Skip to main content
NCBI home page
HAL-INSERM logoLink to HAL-INSERM
. Author manuscript; available in PMC: 2013 Jun 11.
Published in final edited form as: J Epidemiol Community Health. 2010 Jun 27;65(12):1123–1131. doi: 10.1136/jech.2009.102517

Contribution of material, occupational, and psychosocial factors in the explanation of social inequalities in health in 28 countries in Europe

Bénédicte Aldabe 1, Robert Anderson 2, Maija Lyly-Yrjänäinen 2, Agnès Parent-Thirion 2, Greet Vermeylen 2, C Cecily Kelleher 3, Isabelle Niedhammer 1,2,3,*
PMCID: PMC3678208  PMID: 20584725

Abstract

Objectives

To analyse the associations between socioeconomic status (SES), measured using occupation, and self-reported health, and to examine the contribution of various material, occupational, and psychosocial factors to social inequalities in health in Europe.

Methods

This study was based on data from the European Quality of Life Survey (EQLS) carried out in 2003. The total sample consisted of 6038 and 6383 working men and women in 28 countries in Europe (response rates: 30.3–91.2%). Each set of potential material, occupational, and psychosocial mediators included between 8 and 11 variables. Statistical analysis was performed using multilevel logistic regression analysis.

Results

Significant social differences were observed for self-reported health, manual workers being more likely to be in poor health (OR=1.89, 95% CI: 1.46–2.46 for men, OR=2.18, 95% CI: 1.71–2.77 for women). Strong social gradients were found for almost all potential mediating factors, and almost all displayed significant associations with self-reported health. Social differences in health were substantially reduced after adjustment for material, occupational, and psychosocial factors, material factors playing a major role. The four strongest contributions to reducing these differences were found for material deprivation, social exclusion, financial problems, and job reward. Taking all mediators into account led to an explanation of the social differences in health by 78–100% for men and women.

Conclusion

The association between SES and poor health may be attributed to differential distributions of several dimensions of material, occupational, and psychosocial conditions across occupational groups. Interventions targeting different dimensions might result in a reduction of social inequalities in health.

Keywords: Adolescent, Adult, Europe, Female, Health Status Disparities, Health Surveys, Humans, Male, Middle Aged, Occupations, statistics & numerical data, Quality of Life, Regression Analysis, Self Report, Sex Distribution, Sex Factors, Social Class, Social Environment, Social Support, Socioeconomic Factors, Workplace, organization & administration, psychology, Young Adult

Keywords: social inequalities in health, self-reported health, occupation, Europe

Introduction

Social inequalities in health have been described across a range of European countries for many health indicators,[1–13] including self-reported health (SRH) which is considered as a powerful predictor of subsequent morbidity and mortality.[14] The association between socioeconomic status (SES) and health is one of the most well-known findings in the epidemiological literature: the higher the social class, the lower the prevalence and/or incidence of health problems, illness, disease, and death.

Whilst many studies described social inequalities in health in Europe, studies trying to explain these inequalities are still lacking. Identifying the factors that may contribute to these inequalities could be useful to reduce the prevalence of exposure to these factors, especially among the lowest social groups, and thus to reduce inequalities. Determinants of social inequalities in health are complex, and there are likely to be a number of important mediators, such as material, occupational, psychosocial, and behavioural factors.

Various theories have been developed to explain the pathways and mechanisms underlying these inequalities.[15–17] These theories include the materialist explanation, that put the emphasis on material conditions (i.e. access to goods and services, and exposures to material risk factors in the living and working environment), the psychosocial explanation, that focuses on psychosocial and stress related influences with a plethora of risk factors such as those related to social support or sense of control, and the behavioural explanation, that emphasizes the importance of behavioural risk factors in explaining social inequalities in health. The relative contribution of these factors has been very rarely explored to date at national level.[18–20] Furthermore, very few studies explored occupational factors, that include specific material and psychosocial factors of the work environment, among other mediators. As explanatory or mediating factors probably are interrelated, some authors have suggested a simplified causal model to disentangle the direct (independent) effect of mediating factors, and their indirect effect through other factors.[18;19] According to this explanatory model, material factors might affect health inequalities directly or indirectly through psychosocial and behavioural factors, and psychosocial factors might work directly or indirectly through behavioural factors.

To our knowledge, no study has attempted to explain social inequalities in health using harmonised data for Europe as a whole. Most previous studies were based on national or selected samples, and explored only a limited number of mediators that may contribute to these inequalities.[19–40] In addition, some studies describing social inequalities in health in Europe were based on surveys pooling data from several surveys/countries, making comparisons between European countries and generalizations for the whole of Europe difficult.[2;7;11] Mackenbach, in fact, underlined a serious lack of internationally comparable data on the topic of social inequalities in health.[16]

Our study aimed at providing an original overview of social inequalities in health in Europe, and at determining mediating factors for these inequalities. Thus, the objectives of the present study were to analyse the associations between SES, measured using occupation, and health, measured using SRH, and to examine the contribution of various material, occupational, and psychosocial factors to social inequalities in health in the working population of Europe.

Methods

This study was based on the data of the European Quality of Life Survey (EQLS), carried out by the European Foundation for the Improvement of Living and Working Conditions in 2003. Data were collected through face-to-face interviews in the respondent’s home. The basic sampling design used in all countries was a multi-stage, random one.[41] Overall, more than 26000 interviews took place, covering 28 countries in Europe (see Appendix). Total response rate was 58.4% (range: 30.3–91.2% across countries, see Table 2). For the purpose of this study, people who were not working at the time of the survey were excluded in order to concentrate on the role of occupational factors. Our final study sample included 12421 workers (6038 men and 6383 women).

Table 2.

Differences in response rate, self-reported health, occupation, and association between occupation and self-reported health between countries

Country AT BE BG CY CZ DK EE FI FR DE GR HU IE IT LV LT LU MT NL PL PT RO SK SI ES SE TR GB p
Response rate (%) 63.9 70.4 81.0 82.4 75.1 52.4 68.8 91.2 83.4 91.2 39.6 75.1 32.5 63.1 72.8 64.1 66.1 91.0 62.0 66.4 59.8 55.7 51.0 61.6 30.3 53.6 37.0 79.2
MEN
Poor self-reported health (%) 13.8 14.8 18.3 2.4 18.0 8.5 44.5 22.2 13.1 22.6 7.4 26.9 6.7 5.2 42.8 42.5 18.4 11.1 23.1 30.7 30.4 27.5 10.5 22.8 13.6 18.6 18.6 16.0 ****
Occupation (%) ****
Managers, professionals 13.5 7.5 5.3 12.2 8.6 11.3 16.0 9.4 11.6 12.2 7.8 6.0 6.3 18.2 10.9 11.0 14.9 10.8 4.5 6.4 3.7 12.2 7.3 7.4 11.4 17.0 5.4 16.7
Technicians and associate professionals 38.3 46.5 30.3 42.1 29.8 27.6 29.4 33.7 36.0 34.8 41.2 23.6 30.5 38.7 24.0 24.7 30.5 39.7 33.5 24.8 29.4 25.9 28.9 26.2 31.4 33.2 38.4 36.8
Clerical support workers 8.5 5.8 0.8 1.2 4.9 9.6 0.0 7.9 6.0 6.4 8.3 2.8 3.4 12.5 1.1 0.9 16.1 16.7 16.6 3.5 7.3 2.1 7.3 12.1 6.4 3.2 3.5 7.5
Service and sales workers 11.7 12.8 14.4 20.1 19.6 15.1 16.8 21.8 15.4 7.8 14.6 16.5 7.4 15.7 21.7 16.4 10.9 13.2 18.1 5.9 16.1 20.6 17.9 12.1 16.5 21.9 12.8 5.8
Manual workers 28.0 27.4 49.2 24.4 37.1 36.4 37.8 27.2 31.1 38.7 28.1 51.1 52.4 14.9 42.3 47.0 27.6 19.6 27.2 59.4 43.6 39.2 38.6 42.3 34.3 24.7 39.9 33.3
WOMEN
Poor self-reported health (%) 16.2 16.6 28.6 1.8 19.6 13.5 44.1 21.6 18.0 17.3 10.3 28.5 7.3 6.4 55.7 58.1 20.3 24.6 22.3 29.4 40.6 40.4 10.8 25.8 14.9 20.9 27.1 13.4 ****
Occupation (%) ****
Managers, professionals 11.6 7.0 8.4 9.0 8.9 7.8 18.1 7.0 10.1 11.9 4.9 16.8 10.9 13.2 16.9 20.2 6.5 10.0 4.9 7.2 3.8 19.1 10.4 16.3 10.6 7.1 12.9 10.1
Technicians and associate professionals 20.1 50.2 34.7 44.9 18.9 21.1 24.2 26.8 35.3 22.9 36.8 19.4 28.6 39.8 30.5 27.0 24.6 34.2 19.3 29.3 23.1 20.6 24.6 22.4 33.5 43.7 44.3 30.7
Clerical support workers 23.1 12.1 8.8 0.6 19.3 18.4 5.5 23.7 14.0 26.2 15.1 15.7 14.5 24.1 11.0 7.3 26.1 31.7 24.2 18.6 12.8 12.9 19.7 19.7 12.8 10.9 14.3 16.8
Service and sales workers 25.8 14.4 9.6 29.9 24.8 34.0 24.2 34.5 28.4 29.4 15.7 13.4 16.1 9.8 15.1 16.1 13.8 16.7 38.5 4.8 11.5 17.5 21.4 10.9 11.7 27.0 2.9 13.9
Manual workers 19.4 16.3 38.6 15.6 28.2 18.8 28.0 8.0 12.1 9.6 27.6 34.7 29.8 13.2 26.5 29.4 29.0 7.5 13.1 40.1 48.7 29.9 24.0 30.6 31.3 11.2 25.7 28.5
OR of poor healtha 2.58 * 2.03 4.96 * 1.23 2.62 1.21 2.35 2.83 * 1.47 3.24 * 3.65 2.53 * 2.10 8.02 1.51 1.74 10.91 0.52 1.92 3.09 * 1.50 2.85 * 3.10 3.24 * 0.91 1.17 0.80 1.47
Open in a new tab

For country codes, see appendix

Chi-square test/logistic regression analysis:

*

: p<0.05;

**

: p<0.01;

***

p<0.001;

****

p<0.0001

a

OR adjusted for sex and age (manual workers vs managers), each country being studied separately

SRH was the health indicator studied and was assessed using the following question: “In general, would you say your health is …” and response categories were “excellent”, “very good”, “good”, “fair”, and “poor”. The variable was dichotomised as “good” health versus “poor” health (“fair” and “poor”).

Occupation was used as an indicator of SES. Occupational groups were constructed according to the International Standard Classification of Occupations (ISCO-2008). Educational level was also used as an additional SES marker.

Three sets of potential mediating conditions were explored: material, occupational, and psychosocial factors. As we had a specific interest in occupational factors, they were distinguished from the other factors.

Material factors included: household tenure, housing conditions (shortage of space, rot in windows/doors/floors, damp/leaks, and lack of indoor flushing toilet, 4 items), crowding (number of persons per room), material deprivation (not able to afford at least one of these amenities or activities: car, home computer, washing machine, heating, holiday, furniture, meat/fish, clothes, and drinks/meals with friends/family, 9 items), financial problems (payment of bills, food, and/or rent, 4 items), neighbourhood conditions (noise, air pollution, lack of recreational/green areas, water quality, and area safety at night, 5 items), quality of public services (health services, education system, public transport, social services, and state pension system, 5 items) and problems with access to medical services (distance to medical centre, delay in getting appointment, waiting time to see the doctor, and cost of seeing the doctor, 4 items).

Occupational factors included: sector of economy (coded using NACE classification), permanency of work contract, number of hours normally worked per week (including any paid or unpaid overtime), daily commuting time, dangerous/unhealthy working conditions, job insecurity (fear of job loss in the next 6 months), psychological demands (demanding/stressful work, and tight deadlines, 2 items), decision latitude (extent of influence at work and interesting job, 2 items), reward (salary and job prospects, 2 items), responsibility for supervising the work of others, and additional paid job.

Psychosocial factors included: marital status, number of children, unpaid work (housework, care for children and/or elderly/disabled relatives, 3 items), work-life imbalance (difficulties to fulfill work and family responsibilities, 3 items), social support (help from family/friends/others in case of illness, advice, financial problems, or feeling low, 4 items), social network (frequency of contacts with family/friends/neighbours, 4 items), social participation (voluntary work, political activities, and religious services, 5 items), trust level (towards people in general, state pension and social benefit systems, 3 items), and social exclusion (feelings of inferiority or uselessness, and lack of recognition or acceptance, 5 items).

The associations between SES and SRH, between SES and mediating factors, and between mediating factors and SRH were studied using Chi-square test. The associations between SES and SRH were also tested for each country separately, using logistic regression analysis after adjustment for sex and age, SRH being the dependent variable. Managers/professionals were used as the reference group. The associations between SES and SRH in the whole sample were then explored after adjustment for age (model 1) using multilevel logistic regression analysis with the 12421 participants clustered within 28 countries. The potential mediating factors, that displayed social gradients and were associated with SRH, were first introduced separately to model 1. The contribution of each factor (or a set of factors) to the explanation of the social differences in SRH was estimated by the change in OR after inclusion of the variable(s) in the model, according to the formula: (ORmodel 1 − ORextended model)/(ORmodel 1 − 1).[42] Positive % values indicated OR reductions and negative % values indicated OR increases. Only factors reducing social inequalities by more than 5% were retained in subsequent models.[32] These models were adjusted for combinations of two groups of factors, and finally adjusted for all factors simultaneously, allowing us to calculate independent and indirect contributions of mediators. Indirect contributions can be interpretated as the contribution of a group of factors (for example material) through another group (for example psychosocial). Such a method has already been presented and used by others.[18;19] Analyses were carried out using SAS and performed for men and women separately.

Results

The description of the sample is presented in Table 1. Many variables were significantly associated with gender, and almost all variables were also associated with SRH. Only household tenure, work contract, and additional paid job were not associated with SRH. Significant differences were found for SRH and occupation between countries (Table 2). The highest prevalences of poor health were observed in eartern countries, and the proportion of manual workers was the highest in these countries, and the lowest in the scandinavian countries.

Table 1.

Prevalence of self-reported health according to age, occupation, and mediating factors

Men (N=6038) Women (N=6383)
N % % poor health N % % poor health
Self-reported health
 good 4887 81.14 4896 76.91
 poor 1136 18.86 1470 23.09
Age group **** ****
 [18;30[ 1270 21.03 11.45 1369 21.45 13.54
 [30;40[ 1693 28.04 13.27 1897 29.72 17.49
 [40;50[ 1643 27.21 21.86 1727 27.06 26.57
 50 or more 1432 23.72 28.58 1390 21.78 35.84
Socioeconomic status (SES) - Occupation **** ****
 managers, professionals 603 10.06 14.81 675 10.69 20.72
 technicians and associate professionals 1964 32.77 16.36 1843 29.19 19.99
 clerical support workers 410 6.84 13.27 1033 16.36 19.92
 service and sales workers 900 15.02 18.04 1269 20.10 21.98
 manual workers 2116 35.31 23.87 1493 23.65 30.94
Psychosocial factors
Marital status **** ****
 married or living with partner 4128 69.05 20.44 4058 64.06 22.55
 never married and not living with partner 1335 22.33 12.25 1101 17.38 16.92
 separated or widowed 515 8.61 23.69 1176 18.56 30.66
Number of children **** ****
 no child 2031 34.14 13.43 1675 26.57 17.10
 at least one child 3918 65.86 21.77 4628 73.43 25.21
Unpaid work NS ****
 Q1 2106 34.88 17.55 1000 15.67 19.52
 Q2 1697 28.11 19.98 1619 25.36 22.89
 Q3 1042 17.26 20.62 836 13.10 29.46
 Q4 (high) 1193 19.76 18.07 2928 45.87 22.60
Work life imbalance **** ****
 Q1 1439 25.16 15.03 1265 20.75 15.43
 Q2 1490 26.05 14.31 1585 26.00 18.29
 Q3 1604 28.04 20.04 1826 29.95 24.53
 Q4 (high) 1187 20.75 27.48 1420 23.29 32.65
Social support **** ****
 high 4907 86.86 17.19 5304 88.15 19.94
 low 742 13.14 26.22 713 11.85 42.08
Social network NS *
 Q1 1060 18.06 18.21 1456 23.34 21.43
 Q2 1863 31.74 18.84 2019 32.37 22.09
 Q3 1363 23.22 17.90 1394 22.35 22.74
 Q4 (low) 1584 26.98 20.52 1368 21.93 26.54
Social participation * *
 more than one activity 3294 8.03 14.92 467 7.43 19.27
 one activity 2181 36.64 19.93 2455 39.04 22.39
 none 478 55.33 18.81 3366 53.53 24.18
Trust level **** ****
 confident 3817 72.31 16.41 4054 73.05 20.40
 not confident/careful 1462 27.69 23.07 1496 26.95 27.56
Social exclusion **** ****
 Q1 1615 28.78 12.01 1715 29.25 13.00
 Q2 1436 25.59 17.55 1554 26.51 19.18
 Q3 1282 22.84 19.50 1321 22.53 25.89
 Q4 (high) 1279 22.79 26.82 1273 21.71 36.06
Material factors
Housing conditions **** ****
 0 problem 3832 64.22 15.50 3874 61.42 19.44
 1 problem 1351 22.64 21.76 1568 24.86 24.74
 2 problems or more 784 13.14 29.59 865 13.71 36.18
Crowding *** ****
 <1 3253 55.03 17.09 3471 55.39 20.66
 ≥1 2658 44.97 21.14 2796 44.61 25.89
Material deprivation **** ****
 no deprivation 3350 59.71 13.52 3122 53.22 14.07
 1–2 deprivations 1078 19.22 20.48 1243 21.19 24.32
 3–5 deprivations 779 13.89 29.29 1035 17.64 33.24
 6 deprivations or more 403 7.18 32.75 466 7.94 48.37
Financial problems **** ****
 no 4417 76.01 15.96 4481 73.10 18.34
 yes 1394 23.99 28.60 1649 26.90 35.85
Neighbourhood problems *** ****
 Q1 1880 32.20 15.70 1808 29.46 17.41
 Q2 1150 19.70 19.60 1158 18.87 20.35
 Q3 1580 27.06 20.70 1705 27.78 24.87
 Q4 (high) 1229 21.05 20.26 1466 23.89 29.07
Quality of public services **** ****
 Q1 1336 27.12 14.69 1417 27.95 16.58
 Q2 1194 24.23 16.67 1175 23.18 21.08
 Q3 1242 25.21 20.18 1350 26.63 26.21
 Q4 (low) 1155 23.44 23.28 1128 22.25 29.47
Access to medical services **** ****
 Q1 2358 47.26 14.86 2290 41.67 17.13
 Q2 710 14.23 20.14 861 15.67 21.74
 Q3 1036 20.77 25.29 1284 23.37 28.81
 Q4 (low) 885 17.74 28.68 1060 19.29 33.59
Occupational factors
Economic activity **** ****
 fishing/agriculture 307 5.87 26.06 157 2.79 29.49
 industry 1186 22.66 22.03 727 12.93 29.70
 construction 711 13.58 18.48 111 1.97 18.02
 service 3030 57.89 17.23 4628 82.30 22.07
Number of hours per week *** ***
 less than 32 hours 414 7.12 23.49 1420 22.90 21.51
 32–48 3903 67.11 17.56 4063 65.53 22.67
 more than 48 hours 1499 25.77 20.76 717 11.56 28.67
Daily commuting time ** *
 less than 60 minutes 4058 74.60 21.35 4384 76.08 25.40
 60 minutes or more 1382 25.40 17.84 1378 23.92 21.83
Dangerous working conditions **** ****
 no 3535 59.70 15.33 4742 76.20 20.69
 yes 2386 40.30 23.90 1481 23.80 30.56
Job insecurity **** ****
 low 4963 87.21 17.42 5156 86.35 21.13
 high 728 12.79 25.76 815 13.65 35.02
Psychological demands ** ****
 low 3052 51.68 17.49 3590 58.02 20.79
 high 2854 48.32 20.04 2598 41.98 26.00
Decision latitude at work ** ****
 high 3223 54.55 17.03 3329 53.68 20.43
 low 2685 45.45 20.60 2872 46.32 25.94
Reward **** ****
 high 3371 57.64 14.70 2917 47.45 17.11
 low 2477 42.36 24.34 3230 52.55 28.22
Responsibility for supervising others NS **
 no 3547 59.57 19.04 4591 72.87 24.01
 yes 2407 40.43 18.68 1709 27.13 20.87
Open in a new tab

Chi-square test:

*

: p<0.05;

**

: p<0.01;

***

p<0.001;

****

p<0.0001

Q1, Q2, Q3, Q4: quartiles of scores in the whole sample

Other variables (household tenure, work contrct, and additional paid job) were not significantly associated with self-reported health

The association between occupation and SRH was significant for both genders, manual workers being at higher risk of poor health (Table 1). In each country, after adjustment for sex and age (Table 2), the OR of poor health for manual workers was higher than 1 in most cases although not significant for all countries.

All potential mediating factors were associated with occupation, except commuting time for men and marital status among women (Table 3). Strong social gradients were found for many of them, manual workers being more likely to be exposed to housing problems, overcrowding, material deprivation, financial problems, low quality of public services, low access of medical services, temporary contract, dangerous working conditions, job insecurity, low decision latitude, low reward, low social support and network, low social participation, low level of trust, and social exclusion. The factors that displayed inverse social gradients (managers/professionals were more likely to be exposed to long working hours, high psychological demands, responsability for others, and additional paid job), or no clear social gradient were excluded from the analysis. The final selection of potential mediators is presented in Table 4.

Table 3.

Mediating factors according to occupational groups

Men Women
% Professionals, managers Technicians, associate professionals Clerical support workers Service, sales workers Manual workers P-value Professionals, managers Technicians, associate professionals Clerical support workers Service, sales workers Manual workers P-value
Material factors
Tenant (household tenure) 25.25 28.29 28.92 36.82 35.22 **** 27.83 28.42 30.44 38.02 34.93 ****
High level of housing problems 7.24 9.36 10.10 13.79 18.78 **** 10.78 10.59 10.33 14.31 20.73 ****
Overcrowding 33.39 39.13 34.32 51.02 53.61 **** 38.02 39.91 41.76 42.64 56.90 ****
High material deprivation 2.08 3.08 1.29 5.78 14.55 **** 5.28 3.53 4.77 7.37 17.54 ****
Financial problems 15.52 17.91 12.76 25.95 33.38 **** 22.95 21.20 21.34 27.60 39.45 ****
High level of neighbourhood problems 22.11 22.36 22.56 22.30 18.98 * 28.04 26.31 22.24 20.62 23.17 ****
Low quality of public services 20.36 22.11 19.71 23.17 26.89 ** 21.62 21.84 21.73 18.81 27.12 ****
Low access to medical services 12.09 15.27 17.26 20.08 21.26 **** 18.93 18.97 17.58 15.10 25.18 ****
Occupational factors
Economic sectors **** ****
 fishing/agriculture 2.89 2.03 0.85 2.48 12.49 2.08 1.64 1.88 0.26 7.42
 industry 12.16 18.36 18.64 10.19 34.99 6.07 6.03 14.17 3.59 32.55
 construction 10.10 10.89 4.80 4.97 22.08 1.91 1.58 4.76 0.51 1.95
 service 74.85 68.72 75.71 82.36 30.44 89.95 90.74 79.18 95.64 58.08
Temporary contract 14.98 11.16 15.97 25.98 27.06 **** 13.24 16.79 17.12 25.22 27.04 ****
More than 48h per week 32.53 32.98 10.42 23.39 21.38 **** 17.81 14.39 4.22 9.10 12.42 ****
Daily commuting time ≥ 60 min 25.75 23.89 28.72 23.85 26.86 NS 26.51 24.25 27.82 20.61 22.50 ***
Dangerous working conditions 22.22 29.66 14.89 47.40 57.31 **** 18.47 19.25 12.33 28.88 35.81 ****
High job insecurity 9.14 8.65 8.14 13.80 18.48 **** 10.28 11.49 11.89 11.89 21.05 ****
High psychological demands 53.94 52.45 44.67 44.77 45.08 **** 49.85 48.16 38.32 35.20 38.76 ****
Low decision latitude 25.88 28.65 48.14 54.70 62.49 **** 27.49 32.69 49.70 47.45 69.12 ****
Low reward 21.84 31.09 32.27 48.86 57.71 **** 38.40 43.52 44.84 60.31 69.40 ****
Responsibility for supervising others 61.11 61.82 34.32 25.87 21.99 **** 46.03 40.75 20.55 20.62 12.04 ****
Additional paid job 12.61 7.95 6.90 7.42 6.60 **** 9.08 8.6 6.34 6.87 4.29 ****
Psychosocial factors
Separated or widowed 10.15 8.38 7.16 7.87 9.06 **** 18.48 17.91 17.04 20.40 18.87 NS
At least one child 67.84 69.13 54.79 65.35 65.07 **** 70.40 71.14 70.87 76.24 76.81 ****
High unpaid work 20.23 20.32 22.44 21.22 18.10 * 44.74 45.31 43.95 47.52 46.75 **
Work life imbalance 18.84 20.69 12.78 20.80 23.11 **** 25.42 23.47 17.66 21.44 27.69 ****
Low social support 11.07 11.41 8.70 13.38 16.04 **** 9.77 10.03 8.04 11.90 17.44 ****
Low social network 26.35 25.04 27.05 25.58 29.87 * 22.51 21.08 21.71 19.09 25.00 ***
No social participation 43.54 51.68 52.94 59.37 60.73 **** 45.37 46.42 57.21 57.09 60.76 ****
Not confident (trust level) 22.32 23.98 19.29 31.23 33.45 **** 22.65 24.92 24.92 27.36 33.25 ****
Social exclusion 12.52 17.31 18.80 23.99 31.37 **** 13.41 17.12 21.23 22.72 31.11 ****
Open in a new tab

Chi-square test:

*

: p<0.05;

**

: p<0.01;

***

p<0.001;

****

p<0.0001.

Chi-square test performed on variables as described in Table 1.

Table 4.

Contribution of mediating factors to social inequalities in health: results from multilevel logistic regression analysis

Men Women
Poor health Poor health
ORa (95% CI) %b ORa (95% CI) %b
Model 1 1.89 (1.46–2.46) 2.18 (1.71–2.77)
Material factors
Model 1 + housing conditions 1.80 10c 2.08 8c
Model 1 + crowding 1.91 −2 2.16 2
Model 1 + material deprivation 1.39 56c 1.60 49c
Model 1 + financial problems 1.65 27c 1.84 29c
Model 1 + quality of public services 1.76 15c 2.20 −2
Model 1 + access to medical services 1.86 3 2.00 15c
Model 1 + MATERIAL FACTORS 1.21 (0.88–1.65) 76 1.48 (1.11–1.97) 59
Occupational factors
Model 1 + economic activity 1.79 11c 2.09 8c
Model 1 + dangerous working conditions 1.71 20c 2.01 14c
Model 1 + job insecurity 1.86 3 2.08 8c
Model 1 + decision latitude 1.84 6c 2.00 15c
Model 1 + reward 1.66 26c 2.00 15c
Model 1 + OCCUPATIONAL FACTORS 1.43 (1.05–1.95) 52 1.77 (1.33–2.36) 35
Psychosocial factors
Model 1 + social support 1.78 12c 2.04 12c
Model 1 + social network 1.91 -2 2.17 1
Model 1 + social participation 1.88 1 2.16 2
Model 1 + trust level 1.74 17c 2.24 −5
Model 1 + social exclusion 1.61 31c 1.78 34c
Model 1 + PSYCHOSOCIAL FACTORS 1.48 (1.09–2.00) 46 1.66 (1.27–2.17) 44
Model 1 + material and occupational factors 1.08 (0.75–1.56) 91 1.32 (0.94–1.85) 73
Model 1 + material and psychosocial factors 1.11 (0.79–1.56) 88 1.40 (1.03–1.91) 66
Model 1 + occupational and psychosocial factors 1.17 (0.82–1.66) 81 1.51 (1.11–2.07) 57
Model 1 + material, occupational, and psychosocial factors 1.00 (0.67–1.49) 100 1.26 (0.88–1.80) 78
Open in a new tab
a

OR for manual workers adjusted for age (reference group: managers/professionals)

b

(ORmodel 1 − ORextended model)/(ORmodel 1 − 1)

c

Factors with a contribution > 5% included in extended models

Bold OR: significant at 5%

Table 4 presents the changes in ORs for manual workers after inclusion of potential mediating factors in model 1. Among material factors, deprivation, financial problems, and housing conditions contributed to the decrease of social differences for both genders. Quality of public services for men, and access to medical services for women were found to be specific contributors for each sex. Taking these factors into account led to a reduction of the associations between occupation and SRH by 76% and 59% for men and women respectively. Among occupational factors, the biggest contributions to a reduction of social differences were found for reward and dangerous working conditions, and to a lesser extent for economic activity and decision latitude for both genders, as well as job insecurity for women only. Taking these factors into account led to a reduction of the associations between occupation and SRH by 52% and 35% for men and women respectively. Among psychosocial factors, social exclusion and social support contributed to explain social differences in SRH for men and women. Trust level contributed to explain social differences for men only. Taking these factors into account led to a reduction of the associations between occupation and SRH by 46% and 44% for men and women respectively.

The ORs were further reduced when adjusted for combinations of two groups of factors (Table 4), the biggest reductions being observed for the combination of material and occupational factors (91% and 73% for men and women). The independent contribution of material factors in relation to occupational factors was 39% (=91–52) and 38% (=73–35) for men and women, and was higher than the independent contribution of occupational factors in relation to material factors: 15% (=91–76) and 14% (=73–59) for men and women. The final contribution of psychosocial factors in relation to both material and occupational factors was lower, as their independent effect was 9% (=100–91) and 5% (=78–73) for men and women.

Finally, the inclusion of the material, occupational, and psychosocial factors together in model 1 reduced further the ORs by 100% in men and by 78% in women, meaning that all groups of mediators independently contributed to the explanation of social inequalities in SRH. Final ORs were consequently no longer significant.

Additional analyses were performed using educational level as a SES marker. The results were essentially the same, initial ORs of 2.10 (95% CI: 1.65–2.66) and 2.01 (95% CI: 1.61–2.52) for poor health were observed for the less educated men and women (model 1), and the main mediating variables were the same. Only access to medical services was found as an additional mediator for men. Material and psychosocial factors contributed to social inequalities to the same magnitude: 75% and 60% for material factors, and 45% and 54% for psychosocial factors, for respectively men and women. Two occupational factors were not found as mediators: economic sector for both genders and decision latitude for men. The contribution of occupational factors was found to be lower using education as a SES marker: 36% and 22% for men and women.

Discussion

Social inequalities in SRH were observed, with ORs of around 2 for manual workers. Material, occupational, and psychosocial factors contributed to an explanation of these inequalities, and the four biggest contributions were found for material deprivation, social exclusion, financial problems, and reward at work. Material as well as occupational factors played a major role in explaining social inequalities in health. All mediators taken into account together explained social differences in health by 78–100% for men and women, the final ORs being no longer significant for manual workers for both genders.

Comparison with literature

Poor health was more likely to be reported among manual workers than among workers in higher occupational groups. This inverse relationship between SES and SRH is well documented in the literature, which consistently shows a higher level of poor health for people with a lower educational level, occupational category, or income.[10;24;39] The result from the EQLS data covering 28 countries in Europe is in keeping with a previous study investigating educational inequalities in health in 22 European countries and showing ORs of 1.65 and 1.81 for the lowest educated men and women respectively.[13]

Significant differences in SRH were observed between countries. Our study showed that the prevalence of poor health was higher in the eastern countries compared to others for both genders, as other studies had already demonstrated for both morbidity and mortality.[5;43;44] A differential reporting of SRH across countries may also be suspected due to cultural differences for example. Occupational differences in health were found in almost all countries.

Differences between genders were observed for SES, SRH, and mediating factors, and this reinforces the relevance of studying each gender separately. In terms of social inequalities in health, no difference was found between genders, the magnitude of these inequalities being similar for men and women. Some previous studies have reported however that social differences in health were more pronounced in men than in women.[24;37] Our study also showed that gender-specific factors may play a role in explaining social inequalities in health, such as trust level and quality of public services among men, and access to medical services among women.

In this study, strong social gradients were found for almost all mediating factors. This is in agreement with previous studies showing that low SES groups are more likely to be exposed, for example, to high material deprivation, low decision latitude, or high job insecurity than higher SES groups.[22–24] Some exceptions may be worth noting such as the occupational factors of psychological demands, working hours, and responsibility for others. Other studies have underlined the inverse social gradient for psychological demands.[23;32;33;35]

Previous studies have explored material,[19;20;22;24;34;36] psychosocial,[19;20;29] and/or occupational factors[23;24;26;27;31–35;37;40] as mediators for social inequalities in health, but the studies exploring several dimensions at the same time are relatively rare.[19;20;23;24;34;36;37;39]

In our study, material factors were the set of mediators that contributed the most to reduce social inequalities in health, confirming previous results at national levels.[18;19] Material deprivation played the biggest role in the explanation of these inequalities for both genders, and this result was consistent with existing literature.[34] Financial problems also accounted for a substantial part of the observed differences, as previously reported by other authors studying educational differences in mortality in the Netherlands.[19;36] Housing conditions, quality of public services (men only), and access to medical services (women only) were also found to contribute to occupational differences in health in our study.

Occupational factors contributed to the social differences in health for men and women substantially. Important mediators were reward, dangerous working conditions, and economic activity -that may be an indirect marker of working conditions or occupational exposures-, and also decision latitude and job insecurity (women only). Previous studies have demonstrated the contribution of occupational factors, especially decision latitude and various markers of ergonomic/physical/chemical exposures, in the association between social class and health.[24;31–33;35;38] Studies reported that 50–74% of occupational differences in health for men and 38–51% for women were explained by these factors.[32;35] Job insecurity was reported as a mediator of social inequalities in self-reported health in a previous study in Spain.[24] In the present study, reward was also found as a mediator for social inequalities in health, something never reported before.

Psychosocial factors also contributed to reduce social inequalities in health although to a lesser magnitude than the two other sets of mediators. Social exclusion was the factor which contributed the most to the reduction of inequalities, a finding never reported before. Our results were in agreement with studies showing the contribution of social support to social inequalities in health.[37] Our findings suggest that trust level may be an additional mediator in social inequalities in health.

The retained mediating factors in our final models were able to explain social differences in SRH by 100% in men and by 78% for women, the final ORs being finally non significant. A previous study[21] exploring simultaneously physiological, biological, and behavioural factors observed a reduction of the educational differences in SRH by 40% in a population of Danish employees. The contribution of material, psychosocial, and behavioural factors was also explored in a recent study from South Korea, tending to explain 56.5–77.5% of occupational inequalities in all-cause mortality, for men and women together.[20] Material, psychosocial, behavioural, and biomedical factors contributed to reduce social inequalities in mortality by 73–83% in a Norwegian sample.[18] Social inequalities in mortality were explained by 100% by material, psychosocial, and behavioural factors in the Netherlands.[19]

Limitations and strengths

The response rate (58.4%) in the 2003 EQLS may be considered as low,[41] and differed across countries. Consequently, it may be difficult to evaluate the impact on the results. However, previous studies reported that non-participants may be more likely to belong to low social groups, to have poor health behaviours and health outcomes,[45] and such a bias may lead to underestimate the prevalence of health outcomes as well as the association between SES and health outcomes. This cross-sectional study was necessarily restricted to working men and women. A selection bias may have occurred, i.e. workers in poor health may have changed or left their job and/or healthier workers may be more likely to work in more difficult jobs. Thus, the associations between work factors and health outcome may be underestimated as well as the contribution of these factors to social inequalities in health. The relative importance of material and psychosocial mediators might be different if the population studied was the whole population, and not only the working population. This study did not take into account lifetime exposure to the different mediating factors. This may lead to an underestimated contribution of the mediating factors to social inequalities in health.[31] Some mediating factors may have been neglected in this study; this may be especially the case for behavioural factors (e.g. smoking, alcohol consumption, physical activity) which may play a role in the explanation of social inequalities in health.[23;30] However, behavioural factors may also be consequences of material, occupational, and psychosocial mediators, and consequently their independent effects may be low, as underlined by others previously.[19] All variables studied were based on self-reports. The use of SRH might introduce a reporting bias in our study, but a previous paper[46] showed that measurement errors were limited and SRH may have many advantages as a summary health measure. Furthermore, authors have reported that the use of SRH may lead to underestimate social inequalities in health, and not the reverse.[47]

The use of a large scale and comparable data across Europe allowed us to study each gender separately and to avoid problems of data comparability between countries.[2;7] An additional advantage of this study was to be based on face-to-face interviews. Our study was also able to determine the specific mediators of social inequalities for each gender. Our study aimed at describing social inequalities in health and also at determining mediating factors of these inequalities. For this purpose, three different sets of mediating factors were studied, each containing various variables. The relative contributions of material and psychosocial mediators were confirmed in additional analyses using education as a SES marker. However, these analyses were less powerful to identify occupational mediators, maybe because education is a more general SES marker than occupation.

Conclusion

Overall, this study sheds some light on the factors that may mediate social inequalities in health among working men and women in Europe. It emphasises the disadvantages experienced by manual workers and some categories of service workers. This study also underlines the need to address material deprivation, financial problems, social exclusion, -in other words all dimensions of poverty and deprivation-, as well as poor working conditions, to reduce social inequalities in health. More studies are needed to evaluate prospectively the role of a large variety of mediators in these inequalities, to better understand the underlying mechanisms linking these mediators to social inequalities in health, especially as regards specific indicators of morbidity and mortality. Finally, as the mediators and their relative contributions may differ between European countries, forthcoming studies should also focus on contextual factors that may contribute to a better understanding of social inequalities in health in Europe.

What is already known on this subject

  • There are social inequalities in health in various European countries

  • Explanations for social inequalities in health have been reported but studies exploring a large range of mediating factors are still lacking.

  • Explanatory studies using European harmonised data are also missing.

What this study adds

  • Social inequalities in self-reported health have been found for both men and women across Europe.

  • Selections of mediators covering material, occupational, and psychosocial aspects were able to explain these inequalities by 100% and 78% respectively for men and women.

  • The strongest mediators were those related to material deprivation, financial problems, social exclusion, and various working conditions underlying the importance of material and occupational factors.

  • Some specific mediators were found according to gender.

  • Preventive actions focusing on these mediators may contribute to reduce social inequalities in health.

Appendix– Countries and codes in EQLS 2003

Countries Codes
Austria AT
Belgium BE
Bulgaria BG
Cyprus CY
Czech Republic CZ
Denmark DK
Estonia EE
Finland FI
France FR
Germany DE
Greece GR
Hungary HU
Ireland IE
Italy IT
Latvia LV
Lithuania LT
Luxembourg LU
Malta MT
Netherlands NL
Poland PL
Portugal PT
Romania RO
Slovakia SK
Slovenia SI
Spain ES
Sweden SE
Turkey TR
United Kingdom GB
Open in a new tab

Footnotes

Competing interests: None.

Licence for Publication statement

“The Corresponding Author has the right to grant on behalf of all authors and does grant on behalf of all authors, an exclusive licence (or non exclusive for government employees) on a worldwide basis to the BMJ Publishing Group Ltd to permit this article (if accepted) to be published in JECH and any other BMJPGL products and sublicences such use and exploit all subsidiary rights, as set out in our licence (http://jech.bmj.com/site/about/licence.pdf).”

References

  • 1.Avendano M, Kunst AE, Huisman M, et al. Socioeconomic status and ischaemic heart disease mortality in 10 western European populations during the 1990s. Heart. 2006;92(4):461–467. doi: 10.1136/hrt.2005.065532. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Cavelaars AE, Kunst AE, Geurts JJ, et al. Differences in self reported morbidity by educational level: a comparison of 11 western European countries. J Epidemiol Community Health. 1998;52(4):219–227. doi: 10.1136/jech.52.4.219. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Cavelaars AE, Kunst AE, Geurts JJ, et al. Morbidity differences by occupational class among men in seven European countries: an application of the Erikson-Goldthorpe social class scheme. Int J Epidemiol. 1998;27(2):222–230. doi: 10.1093/ije/27.2.222. [DOI] [PubMed] [Google Scholar]
  • 4.Dalstra JA, Kunst AE, Borrell C, et al. Socioeconomic differences in the prevalence of common chronic diseases: an overview of eight European countries. Int J Epidemiol. 2005;34(2):316–326. doi: 10.1093/ije/dyh386. [DOI] [PubMed] [Google Scholar]
  • 5.Eikemo TA, Huisman M, Bambra C, et al. Health inequalities according to educational level in different welfare regimes: a comparison of 23 European countries. Sociol Health Illn. 2008;30(4):565–582. doi: 10.1111/j.1467-9566.2007.01073.x. [DOI] [PubMed] [Google Scholar]
  • 6.Eikemo TA, Bambra C, Joyce K, et al. Welfare state regimes and income-related health inequalities: a comparison of 23 European countries. Eur J Public Health. 2008;18(6):593–599. doi: 10.1093/eurpub/ckn092. [DOI] [PubMed] [Google Scholar]
  • 7.Kunst AE, Groenhof F, Mackenbach JP. Mortality by occupational class among men 30–64 years in 11 European countries. EU Working Group on Socioeconomic Inequalities in Health. Soc Sci Med. 1998;46(11):1459–1476. doi: 10.1016/s0277-9536(98)00041-0. [DOI] [PubMed] [Google Scholar]
  • 8.Kunst AE, Groenhof F, Andersen O, et al. Occupational class and ischemic heart disease mortality in the United States and 11 European countries. Am J Public Health. 1999;89(1):47–53. doi: 10.2105/ajph.89.1.47. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Kunst AE, Bos V, Lahelma E, et al. Trends in socioeconomic inequalities in self-assessed health in 10 European countries. Int J Epidemiol. 2005;34(2):295–305. doi: 10.1093/ije/dyh342. [DOI] [PubMed] [Google Scholar]
  • 10.Mackenbach JP, Kunst AE, Cavelaars AE, et al. Socioeconomic inequalities in morbidity and mortality in western Europe. The EU Working Group on Socioeconomic Inequalities in Health. Lancet. 1997;349(9066):1655–1659. doi: 10.1016/s0140-6736(96)07226-1. [DOI] [PubMed] [Google Scholar]
  • 11.Mackenbach JP, Stirbu I, Roskam AJ, et al. Socioeconomic inequalities in health in 22 European countries. N Engl J Med. 2008;358(23):2468–2481. doi: 10.1056/NEJMsa0707519. [DOI] [PubMed] [Google Scholar]
  • 12.van Doorslaer E, Wagstaff A, Bleichrodt H, et al. Income-related inequalities in health: some international comparisons. J Health Econ. 1997;16(1):93–112. doi: 10.1016/s0167-6296(96)00532-2. [DOI] [PubMed] [Google Scholar]
  • 13.Von Dem Knesebeck O, Verde PE, Dragano N. Education and health in 22 European countries. Soc Sci Med. 2006;63(5):1344–1351. doi: 10.1016/j.socscimed.2006.03.043. [DOI] [PubMed] [Google Scholar]
  • 14.Idler EL, Benyamini Y. Self-rated health and mortality: a review of twenty-seven community studies. J Health Soc Behav. 1997;38(1):21–37. [PubMed] [Google Scholar]
  • 15.Goldman N. Social inequalities in health disentangling the underlying mechanisms. Ann N Y Acad Sci. 2001;954:118–139. [PubMed] [Google Scholar]
  • 16.Mackenbach J-P. Health inequalities: Europe in profile. European Commission; 2006. http://www.dh.gov.uk/prod_consum_dh/groups/dh_digitalassets/@dh/@en/documents/digitalasset/dh_4121584.pdf. [Google Scholar]
  • 17.Wilkinson R, Marmot M. Social determinants of health: The solid facts. 2. World Health Organization; 2003. [Google Scholar]
  • 18.Skalicka V, Van Lenthe F, Bambra C, et al. Material, psychosocial, behavioural and biomedical factors in the explanation of relative socio-economic inequalities in mortality: evidence from the HUNT study. Int J Epidemiol. 2009;38(5):1272–1284. doi: 10.1093/ije/dyp262. [DOI] [PubMed] [Google Scholar]
  • 19.van Oort FV, van Lenthe FJ, Mackenbach JP. Material, psychosocial, and behavioural factors in the explanation of educational inequalities in mortality in The Netherlands. J Epidemiol Community Health. 2005;59(3):214–220. doi: 10.1136/jech.2003.016493. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Khang YH, Lynch JW, Yang S, et al. The contribution of material, psychosocial, and behavioral factors in explaining educational and occupational mortality inequalities in a nationally representative sample of South Koreans: Relative and absolute perspectives. Soc Sci Med. 2009;68(5):858–866. doi: 10.1016/j.socscimed.2008.12.003. [DOI] [PubMed] [Google Scholar]
  • 21.Arendt JN, Lauridsen J. Do risk factors explain more of the social gradient in self-reported health when adjusting for baseline health? Eur J Public Health. 2008;18(2):131–137. doi: 10.1093/eurpub/ckm096. [DOI] [PubMed] [Google Scholar]
  • 22.Bobak M, Pikhart H, Rose R, et al. Socioeconomic factors, material inequalities, and perceived control in self-rated health: cross-sectional data from seven post-communist countries. Soc Sci Med. 2000;51(9):1343–1350. doi: 10.1016/s0277-9536(00)00096-4. [DOI] [PubMed] [Google Scholar]
  • 23.Borg V, Kristensen TS. Social class and self-rated health: can the gradient be explained by differences in life style or work environment? Soc Sci Med. 2000;51(7):1019–1030. doi: 10.1016/s0277-9536(00)00011-3. [DOI] [PubMed] [Google Scholar]
  • 24.Borrell C, Muntaner C, Benach J, et al. Social class and self-reported health status among men and women: what is the role of work organisation, household material standards and household labour? Soc Sci Med. 2004;58(10):1869–1887. doi: 10.1016/S0277-9536(03)00408-8. [DOI] [PubMed] [Google Scholar]
  • 25.Fuhrer R, Shipley MJ, Chastang JF, et al. Socioeconomic position, health, and possible explanations: a tale of two cohorts. Am J Public Health. 2002;92(8):1290–1294. doi: 10.2105/ajph.92.8.1290. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Hemstrom O. Health inequalities by wage income in Sweden: the role of work environment. Soc Sci Med. 2005;61(3):637–647. doi: 10.1016/j.socscimed.2004.12.028. [DOI] [PubMed] [Google Scholar]
  • 27.Huisman M, Van LF, Avendano M, et al. The contribution of job characteristics to socioeconomic inequalities in incidence of myocardial infarction. Soc Sci Med. 2008;66(11):2240–2252. doi: 10.1016/j.socscimed.2008.01.049. [DOI] [PubMed] [Google Scholar]
  • 28.Marmot MG, Bosma H, Hemingway H, et al. Contribution of job control and other risk factors to social variations in coronary heart disease incidence. Lancet. 1997;350(9073):235–239. doi: 10.1016/s0140-6736(97)04244-x. [DOI] [PubMed] [Google Scholar]
  • 29.Marmot MG, Fuhrer R, Ettner SL, et al. Contribution of psychosocial factors to socioeconomic differences in health. Milbank Q. 1998;76(3):403–48. 305. doi: 10.1111/1468-0009.00097. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Monden CW. Current or lifetime smoking? Consequences for explaining educational inequalities in self-reported health. Prev Med. 2004;39(1):19–26. doi: 10.1016/j.ypmed.2004.02.016. [DOI] [PubMed] [Google Scholar]
  • 31.Monden CW. Current and lifetime exposure to working conditions. Do they explain educational differences in subjective health? Soc Sci Med. 2005;60(11):2465–2476. doi: 10.1016/j.socscimed.2004.11.017. [DOI] [PubMed] [Google Scholar]
  • 32.Niedhammer I, Chastang JF, David S, et al. The contribution of occupational factors to social inequalities in health: findings from the national French SUMER survey. Soc Sci Med. 2008;67(11):1870–1881. doi: 10.1016/j.socscimed.2008.09.007. [DOI] [PubMed] [Google Scholar]
  • 33.Rahkonen O, Laaksonen M, Martikainen P, et al. Job control, job demands, or social class? The impact of working conditions on the relation between social class and health. J Epidemiol Community Health. 2006;60(1):50–54. doi: 10.1136/jech.2005.035758. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Sacker A, Bartley M, Firth D, et al. Dimensions of social inequality in the health of women in England: occupational, material and behavioural pathways. Soc Sci Med. 2001;52(5):763–781. doi: 10.1016/s0277-9536(00)00176-3. [DOI] [PubMed] [Google Scholar]
  • 35.Schrijvers CT, van de Mheen HD, Stronks K, et al. Socioeconomic inequalities in health in the working population: the contribution of working conditions. Int J Epidemiol. 1998;27(6):1011–1018. doi: 10.1093/ije/27.6.1011. [DOI] [PubMed] [Google Scholar]
  • 36.Schrijvers CT, Stronks K, van de Mheen HD, et al. Explaining educational differences in mortality: the role of behavioral and material factors. Am J Public Health. 1999;89(4):535–540. doi: 10.2105/ajph.89.4.535. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Sekine M, Chandola T, Martikainen P, et al. Socioeconomic inequalities in physical and mental functioning of Japanese civil servants: explanations from work and family characteristics. Soc Sci Med. 2006;63(2):430–445. doi: 10.1016/j.socscimed.2006.01.020. [DOI] [PubMed] [Google Scholar]
  • 38.Toivanen S, Hemstrom O. Income differences in cardiovascular disease: is the contribution from work similar in prevalence versus mortality outcomes? Int J Behav Med. 2006;13(1):89–100. doi: 10.1207/s15327558ijbm1301_11. [DOI] [PubMed] [Google Scholar]
  • 39.van Lenthe FJ, Schrijvers CT, Droomers M, et al. Investigating explanations of socioeconomic inequalities in health: the Dutch GLOBE study. Eur J Public Health. 2004;14(1):63–70. doi: 10.1093/eurpub/14.1.63. [DOI] [PubMed] [Google Scholar]
  • 40.Warren JR, Hoonakker P, Carayon P, et al. Job characteristics as mediators in SES-health relationships. Soc Sci Med. 2004;59(7):1367–1378. doi: 10.1016/j.socscimed.2004.01.035. [DOI] [PubMed] [Google Scholar]
  • 41.European Foundation for the Improvement of Living and Working Conditions. 2007. First European Quality of Life Survey: Methodological review. [Google Scholar]
  • 42.Lynch JW, Kaplan GA, Cohen RD, et al. Do cardiovascular risk factors explain the relation between socioeconomic status, risk of all-cause mortality, cardiovascular mortality, and acute myocardial infarction? Am J Epidemiol. 1996;144(10):934–942. doi: 10.1093/oxfordjournals.aje.a008863. [DOI] [PubMed] [Google Scholar]
  • 43.Bonneux LG, Huisman CC, de Beer JA. Mortality in 272 European regions, 2002–2004. An update. Eur J Epidemiol. 2010;25(2):77–85. doi: 10.1007/s10654-009-9415-y. [DOI] [PubMed] [Google Scholar]
  • 44.Eikemo TA, Bambra C, Judge K, et al. Welfare state regimes and differences in self-perceived health in Europe: a multilevel analysis. Soc Sci Med. 2008;66(11):2281–2295. doi: 10.1016/j.socscimed.2008.01.022. [DOI] [PubMed] [Google Scholar]
  • 45.Goldberg M, Chastang JF, Leclerc A, et al. Socioeconomic, demographic, occupational, and health factors associated with participation in a long-term epidemiologic survey: a prospective study of the French GAZEL cohort and its target population. Am J Epidemiol. 2001;154(4):373–384. doi: 10.1093/aje/154.4.373. [DOI] [PubMed] [Google Scholar]
  • 46.Crossley TF, Kennedy S. The reliability of self-assessed health status. J Health Econ. 2002;21(4):643–658. doi: 10.1016/s0167-6296(02)00007-3. [DOI] [PubMed] [Google Scholar]
  • 47.Delpierre C, Lauwers-Cances V, Datta GD, et al. Using self-rated health for analysing social inequalities in health: a risk for underestimating the gap between socioeconomic groups? J Epidemiol Community Health. 2009;63(6):426–432. doi: 10.1136/jech.2008.080085. [DOI] [PubMed] [Google Scholar]

RESOURCES