The European Journal of Public Health Advance Access originally published online on June 3, 2006
The European Journal of Public Health 2007 17(2):125-133; doi:10.1093/eurpub/ckl070
Health inequalities |
Socioeconomic conditions, lifestyle factors, and self-rated health among men and women in Sweden
Anu Molarius1, Kenneth Berglund2, Charli Eriksson3,4, Mats Lambe2,5, Eva Nordström6, Hans G. Eriksson6 and Inna Feldman2
1 Västmanland County Council, Department of Community Medicine Västerås, Sweden
2 Uppsala County Council, Department of Community Medicine Uppsala, Sweden
3 Örebro County Council, Department of Community Medicine Örebro, Sweden
4 Department of Caring Sciences, Örebro University Örebro, Sweden
5 Department of Medical Epidemiology and Biostatistics, Karolinska Institute Stockholm, Sweden
6 Sörmland County Council, Department of Community Medicine Eskilstuna, Sweden
Correspondence: A. Molarius, Västmanland County Council, Department of Community Medicine, 721 51 Västerås, Sweden, tel: +46 21 17 45 83, fax: +46 21 17 45 09, e-mail: anu.molarius{at}ltv.se
Received June 10, 2005, accepted April 4, 2006
 |
Abstract |
|---|
 Top Abstract Material and methodsResultsDiscussionReferences |
|---|
Background: Socioeconomic conditions and lifestyle factors have been found to be related to self-rated health, which is an established predictor of morbidity and mortality. Few studies, however, have investigated the independent effect of material and psychosocial conditions as well as lifestyle factors on self-rated health. Methods: The association between socioeconomic conditions, lifestyle factors, and self-rated health was investigated using a postal survey questionnaire sent to a random population sample of men and women aged 18–79 years during March–May 2000. The overall response rate was 65%. The area investigated covers 58 municipalities in the central part of Sweden. Multivariate odds ratios for poor self-rated health were calculated for a range of variables. A total of 36 048 subjects with full data were included in the analysis. Similar analyses of the influence of working conditions were conducted among those employed aged 18–64 years (17 820 subjects). Results: The overall prevalence of poor self-rated health was 7% among men and 9% among women. Poor self-rated health was most common among persons who had been belittled, who had experienced economic hardship, who lacked social support, or who had retired early. A low educational level was independently associated with poor self-rated health among men, but not among women. Physically inactive as well as underweight and obese subjects were more likely to have poor self-rated health than other subjects. Working conditions associated with poor self-rated health were dissatisfaction with work, low job control and worry about losing one's job. Conclusion: While a cross-sectional study does not allow definite conclusions as to which factors are determinants and which are consequences of poor self-rated, the present findings support the notion that both psychosocial and material conditions as well as lifestyle factors are independently related with poor self-rated health.
Keywords: health inequalities, lifestyle, population studies, self-rated health, socioeconomic conditions, Sweden
Self-rated health has been found to be a good predictor of morbidity and mortality.1–4 Large socioeconomic differences have been observed in self-rated health. Persons with low socioeconomic status have in general poorer self-rated health than persons with high socioeconomic status.5–9
Several theories have been put forward to explain observed social gradients in health.10 The materialist/structural theory suggests an important role of the physical environment e.g. working conditions, material conditions, and housing environment. The behavioural/lifestyles explanation emerged when individual risk factors such as smoking, alcohol use, and physical inactivity were identified as determinants of health. According to this theory, social differences in health are explained by an unhealthier lifestyle among those with disadvantaged positions. The psychosocial theory proposes that unhealthy habits are a reaction to stress and a way to alleviate frustration and that social capital, social support, and autonomy represent key elements for good health.11,12 However, none of these theories have been able to completely explain social differences in health.
It is important to disentangle the roles played by lifestyle, material factors, and psychosocial factors in health disparities. This will enable us to understand whether interventions should be aimed mainly at changes in lifestyle, in material conditions, or in the psychosocial environment.
The aim of this study was to investigate the association between socioeconomic status, material and psychosocial conditions, lifestyle factors, and self-rated health in the general population using a wide range of variables covering these factors. The effects of working conditions were examined separately.
|
Material and methods |
|---|
|
TopAbstractMaterial and methodsResultsDiscussionReferences |
|---|
The association between socioeconomic conditions, lifestyle factors, and self-rated health was investigated in a population sample of men and women aged 18–79 years. The data were obtained during March–May 2000 using a postal survey questionnaire. The sampling was random and stratified by gender, age group, county, and municipality. The data collection was completed after two postal reminders. A total of 46 646 subjects answered the questionnaire. The overall response rate was 65%. The area investigated covered 58 municipalities in 6 counties with about one million inhabitants in Central Sweden.
Self-rated health was assessed with the question ‘How do you rate your general health?’ with the options ‘very good’, ‘good’, ‘neither good nor poor’, ‘poor’, and ‘very poor’. For the analysis the categories ‘very good’ and ‘good’ were combined, as were the categories ‘poor’ and ‘very poor’.
Socioeconomic status
Educational level was obtained through record linkage to information from a national education register and was categorized into three classes: low (elementary school), medium (upper secondary school), and high (at least 3 years of university or corresponding education). Since educational level was not available for respondents aged 75 years and older, these were omitted from the analyses based on educational attainment. Following record linkage to a national population register, where information on country of origin was available, the respondents were categorized into those born in Sweden, in other Nordic countries, in other European countries, and outside Europe.
Employment status was derived from a survey question about whether the respondent was employed, a student, on parental leave, unemployed, working at home, retired early, retired due to age, or other.
Material conditions
Economic hardship was assessed by asking whether the respondent had had problems with paying running bills during the last 3 months (no problems, 1–2 months, 3 months).
Physical environment was described by a question: ‘How often do you have disturbance in or around your house from the following sources?’ with the alternatives: noise or exhaust from outside, disturbing neighbours, bad smell, poor quality of drinking water, littered environment, damage or graffiti and other disturbance with the options ‘often’, ‘sometimes’, ‘seldom’, and ‘never’. The answers were then summarized to an index and divided into three categories: good, less good, and poor physical environment.
Psychosocial factors
Social support was assessed with the question ‘Do you have any persons in your surroundings you can get support from in emotional crises or problems?’ with the answer options ‘yes, definitely’, ‘yes, probably’, ‘probably not’, and ‘definitely not’. The participants were also asked whether they had experienced that someone had belittled them during the last 3 months. The answer categories were never, once or twice, and several times.
The statement: ‘One can trust the people living in this neighbourhood’ was used to evaluate neighbourhood social coherence where agreement was coded as good, partial disagreement as less good, and total disagreement as poor social coherence. Participation in associations was assessed by a question about whether the respondent was an active member of a trade union, political party, administrative board, sports club, cultural association, religious association, or any other association.
Lifestyle factors
Body mass index (BMI), a measure of relative weight, was calculated from self-reported weight and height as weight divided by height squared (kg/m2). The participants were categorized according to the WHO guidelines13 as underweight when BMI was <18.5 kg/m2, normal weight when BMI was between 18.5 and 24.9 kg/m2, overweight when BMI was between 25 and 29.9 kg/m2, and obese when BMI was equal to or over 30 kg/m2.
The level of physical activity was determined with the question: ‘How much do you exercise physically in your leisure time?’ with the options little exercise (walking, bicycling, or other light exercise <2 h a week), moderate exercise (walking, bicycling, or other light exercise >2 h a week), moderate regular exercise (exercising 1–2 times a week at least for half an hour at a time in jogging, playing tennis, bicycling, exercising at a gym, or other moderate exercise that makes one sweat), and vigorous exercise and training (exercising or competing at least three times a week at least for half an hour at a time in team sports, jogging, playing tennis, swimming, or other intensive physical activity). The two middle categories were combined into moderate exercise.
Smoking habits were assessed by the question: ‘Do you smoke?’ with the answer options ‘No, I have never smoked regularly’, ‘No, I have stopped smoking’, ‘Yes, occasionally’, and ‘Yes, daily’.
Physical working conditions
Working conditions were assessed among those who were 18–64 years old and employed. Occupational status was obtained from a question about the respondent's occupation and then coded according to a national classification of occupations.
The physical working environment was assessed by how often the participant was exposed to the following three elements in her/his work: ‘My work includes heavy lifting (over 20 kg)’, ‘I am exposed to noise (have to raise my voice when speaking)’ and ‘I am exposed to chemicals, steams or gases’ with the answer categories ‘every day’, ‘some days a week’, ‘more seldom’, and ‘never’. The sum of these answers was then categorized into good, less good, and poor physical working environment. Monotonous work was derived from a question on how often the participant had the following elements in her/his work: ‘I perform monotonous typing at a data terminal’ and ‘I perform repetitive one-sided movements in assembly or manufacturing work’.
Psychosocial working conditions
Satisfaction with work was categorized into very good, good, neither good nor poor, poor, and very poor. Job control was derived from agreement with the following three statements: ‘I can decide the pace of work myself’, ‘I can take a break when I want’, and ‘I can influence the way my work is carried out’. The answers were categorized into none, low, moderate, and high job control. In addition, the participants were asked whether they were worried about losing their job during the next year with the answer categories ‘not at all’, ‘not especially’, ‘quite worried’, and ‘very worried’.
Statistical analyses
Odds ratios for poor self-rated health were calculated for a range of variables. A binomial logistic regression analysis was performed in four steps following the theoretical framework presented in the introduction. First, background variables (age, gender, and country of origin) and socioeconomic status (educational level and employment status) were introduced into the model. Next, material conditions (economic hardship and physical environment) were added, and thereafter lifestyle factors (physical activity, BMI category, and smoking). In a final step, psychosocial factors (being belittled, social support, social coherence in the neighbourhood, and participation in associations) were introduced. A total of 36 048 subjects with full data for all the variables considered were included in this analysis. Since the results for men and women were fairly similar, the combined analyses are reported in this paper. The variables that were statistically significantly associated with self-rated health in univariate analyses were included in the analyses.
Similar analyses were conducted among those aged 18–64 years and employed (17 820 respondents). This analysis included the same variables as used for the total population with the exception that occupational status was used instead of employment status and with working conditions added into the model. The working conditions were added into the model in two steps: first the physical working conditions (physical working environment and monotonous work) and then the psychosocial working conditions (work satisfaction, job control, and worry about losing one's job).
In addition, multinomial logistic regression analyses were conducted in the same way as the binomial logistic regression analyses in order to reassure that the same pattern between socioeconomic conditions, lifestyle factors, and self-rated health emerged.
|
Results |
|---|
|
TopAbstractMaterial and methodsResultsDiscussionReferences |
|---|
Table 1 gives the number of participants in the different age categories and the distributions of the explanatory variables both for the total population (18–79 years) and for those employed (18–64 years). The distribution of working conditions is reported only for those employed.
|
The overall prevalence of poor self-rated health was 7% among men and 9% among women (figure 1). The prevalence of poor self-rated health increased with age up to the age of retirement (65 years). The proportion of neither good nor poor health increased consistently with age and was highest in the oldest age group 65–79 years.
|
Employment status had a strong association with self-rated health whereas educational level had a weaker association (table 2). Those retired early, unemployed, and the category ‘other’ were more likely to have poor self-rated health than employed subjects. Those born in other Nordic countries and outside Europe had poorer self-rated health than native Swedes. When economic hardship was introduced into the model, the risks were attenuated for the unemployed and those retired early, suggesting that economic hardship affects these groups in particular. The odds ratios remained, however, highly significant.
|
Lifestyle factors were also strongly related to self-rated health (table 2). Poor self-rated health was more common among physically inactive than among physically active subjects. There was a U-shaped association between relative weight and poor self-rated health. The effect of lifestyle factors was mainly independent of other factors and did not change when psychosocial factors were introduced into the model.
Adding psychosocial factors into the model attenuated the association between economic hardship and self-rated health, but it still remained statistically significant (table 2). Adjustment for lifestyle factors and psychosocial factors explained the difference in self-rated health between those born outside Europe and native Swedes, whereas those born in other Nordic countries still had poorer self-rated health than native Swedes after adjustment.
In the final model, the strongest association with poor self-rated health was found for experiences of being belittled when adjusted for the other variables included in the analysis. Those who had been belittled several times during the last 3 months were four times more likely to report poor self-rated health than those who had not been belittled. Economic hardship, lack of social support, and employment status were also important factors. Of lifestyle factors, physical inactivity in particular but also underweight and obesity were associated with poor self-rated health when adjusted for all other factors.
There were two variables for which the association with poor self-rated health differed between men and women: educational level and age (data not shown). In women, educational level was not statistically significantly associated with self-rated health after adjustment for the other factors, whereas in men low educational level was independently associated with poor self-rated health. In women, the poorest self-rated health was found in the age group 55–64 years after adjustment for covariates, whereas in men it was in the oldest age group 65–79 years.
The prevalence of poor self-rated health was lower among those employed (18–64 years) than among the total population: 4% in men and 6% in women. Occupational status was only weakly associated with self-rated health when adjusted for age, gender, ethnicity, socioeconomic status, material conditions, lifestyle factors, and psychosocial factors (table 3). The psychosocial working conditions, satisfaction with work in particular, showed a strong association with self-rated health. Those who were very dissatisfied with their work were almost seven times more likely to have poor self-rated health compared with those who were very satisfied with their work. Physical working environment and monotonous work were less important for self-rated health. Of the covariates, current smoking had a stronger association with poor self-rated health (odds ratio = 1.6, 95% confidence interval: 1.3–1.9) among those employed than in the total population.
|
|
Discussion |
|---|
|
TopAbstractMaterial and methodsResultsDiscussionReferences |
|---|
In this cross-sectional study, the strongest association with poor self-rated health was found for experiences of being belittled, economic hardship, lack of social support, and employment status following adjustment for age group, gender, ethnicity, other socioeconomic conditions, and lifestyle factors. Of lifestyle factors, physical inactivity as well as underweight and obesity were independently related with poor self-rated health. Among those employed, dissatisfaction with work, low job control, and worry about losing one's job also showed strong associations with poor self-rated health.
It remains debatable whether socioeconomic inequalities in health can be explained by economic and material conditions or psychosocial factors such as social support and job control.11,12,14,15 Other researchers have also emphasized the importance of lifestyle factors.16,17 The different views are based on diverging theories on the causes of ill health (material/psychosocial factors) and whether the responsibility for ill health lies with the individuals (lifestyle) or on society (structural factors). It has also been suggested that all these factors play a role in explaining socioeconomic inequalities in health.9 In this study, our aim was to investigate all these factors using a stepwise approach.
Social relations in the form of social capital, support, and networks have been found to be important determinants of self-rated health.5,18,19 It is assumed that the quality of social interaction results in psychological reactions, which in turn affect health. In our study, experiences of being belittled can be seen as a part of social relations. This was, together with lack of social support, strongly associated with poor self-rated health. In previous studies, neighbourhood problems20 and participation in associations21 have been found to be related to self-rated health. In our study, neighbourhood social coherence and participation in associations were only weakly associated with poor self-rated health when controlled for other socioeconomic conditions.
Psychosocial pathways are likely to only partly explain social differences in self-rated health. Pathways based on material indicators, such as economic hardship and financial insecurity, are also important.7,22,23 In our study, the association between economic hardship and poor self-rated health became attenuated, but remained strong even after adjustment for psychosocial factors. Respondents who had had difficulties in paying their running bills were twice as likely to report poor self-rated health compared with those without financial difficulties. Similar results where poor health has been found to be associated with both material (economic hardship) and non-material (shame) indicators have been reported among the unemployed.24 We did not measure individual income in our study. However, some researchers have argued that measures reflecting income over decades and over generations, such as household wealth, may represent a more appropriate measure of material circumstances than individual income.25 In addition, there seems to be differences between countries as to what extent income can explain social differences in health. Yngwe et al.7 found that income distribution explained a larger part of health inequalities in Britain than in Sweden.
In previous studies, high job demands in combination with low job control have been found to be detrimental to health.26,27 In addition, job insecurity appears to have a negative effect on health.28,29 A Danish study found that working conditions and lifestyle factors explained two-thirds of social differences in health among employed people.30 Of these, working conditions such as monotonous work, low job control, job insecurity, and poor physical working environment explained most of the socioeconomic differences in health among the employed. In our study, especially dissatisfaction with work, but also low job control and worry about losing one's job were strongly associated with poor self-rated health. Occupational class was not associated with self-rated health when these other factors were accounted for.
In the present study, employment status was strongly associated with self-rated health. As expected, individuals who had retired at an early age were more likely to report poor self-rated health, a finding that reflects that a majority in this group receives a disability pension and that the social security system requires a certain level of illness before this benefit is granted. Unemployed persons had also significantly poorer health than employed persons, corroborating findings in other studies.31 Among unemployed persons and persons who had retired early, economic hardship explained a part of the association with poor self-rated health suggesting that economic hardship affects these groups in particular. We found that educational level was independently associated with self-rated health among men, but not among women. This suggests that educational level has a stronger influence on health in men than in women.
Previous studies have found that lifestyle factors such as smoking, physical activity, alcohol intake, and obesity are strongly related to impaired self-rated health.32–36 In addition, physical inactivity and smoking have been shown to predict poor self-rated health in longitudinal studies.34–36 For BMI and alcohol intake a U-shaped relationship with ill health has been reported.35 In our study, physical activity and BMI were strongly associated with self-rated health independently of other factors. Smoking was a more important factor among those employed than in the total population. However, the lifestyle factors could not explain the socioeconomic differences in self-rated health, a finding that is consistent with previous studies.6,30 In the present study, we were unable to assess the association between alcohol use and self-rated health since the proportion of non-respondents to the questions about alcohol use was high (29%).
The response rate in our study was acceptable (65%) for a population-based study. The response rate was lower among younger than older subjects and among men compared with women. The respondents had also somewhat higher educational level than the general population of the same age. It is, however, unlikely that response bias would explain the results obtained. It is also unlikely that exclusion of subjects with missing data for one or several of the variables used in the multivariate analyses would have influenced the results to any larger extent, since the distribution of the explanatory variables was the same among those included as in the total study population.
The present study is based on cross-sectional data. It is, therefore, not possible to draw conclusions about which factors are determinants and which consequences of poor self-rated health. Persons with poor self-rated health can be underweight or overweight owing to their illness, they can have difficulties to perform physical exercise, or they can be dissatisfied with their work as a result of ill health. Likewise is it not clear whether people with poor health are belittled or people who are belittled become ill.
Our findings suggest that the association between traditional measures of socioeconomic status, such as educational level and occupational status, and poor self-rated health is mediated through other socioeconomic conditions such as experiences of being belittled, economic hardship, and lack of social support. The strength of our study is that it is based on a very large population and that a wide range of variables covering socioeconomic conditions, including working conditions, and lifestyle factors were investigated. Our findings support the notion that socioeconomic conditions, both psychosocial and material, as well as lifestyle factors are independently related with poor self-rated health.
|
Acknowledgments |
|---|
This study was funded by the County Councils of Västmanland, Uppsala, Örebro, Sörmland and Värmland. We thank Dr Cate Burns for her comments on the text.
| Key points
|
|
References |
|---|
|
TopAbstractMaterial and methodsResultsDiscussionReferences |
|---|
1 Idler EL and Benyamini Y. (1997) Self-rated health and mortality: a review of twenty-seven community studies. J Health Soc Behav 38:21–7.[CrossRef][ISI][Medline]
2 Kaplan GA, Goldberg DE, Everson SA, et al. (1996) Perceived health status and morbidity and mortality: evidence from the Kuopio Ischeamic Heart Disease Risk Factor Study. Int J Epidemiol 25:259–65.
3 Mansson NO and Råstam L. (2001) Self-rated health as a predictor of disability pension and death—a prospective study of middle-aged men. Scand J Public Health 29:151–8.[CrossRef][ISI][Medline]
4 Burström B and Fredlund P. (2001) 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 55:836–40.
5 Kawachi I, Kennedy BP, Glass R. (1999) Social capital and self-rated health: a contextual analysis. Am J Public Health 89:1187–93.
6 Lantz PM, Lynch JW, House JS, et al. (2001) Socioeconomic disparities in health change in a longitudinal study of US adults: the role of health-risk behaviors. Soc Sci Med 53:29–40.[CrossRef][ISI][Medline]
7 Yngwe MA, Diderichsen F, Whitehead M, et al. (2001) The role of income differences in explaining social inequalities in self rated health in Sweden and Britain. J Epidemiol Community Health 55:556–61.
8 Knesebeck O, Luschen G, Cockerham WC, Siegrist J. (2003) Socioeconomic status and health among the aged in the United States and Germany: a comparative cross-sectional study. Soc Sci Med 57:1643–52.[CrossRef][ISI][Medline]
9 van Lenthe FJ, Schrijvers CTM, Droomers M, et al. (2004) Investigating explanations of socio-economic inequalities in health: the Dutch GLOBE study. Eur J Public Health 14:63–70.
10 Elstad J. (2000) Social inequalities in health and their explanations. Norwegian Social Research NOVA Rapport 9/00(NOVA, Oslo).
11 In Marmot M and Wilkinson RG (Eds.). Social determinants of health (1999) (Oxford University Press, Oxford).
12 Marmot M and Wilkinson RG. (2001) Psychosocial and material pathways in the relation between income and health: a response to Lynch et al. BMJ 322:1233–6.
13 Obesity–preventing and managing the global epidemic. Report of a WHO Consultation on Obesity. WHO. WHO/NUT/NCD/98.1. Geneva: WHO, 3–5 June 1997.
14 Lynch JW, Davey Smith G, Kaplan GA, House JS. (2000) Income inequality and mortality: importance to health of individual income, psychosocial environment, or material conditions. BMJ 320:1200–4.
15 Lynch JW, Smith GD, Hillemeier M, et al. (2001) Income inequality, the psychosocial environment, and health: comparison of wealthy nations. Lancet 358:194–200.[CrossRef][ISI][Medline]
16 Alvarez-Dardet C, Montahud C, Ruiz MT. (2001) The widening social class gap of preventive health behaviours in Spain. Eur J Public Health 11:225–6.
17 Denton M and Walters V. (1999) Gender differences in structural and behavioural determinants of health: an analysis of the social production of health. Soc Sci Med 48:1221–35.[CrossRef][ISI][Medline]
18 Marmot MG, Fuhrer R, Ettner SL, et al. (1998) Contribution of psychosocial factors to socioeconomic differences in health. Milbank Q 76:403–48.[CrossRef][ISI][Medline]
19 Melchior M, Berkman LF, Niedhammer I, et al. (2003) Social relations and self-reported health: a prospective analysis of the French Gazel cohort. Soc Sci Med 56:1817–30.[CrossRef][ISI][Medline]
20 Steptoe A and Feldman PJ. (2001) Neighbourhood problems as sources of chronic stress: development of a measure of neighbourhood problems, and associations with socioeconomic status and health. Ann Behav Med 23:177–85.[CrossRef][ISI][Medline]
21 Veenstra G. (2000) Social capital, SES and health: an individual-level analysis. Soc Sci Med 50:619–29.[CrossRef][ISI][Medline]
22 Ecob R and Smith GD. (1999) Income and health: what is the nature of the relationship? Soc Sci Med 48:693–705.[CrossRef][ISI][Medline]
23 Ferrie JE, Shipley MJ, Stansfeld SA, et al. (2003) Future uncertainty and socioeconomic inequalities in health: the Whitehall II study. Soc Sci Med 57:637–46.[CrossRef][ISI][Medline]
24 Starrin B, Jönsson B, Forsberg E, Rantakeisu U. (1998) Varför blir arbetslösheten till en plåga för vissa men inte för andra? Socialmedicinsk tidskrift 75:291–6.
25 Martikainen P, Adda J, Ferrie JE, et al. (2003) Effects of income and wealth on GHQ depression and poor self rated health in white collar women and men in the Whitehall II study. J Epidemiol Community Health 57:718–23.
26 Karasek R. and Theorell T. (1990) Healthy work, stress, productivity and the reconstruction of working life(Basic Books, New York).
27 Marmot MG, Bosma H, Hemingway H, et al. (1997) Contribution of job control and other risk factors to social variations in coronary heart disease incidence. Lancet 350:235–9.[CrossRef][ISI][Medline]
28 McDonough P. (2000) Job insecurity and health. Int J Health Serv 30:453–76.[CrossRef][ISI][Medline]
29 Ferrie JE, Shipley MJ, Stansfeld SA, Marmot MG. (2002) Effects of chronic job insecurity and change in job security on self reported health, minor psychiatric morbidity, physiological measures, and health related behaviours in British civil servants: the Whitehall II study. J Epidemiol Community Health 56:450–4.
30 Borg V and Kristensen TS. (2000) Social class and self-rated health: can the gradient be explained by differences in life style or work environment? Soc Sci Med 51:1019–30.[CrossRef][ISI][Medline]
31 Dahl E and Elstad JI. (2001) Recent changes in social structure and health inequalities in Norway. Scand J Public Health Suppl 55:7–17.[Medline]
32 Mackenbach JP, van den Bos J, Joung IM, et al. (1994) The determinants of excellent health: different from determinants of ill-health? Int J Epidemiol 23:1273–81.
33 Williamson DL. (2000) Health behaviours and health: evidence that the relationship is not conditional on income adequacy. Soc Sci Med 51:1741–54.[CrossRef][ISI][Medline]
34 Johansson SE and Sundquist J. (1999) Change in lifestyle factors and their influence on health status and all-cause mortality. Int J Epidemiol 28:1073–80.
35 Ostbye T, Taylor DH, Jung SH. (2002) A longitudinal study of the effects of tobacco smoking and other modifiable risk factors on ill health in middle-aged and old Americans: results from the Health and Retirement Study and Asset and Health Dynamics among the Oldest Old survey. Prev Med 34:334–45.[CrossRef][ISI][Medline]
36 Haveman-Nies A, De Groot LC, Van Staveren WA. (2003) Survey in Europe on Nutrition and the Elderly: a concerted action. Relation of dietary quality, physical activity, and smoking habits to 10-year changes in health status in older Europeans in the SENECA study. Am J Public Health 93:318–23.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||