Skip to main content

The Gutenberg Health Study: measuring psychosocial factors at work and predicting health and work-related outcomes with the ERI and the COPSOQ questionnaire

Abstract

Background

Several instruments have been developed to assess psychosocial workload. We compared two of these instruments, the Effort-Reward Imbalance (ERI) model and the Copenhagen Psychosocial Questionnaire (COPSOQ) with regard to congruent validity and internal validity.

Methods

This analysis is based on a population-based sample of the baseline examination of 2,783 employees from the Gutenberg Health Study (GHS). About half of the participants completed the ERI questionnaire (n = 1,342), the other half completed the COPSOQ (n = 1,441). First, the two samples were compared and descriptive analyses were carried out calculating mean values for both instruments in general, then separately for age, gender and main occupational groups. Second, we analyzed the relationship between ERI and COPSOQ scales on the workplace situation and on the workplace outcomes: job satisfaction, general health, burnout, satisfaction with life, by applying stepwise logistic regression analysis.

Results and discussion

For the majority of occupations, high effort as reflected by the ERI corresponded with high demands as reflected by the COPSOQ. Comparably, high reward (according to ERI) yielded a good agreement with high “influence and development” (according to COPSOQ). However, we could also find differences between ERI and COPSOQ concerning the intensity of psychosocial workload in some occupations (e.g., physicians/pharmacists or warehouse managers/warehousemen/transport workers). These differences point to differing theoretical concepts of ERI and COPSOQ. When the ability of ERI and COPSOQ was examined to determine the associations with health and work outcomes, burnout could be better predicted by the COPSOQ; this might be due to the fact that COPSOQ comprises the constructs “work-privacy conflict” and “emotional demand”, which are closely related to burnout. However, methodological differences between these instruments limit their direct comparability.

Conclusions

The ERI and COPSOQ instrument yielded similar results for most occupational groups. The slightly stronger association between psychosocial workload as assessed by COPSOQ and burnout might be explained by its broader approach. The ability of the ERI and COPSOQ instrument to reflect relevant risk factors for clinically manifest disorders (e.g., coronary heart disease) will be derived from subsequent prospective analyses of the GHS with the follow-up data.

Peer Review reports

Background

Working conditions in western industrial countries have experienced numerous, partly fundamental changes over the last decades. Alongside with comprehensive changes of production conditions and realities in industry, administration and service, demands on the employees are also changing. Flexibility regarding time and location, endurance/robustness and social competence are becoming more and more key qualifications. As a consequence of these developments, the psychological stress has markedly increased [1].

A frequently applied concept in occupational health research distinguishes between work load or stressors (the entirety of measurable external influences) and strain (on the stress response of the employee depending on his/her individual conditions) as well as consequences of chronic strain (e.g. disease).

Models and measurement of psychosocial stress at work

Many instruments have been developed which attempt to quantitatively assess the amount of stress experienced by employees. For an effective workplace health promotion process, adequate assessment of psychosocial factors is indispensable in risk assessment. Two models dealing with the relationship between stress factors and the consequences of stress as health complaints and clinical disorders are leading in European occupational health research: First, the “demand-control model”, originally formulated by Karasek [2] which has later been expanded to the “demand-control-support model” [3] by adding the dimension of social support. This model assumes working situations to have negative psychological or physical consequences especially when high demands concur with limited decision latitude (and low social support at the workplace in the extended model).

Second, the “effort-reward imbalance model” has been developed by Siegrist [4, 5]. It postulates that the concurrence of highly extrinsic and intrinsic efforts with low chances of reward has particularly negative effects such as poor health.

On the basis of the “effort-reward imbalance model” Siegrist developed the Effort-Reward Imbalance (ERI) questionnaire [4, 5]. This is a validated and widely used instrument. There are a number of studies examining the association between the ERI and diverse health outcomes, for instance cardiovascular diseases. In recently conducted systematic reviews [6, 7], the authors found moderate evidence for a relation between psychosocial stress at work and cardiovascular morbidity and mortality. Particularly the ERI coefficient assessing the imbalance seemed to be a consistent predictor of cardiovascular diseases [6]. However, for women work stress tends to be a less powerful predictor of cardiovascular diseases than for men [7].

Another questionnaire assessing psychosocial stress at work, which has been developed more recently, is the Copenhagen Psychosocial Questionnaire (COPSOQ). The COPSOQ has been developed and validated by Kristensen and Borg of the Danish National Institute for Occupational Health in Copenhagen [8]. The questionnaire was aimed to be “theory-based without being based on one specific theory”. Therefore, the COPSOQ encompasses a broad range of aspects of currently leading concepts and theories. The following are mentioned by Kristensen et al. [9]: 1. the job characteristics model; 2. the Michigan organizational stress model; 3. The demand-control-(support) model; 4. the sociotechnical approach; 5. The action-theoretical approach; 6. the effort-reward-imbalance model; and the 7. the vitamin model. The COPSOQ tries to deal with the broadness or rather the indefiniteness of the construct “psychosocial factors” by applying a multidimensional approach with a very wide spectrum of ascertained aspects [9]. Most COPSOQ questions were taken from already existing and well approved and validated instruments, for instance from the “Setterlind Stress Profile” [10] and the “Job Content Questionnaire” [11] or from a large study (“Whitehall II Study” [12]). Only a small portion of items has been newly developed.

Aims of the current study

This study is part of the baseline examination of the Gutenberg Health Study (GHS). The aims are (1) to present and compare psychosocial factors at work separately for different occupational groups and both genders; (2) to examine how these psychosocial factors are related to health and work related outcomes (internal validity); and (3) to compare the ERI and the COPSOQ with respect to their ability to predict these health and work related outcomes (congruent validity).

Methods

Design and participants

The GHS is designed as a population-based, prospective, observational, single-centre cohort study in the Rhine-Main region in western Germany [1315] with a baseline examination and follow-up examinations after 2.5 years and 5 years. The primary aim is to evaluate and improve cardiovascular risk stratification. The sample was randomly drawn from the governmental local registry offices in the city of Mainz and the district of Mainz-Bingen. The sample was stratified 1:1 for sex and residence (urban and rural) and in equal strata for decades of age. Individuals between 35 and 74 years of age were enrolled, and written informed consent was obtained from all participants. Exclusion criteria were insufficient knowledge of the German language and physical or psychological inability to participate in the examinations at the study centre. The study protocol and sampling design were approved by the local ethics committee and by the local and federal data safety commissioners.

In the current analysis, we investigated the cross-sectional data of the baseline examination and included the first 5,000 subjects enrolled into the GHS between April 2007 and October 2008. The GHS has a response rate of 60% [16].

Measures

The (expanded) ERI questionnaire is comprised of three dimensions (in the following also called “scales”: “Effort” (6 items), “Reward” (11 items) and “Overcommitment” (6 items), all 4 point Likert scales. Examples are: “My job is physically demanding” (Effort), “I receive the respect I deserve from my work colleagues” (Reward), “I have difficulties to relax and switch off work in the evenings” (Overcommitment).

All three scale values are calculated as sum scores. They can be given as raw values with a range from 6 to 30 for effort, from 11 to 55 for reward and from 6 to 24 for overcommitment or (to facilitate comparisons with COPSOQ) transposed to a 0 to 100 range, where 0 represents the minimum and 100 the maximum (see Table 1).

Table 1 Distribution of ERI and COPSOQ scales

In order to determine the balance or imbalance between effort and reward, the ERI ratio was constructed [17] with the effort score in the nominator and the reward score in the denominator. Hence, higher values of the ratio express a higher level of imbalance between (high) effort and (low) reward.

The German version of the COPSOQ used in this study consists of 5 thematic domains including 25 scales (see Figure 1). The first four thematic domains are the psychosocial factors at work: “Demands” (4 scales), “Influence and development” (5 scales), “Interpersonal relations and leadership” (9 scales) and “Further parameters” (1 scale on insecurity at work in the present study). “Strain” represents the fifth domain with 6 scales assessing the reactions of the employees on the workplace situation as the internal outcome parameters. In addition to the 25 scales we also calculated three short scales (“Demands short scale”, “Influence short scale” and “Social support/Leadership short scale”). Each of these short scales represents one of the major predictor domains in an abbreviated manner [9, 18, 19]. We compare the mean values of these short scales to those of the three ERI scales for different occupational groups.

Figure 1
figure 1

Depicts the structure of the German COPSOQ questionnaire. The COPSOQ consists of 5 thematic domains including 25 scales. The first four thematic domains are the psychosocial factors at work and predict occupational strain. “Strain” represents the fifth domain and the outcome of the psychosocial factors.

To compare how well the ERI and COPSOQ are able to predict the presence of health and work related outcomes, we supplemented the ERI questionnaire with 4 of the 6 strain scales of the COPSOQ: “Job satisfaction”, “General health”, “Burnout” and “Satisfaction with life”.

Statistical analysis

Descriptive analyses were carried out in order to obtain range, means and standard deviations of all ERI and COPSOQ scales including the outcome variables and to obtain information about employment status, occupation, age and sex distribution. Occupations were manually double-coded according to the classification of occupations of the Federal Statistical Office Germany (KldB 92, Klassifikation der Berufe) [20]. In the analyses related to occupations, we included all professions which had a frequency of at least 20 in either the ERI or COPSOQ questionnaire. For all other analyses all cases available were included.

To determine which questionnaire and which of their respective scales predict best the health and work related outcomes we carried out multiple linear forward stepwise and backward stepwise regression analyses for each of the four outcome variables. Of the predictors with a p-value < 0.05 we kept up to five scales (COPSOQ) (those with the highest predictive power included first into the models in the forward procedure resp. excluded at last in the backward analysis) in the final model of each regression analysis. With the study-aim to compare the predictive capacity of the two instruments in groups with almost identical age and gender distribution (see below) adjustment for these factors was not necessary.

Results

Descriptive analysis

Of the first 5,000 eligible cases (this is the base for the analysis), 2,460 (49.2%) were female and 2,540 (50.8%) were male (Table 2). Mean age was 55.49 years (SD = 10.95), distribution of grouped age was as follows: 35 – 44 years: 20%, 45 – 54 years: 27.4%, 55 – 64 years: 27.1%, 65 – 74 years: 25.5%. In total, 2,783 (55.7%) of the baseline participants of GHS were currently employed (or had been employed within the last 6 weeks prior to the interview) and therefore received and completed the ERI or COPSOQ questionnaire. Of these, 1,342 (48%) completed the ERI and 1,441 (52%) completed the COPSOQ questionnaire. For two persons no data were available (see Table 3).

Distribution of the ERI or the COPSOQ questionnaire alternated weekly. No significant differences regarding age and gender of the participants were found between the weeks when either COPSOQ or ERI were distributed: age in years 55.59 (SD = 10.95) for persons in the ERI-periods and 55.39 (SD = 10.95) for COPSOQ-periods (n.s.); 48.5% females for ERI and 49.9% for COPSOQ (n.s.). There were also no differences observable between these two groups concerning participation rate with 55.1% for ERI and 56.2% for COPSOQ (n.s.).

Compared to the whole cohort, the mean age in years was significantly lower for the participants who received and completed the COPSOQ / ERI questionnaires (mean of persons with questionnaire = 49.2, standard deviation = 8.0 vs. mean = 63.4, SD = 8.4 for persons not currently employed; p < 0.001) as they needed to be employed to be included in this study part. Furthermore, there were fewer females (45.7%, p < 0.001) among the participants who completed the questionnaires, as fewer women were employed than men (Table 2).

Table 2 Demographic data on participants with COPSOQ / ERI (employed persons only)
Table 3 Demographic data on participants in GHS

Comparison of methods

Occupational groups

Table 1 shows the distribution of all ERI and COPSOQ scales without the outcome scales. Tables 4 and 5 give the distribution of the scales separately for different occupational groups. These two tables provide ample information. In the following, for the sake of simplicity we compare the ERI scales with the COPSOQ short scales instead of comparing them to each of the 19 workplace scales. We compared “Effort” (ERI) with the “Demands short scale” (COPSOQ) on the one hand, and “Reward” (ERI) and “Overcommitment” (ERI) with the “Influence short scale” (COPSOQ) and the “Social support/leadership short scale (COPSOQ)” on the other hand. This classification yields relatively good conformity in the mean values of ERI and COPSOQ for the different occupational groups between the ERI and the COPSOQ questionnaire for the majority of occupations.

Table 4 Means and standard deviation of ERI scales for different occupations
Table 5 Means and standard deviation of COPSOQ scales for different occupations

For instance, according to the ERI, agricultural occupations are rated as the occupations involving the highest effort (M = 15.6, SD = 4.2). They also yield high reward (M = 46.6, SD = 9.8). Similarly, according to the COPSOQ agricultural occupations score highest on both the “Influence short scale” (M = 82.9, SD = 15.6) and “Social support/leadership short scale” (M = 70.5, SD = 12.2). They are likewise in the top half of the “Demands short scale” (M = 50.6, SD = 15.0).

Physicians/pharmacists involve the highest demands according to the COPSOQ (M = 59.9, SD = 12.7). Physicians/pharmacists yield high scores on both the “Influence short scale” (M = 76.9, SD = 12.6) and “Social support/leadership short scale” (M = 63.8, SD = 14.5). The ERI shows similar results, as this occupational group has both a high “Effort” (M = 14.8, SD = 3.6) and “Reward” score (M = 49.3, SD = 7.2) (examples of similar results between ERI and COPSOQ are highlighted with bold letters in Tables 4 and 5).

Nonetheless, there are examples for deviances between ERI and COPSOQ, too. Scientific occupations in the humanities or nature sciences, e.g., have a high score on the “Demands short scale” in the COPSOQ (M = 54.0, SD = 15.8). In the ERI, however, they score rather low on “Effort” (M = 11.9, SD = 3.4). On the contrary, warehouse managers, warehousemen and transport workers score high on “Effort” (M = 15.5, SD = 4.3), whereas they score low on “Demands” in the COPSOQ questionnaire (M = 42.5, SD = 17.7). This may be due to the fact, that the ERI-construct “Effort” includes an item on physical demands but the COPSOQ-scale “Demand” does not (examples of dissimilar results between ERI and COPSOQ are underscored and highlighted with italic letters in Tables 4 and 5).

Gender differences

Comparing the means of the psychosocial factors at work, some gender differences were found. Concerning the ERI, men scored higher on “Effort” and “Overcommitment”. Concerning the COPSOQ, men scored higher on some scales from different domains: “Quantitative demands”, “Work-privacy-conflict”, “Influence at work”, “Degree of freedom at work”, “Possibilities for development”, “Meaning of work”, “Workplace commitment”, “Predictability”, “Role conflicts” and “Feedback”.

Comparing the means of the outcome scales, gender differences were found only for “Burnout”. On this scale, women had higher scores.

With the gender differences being relatively small and pointing predominantly to the same direction in both instruments - and since the aim of this paper to compare the two models - gender specific models are not presented in this baseline analysis.

Regression models

Table 6 shows the distribution of the four outcome scales of COPSOQ included in both questionnaires. Of the four health and work related outcomes examined in regression models, “Job satisfaction” was the outcome with the largest explained variance (R2 = 0.46 and 0.51 for ERI and COPSOQ respectively) (Table 7). “Burnout” (R2 = 0.26 vs. 0.35 respectively) and “Satisfaction with life” (R2 = 0.21 vs. 0.18 respectively) were also clearly associated with the psychosocial factors at work. The explained variance for the more “distal” outcome “General health” was somewhat lower (R2 = 0.10 vs. 0.11 respectively). Nonetheless, also some of the variance of “General health” could be predicted by psychosocial working conditions. Thus, all health and work-related outcomes are related to workplace factors, however the COPSOQ was somewhat more strongly associated with two of the four outcomes (“Job satisfaction” and “Burnout”). Concerning the ERI, the “Reward”-dimension proved to be the most important predictor for the regarded outcomes. This scale was related to each of the four outcomes. In addition, the effect size of “Reward” (the beta coefficient) was highest for all outcome scales except “Burnout”. “Burnout” was best predicted by “Overcommitment”. This scale proved to be an important factor, too, as it was related to three out of four outcome scales (not to “Job satisfaction”). The ERI-ratio (effort divided by reward) did not enter into the models, which means that this ratio in our data had no supplementary effect when tested simultaneously with the original scales “Effort” and “Reward”. We also tested whether R2 would increase by including the non-significant predictors (“Overcommitment” for “Job satisfaction” and “Effort” for “General health” and “Satisfaction with life”; results not shown). This was not the case, and hence they were not included into the models (Table 7).

Table 6 Distribution of four outcome scales presented in ERI and COPSOQ
Table 7 Regression analyses

Regarding the COPSOQ, many more scales could potentially be included in the regression models as the COPSOQ questionnaire is more comprehensive and longer than the ERI instrument. Hence, a variety of scales predicted the different outcomes. A number of predictors were included in only one of the four regression models (“Meaning of work”, “Sense of community”, “Mobbing”, “Workplace commitment” and “Quantitative demands”). However, “Work-privacy conflict” proved to be a potent predictor and was related to each of the four outcomes. For both “Burnout” and “Satisfaction with life” it was the most important predictor of all COPSOQ scales. Also “Job insecurity” was related to three of four outcomes (not to “Job satisfaction”).

For both instruments we also tested whether there were gender differences in our regression models. Many of the same predictors proved to be important for men and women even though the order in which they entered the forward stepwise regression analysis differed sometimes. However, we decided to not present gender specific models in this baseline analysis – this will be done in the longitudinal assessment later.

Discussion

This study of a population-based sample of employees from the GHS was carried out using the COPSOQ and the ERI in two comparable sub-samples. Analyses revealed a similar distribution of the corresponding ERI scales and the COPSOQ short scales for most occupational groups. When the ability of ERI and COPSOQ was examined to predict health and work outcomes, particularly burnout could be slightly better predicted by the COPSOQ.

Self-rated psychosocial workload of occupational groups according to ERI and COPSOQ

For the majority of occupations, the mean values of high effort as reflected by the ERI were relatively comparable to the mean values of high demands as reflected by the COPSOQ “Demands short scale”. Comparably, high reward (according to ERI) yielded a good agreement with high influence and development (according to the “Influence short scale” and the “Social support/leadership short scale” of the COPSOQ). This points to a high congruent validity between the two questionnaires. However, considerable differences could also be observed between ERI and COPSOQ in some occupational groups. For instance, scientists reported a high degree of “Demands short scale” in the COPSOQ questionnaire but scored rather low on “Effort” (ERI). As a potential explanation, we would like to point out that the ERI “Effort” dimension and the COPSOQ “Demand short scale” represent different theoretical constructs. For instance, the COPSOQ “Demand short scale” contains the dimensions “Demands to hide emotions” and “Emotional demands”, whereas the ERI does not cover these issues. It is conceivable that career pressure and a high degree of flexibility (home office solutions, etc.) have a potential for increased demands on these dimensions among scientists. Jiang and colleagues, for example, found that the negative spillover between work and family significantly predicted poor sleep quality among scientists [21]. On the contrary, warehouse managers, warehousemen and transport workers scored high on “Effort”, whereas they scored low on the “Demands short scale” in the COPSOQ questionnaire. Obviously these occupations involve high effort as measured by ERI or – comparably – high quantitative demands as measured by COPSOQ, but less “Demands to hide emotions”and “Emotional demands”.

Gender differences

Both in the ERI as well as in the COPSOQ, men scored higher on a number of workplace scales. Still, regarding strain, men and women scored equally high on three of the four of the outcome scales. A reason for this finding may be that men did not only score higher on psychosocial factors with negative impact (such as “Quantitative demands”, “Work-privacy-conflict” and “Role conflicts”) but also on the ones which entail status and a potentially positive impact (such as “Influence at work”, “Degree of freedom at work” and “Possibilities for development”). “Burnout” was the only outcome scale for which we found gender differences. Women reported higher levels of burnout. This has been found in earlier studies, too [1].

The ERI regression models were very similar for both men and women. Also the COPSOQ regression models were similar, although we found some differences, too. In a prospective analysis of the GHS when follow-up data will be available potential gender differences and their differential impact will be explored in more detail.

Association of health and work-related outcomes with ERI and COPSOQ scales (internal validity)

Both instruments appeared to be potent predictors of self-reported health and work-related outcomes. However, some of the 19 COPSOQ scales tended to be particularly associated with the presence of “Burnout” more so than the three ERI dimensions (“scales”). In a similar vein, Burr and colleagues found that COPSOQ was better suited to predict mental health than ERI [22]. The strongest burnout predictor among the COPSOQ scales was “Work-privacy conflict”, while among the ERI dimensions “Overcommitment” constituted the strongest burnout predictor. As the “Overcommitment” scale contains items like “Overwhelmed by pressure” and “Trouble sleeping at night” [23], the association with “Burnout” seems logical. Similarly, being torn apart between work and home environment can lead to feeling burnout [24, 25]. “Overcommitment” has been referred to as an individual’s exhaustive coping style and also in earlier studies it has been found to predict adverse health effects [26, 27]. However, in the multivariate model in our study it was not related to “Job satisfaction”. It is conceivable that high involvement in work may on the one hand lead to fading job satisfaction. On the other hand great involvement may also be an expression of great joy and positive engagement, rather than dissatisfaction, at work [28]. From a theoretical point of view, “Job satisfaction” was presumably the outcome most closely related to the actual work among the four aspects assessed. Therefore it makes sense that the psychosocial factors at work in both models predicted this outcome well and that “Job satisfaction” was the outcome with the largest explained variance. “Burnout” was also closely related to workplace factors for both instruments, but the proportion of variance explained was expectedly lower than for “Job satisfaction”. But even “General health”, which is a rather “distal” outcome and predicted by many other variables [29], could be predicted in part by psychosocial working conditions. Lastly, “Satisfaction with life” was moderately predicted by psychosocial factors at work. Work is an important part of life. Conditions at work contribute clearly to satisfaction with life [3033]. Particularly “Work-privacy conflict” and “Job insecurity” seem to hamper life satisfaction.

Hence, indicating internal validity, our results show a clear association between psychosocial workload and stress on the one hand and strain on the other hand, as would be expected and postulated by theory models in occupational health research [35, 9, 11]. Moreover, the results highlight the importance of the psychosocial environment at work for the individual’s health and well-being.

Comparison (and comparability) of ERI and COPSOQ

The COPSOQ encompasses more scales and aspects than the ERI. It aims at covering a broad range of aspects in order to catch the complexity of psychosocial factors at work predicting strain [9]. Our results indicate that the higher number of scales led to increased predictive power, albeit the difference was not that large. As the 19 regular scales of the COPSOQ predicting strain (Figure 1) are more fine-grained than the three ERI dimensions, they yielded a more nuanced picture predicting the health and work-related outcomes than the ERI scales. While the ERI dimension “Reward” was a predictor in each of the four regression analyses (focusing on burnout, job satisfaction, general health, and satisfaction with life as outcomes), different COPSOQ “Reward” subscales constituted the best predictors of these four outcomes: “Degree of freedom” predicted burnout and job satisfaction; “Meaning of work” and “Sense of community” predicted job satisfaction; “Possibilities of development” predicted general health; and “Workplace commitment” predicted satisfaction with life. The explanatory power of the ERI might have increased, if we had not based our analyses on the three ERI dimensions, but on the 23 single ERI items. However, this approach would not have been in accordance with the fundamental conceptual approach of the ERI which does not allow or suggest an analysis using the single items. Moreover, independent from the number of items, the COPSOQ covers a far much broader spectrum of work-related psychosocial factors. The ERI, on the other hand, provides a ratio which, as a single variable, has proven to constitute a potent predictor of health risks and cardiovascular morbidity and mortality in particular [6, 7]. The COPSOQ does not have this concept and an equivalent ratio or single coefficient. We would like to point out that psychosocial predictors of general health and work-related outcomes might differ from the psychosocial predictors of specific clinical disorders as hypertension, diabetes, or coronary heart disease. Therefore, further analyses using the longitudinal data of GHS will focus on clinically diagnosed illnesses rather than on self-reported health outcomes.

Strengths and limitations

This is the first investigation that enables a comparison between the COPSOQ instrument and the ERI model in a large population-based study based on the first 5,000 cases of the GHS. The 2.783 currently employed participants of the 5.000 baseline persons filled in the questionnaires concerning psychosocial workload.

Half of the participants of the GHS baseline completed the ERI questionnaire and the other half the COPSOQ questionnaire. As a consequence, we were not able to compare the association between ERI and COPSOQ on an individual level but only on an aggregate level. We have, however, shown that the two groups do not differ by age, gender, and rate participation. Further, we observed no difference in the participation rate between the participants.

Moreover, the present study has a cross-sectional design. Hence causal claims cannot be made and reversed causality is possible. It is conceivable, for instance, that poor satisfaction with life or having been burned out before influences the way working conditions are perceived [34, 35]. In the COPSOQ model however (and also in the way ERI and Demand Control Support are often used together with strain-outcomes), the health and work-related factors are conceptualized as being outcomes and predicted by the psychosocial factors at work (Figure 1) [9]. Currently, the GHS has completed its baseline and started the next phase assessing 5-year follow-up data. These data will give the opportunity to control for initial states and to prospectively analyse the psychosocial risk factors for the above mentioned health and work-related outcomes. Further, it is planned to prospectively analyse psychosocial risk factors for clinical outcomes such as cardiovascular events.

The ERI and the COPSOQ have somewhat different approaches to measuring psychosocial factors at work. Because of their different scales it is not straightforward to compare their relation to the various occupations. However, focusing on the COPSOQ short scales and comparing them with the ERI scales proved to be a fruitful way to solve this challenge.

In the regression analyses, we did not control for confounding variables. The focus of this study was rather on a comparative prediction of outcomes. However, future studies ought to include potential confounders in order to determine how much of the proportion of reported strain is indeed due to the psychosocial conditions at work.

Conclusion

In the present study, the split application of two instruments assessing psychosocial workload (ERI and COPSOQ) in a population-based study (GHS) yielded similar psychosocial workplace exposures for most occupational groups. Various occupational groups showed similar strain patterns on equivalent ERI and COPSOQ scales, thus indicating congruent validity. Some gender differences could be found regarding psychosocial workload. However regarding strain, we could find gender differences only for the outcome scale “Burnout”. In line with earlier findings [1], women reported higher burnout levels.

Indicating internal validity, both questionnaires appeared to be potent indicators of the presence of self-reported health and work-related outcomes. However, due to its broader spectrum of included psychosocial workplace factors and – potentially – due to its higher number and variety of analysed scales, the COPSOQ was somewhat more strongly associated with two of the four outcomes. The results underline the importance of the psychosocial environment at work for the individual’s health and well-being. The prospective analysis of the GHS will allow us to further determine the ability of the ERI and COPSOQ instrument to reflect relevant risk factors not only for self-reported health outcomes, but for diagnosed clinical disorders.

References

  1. Parent-Thirion A, Vermeylen G, van Houten G, Lyly-Yrjänäinen M, Biletta I, Cabrita J: Fifth European Working Conditions Survey. 2012, Luxembourg: Publications Office of the European Union. Eurofound

    Google Scholar 

  2. Karasek RA: Job demands, job decision latitude and mental strain: Implications for job redesign. Admin Sci Quart. 1979, 24: 285-308. 10.2307/2392498.

    Article  Google Scholar 

  3. Johnson JV: Control, collectivity and the psychosocial work environment. Job control and worker health. Edited by: Sauter SL, Hurrell JJJr, Cooper CL. 1989, Chichester: Jon Wiley & Sons

    Google Scholar 

  4. Siegrist J: Adverse health effects of high effort - low reward conditions at work. J Occupat Health Psychol. 1996, 1: 27-43.

    Article  CAS  Google Scholar 

  5. Siegrist J: A theory of occupational stress. Stress in the workplace. Edited by: Dunham J. 2001, London/Philadelphia: Whurr Publishers, 52-66.

    Google Scholar 

  6. Backé E-M, Seidler A, Latza U, Rossnagel K, Schumann B: The role of psychosocial stress at work for the development of cardiovascular diseases: a systematic review. Int Arch Occup Environ Health. 2012, 85: 67-79. 10.1007/s00420-011-0643-6.

    Article  PubMed  Google Scholar 

  7. Kivimäki M, Virtanen M, Elovainio M, Kouvonen A, Vaananen A, Vahtera J: Work stress in the etiology of coronary heart disease–a meta-analysis. Scand J Work Environ Health. 2006, 32: 431-442. 10.5271/sjweh.1049.

    Article  PubMed  Google Scholar 

  8. Kristensen TS, Borg V: AMI's spørgeskema om psykisk arbejdsmiljø. 2000, Copenhagen: National Institute of Occupational Health

    Google Scholar 

  9. Kristensen TS, Hannerz H, Høgh H, Borg V: The Copenhagen Psychosocial Questionnaire (COPSOQ) - a tool for the assessment and improvement of the psychosocial work environment. Scand J Work Environ Health. 2005, 31: 438-449. 10.5271/sjweh.948.

    Article  PubMed  Google Scholar 

  10. Setterlind S, Larsson G: The stress profile: A psychosocial approach to measuring stress. Stress Medicine. 1995, 11: 85-92. 10.1002/smi.2460110116.

    Article  Google Scholar 

  11. Karasek RA, Brisson C, Kawakami N, Houtman I, Bongers P, Amick B: The Job Content Questionnaire (JCQ): An Instrument for internationally comparative assessments of psychosocial job characteristics. J Occupat Health Psychol. 1998, 3: 322-55.

    Article  CAS  Google Scholar 

  12. Marmot MG, Smith GD, Stansfeld S, Patel C, North F, Head J, et al: Health inequalities among British civil servants. The Whitehall II Study. Lancet. 1991, 337: 1387-93. 10.1016/0140-6736(91)93068-K.

    Article  CAS  PubMed  Google Scholar 

  13. Schnabel RB, Wilde S, Wild PS, Munzel T, Blankenberg S: Atrial Fibrillation: Its Prevalence and Risk Factor Profile in the German General Population. Dtsch Arztebl Int. 2012, 109: 293-299.

    PubMed  PubMed Central  Google Scholar 

  14. Beutel ME, Wiltink J, Till Y, Wild PS, Munzel T, Ojeda FM, et al: Type D personality as a cardiovascular risk marker in the general population: results from the Gutenberg health study. Psychother Psychosom. 2012, 81: 108-117. 10.1159/000331776.

    Article  CAS  PubMed  Google Scholar 

  15. Wild PS, Zeller T, Beutel M, Blettner M, Dugi KA, Lackner KJ, et al: Die Gutenberg Gesundheitsstudie. Bundesgesundheitsblatt. 2012, 55: 824-830. 10.1007/s00103-012-1502-7.

    Article  CAS  Google Scholar 

  16. Sinning C, Wild PS, Echevarria FM, Wilde S, Schnabel R, Lubos E, et al: Sex differences in early carotid atherosclerosis (from the community-based Gutenberg-Heart Study). Am J Cardiol. 2011, 107: 1841-1847. 10.1016/j.amjcard.2011.02.318.

    Article  PubMed  Google Scholar 

  17. Rugulies R, Norborg M, Sorensen TS, Knudsen LE, Burr H: Effort-reward imbalance at work and risk of sleep disturbances. Cross-sectional and prospective results from the Danish Work Environment Cohort Study. J Psychosom Res. 2009, 66: 75-83. 10.1016/j.jpsychores.2008.05.005.

    Article  PubMed  Google Scholar 

  18. Nuebling M, Stößel U, Hasselhorn H-M, Michaelis M, Hofmann F: Methoden zur Erfassung psychischer Belastungen - Erprobung eines Messinstrumentes (COPSOQ). 2005, Bremerhaven: Wirtschaftsverlag NW. Schriftenreihe der Bundesanstalt für Arbeitsschutz und Arbeitsmedizin, Fb 1058

    Google Scholar 

  19. Nuebling M, Stößel U, Hasselhorn H-M, Michaelis M, Hofmann F: Measuring psychological stress and strain at work: Evaluation of the COPSOQ Questionnaire in Germany. GMS Psycho-Social-Medicine. 2006, 3: ISSN 1860-5214

    Google Scholar 

  20. Bundesamt S: Personensystematik. Klassifizierung der Berufe - Systematisches und alphabetisches Verzeichnis der Berufsbenennungen. 1992, Stuttgart: Metzler-Poeschel

    Google Scholar 

  21. Jiang L, Guo F, Zhang J, Huang Z, Chen ZY: Relation between work-family spillover and sleep quality among middle-aged scientific staff. Chinese J Clin Psychol. 2010, 639-641.

    Google Scholar 

  22. Burr H, Albertsen K, Rugulies R, Hannerz H: Do dimensions from the Copenhagen Psychosocial Questionnaire predict vitality and mental health over and above the job strain and effort-reward imbalance models?. Scand J Public Health. 2010, 38: 59-68. 10.1177/1403494809353436.

    Article  PubMed  Google Scholar 

  23. Siegrist J, Starke D, Chandola T, Godin I, Marmot M, Niedhammer I, et al: The measurement of effort-reward imbalance at work: European comparisons. Soc Sci Med. 2004, 58: 1483-1499. 10.1016/S0277-9536(03)00351-4.

    Article  PubMed  Google Scholar 

  24. Braunstein-Bercovitz H, Frish-Burstein S, Benjamin BA: The role of personal resources in work-family conflict: Implications for young mothers' well-being. J Vocat Behav. 2012, 80: 317-325. 10.1016/j.jvb.2011.10.003.

    Article  Google Scholar 

  25. Magee CA, Stefanic N, Caputi P, Iverson DC: The association between job demands/control and health in employed parents: The mediating role of work-to-family interference and enhancement. J Occup Health Psychol. 2012, 17: 196-205.

    Article  PubMed  Google Scholar 

  26. Lau B: Effort-reward imbalance and overcommitment in employees in a Norwegian municipality: a cross sectional study. J Occup Med Toxic. 2008, 3:

    Google Scholar 

  27. van Vegchel N, de Jonge J, Bosma H, Schaufeli W: Reviewing the effort-reward imbalance model: drawing up the balance of 45 empirical studies. Soc Sci Med. 2005, 60: 1117-1131. 10.1016/j.socscimed.2004.06.043.

    Article  PubMed  Google Scholar 

  28. Hirschfeld RR, Feild HS: Work centrality and work alienation: distinct aspects of a general commitment to work. J Organ Behav. 2000, 21: 789-800. 10.1002/1099-1379(200011)21:7<789::AID-JOB59>3.0.CO;2-W.

    Article  Google Scholar 

  29. Zatonski WA, HEM project team: Epidemiological analysis of health situation development in Europe and its causes until 1990. Ann Agric Environ Med. 2011, 18: 194-202.

    PubMed  Google Scholar 

  30. Wilcock AA: An occupational perspective of health. Thorofare, NJ: SLACK, /20-

  31. Eklund M, Leufstadius C: Relationships between occupational factors and health and well-being in individuals with persistent mental illness living in the community. Can J Occup Ther. 2007, 303-313.

    Google Scholar 

  32. Hakansson C, Dahlin-lvanoff S, Sonn U: Achieving Balance in Everyday Life. J Occup Sci. 2006, 13: 74-82. 10.1080/14427591.2006.9686572.

    Article  Google Scholar 

  33. Hakansson C, Bjorkelund C, Eklund M: Associations between women's subjective perceptions of daily occupations and life satisfaction, and the role of perceived control. Aust Occup Ther J. 2011, 58: 397-404. 10.1111/j.1440-1630.2011.00976.x.

    Article  PubMed  Google Scholar 

  34. Qu H, Zhao X: Employees' work-family conflict moderating life and job satisfaction. J Bus Res. 2012, 65: 22-28. 10.1016/j.jbusres.2011.07.010.

    Article  Google Scholar 

  35. Millear PMR, Liossis PL: Longitudinal modelling of the influence of individual differences and the workplace on wellbeing and work engagement. Personality and individual differences: current directions. Edited by: Hicks RE. 2010, Brisbane, Queensland: Australian Academic Press

    Google Scholar 

Pre-publication history

Download references

Acknowledgements

The GHS is funded through the government of Rheinland-Pfalz (“Stiftung Rheinland Pfalz für Innovation”, contract number AZ 961-386261/733), the research programs “Wissen schafft Zukunft” and “Schwerpunkt Vaskuläre Prävention” of the Johannes Gutenberg-University of Mainz and its contract with Boehringer Ingelheim, PHILIPS Medical Systems including an unrestricted grant for the GHS. The Federal Institute of Occupational Safety and Health (BAuA) funded the collection of occupational variables (including the psychosocial questionnaires) within the GHS and this analysis. Philipp S. Wild is funded by the Federal Ministry of Education and Research (BMBF 01EO1003).

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Matthias Nuebling.

Additional information

Competing interests

The authors declare that they have no competing interests.

Authors’ contributions

MN was involved in composition of the questionnaires, the organisation of data collection and analysis and the writing of the manuscript. AS was involved in the study organization, instruction of interviewers, composition of the questionnaires, the data analysis, and the writing of the manuscript. SGN was involved in interpreting the data and the writing of the manuscript. UL, SJ, and FL were involved in the instruction of interviewers, classification of occupation, the data analysis, and the writing of the manuscript. MW, JH and IZ were involved in the data analysis and the writing of the manuscript. PW was involved in organisation of the study, the data collection and the writing of the manuscript. SL was involved in the study organisation, the data analysis and the writing of the manuscript. All authors approved the final manuscript.

Matthias Nuebling, Andreas Seidler contributed equally to this work.

Authors’ original submitted files for images

Below are the links to the authors’ original submitted files for images.

Authors’ original file for figure 1

Rights and permissions

This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Reprints and permissions

About this article

Cite this article

Nuebling, M., Seidler, A., Garthus-Niegel, S. et al. The Gutenberg Health Study: measuring psychosocial factors at work and predicting health and work-related outcomes with the ERI and the COPSOQ questionnaire. BMC Public Health 13, 538 (2013). https://doi.org/10.1186/1471-2458-13-538

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/1471-2458-13-538

Keywords