Web Info

Rawd Alach

Summary of Article: “Work Stress, Mental Health and Antidepressant Medication Finding from the Health 200 Study”

During recent years, there has been increasing concern about the effects of work stress on mental health.

There is speculation that work with high job demands and low job control, leads to depression and anxiety disorders.

Several theories have come to light attempting to detail the effects of work stress on individuals.

One such theory is the Job Strain Model, also called the Demand-Control Model. This theory endorses the idea that high job demand and low job control, i.e., high job strain, leads to mental health problems.

In 2000, an epidemiologic study sought to test the Demand-Control Model using 8,000 participants consisting of both men and women.

The study was to determine whether or not degree of job strain had an effect mental health.

Two methods were used to determine such conclusions. One was a standardized psychiatric interview of the participant. Another was the meansure of the use antidepressant medication determine possible existence of a depressive and or anxiety disorder.

During the study participants were asked to participate in interviews, questionnaires, and health examinations. Medical health was assessed based on the Composite International Diagnostic Interview (CIDI).

This interview is a common tool for measure of mental disorders. The questionnaires given the participants implemented the use of the Demand-Control Model, utilizing a scale of job demands and job control to measure the intensity of work stress.

Individuals would use a five point scale to measure their perceived intensity of job demand and job control. Using statistical analysis methods, the data was collected and then analyzed.

Since results showed a significant relation between sex and job control associated with mental disorder, men and women’s results were separated to show degree of correlation. Also, during analysis of the study data was adjusted for leading factors such as age, martial status, income, etc.

Such factors were highly influential in the outcome measures. Household income was eliminated from the analysis because it showed no related to antidepressant use. While the results of antidepressant use were taken into consideration during the study, they were controlled for those who had a history of lifetime mental disorders.

Results of the study showed that women reported lifetime mental disorders much more often then men. They tended to be placed in active jobs with high job strain unlike men who had high job control and low job strain. In men, occupational grade highly impacted job strain which lead to mental disorder. Men placed in jobs that were of higher-grade and non-manually oriented showed greater job strain and depressive or anxiety disorders than men in other types of jobs.

In regard to antidepressant medication, women medicated two times as much as men. Working married women tended to implement the use of antidepressants more often then their non-married counterparts. However, there was no solid relation found between women’s use of antidepressants with work characteristics. In men, antidepressant use was highly related to job strain.

Therefore, the use of antidepressants only in men was due to high job strain and pointed to mental disorders.

Although there was a difference between results for men and women, overall the results of the study showed that in both sexes high job demand and low job control resulted in a prevalence of mental disorders. Hence, the study has further enforced the Demand-Control Model, showing that high job strain leads to mental disorders in both men and women.

Article:

Research report

Work stress, mental health and antidepressant medication findings from the Health 2000 Study

Marianna Virtanena, Teija Honkonena, Mika Kivimäkia, Kirsi Aholaa, Jussi Vahteraa, Arpo Aromaac and Jouko Lönnqvistd

aFinnish Institute of Occupational Health, Topeliuksenkatu 41 aA, FIN-00250 Helsinki, Finland
bUniversity of Helsinki, Department of Psychology, P. O. Box 9, FI-00014 University of Helsinki, Finland

cNational Public Health Institute, Department of Health and Functional Capacity, Mannerheimintie 166, FI-00300, Helsinki, Finland

dNational Public Health Institute, Department of Mental Health and Alcohol Research, Mannerheimintie 166, FI-00300 Helsinki, Finland, and University of Helsinki, Department of Psychiatry, P.O. Box 320, FI-00029 Helsinki University Central Hospital (HUCH), Helsinki, Finland

Received 3 March 2006; revised 30 May 2006; accepted 31 May 2006. Available online 19 December 2006.

Abstract

Background

Population-based studies on the association between work stress and mental disorders are scarce, and it is not known whether work stress predicts mental disorders requiring treatment.

Aims

To examine the associations of work stress with DSM-IV mental disorders and subsequent antidepressant medication.

Methods

3366 participants from a representative sample of the Finnish working population responded to a survey (The Health 2000 Study). 12-month prevalence of depressive or anxiety disorders was examined with the Composite International Diagnostic Interview. Data on antidepressant prescriptions with a 3-year follow-up period were collected from a nationwide register of Social Insurance Institution.

Results

In men and women, high job demands, low job control and high job strain were associated with 12-month prevalence of depressive or anxiety disorders. After adjustment for lifetime and baseline mental disorders, men with high job demands and high job strain had increased risk of future antidepressant medication.

Conclusions

Work stress is associated with mental disorders among both sexes and among men it is a risk factor for mental disorders treated with antidepressant medication.

Keywords: Antidepressants; CIDI; Demand–Control model; Mental disorders; Work stress

Article Outline

1. Introduction
2. Materials and methods
2.1. Study sample
2.2. Measurements
2.3. Statistical analyses
3. Results
4. Discussion
5. Conclusion
References

1. Introduction

Studies from several countries have reported an increase in work stress (Schaufeli and Enzmann, 1998), and that mental disorders, especially depression, are growing reasons for work disability and early retirement (Kruijshaar et al., 2003, Stewart et al., 2003 and Gould and Nyman, 2004). Along with such evidence, concern is growing about the adverse effects that work stress may have on mental health. The Job Strain Model (Karasek, 1979, Karasek and Theorell, 1990, Van der Doef and Maes, 1999, De Lange et al., 2003 and Theorell, 2003), also known as the Demand–Control Model, has been one of the most influential theories in research on psychosocial work characteristics and health.

The model posits that a combination of high job demands and low job control, referred to as job strain, is a risk factor for health problems. Although most previous research has focused on the relationship between work strain and cardiovascular diseases, there is some evidence that high job strain, high demands and low control are also associated with mental health problems (Karasek, 1979, Bromet et al., 1988, Karasek and Theorell, 1990, Stansfeld et al., 1997, Stansfeld et al., 1999, Cropley et al., 1999, Van der Doef and Maes, 1999, Mausner-Dorsch and Eaton, 2000, Paterniti et al., 2002, Stansfeld, 2002, De Lange et al., 2003, Theorell, 2003 and Wang, 2005) and self-reported use of psychopharmacological medication (Karasek, 1979 and Moisan et al., 1999).

However, most research has been cross-sectional and few published longitudinal studies have used symptoms or self-certified psychiatric sickness absences as an outcome measure. Some of these studies have reported null findings (Van der Doef and Maes, 1999 and Ylipaavalniemi et al., 2005). Two used a clinical or standardised interview and reported that high job demands and high work stress were associated with new affective disorder (Bromet et al., 1988), and new major depressive episode (Wang, 2005). However, these studies involved a selected vocational group of men (Bromet et al., 1988), or did not report results based on the original classification of the Job Strain Model (Bromet et al., 1988 and Wang, 2005). A major limitation in earlier research is related to common method variance, i.e. both work stress and mental health have been based on subjective assessments.

Register data on antidepressant medication offers an opportunity to avoid such a bias since prescriptions are based on physicians’ diagnoses. Antidepressant medication can also be considered as an indicator of mental disorders requiring pharmacological treatment. According to clinical practice guidelines on managing depression, in depressive disorders with significant disability, treatment with antidepressant medication is recommended (Finnish Psychiatric Association, 2004 and National Institute for Clinical Excellence, 2004). Prospective evidence on the relationship between work stress and antidepressant use would therefore be important in terms of prevention of disabling mental disorders. Thus far, it is not known whether work stress predicts antidepressant use.

In sum, limitations in previous studies include lack of structured psychiatric interviews in the assessment of mental disorders and bias due to common-method variance, recall problems and non-generalisability to the general population. In this population-based study, we examined whether high job demands, low job control and high job strain are associated with prevalence of depressive or anxiety disorders assessed by a standardised psychiatric interview. We also examined whether these stress indicators predict clinically significant depressive or anxiety disorders as measured by register-based antidepressant medication.

2. Materials and methods

2.1. Study sample

A multidisciplinary epidemiologic health survey, the Health 2000 Study, was carried out in the years 2000–2001 in Finland.

The two-stage stratified cluster sample was representative of the Finnish population (0.24% sample) and included 8028 subjects aged 30 years or over (Aromaa and Koskinen, 2004). Stratification and sampling were conducted as follows: The strata were the five university hospital districts, each serving about 1 million inhabitants and differing in several features related to geography, economic structure, health services and the socio-demographic characteristics of the population. First, the 15 largest cities were included with a probability of one.

Next, within each of the five districts all 65 other areas were sampled applying the probability proportional to population size (PPS) method. Finally, from each of these 80 areas a random sample of individuals was drawn from the National Population Register.

The detailed methodology of the project has been published (Aromaa and Koskinen, 2004).

The data collection phase started in August 2000 and was completed in March 2001, during which a total of 7419 subjects (93% of the 7977 subjects alive on the day the first phase of the survey began) attended at least one phase of the study. The subjects were interviewed at home, where they were given a questionnaire to be returned at the clinical health examination.

During the interview the respondents received an information leaflet and their written informed consent was obtained. Approval of the Ethics Committee of Epidemiology and Public Health in the Hospital District of Helsinki and Uusimaa was obtained for this study.

Of the total sample, 5871 persons were of working age (30 to 64 years). Of this base population, 5152 persons were interviewed (87.8%), 4935 persons returned the questionnaire (84.1%), and 4886 (83.2%) participated in the health examination, including the structured mental health interview (CIDI). The final cohort of the present study comprised the 3366 participants (1662 men, 1704 women) who were employed at the time of the interview.

2.2. Measurements

We used self-assessment scales to measure the components of the Demand–Control Model of job strain (Karasek, 1979, Karasek and Theorell, 1990 and Theorell, 2003). The scale of job demands comprises five items (α = 0.79) (e.g. “My job requires working very fast”).

The scale of job control comprises nine items (α = 0.85) (e.g. “My job allows me to make a lot of decisions on my own”; “My job requires a high level of skills”).

Responses are given on a 5-point scale ranging from 1 (strongly disagree) to 5 (strongly agree).

In job-strain literature, many methods for testing interactions between job demand and job control are widely used and accepted (Landsbergis et al., 1994, Schnall et al., 1994 and Hintsanen et al., 2005). Mean scores of job demands and job control were standardised and treated as continuous variables.

A continuous quotient term of job strain was calculated by dividing job demands score by job control score and then standardised (Landsbergis et al., 1994). An alternative job strain formulation, a quadrant term, was calculated for comparison. As presented in the original work of Karasek (1979), we dichotomised the job demand and job control scales at their median and formulated the following four subgroups: low-strain (low demands and high control), active (high demands and high control), passive (low demands and low control), and high-strain (high demands and low control).

Lifetime mental disorders were assessed by a single-item question asking whether a doctor had ever confirmed a diagnosis of mental disorder (yes/no).

Mental health status at baseline was based on a computerized version of the WHO Composite International Diagnostic Interview (M-CIDI) as a part of a comprehensive health examination. The standardized CIDI interview has been shown to be a valid assessment measure of common mental non-psychotic disorders (Jordanova et al., 2004).

The program uses operationalized criteria for DSM-IV diagnoses and allows the estimation of DSM-IV diagnoses for major mental disorders.

The 21 interviewers were mostly non-psychiatric health care professionals. They were trained for the CIDI interview for 3–4 days by psychiatrists and physicians who had been trained by a WHO authorised trainer. Mental disorders were assessed using DSM-IV definitions and criteria.

The participant was identified as a case if he/she fulfilled the criteria for a depressive or anxiety disorder. Depressive disorder meant a diagnosis during the previous 12 months of major depressive disorder or dysthymia, while anxiety disorders included diagnoses of panic disorder (with or without agoraphobia), generalized anxiety disorder, social phobia, phobia NOS and agoraphobia (without panic disorder).

Data on antidepressant medication were obtained from the National Prescription Register, managed by the Social Insurance Institution of Finland.

The national sickness insurance scheme covers the entire population and reimburses the costs of prescribed medication for virtually all patients. Information on drug prescriptions was linked to the data by means of each participant’s personal identification number (a unique number that all Finns receive at birth and that is used for all contacts with the social welfare and health care systems).

The prescription register of the Social Insurance Institution contains outpatient prescription data based on the WHO’s Anatomical Therapeutic Chemical (ATC) classification code (WHO Collaborating Centre for Drug Statistics Methodology, 2004). We extracted all the prescriptions coded as N06A, which is the ATC code for antidepressants, from January 1st, 2001 to December 31, 2003.

Information on demographic factors was collected in the home interview: sex, age, marital status, occupational grade, type of business, household income (range from 1 = less than 2500 FIM (420 €) per month to 13 = more than 50 000 FIM (8409 €) per month, and employment status (employed, unemployed or economically inactive).

Marital status was divided into two groups: those who were married or cohabiting and those who were divorced, widowed or single. Occupational grade was formed based on occupation and type of business: upper grade non-manual, lower grade non-manual, manual workers and self-employed (Statistics Finland, 1999).

2.3. Statistical analyses

Binary logistic regression models were used to calculate adjusted odds ratios and their 95% confidence intervals for having 12-month mental disorders, and at least one antidepressant prescription during the 3-year follow-up. We found a significant interaction between sex and job control associated with mental disorder (p = 0.024). We therefore present the results separately for men and women.

Analyses were adjusted for potential confounding factors: age, marital status, occupational grade and household income, which all showed some association with the outcome measures. Because household income was not related to antidepressant use it was excluded from that part of the analyses. To evaluate the association between work stress and antidepressant medication, lifetime and baseline mental disorders were additionally controlled for.

In addition to the main effects, the cross-product terms (Cohen and Cohen, 1983) of work stress indicator with each demographic factor were entered in the models to assess whether the association of work stress with mental disorders or antidepressant use was modified by any demographic factor.

Test of curvilinearity was performed by entering the cross-product term of the work stress variable itself after the main effect, e.g. ‘job demands × job demands’. In statistical analyses, the data were weighted to take into account the sampling design and to reduce the bias due to non-response. We used the SAS/Sudaan 9.0.1 statistical program package to perform the analyses.

3. Results

Table 1 presents the characteristics of the study participants for men and women. Men were younger, had higher household income, were more often in manual occupations or self-employed, and were more likely to be married or co-habiting than women. Men also reported lifetime mental disorders less often and had a lower prevalence of 12-month mental disorders than women. Men had higher job control and were more often in low-strain and active jobs whereas women were more often in passive and high-strain jobs.

a P-value for difference between men and women in ANOVA and χ2 tests.
b Self-reported information on doctor-diagnosed mental disorder.
c Diagnosis based on the CIDI interview.
d Job demands/Job control.

The findings regarding the association of work characteristics with 12-month mental disorders appear in Table 2.

In both sexes, high job demands (Model I) and high job strain (Model IV) were related to a higher prevalence and high job control (Model II) was associated with lower prevalence of 12-month mental disorder.

When job demands and job control were entered simultaneously in the model (Model III), the associations of both variables with mental disorders remained statistically significant. We found that among men and women, the association between job demands and mental disorder was curvilinear (p = 0.04 and 0.01, respectively).

This association is adjusted in Models I and III. We split the job demands score into tertiles and found that among men, the association between job demands and mental disorder was U-shaped (odds ratio for intermediate job demands = 0.58, 95% CI = 0.35–0.99 and for high job demands = 1.44, 95% CI = 0.89–2.33 when compared with low job demands; data not shown). Among women, only high job demands but not intermediate job demands was associated with mental disorder (OR = 1.74, 95% CI = 1.24–2.44 and OR = 1.02, 95% CI = 0.70–1.48, respectively).

Among men, we also found a significant interaction effect between occupational grade and job strain associated with mental disorder (p = 0.04). A separate analysis revealed that among higher-grade non-manual workers, the association between job strain and mental disorder was 4.10 (95% CI = 1.78–9.43).

Corresponding odds ratios among lower grade non-manual, manual and self-employed men were 1.31 (95% CI = 0.76–2.26), 1.60 (95% CI = 1.26–2.04), and 0.77 (95% CI = 0.39–1.54).

a Adjusted for age, marital status, household income and occupational grade.
b OR refers to change in probability of depressive or anxiety disorder per standard deviation increase in job demands, job control and job strain.
c Indicates curvilinear effect of job demands.

During the follow-up period, 96 (6%) men and 199 (12%) women had antidepressant medication. Findings regarding the association between work stress indicators and antidepressant medication are shown in Table 3.

No curvilinearity was found with regard to the association between work characteristics and antidepressant use. High job demands were associated with a greater risk of receiving antidepressant medication among men (Model I).

Job control was not significantly related to antidepressant use (Model II). After adjustment for job control, odds ratio for high job demands was 1.30 (95% CI = 1.04–1.62) among men (Model III). Also in men, high job strain was related to an odds ratio of 1.30 (95% CI = 1.08–1.57) with antidepressant use (Model IV).

Among women, work characteristics were not significantly associated with antidepressant use. However, we found a significant interaction between marital status and job demands predicting antidepressant use among women (p = 0.03).

In a stratified analysis, job demands were associated with an odds ratio of 1.19 (95% CI = 0.98–1.45) for antidepressant use among married women. Among non-married women, the corresponding odds ratio was 0.79 (95% CI = 0.59–1.07).

a Adjusted for age, marital status, occupational grade, lifetime mental disorder and baseline DSM-IV depressive or anxiety disorder.

b OR refers to change in probability of depressive or anxiety disorder per standard deviation increase in job demands, job control and job strain.

As shown in Table 4, test of interaction between job demands and job control associated with DSM-IV depressive or anxiety disorders and antidepressant use resulted in no statistically significant effects. We also tested the interaction using job demands and job control as categorical variables (tertiles) and found no statistically significant interactions.

a Adjusted for age, marital status, household income, occupational grade and the main effects of work characteristics.

b Adjusted for age, marital status, occupational grade, baseline depressive or anxiety disorders and lifetime mental disorders, and the main effects of work characteristics.
c Indicates curvilinear effect of job demands.

In a four-category model of job strain, high job strain was related to odds ratio of 2.54 for mental disorder compared with low job strain among men and 1.68 among women (Table 5). High job strain was associated with higher probability of antidepressant use among men (OR = 1.95) but not among women (OR = 1.16). A subgroup analysis of 1442 healthy men (with no lifetime or 12-month mental disorder) revealed a similar although statistically non-significant association between job strain and antidepressant medication (OR = 1.20, 95% CI = 0.90–1.58 for linear job strain and OR = 2.09, 95% CI = 0.80–5.50 for quadrant job strain, results not shown in the table).

a Adjusted for age, marital status, occupational grade and household income.

b Adjusted for age, marital status, occupational grade, and DSM-IV depressive or anxiety disorders and lifetime mental disorder at baseline.

4. Discussion

This population-based study of 3366 men and women showed that high work stress, as indicated in the Job Strain Model, was associated with DSM-IV diagnoses of depressive or anxiety disorders.

In men, high job demands and high job strain were also associated with increased risk of antidepressant medication at follow-up. Because we adjusted the analyses for lifetime mental disorders and 12-month depressive and anxiety disorders, we were largely able to control for their possible confounding effects on the perception of work stress. Since DSM-IV diagnoses also included dysthymia, we could substantially control for confounding by participant’s milder depression at baseline.

Job strain was associated with the 12-month prevalence of depressive or anxiety disorders in men and women.

This accords with earlier findings of an association between work stress and mental health problems (Karasek, 1979, Bromet et al., 1988, Karasek and Theorell, 1990, Stansfeld et al., 1997, Cropley et al., 1999, Stansfeld et al., 1999, Van der Doef and Maes, 1999, Mausner-Dorsch and Eaton, 2000, Paterniti et al., 2002, Stansfeld, 2002, De Lange et al., 2003, Theorell, 2003 and Wang, 2005).

However, only in two previous studies (Cropley et al., 1999 and Mausner-Dorsch and Eaton, 2000) was the Job Strain Model tested using a standardised psychiatric interview to define mental disorder.

In the former, the outcome measure was the prevalence of neurotic disorder, and in the latter, job strain was calculated as high demands and low control versus the other three combinations.

In the present study, we used the continuous job strain score as well as all items of the quadrant term (low-strain, active, passive and high-strain) as defined in the original model of Karasek (1979).

Our findings suggest that job control, i.e. having influence over one’s job is strongly associated with DSM-IV diagnosis of depressive or anxiety disorders, particularly among men. Among men and women, the association between job demands and these disorders was curvilinear.

Among men, high and low job demands were associated with higher probability of mental disorders than intermediate job demands. Because this part of the study was cross-sectional it is possible that this finding reflects reversed causality, i.e. men with mental disorders have changed to less demanding jobs. However, it is also likely that an adequate level of job demands is a prerequisite for well-being of men. Among women, only very high job demands were related to mental disorders, indicating a threshold effect.

In our study, the association between work stress and depressive or anxiety disorders was stronger for men than women and the association with future antidepressant use was evident only among men. In some earlier longitudinal studies, high job demands have predicted subsequent psychological symptoms in both sexes (Stansfeld et al., 1997, Stansfeld et al., 1999 and Paterniti et al., 2002).

In the Whitehall II study of white-collar civil servants, high job demands and low job control were also associated with sickness absence, especially among men (North et al., 1996). A study on female-dominated hospital personnel produced no support for the job strain model as a predictor of depression (Ylipaavalniemi et al., 2005).

Considering the present findings and earlier evidence, the effects of work strain on health may be weaker in women than men (Kessler, 2003 and Theorell, 2003).

One possible explanation for this is that the etiology of mental disorders may be related to different psychosocial factors in men and women. Work may be a dominant factor for men, whereas for women the psychosocial etiological factors may be distributed across several spheres, including domestic factors and social relations.

In fact, we found that high job demands were associated with an odds ratio of 1.2 for future antidepressant use among married women whereas the corresponding odds ratio among non-married women was 0.8. This finding may be related to accumulation of total burden among married women (Denton et al., 2004).

In men, the association between job strain and depressive or anxiety disorders was stronger among higher-grade non-manual men (OR = 4.1) than among men in other occupational groups.

In manual occupations, work stress seemed not to be among the factors that strongly contribute to mental health.

One potential explanation for this is higher work and career orientation and thus higher emotional commitment to work among non-manual men. However, the reasons for socioeconomic as well as sex differences in the relationship between psychosocial work environment and health are poorly understood and require further investigation.

The association between work stress and future antidepressant use was evident in men after the adjustment for baseline mental health status.

Antidepressant medication can be considered, first, as a proxy measure for clinically significant depressive or anxiety disorders. Onset of new mental disorder may have been more likely among men with high work stress. However, it is also possible that those men did not have a mental disorder but, on the contrary, their disturbing work stress symptoms may have been misinterpreted as depressive or anxiety disorder and treated with antidepressants (see Heath, 1999 and Kessler et al., 1999). Whichever is the case, the present evidence on the relationship between work stress and antidepressant use is important in terms of promotion of mental health at workplaces.

As in the majority of previous studies (Stansfeld et al., 1997, Cropley et al., 1999, Stansfeld et al., 1999, Van der Doef and Maes, 1999, Mausner-Dorsch and Eaton, 2000, Paterniti et al., 2002, De Lange et al., 2003 and Wang, 2005; discussion, see Warr, 1990), no evidence of multiplicative interaction between job demands and job control was found in this study, suggesting an additive rather than a synergistic effect. Also in his original paper Karasek (1979, p. 293) points out that “there is only moderate evidence for an interaction effect, understood as a departure from a linear additive model”.

The main limitation of this study was a cross-sectional analysis of work stress and DSM-IV mental disorders. In this design, the association between mental disorder and perceived work stress may actually reflect the association between a disorder and its symptoms.

The standardized CIDI interview is a valid measure of DSM-IV non-psychotic disorders (Jordanova et al., 2004). The validity of the measure of lifetime mental disorder used in our study is, however, unknown.

Corresponding with earlier research (Young et al., 1990), our results showed that women were more likely than men to have a history of lifetime mental disorders.

The reported lifetime mental disorder 9.7%, however, was lower than that reported by e.g. in the National Comorbidity Study (19% for affective disorders, 25% for anxiety disorders, Kessler et al., 1994). Our measurement of past doctor-diagnosed mental disorders is likely to exclude individuals who had not sought help for their mental health problems from a physician.

Data on antidepressant prescriptions covered a 3-year follow-up period and the adjustments were made for baseline DSM-IV mental disorders and mental health history. Register data on prescriptions are based on a visit to the physician and cover virtually all prescriptions for the cohort. Uneven treatment practice between physicians may affect the prescriptions but such variability is likely to be random in relation to work stress. Using antidepressant medication as an indicator of clinically significant depressive and anxiety disorders is likely to have resulted in an underestimate rather than overestimate of these disorders, because this measure did not cover persons with unrecognized or under-treated disorders or those treated with non-pharmacological methods.

5. Conclusion

Psychosocial work stress is associated with DSM-IV depressive or anxiety disorders among both sexes and among men it is a risk factor for mental disorders treated with antidepressant medication.

As mental disorders account for a considerable proportion of the disease burden and are a major cause of work disability, psychosocial factors at work should be regarded as a target worthy of priority in the promotion of mental health at workplaces.

Journal of Affective Disorders

Volume 98, Issue 3, March 2007, Pages 189-197

Article Title:Work stress, mental health and antidepressant medication findings from the Health 2000 Study Source:JOURNAL OF AFFECTIVE DISORDERS (0165-0327); Volume: 98; Issue: 3; Date: 2007