subjective health. Indeed, a range of conditions are already managed by physical rehabilitation (eg, myalgic encephalitis and myocardial infarction), consistent with the relationships between SRH and physical functioning confirmed here. Further work is needed to determine in which conditions such an approach may be useful. The strong association of physical functioning to subjective health in those with apparently no significant medical condition, however, raises other important questions. This is particularly so, in view of observed associations between physical functioning and mortality in some studies, even in the absence of associated disease or risk factors and after correction for confounders.46 47 If SRH is associated with an individual's perception of their abilities to perform physical activities, what factors impact on the capacity to perform such activities, what is the role of the expectation of functional ability and of vitality and positive affect, and how do physical, mental and social dimensions interact in determining functional capacity?48 49 Indeed, is poor SRH itself a potential factor impacting upon physical functioning as well as a reflection of it? Addressing questions such as these will contribute to the formulation of holistic theories placing assessment of the individual's health in the context of physical, mental and social factors, and informing the development of wider strategic approaches to the promotion of health and well-being.

    This study was limited by its cross-sectional nature which precludes conclusions on causation. Although a gap existed between demographic and disease, and subjective health measures, the data were considered as cross-sectional for the purposes of analyses. New instances of disease may have arisen during this time. Further, due to the method of eliciting prevalent conditions, some may have been over-reported (eg, those considered depressed may not have been currently so) or under- or not reported (eg, those relating to neurological illness or injury in younger adults, which were not enquired about). Also, no measure of severity of illness or other measure of function was available, the SF-36 itself being limited by its inability to address all areas of quality of life and functioning that may correlate with self-rated health. The population we studied also did not include the younger and more elderly, and was predominantly homogenous in relation to ethnicity. Nevertheless, it is likely that measurement errors arising due to such truncation of distribution of the population or misclassification with respect to history of illness are likely to have only attenuated our observed associations. We therefore, believe that this large population study in the Norfolk, UK population adds to the growing literature on SRH, informs of the relative associations of physical, mental and social dimensions with SRH and helps focus further research on its' determinants.

    SRH is represented as a composite response to inputs from the organism and its environment; these can be seen as having physical, mental and social dimensions. Predispositions to physical or emotional resilience or vulnerability interact with external insults, or enablers to mediate SRH through behavioural, lifestyle, physiological, immunological and coping responses. The subjective assessment of health may in turn impact upon each dimension, forming a feedback loop.

    What is already known on this subject

    Self-Rated Health (SRH) as assessed by a single-item measure is an independent predictor of health outcomes. Studies suggest that SRH is a global construct. However, which elements of the subjective health experience (physical, mental or social) SRH most strongly captures requires further study.

    Policy Implications

    These findings are potentially important in the planning and provision of services for those reporting poor subjective health.

    Acknowledgments

    This report presents independent commissioned research by the National Institute for Health Research. The views expressed in this publication are those of the author(s) and not necessarily those of the NHS, the NIHR or the Department of Health. We would also like to thank M Hotopf for his helpful comments.

    References

    1. ↵
      1. Desalvo KB,
      2. Fan VS,
      3. McDonell MB,
      4. et al
      . Predicting mortality and healthcare utilization with a single question. Health Serv Res 2005;40:1234–46.
      OpenUrlCrossRefPubMedWeb of Science
    2. ↵
      1. Idler EL,
      2. Benyamini Y
      . Self-rated health and mortality: a review of twenty-seven community studies. J Health Soc Behav 1997;38:21–37.
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      1. Bosworth HB,
      2. Siegler IC,
      3. Brummett BH,
      4. et al
      . The association between self-rated health and mortality in a well-characterized sample of coronary artery disease patients. Med Care 1999;37:1226–36.
      OpenUrlCrossRefPubMedWeb of Science
    4. ↵
      1. Desalvo KB,
      2. Bloser N,
      3. Reynolds K,
      4. et al
      . Mortality prediction with a single general self-rated health question. A meta-analysis. J Gen Intern Med 2006;21:267–75.
      OpenUrlCrossRefPubMedWeb of Science
    5. ↵
      1. Froom P,
      2. Melamed S,
      3. Triber I,
      4. et al
      . Predicting self-reported health: the CORDIS study. Prev Med 2004;39:419–23.
      OpenUrlCrossRefPubMedWeb of Science
    6. ↵
      1. Idler EL,
      2. Kasl S
      . Health perceptions and survival: do global evaluations of health status really predict mortality? J Gerontol 1991;46:S55–65.
      OpenUrlAbstract
    7. ↵
      1. Kaplan GA,
      2. Camacho T
      . Perceived health and mortality: a nine-year follow-up of the human population laboratory cohort. Am J Epidemiol 1983;117:292–304.
      OpenUrlAbstract/FREE Full Text
    8. ↵
      1. Shadbolt B,
      2. Barresi J,
      3. Craft P
      . Self-rated health as a predictor of survival among patients with advanced cancer. J Clin Oncol 2002;20:2514–19.
      OpenUrlAbstract/FREE Full Text
    9. ↵
      1. Kaplan G,
      2. Baron-Epel O
      . What lies behind the subjective evaluation of health status? Soc Sci Med 2003;56:1669–76.
      OpenUrlCrossRefPubMedWeb of Science
    10. ↵
      1. Moum T
      . Self-assessed health among Norwegian adults. Soc Sci Med 1992;35:935–47.
      OpenUrlCrossRefPubMedWeb of Science
    11. ↵
      1. Singh-Manoux A,
      2. Martikainen P,
      3. Ferrie J,
      4. et al
      . 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.
      OpenUrlAbstract/FREE Full Text
    12. ↵
      1. Bailis DS,
      2. Segall A,
      3. Chipperfield JG
      . Two views of self-rated general health status. Soc Sci Med 2003;56:203–17.
      OpenUrlCrossRefPubMedWeb of Science
    13. ↵
      1. Boardman JD
      . Self-rated health among US adolescents. J Adolesc Health 2006;38:401–8.
      OpenUrlCrossRefPubMedWeb of Science
    14. ↵
      1. Eriksson M,
      2. Lindstrom B,
      3. Lilja J
      . A sense of coherence and health. Salutogenesis in a societal context: Aland, a special case? J Epidemiol Community Health 2007;61:684–8.
      OpenUrlAbstract/FREE Full Text
    15. ↵
      1. Hooker K,
      2. Siegler IC
      . Life goals, satisfaction, and self-rated health: preliminary findings. Exp Aging Res 1993;19:97–110.
      OpenUrlPubMedWeb of Science
    16. ↵
      1. Kosloski K,
      2. Stull DE,
      3. Kercher K,
      4. et al
      . Longitudinal analysis of the reciprocal effects of self-assessed global health and depressive symptoms. J Gerontol B Psychol Sci Soc Sci 2005;60:296–303.
      OpenUrl
    17. ↵
      1. McFadden E,
      2. Luben R,
      3. Bingham S,
      4. et al
      . Social inequalities in self-rated health by age: cross-sectional study of 22,457 middle-aged men and women. BMC Public Health 2008;8:230.
      OpenUrlCrossRefPubMed
    18. ↵
      1. Ruo B,
      2. Bertenthal D,
      3. Sen S,
      4. et al
      . Self-rated health among women with coronary disease: depression is as important as recent cardiovascular events. Am Heart J 2006;152:921–7.
      OpenUrlPubMed
    19. ↵
      1. Steptoe A,
      2. Wright C,
      3. Kunz-Ebrecht SR,
      4. et al
      . Dispositional optimism and health behaviour in community-dwelling older people: associations with healthy ageing. Br J Health Psychol 2006;11(Pt 1):71–84.
      OpenUrlCrossRefPubMedWeb of Science
    20. ↵
      1. Tessler R,
      2. Mechanic D
      . Psychological distress and perceived health status. J Health Soc Behav 1978;19:254–62.
      OpenUrlCrossRefPubMedWeb of Science
    21. ↵
      1. von Bothmer MI,
      2. Fridlund B
      . Self-rated health among university students in relation to sense of coherence and other personality traits. Scand J Caring Sci 2003;17:347–57.
      OpenUrlCrossRefPubMedWeb of Science
    22. ↵
      1. Jamoom EW,
      2. Horner-Johnson W,
      3. Suzuki R,
      4. et al
      . Age at disability onset and self-reported health status. BMC Public Health 2008;8:10.
      OpenUrlCrossRefPubMed
    23. ↵
      1. Jenkinson C,
      2. Wright L,
      3. Coulter A
      . Criterion validity and reliability of the SF-36 in a population sample. Qual Life Res 1994;3:7–12.
      OpenUrlCrossRefPubMedWeb of Science
    24. ↵
      1. Jylha M,
      2. Volpato S,
      3. Guralnik JM
      . Self-rated health showed a graded association with frequently used biomarkers in a large population sample. J Clin Epidemiol 2006;59:465–71.
      OpenUrlCrossRefPubMedWeb of Science
    25. ↵
      1. Krahn GL,
      2. Suzuki R,
      3. Horner-Johnson W
      . Self-rated health in persons with spinal cord injury: relationship of secondary conditions, function and health status. Qual Life Res 2009;18:575–84.
      OpenUrlCrossRefPubMedWeb of Science
    26. ↵
      1. Ware JE,
      2. Snow K,
      3. Kosinski M,
      4. et al
      . SF-36 Health Survey: Manual and Interpretation Guide. Boston, MA: New England Medical Center, The Health Institute, 1993.
    27. ↵
      1. Riboli E,
      2. Kaaks R
      . The EPIC Project: rationale and study design. European Prospective Investigation into Cancer and Nutrition. Int J Epidemiol 1997;26 (Suppl 1):S6–14.
      OpenUrlAbstract/FREE Full Text
    28. ↵
      1. Day N,
      2. Oakes S,
      3. Luben R,
      4. et al
      . EPIC-Norfolk: study design and characteristics of the cohort. European Prospective Investigation of Cancer. Br J Cancer 1999;80 (Suppl 1):95–103.
      OpenUrlPubMed
    29. ↵
      1. Brazier JE,
      2. Harper R,
      3. Jones NM,
      4. et al
      . Validating the SF-36 health survey questionnaire: new outcome measure for primary care. BMJ 1992;305:160–4.
      OpenUrlAbstract/FREE Full Text
    30. ↵
      1. Surtees PG,
      2. Wainwright NW,
      3. Khaw KT
      . Obesity, confidant support and functional health: cross-sectional evidence from the EPIC-Norfolk cohort. Int J Obes Relat Metab Disord 2004;28:748–58.
      OpenUrlCrossRefPubMedWeb of Science
    31. ↵
      1. Elias P,
      2. Halstead K,
      3. Prandy K
      . CASOC: Computer-assisted standard occupational coding. London: HMSO, 1993.
    32. ↵
      1. Shohaimi S,
      2. Luben R,
      3. Wareham N,
      4. et al
      . Residential area deprivation predicts smoking habit independently of individual educational level and occupational social class. A cross sectional study in the Norfolk cohort of the European Investigation into Cancer (EPIC-Norfolk). J Epidemiol Community Health 2003;57:270–6.
      OpenUrlAbstract/FREE Full Text
    33. ↵
      1. Andersen FK,
      2. Christensen K,
      3. Frederiksen H
      . Self-rated health and age: a cross-sectional and longitudinal study of 11,000 Danes aged 45-102. Scand J Public Health 2007;35:164–71.
      OpenUrlAbstract/FREE Full Text
    34. ↵
      1. Burstrom B,
      2. Fredlund P
      . Self rated health: Is it as good a predictor of subsequent mortality among adults in lower as well as in higher social classes? J Epidemiol Community Health 2001;55:836–40.
      OpenUrlAbstract/FREE Full Text
    35. ↵
      1. Deeg DJ,
      2. Kriegsman DM
      . Concepts of self-rated health: specifying the gender difference in mortality risk. Gerontologist 2003;43:376–86.
      OpenUrlAbstract/FREE Full Text
    36. ↵
      1. Hoeymans N,
      2. Feskens EJ,
      3. van den Bos GA,
      4. et al
      . Age, time, and cohort effects on functional status and self-rated health in elderly men. Am J Public Health 1997;87:1620–5.
      OpenUrlPubMedWeb of Science
    37. ↵
      1. Radley A,
      2. Grove A,
      3. Wright S,
      4. et al
      . Gender-role identity after heart attack: Links with sex and subjective health status. Psychol Health 2000;15:123–33.
      OpenUrlCrossRefWeb of Science
    38. ↵
      1. Lindeberg SI,
      2. Ostergren PO,
      3. Lindbladh E
      . Exhaustion is differentiable from depression and anxiety: evidence provided by the SF-36 vitality scale. Stress 2006;9:117–23.
      OpenUrlPubMedWeb of Science
    39. ↵
      1. Ryan RM,
      2. Frederick C
      . On energy, personality, and health: subjective vitality as a dynamic reflection of well-being. J Pers 1997;65:529–65.
      OpenUrlCrossRefPubMedWeb of Science
    40. ↵
      1. Wood C,
      2. Magnello ME,
      3. Jewell T
      . Measuring vitality. J R Soc Med 1990;83:486–9.
      OpenUrl
    41. ↵
      1. Krause NM,
      2. Jay GM
      . What do global self-rated health items measure? Med Care 1994;32:930–42.
      OpenUrlCrossRefPubMedWeb of Science
    42. ↵
      1. Ware JE,
      2. Kosinski M,
      3. Keller S
      . SF-36 Physical and mental health summary scales: a user's manual. Boston, MA, USA: New England Medical Center, The Health Institute, 1994.
    43. ↵
      1. Bailis DS,
      2. Segall A,
      3. Mahon MJ,
      4. et al
      . Perceived control in relation to socioeconomic and behavioral resources for health. Soc Sci Med 2001;52:1661–76.
      OpenUrlCrossRefPubMedWeb of Science
    44. ↵
      1. Fayers PM,
      2. Sprangers MA
      . Understanding self-rated health. Lancet 2002;359:187–8.
      OpenUrlCrossRefPubMedWeb of Science
    45. ↵
      1. Fayers PM,
      2. Langston AL,
      3. Robertson C
      . Implicit self-comparisons against others could bias quality of life assessments. J Clin Epidemiol 2007;60:1034–9.
      OpenUrlCrossRefPubMedWeb of Science
    46. ↵
      1. Cesari M,
      2. Onder G,
      3. Zamboni V,
      4. et al
      . Physical function and self-rated health status as predictors of mortality: results from longitudinal analysis in the ilSIRENTE study. BMC Geriatr 2008;8:34.
      OpenUrlCrossRefPubMed
    47. ↵
      1. Myint PK,
      2. Luben RN,
      3. Surtees PG,
      4. et al
      . Relation between self-reported physical functional health and chronic disease mortality in men and women in the European Prospective Investigation into Cancer (EPIC-Norfolk): a prospective population study. Ann Epidemiol 2006;16:492–500.
      OpenUrlCrossRefPubMedWeb of Science
    48. ↵
      1. Fylkesnes K,
      2. Forde OH
      . The Tromso Study: predictors of self-evaluated health—has society adopted the expanded health concept? Soc Sci Med 1991;32:141–6.
      OpenUrlCrossRefPubMedWeb of Science
    49. ↵
      1. Fylkesnes K,
      2. Forde OH
      . Determinants and dimensions involved in self-evaluation of health. Soc Sci Med 1992;35:271–9.
      OpenUrlCrossRefPubMedWeb of Science

    Supplementary materials

    Footnotes

    • Funding This work was undertaken by the General Practice and Primary Care Research Unit, University of Cambridge, which receives core funding from the National Institute for Health Research (NIHR) and is part of the National School for Primary Care Research. EPIC-Norfolk is supported by programme grants from Medical Research Council UK (G9502233,G0300128) and Cancer Research UK (C865/A2883) with additional support from the European Union, Stroke Association, Research into Ageing, British Heart Foundation, Department of Health and the Wellcome Trust.

    • Competing interests None.

    • Patient consent Obtained.

    • Ethics approval Ethics approval was provided by the Norfolk Local Research Ethics Committee, UK.

    • Provenance and peer review Not commissioned; externally peer reviewed.