Objectives: We measured health inequalities among employed Luxembourg residents over time and the socio-economic and work-related determinants. Design and Setting: Longitudinal data were obtained from the Socio-economic Liewen zu L ëtzebuerg/European Union Survey on Income and Living Conditions, which has been conducted each year since 2003 in Luxembourg. Participants: Participants comprised 727 Luxembourg residents (58% men), aged between 21 and 55 years in 2003, who were employed between 2003 and 2012. Primary and Secondary Outcomes Measured: The variable of interest was self-reported health. We used transition indicators on work-related factors to consider changes that individuals may have experienced in their job over this period. Results: People who moved from a part-time to a full-time contract (odds ratio (OR): 5.52, confidence interval (CI): 1.55 - 19.73), and those who moved from the 3rd or 4th quartile of earnings to the 1st or 2nd quartile (OR: 2.48, CI: 1.02 - 6.05) between 2003 and 2012, had a higher risk of being in poor health in 2012. The risk of deterioration in self-reported health in 2012 among people who were healthy in 2003 was associated with the type of contract, economic activity, and occupation. Conclusion: Health inequalities occur among employed people in Luxembourg. Their importance varies according to work-related characteristics and economic activity. Our findings showed that declined health status was associated with contract type, profession, and economic activity. This suggests that measures should be taken to maintain good health for people working in these specific occupations or economic sectors (e.g. preventive action, reduction of risk exposure, change of occupation in the same company, and so on).
Many studies have shown links between health and employment [
In the European Union (EU), the percentage of employees with fixed-term contracts or other temporary contracts was lower and stable between 2003 and 2015, respectively 13% and 14% [
Specifically, in Luxembourg, with 10% of employees in temporary employment in 2015 [
Data were derived from the Socio-economic Liewen zu Lëtzebuerg/European Union Survey on Income and Living Conditions (PSELL-3/EU-SILC), which has been conducted each year since 2003. However, since 2009, EUROSTAT (The Statistical Office of the European Communities) requires that one-quarter of the sample is renewed. This face-to-face survey focuses on a sample of households (5750 households, representing approximately 10,000 people) who live in Luxembourg and who have an affiliation with the Luxembourg social security system. All interviewers were able to speak the three official languages of Luxembourg (Luxembourgish, French and German). On average, an interview takes place for one hour; but it depends mostly on the number of people in the household. This survey has the advantage of providing both cross-sectional and longitudinal data. The cross-sectional data pool of the EU-SILC allows researchers to track changes in the income and living conditions of households for use in social protection policies [
In this study, the longitudinal sample included people aged 21 - 55 years who were present and employed in both 2003 and 2012. Therefore, 727 Luxembourg residents (58% men) were analysed.
The self-reported health was considered as a dependent variable in this study. The variable regarding the self-reported health of individuals at the time of the survey was derived from the question: “How do you qualify your overall health?” The answer has five components: very good, good, fair, bad, and very bad. Health status was poor when a person declares fair, bad, or very bad health, and is considered as good when the individual reports to be in good or very good health. Self-reported health is subjective information, which can be affected by the past health of an individual, his actual health (outside of the framework of employment) or other sociocultural factors. However, this indicator is widely used in many studies [
The presence of any chronic (long-standing) illness and of any limitation in activities because of health problems are other parameters that can influence the overall perception of individuals of their self-reported health and this can reflect an overall health status running for several years. These health variables were assessed respectively by follow-up questions: “Do you suffer from/have any chronic (long-standing) illness or condition (health problem)?” and “For at least the last 6 months, have you been limited in activities people usually do because of a health problem?” (If limited, specify whether strongly limited or limited). Demographic variables examined included sex, age, nationality (Luxembourgish, Portuguese, and others), and marital status (never married, married, and separated/divorced/widower).
The socio-economic and work-related variables comprised the following:
1) Education level (primary, secondary, and tertiary).
2) Occupation (level 1: legislators, senior officials, managers, and professionals; level 2: technicians and associate professionals, and armed forces; level 3: clerks, service workers, and shop and market sales workers; level 4: skilled agricultural workers, craft, and skilled workers; and level 5: elementary occupation) [
3) Contract type (permanent or fixed-term): derived from the question “Please, could you tell me what is the type of M.’s contract?”
4) Working-time (full-time or part-time): derived from the question “Does M. have a part-time or a full-time job?”
5) Economic activity (manufacturing, construction, wholesale, hotels, restaurants, transport, financial intermediation, real estate, lettings, public administration, education, health and social work, and others): derived from the question “Please describe the main activity of the local unit of the business or organisation where M. works?”
6) Gross wage distribution by quartile (in 2003:
Transition indicators on worked-related factors
Transition indicators on work-related factors were created to consider the changes that individuals may have experienced in their job from 2003-2012 and to assess the consequences of these changes in self-reported health [
Changes in self-reported health
It is considered positive when the individual declared a very bad, bad, or fair health status in 2003 and a good or very good health status in 2012 and vice versa. If the individual declared the same answer in 2003 and 2012, it means that there was no change in their self-reported health.
Change in contract type
It is considered as positive if there is a transition from a fixed-term contract to an open-ended contract (−/+), as negative if the individual loses/quits his/her open-ended contract and gets a fixed-term contract (+/−), and as stable when the contract does not change (stable and positive if the individual keeps his/her open-ended contract (+/+), and stable and negative if the individual keeps his/her fixed-term contract (−/−).
Change in working-time
It is positive when the individual had a full-time contract in 2012 when he/she had a part-time contract in 2003 (−/+) and vice versa. If the working-time remained stable between 2003 and 2012, there was no change: we distinguished stability with a full-time job (+/+) from stability with a part-time job (−/−).
In the first step, we analysed the relationship between the trajectory of self-reported health status and the individual characteristics and worked-related factors. The p-values were calculated by using the Chi-square test (threshold p-value < 5%).
In the second step, two logistic regression models groups were used to assess the influence of education and work-related factors on the inequality in self-reported health. The first measured the association between self-reported health in 2012 and the evolution of employment characteristics (transition indicators) adjusted for demographics (sex, age, nationality, and marital status), socioeconomic (education level), and health (chronic conditions and activity limitation) characteristics in 2012. The second measured the probability of self-reported deterioration in 2012 amongst people who considered themselves in good or very good health in 2003, depending on the characteristics of employment in 2003 and was controlled for by using the demographic characteristics from 2003. For each logistic regression model group, we have calculated the univariable (Model 1) and multivariable (Model 2) odds-ratios to measure both the crude and adjusted values of associations. The univariable odds-ratio of the model 1 assess the probability of being in poor health by using only one independent variable (or covariate), while the multivariable odds-ratio in the model 2 includes all covariates. The results are significant to the error threshold of 5%. Furthermore, due to the small sample, the generalised estimation equation (GEE) models were not used. All statistical analyses were performed using SAS 9.4 software (SAS Institute Inc., USA).
Descriptive analyses showed links between education level, change in economic activity, occupation, type of contract, working time, and change in earnings on the one hand, and an individual’s self-reported health and its evolution over time on the other (
In
In
Poor self-reported health in 2012 | Trajectory of self-reported health status between 2003 and 2012 | ||||||
---|---|---|---|---|---|---|---|
N | % | p-value* | Positive evolution (%) | No evolution (%) | Negative evolution (%) | p-value* | |
Total | 727 | 19.8 | 6.7 | 80.3 | 13.0 | ||
Gender | |||||||
Male | 424 | 20.1 | 0.8014 | 6.7 | 78.7 | 14.6 | 0.3230 |
Female | 303 | 19.4 | 6.8 | 82.5 | 10.8 | ||
Age (years) | |||||||
[30 - 40] | 211 | 13.9 | 0.0084 | 3.8 | 87.6 | 8.6 | 0.0222 |
[40 - 50] | 301 | 19.7 | 8.7 | 77.7 | 13.6 | ||
[50 - 64] | 215 | 25.8 | 6.9 | 76.8 | 16.3 | ||
Nationality | |||||||
Luxembourgish | 423 | 16.1 | 0.0001 | 5.0 | 84.7 | 10.3 | 0.0054 |
Portuguese | 139 | 32.5 | 9.4 | 71.0 | 19.6 | ||
Other | 165 | 18.6 | 9.0 | 76.7 | 14.4 |
Marital status | |||||||
---|---|---|---|---|---|---|---|
Never married | 139 | 21.5 | 0.5770 | 9.1 | 81.8 | 9.1 | 0.3242 |
Married | 514 | 18.8 | 6.6 | 79.6 | 13.7 | ||
Separated/divorced/widowed | 74 | 23.2 | 3.2 | 81.8 | 15.0 | ||
Education level | |||||||
Primary | 154 | 31.7 | <0.0001 | 6.6 | 74.5 | 19.0 | 0.0226 |
Secondary | 356 | 19.2 | 7.7 | 79.4 | 12.9 | ||
Tertiary | 214 | 12.4 | 4.2 | 86.9 | 8.9 | ||
Economic activity | |||||||
Manufacturing | 87 | 21.7 | 0.2778 | 10.9 | 77.1 | 12.0 | 0.0065 |
Construction | 54 | 28.9 | 14.9 | 61.7 | 23.4 | ||
Wholesale, hotels, restaurants, transport | 146 | 13.9 | 5.7 | 85.8 | 8.5 | ||
Financial intermediation, real estate, lettings | 78 | 20.1 | 1.9 | 80.0 | 18.1 | ||
Public administration, education, health and social work | 294 | 20.4 | 5.1 | 82.9 | 12.0 | ||
Others | 68 | 19.7 | 9.7 | 76.5 | 13.9 | ||
Economic activity change | |||||||
Yes | 178 | 15.0 | 0.0631 | 10.4 | 81.7 | 7.9 | 0.0096 |
No | 549 | 21.4 | 5.6 | 79.8 | 14.6 | ||
Occupation | |||||||
Stable positive (+/+) | 249 | 14.0 | 0.0016 | 4.4 | 84.9 | 10.7 | 0.0015 |
Stable negative (−/−) | 301 | 25.9 | 8.2 | 77.5 | 14.3 | ||
Negative (+/−) | 119 | 20.8 | 2.6 | 81.2 | 16.3 | ||
Positive (−/+) | 58 | 11.0 | 18.0 | 72.9 | 9.2 | ||
Employment contract (permanent contract/others) | |||||||
Stable positive (+/+) | 636 | 18.0 | 0.0044 | 6.2 | 81.1 | 12.7 | 0.1718 |
Stable negative (−/−) | 10 | 54.6 | 6.4 | 93.6 | 0.0 | ||
Negative (+/−) | 16 | 31.2 | 5.7 | 79.9 | 14.4 | ||
Positive (−/+) | 56 | 28.6 | 14.5 | 67.7 | 17.8 | ||
Working-time (Part-time/Full time) | |||||||
Stable positive (+/+) | 548 | 18.2 | 0.0854 | 6.8 | 80.7 | 12.5 | 0.0545 |
Stable negative (−/−) | 99 | 23.1 | 11.0 | 72.3 | 16.7 | ||
Negative (+/−) | 57 | 22.8 | 0.8 | 83.1 | 16.0 | ||
Positive (−/+) | 21 | 38.7 | 0.3 | 98.1 | 1.6 | ||
Unemployment period between 2003 and 2012 | |||||||
At least one | 78 | 15.9 | 0.3547 | 14.6 | 73.6 | 11.9 | 0.0137 |
None | 649 | 20.3 | 5.8 | 81.1 | 13.1 | ||
---|---|---|---|---|---|---|---|
Frequency of job changes between 2003 and 2012 | |||||||
None | 494 | 21.2 | 0.3890 | 4.0 | 82.6 | 13.3 | 0.0011 |
One | 151 | 16.4 | 12.8 | 74.0 | 13.2 | ||
2 or more | 82 | 17.8 | 11.9 | 77.6 | 10.4 | ||
Evolution of wage quartile | |||||||
Stable positive (+/+) | 306 | 16.1 | 0.0515 | 6.0 | 81.9 | 12.2 | 0.2128 |
Stable negative (−/−) | 309 | 21.4 | 9.2 | 77.1 | 13.8 | ||
Negative (+/−) | 49 | 31.9 | 2.6 | 87.1 | 10.2 | ||
Positive (−/+) | 61 | 21.4 | 1.6 | 82.6 | 15.9 |
Source: PSELL-EU-SILC 2003-2012; authors’ calculations. *: p-values were calculated using Chi-square test.
Model 1 | Model 2 | |||
---|---|---|---|---|
ORu | 95% CI | OR adjusted | 95% CI | |
Gender | ||||
Female | ref. | ref. | ||
Male | 1.05 | [0.72 - 1.52] | 2.50** | [1.31 - 4.79] |
Age | ||||
[30 - 39] | ref. | ref. | ||
[40 - 49] | 1.52 | [0.94 - 2.47] | 1.41 | [0.76 - 2.61] |
[50 - 64] | 2.16** | [1.32 - 3.55] | 2.83* | [1.47 - 5.43] |
Nationality | ||||
Luxembourgish | ref. | ref. | ||
Portuguese | 2.51*** | [1.62 - 3.90] | 6.11*** | [2.63 - 14.19] |
Other | 1.19 | [0.74 - 1.90] | 2.74** | [1.40 - 5.36] |
Marital Status | ||||
Married | ref. | ref. | ||
Never married | 1.18 | [0.74 - 1.87] | 1.64 | [0.88 - 3.04] |
Separated/divorced/widowed | 1.30 | [0.73 - 2.33] | 1.47 | [0.70 - 3.11] |
Chronic disease | ||||
No | ref. | ref. | ||
Yes | 6.06*** | [3.97 - 9.25] | 4.87*** | [2.90 - 8.18] |
Activity limitations | ||||
---|---|---|---|---|
No | ref. | ref. | ||
Yes | 6.44*** | [4.09 - 10.13] | 5.92*** | [3.39 - 10.34] |
Level of education | ||||
Tertiary | ref. | ref. | ||
Primary | 3.27*** | [1.93 - 5.55] | 2.51 | [0.88 - 7.12] |
Secondary | 1.67* | [1.03 - 2.71] | 2.62* | [1.23 - 5.62] |
Occupation | ||||
Stable positive (+/+) | ref. | ref. | ||
Stable negative (−/−) | 2.15*** | [1.39 - 3.35] | 1.35 | [0.63 - 2.92] |
Negative (+/−) | 1.61 | [0.91 - 2.85] | 1.42 | [0.69 - 2.92] |
Positive (−/+) | 0.76 | [0.31 - 1.87] | 0.69 | [0.22 - 2.20] |
Employment contract (permanent contract/others) | ||||
Stable positive (+/+) | ref. | ref. | ||
Stable negative (−/−) | 5.49** | [1.56 - 19.39] | 3.23 | [0.66 - 15.83] |
Negative (+/−) | 2.07 | [0.71 - 6.07] | 2.08 | [0.50 - 8.59] |
Positive (−/+) | 1.83 | [0.99 - 3.38] | 1.56 | [0.68 - 3.59] |
Working-time (Part-time/Full-time) | ||||
Stable positive (+/+) | ref. | ref. | ||
Stable negative (−/−) | 1.35 | [0.81 - 2.25] | 2.20 | [0.95 - 5.09] |
Negative (+/−) | 1.33 | [0.69 - 2.56] | 2.16 | [0.79 - 5.93] |
Positive (−/+) | 2.84* | [1.15 - 6.99] | 5.41** | [1.52 - 19.23] |
Unemployment period between 2003 and 2012 | ||||
None | ref. | ref. | ||
At least one | 0.74 | [0.39 - 1.40] | 0.82 | [0.32 - 2.07] |
Frequency of job changes between 2003 and 2012 | ||||
None | ref. | ref. | ||
One | 0.73 | [0.45 - 1.18] | 1.29 | [0.68 - 2.46] |
Two or more | 0.81 | [0.44 - 1.48] | 1.23 | [0.52 - 2.86] |
Evolution of wage quartile | ||||
Stable positive (+/+) | ref. | ref. | ||
Stable negative (−/−) | 1.43 | [0.95 - 2.14] | 0.51 | [0.23 - 1.12] |
Negative (+/−) | 2.45** | [1.24 - 4.81] | 2.46* | [1.01 - 6.03] |
Positive (−/+) | 1.42 | [0.72 - 2.81] | 1.67 | [0.66 - 4.24] |
R2 | 0.2212 |
*p < 0.05, **p < 0.01, ***p < 0.0001. Model 1: a model with one independent variable; ORu: Univariable Odds Ratio (or unadjusted OR); Model 2: a model with all independent variables in the table; OR adjusted: Adjusted Odds Ratio (or multivariable OR). Source: PSELL-EU-SILC 2003-2012; authors’ calculations.
Model 1 | Model 2 | |||
---|---|---|---|---|
ORu | 95% CI | OR adjusted | 95% CI | |
Gender | ||||
Female | ref. | ref. | ||
Male | 1.36 | [0.86 - 2.16] | 1.45 | [0.74 - 2.84] |
Age (years) | ||||
[21 - 29] | ref. | ref. | ||
[30 - 39] | 1.51 | [0.82 - 2.80] | 1.59 | [0.75 - 3.37] |
[40 - 49] | 2.12* | [1.13 - 3.98] | 2.57* | [1.17 - 5.64] |
[50 - 64] | 2.51 | [0.91 - 6.92] | 3.40 | [0.99 - 11.68] |
Nationality | ||||
Luxembourgish | ref. | ref. | ||
Portuguese | 1.89* | [1.11 - 3.21] | 1.85 | [0.77 - 4.45] |
Other | 1.10 | [0.64 - 1.88] | 1.27 | [0.64 - 2.54] |
Marital status | ||||
Married | ref. | ref. | ||
Never married | 0.64 | [0.38 - 1.09] | 0.91 | [0.45 - 1.83] |
Separated/divorced/widowed | 1.23 | [0.60 - 2.53] | 1.43 | [0.61 - 3.32] |
Chronic disease | ||||
No | ref. | ref. | ||
Yes | 1.47 | [0.77 - 2.80] | 1.37 | [0.66 - 2.86] |
Activity limitations | ||||
No | ref. | ref. | ||
Yes | 3.18*** | [1.74 - 5.83] | 2.37* | [1.17 - 4.79] |
Level of education | ||||
Tertiary | ref. | ref. | ||
Primary | 2.00* | [1.07 - 3.73] | 0.57 | [0.19 - 1.69] |
Secondary | 1.64 | [0.93 - 2.91] | 1.10 | [0.50 - 2.46] |
Economic activity | ||||
Public administration, education, health and social work | ref. | ref. | ||
Manufacturing | 1.05 | [0.45 - 2.43] | 1.43 | [0.52 - 3.97] |
Construction | 3.08** | [1.39 - 6.83] | 5.14** | [1.53 - 17.31] |
Wholesale, hotels, restaurants, transport | 0.87 | [0.44 - 1.71] | 1.15 | [0.49 - 2.70] |
Financial intermediation, real estate, lettings | 1.11 | [0.59 - 2.11] | 1.57 | [0.74 - 3.33] |
Others | 1.60 | [0.77 - 3.30] | 1.42 | [0.60 - 3.34] |
Occupation |
Level 1: Legislators, senior official, managers, and professionals | ref. | ref. | ||
---|---|---|---|---|
Level 2: Technicians and associate professionals, armed forces | 2.03 | [0.94 - 4.40] | 2.63* | [1.03 - 6.71] |
Level 3: Clerks, service workers, shop and market sales workers | 1.92 | [0.87 - 4.21] | 3.15* | [1.10 - 9.01] |
Level 4: Skilled agricultural workers, craft, and skilled workers | 2.52* | [1.11 - 5.71] | 2.10 | [0.63 - 6.96] |
Level 5: Elementary occupation | 4.06*** | [1.77 - 9.30] | 5.92** | [1.57 - 22.28] |
Employment contract (permanent contract/others) | ||||
Permanent contracts | ref. | ref. | ||
Others | 1.65 | [0.79 - 3.46] | 2.74* | [1.15 - 6.57] |
Working-time (Part-time/Full time) | ||||
Full-time | ref. | ref. | ||
Part-time | 1.27 | [0.71 - 2.26] | 1.33 | [0.60 - 2.97] |
Wage quartile | ||||
3rd or 4th quartile | ref. | ref. | ||
1st quartile | 1.24 | [0.73 - 2.11] | 0.73 | [0.31 - 1.74] |
2nd quartile | 1.32 | [0.77 - 2.25] | 1.14 | [0.57 - 2.26] |
R2 | 0.0776 |
*p < 0.05, **p < 0.01, ***p < 0.0001. Model 1: a model with one independent variable; ORu: Univariable Odds Ratio (or unadjusted OR); Model 2: a model with all independent variables in the table; OR adjusted: Adjusted Odds Ratio (or multivariable OR). Source: PSELL-EU-SILC 2003-2012; authors’ calculations.
The results of this study showed health inequalities among the employed cohort between 2003 and 2012. These inequalities in self-reported health were associated with wage level, economic activity, occupation, type of contract, and evolution in the type of job.
Our findings showed that people who experienced a negative evolution of wage quartile between 2003 and 2012 perceived poor self-reported health. Previous studies have generally sought to identify the link between income inequalities and self-reported health rather than the link between the distribution of income within a population and self-reported health [
No significant relationship was found between self-reported health in 2012 and the frequency of changes in employment between 2003 and 2012, which was consistent with other studies [
About the deterioration of self-reported health between 2003 and 2012, our results showed that individuals who worked in 2003 in the construction sector had the greatest odds of declining health in 2012 compared to those working in the administration, education, or the health sector. Debrand and Lengagne also studied the link between difficult working conditions such as physically arduous work and self-reported health among older workers [
The association between occupation and self-reported health shows that clerical support workers, service and sales workers, and workers in elementary occupations had a higher risk of perceiving their health as poor or very poor compared to managers and professionals. This result is consistent with Ortiz, who highlights that being workers or employees, as opposed to managers, regardless of their qualifications, increased the probability of poor self-reported health [
People who had moved from a part-time to a full-time job between 2003 and 2012 were more likely to report poor health than others were. This working-time effect is considered to have detrimental effects on the health and psychological being-well of workers [
Kim combined these two work-related factors in one of his studies and showed that people with a precarious job (i.e. in temporary work or with part-time and or fixed-term contracts) declared more frequently than others to be in poor health [
Moreover, Tsurugano and colleagues found links between precarious contracts and fair or poor self-reported health, subjective symptoms, and unhealthy behaviour [
Limitations and strengths
Our study was based on a longitudinal sample combining ten waves from the EU-SILC. It was a significant advantage to be able to analyse the health changes of employed people. Nevertheless, we did not consider that some of the individuals had exited the labour market and did not return during this period. Consequently, this has greatly reduced the sample size. It has also prevented us from categorising some explanatory variables more accurately. The lack of some important variables related to employment, which may have a direct or indirect link with self-reported health such as autonomy in tasks or latitude in decision-making, stressful or arduous working conditions, or even the feeling of either job insecurity and instability or job satisfaction is also a study limitation [
Due to the small sample, it was not possible to conduct analyses by sex like Schrijvers [
Finally, this study presents a selection bias. Indeed, during the study period, some employees with more severe health problems were forced to leave the labour market. There is also a problem of endogeneity in the relationship between employment and health; however, this endogeneity problem is sometimes useful for control purposes, especially in the framework of career-path issues.
Health inequalities occur among employed people, and their importance depends on both work-related characteristics and the type of economic activity/sector of the individual. Indeed, our findings showed that declined health status in 2012 in people with good health in 2003 was associated with the type of contract, profession, and economic activity. Therefore, measures should be taken to reduce exposure to occupational risks and arduous working conditions across varied economic activities and occupations. Specifically, it is necessary to implement management policies concerning, amongst others, improvement in working conditions, the prevention of difficult/hard working conditions, and other preventative actions to protect human health.
We thank the editorial team and anonymous referees for helpful comments. So, the views expressed herein are of the authors.
Authors’ ContributionsA.T and N.L. conducted the statistical analysis and wrote the manuscript. K.L. wrote the manuscript. All authors read and approved the final manuscript.
FundingThis study received no specific grant from any funding agency in public, commercial or not-for-profit sectors.
Conflicts of InterestThe authors declare no conflict of interest.
Consent for PublicationNot applicable.
Ethics Approval and Consent to ParticipateAll of the households that were selected for participation in the PSELL/EU-SILC survey received a pre-notification letter in which the survey topic and contents, interview, and voluntary nature of the study were described. Therefore, signed informed consent was not necessary. The National Commission approved the survey design and questionnaires for Data Protection of Luxembourg.
Availability of Data and MaterialsData used from the PSELL/EU-SILC survey (Panel Socio-Economique―Liewen zu Letzerbuerg/European Union―Survey on Income and Living Conditions) which is part of the Luxembourg Socio-economic panel, conducted by LISER and Statec, and part of the EU-SILC is managed by Eurostat. Currently, the authors cannot make data publicly available because they have not the right to do it. Data supporting the conclusions of this article are included in the article.
Tchicaya, A., Lorentz, N. and Leduc, K. (2018) What Is the Role of Work-Related Factors in Self-Reported Health Inequalities among Employed Individuals? A Longitudinal Study in Luxembourg. Health, 10, 1141-1158. https://doi.org/10.4236/health.2018.109087
Typology of the trajectory change of work-related factors (occupation, employment contract, working-time, wage) between 2003 and 2012.
・ NB: Level 1 was the top position and level 5 was the bottom position; (+) and (−) were considered as positive or negative position, respectively.
・ Change in economic activity between 2003 and 2012.
・ Change in economic activity was assessed by one question (answer items: yes or no).
・ Change in occupation.
・ This was considered positive when the individual held a more qualified occupation in 2012 than he/she had in 2003 (−/+: level of occupation in 2003 < level of occupation in 2012). The change was considered negative if the occupation in 2012 required less qualification than the occupation in 2003 (+/−: level of occupation in 2003 > level of occupation in 2012). There was no change if the occupation required the same qualification in 2003 and 2012: a distinction is, however, made between stability in a highly qualified occupation (+/+: level of occupation 1 or 2 in 2003 and 2012) and stability in a low-skilled occupation (−/−: level of occupation 3, 4 or 5 in 2003 and 2012). Sign (−) represents the low-skilled occupations (level of occupation: 3, 4, or 5).
・ Development of the professional path.
・ This represented the continuity of the career through the number of periods of unemployment or inactivity (none or at least one).
・ Wage change.
・ This occurred when the individual was not in the same quartile in 2003 as in 2012; it was defined as stable otherwise. The wage changed positively if he/she reached an upper quartile (−/+) and negatively if he/she decreased (+/−). There was negative stability (−/−) if the wage remained the same, but within the lowest quartile, i.e. in the 1st and 2nd, and there was positive stability (+/+) if the wage remained the same, but in the highest quartiles (3rd and 4th quartiles) between 2003 and 2012.