Psychosocial risk factors at work and mental ill health

Background

At the level of the EU, the Framework Directive on Safety and Health at Work (89/391/EEC) covers all kinds of risks, including psychosocial ones, and lays down the main principles of prevention and protection against them. In line with the Framework Directive, employers must assess the risks to which all workers are exposed and put in place preventive and protective measures. It is essential to apply and enforce existing provisions properly and effectively. 

A peer review hosted by Sweden in 2019^[1] concluded that EU countries have different legislative and enforcement approaches to address psychosocial risks at work: some have detailed binding legislation, others general legislation.

The EU’s OSH strategic framework 2021-2027^[2] announced an OSH summit in 2023 to take stock of progress since the adoption of the framework. It will focus, among others, also on psychosocial risks and mental health at the workplace.

President Von der Leyen’s initiative on a comprehensive approach to mental health initiative, announced during her address at the State of the Union 2022^[3], and included in the Commission Work Programme 2023^[4], will cover psychosocial risks at work among other policy areas and the results of the OSH summit will be an integral part of it.

This paper is produced for the OSH summit on the topic of psychosocial risks and mental health at work. It contains a summary of research-based knowledge regarding the subject.

Research development in the field and the need for communicating new knowledge to social partners

Since there is a lag in all research fields between new findings on the one hand and acceptance and applications of such findings in society on the other and since that lag is different in different EU countries, the social partners have an important task in coordinating the spread of such new research findings. For several decades, it has been stated by administrators, managers and trade unions that there is insufficient knowledge about stress and psychosocial factors and that even in areas where it can be found, principles for prevention are not established. That attitude is becoming less prevalent with the development of research on psychosocial risk factors at work. Such research has been expanded and sharpened during the six recent decades. One of the most important reasons for this is a dramatic improvement of research techniques in endocrinology (catecholamines adrenalin and noradrenalin, adrenocortical hormones like the classical stress hormone cortisol and the “protective” stress hormone DHEA-s), immunology (substances that stimulate inflammation and coagulation) and heart activity (continuous recordings of electrocardiogram). Thus, it is important to communicate to all the Occupational Safety and Health (OSH) stakeholders that there is growing knowledge about psychosocial work risk factors and that techniques for the reduction of psychosocial risk factors are developing rapidly.

Like other occupational safety and health topics, there are several possibilities to reduce psychosocial risk factors at work in the EU, such as promoting the better application of the law, financial incentives, and educational input to social partners. A necessary requirement for successful work in this area is to engage managers. Tools for the assessment of the company´s level of engagement in psychosocial risk protection – psychosocial safety climate (PSC) - have been developed and are used in many countries. One of the first such questionnaire tools was developed by Bailey, Dollard and Richards^[5]. In most EU Member States, occupational health care units are engaged in the exploration of psychosocial working conditions. Such explorations may include standardized questionnaires, as well as questionnaires tailored to the specific company, interviews with employees and managers and group interviews. The advantages and disadvantages of different techniques vary with branch, company, country, etc. It is often wise to combine different techniques. Since explorations raise expectations in employees, they should always be followed up. When problems have been identified, the company’s management is obliged to intervene.

The task formulated in the present project for the European Commission and all the relevant stakeholders is primarily the promotion of and/or the reduction of psychosocial risk factors at work. In the complex web of factors through which job conditions can influence employees’ health, there are also health-promoting factors which may counteract the adverse effects of psychosocial risk factors. Examples of such factors are listening and constructively supportive management, and organization of health-promoting activities for employees (resources for physical training at work and cultural activities such as joint attendance at concerts and theatre performances). The latter health-promoting factors have been shown to be beneficial to employees’ health[8]. The management´s involvement in the PSC is central to the work with psychosocial risk factors and can be assessed by means of an assessment tool (for instance Bailey, Dollard and Richards^[5]). It has been shown in Swedish longitudinal studies that manager attitudes to employees are hardening during societal crisis periods with increasing unemployment^[6] and that interest in organizing cultural activities associated with work decreases. According to results in Swedish studies, during such periods, managers also tend to be less willing to listen to their own employees^[7]. It is expected that the reduction of psychosocial risk factors is facilitated by a constructively communicating leadership and by discussions stimulated by cultural activities in the workplace.

This presentation is focused on the reduction of harmful psychosocial risk factors at work in relation to mental health. The concept of mental health should be perceived broadly, however. There is accumulating evidence showing that working conditions that increase the risk of developing depression, anxiety and related mental disorders are also associated with somatic ones such as cardiovascular disease, stroke, diabetes, and musculoskeletal disorders. Modern science has shown that there are common biological pathways, such as immunological and endocrinological ones, to mental and somatic disorders. In addition, some adverse health behaviors such as cigarette smoking and lack of physical activity are aggravated by poor working conditions. When the concept of mental health is used in the present document, we are referring to its broad definition which includes psychological as well as somatic conditions with broad implications for health.

It should also be pointed out that psychosocial factors interact with physical, chemical, or other risk factors. Accordingly, it has been shown, for instance, that exposure to noise (at home or at work) amplifies the effect on the mental health of a poor psychosocial work situation (see below).

Finally, in biological terms “negative” stress factors^[8], such as psychosocial risk factors at work, are operating in a complex interplay with positive factors. In Figure 1, psychosocial risk factors at work should be considered stressors – factors that induce stress reactions. However, stressors to the left in the figure could always be balanced by anti-stressors. The variation from bottom black to top white illustrates that stressors could turn into anti-stressors and vice versa. A working condition that was initially regarded as a stressor could be transformed into a positive protective factor. A psychosocial intervention which was planned as a beneficial process could unexpectedly turn into a bad stressor. This dynamic view is also mirrored in the other parts of the diagram. Stressors and anti-stressors induce either catabolism (corresponding to the release of energy) or anabolism (corresponding to regeneration and build-up of resources) which is displayed to the right. In the middle of the diagram, coping could be either destructive (black) or constructive (white). Our coping is constantly changing throughout life, and it is influenced by the experiences we have and by our genetic makeup. Thus, our coping is to some extent influenced positively or negatively by the experiences we have at work. In addition, a rapidly developing scientific field, epigenetics, has shown that the activity levels of the genes are to some extent regulated by conditions in the environment, making us more or less sensitive to stressors. It has been shown that genes can be methylated and de-methylated, for instance – which is part of the chemical background. The conclusion is that stressors act in a dynamic and complex system and that of course, positive conditions can counterbalance the effects of bad ones.

Physiologically we mostly define stress as energy mobilization which is under normal conditions a healthy reaction which we need when there is a challenge or threat. It only becomes harmful when it lasts for long periods and is not interfoliated by periods of regeneration. We are all biologically constructed for a life with variations between effort and rest.

Educational aspects

In Nordic countries, psychosocial factors at work have been in focus for a long time. For instance, in Sweden, discussions between trade unions and employer organizations resulted in laws that regulate working hours, shift work, the right to influence through trade unions and several other factors (1977). As a result of recent discussions regarding a growing problem with workplace bullying and harassment, there is also a special national directive on the handling and prevention of bullying at work (2015). There are similar laws in other Nordic countries. In comparison to many other comparable countries, the Nordic countries have also allocated substantial financial resources to work environment research, including the psychosocial environment. The situation varies across Member States. In Western Europe (for instance Germany, France, the Netherlands, and Belgium) there are also strong traditions in psychosocial work environment research but the discussions among social partners differ substantially between EU countries.

Financial arguments of importance for management

There are many financial arguments for the management to engage in these psychosocial discussions. High-level managers may not always be physically present during such discussions at the workplace. There are several reasons for this, apart from lack of time. One is that managers are often not aware of the many layers of financial losses that may arise when the psychosocial environment is malfunctioning in a company. Thus, if such risk factors can be reduced, substantial financial gains could be achieved, as shown in Table 1.

For some of these costs, in particular sick leave, calculations have been made which show that substantial savings could be made both for companies and society at large if sick leave due to psychosocial risk factors at work could be reduced^{[9],[10],[11]}.

The list in Table 1 should of course not only include financial costs but also suffering. However, if employers become aware of all the possible sources of financial losses due to a poor psychosocial environment, this may contribute to active participation in improving it. We are also aware that there are many employers and managers who rate the quality-of-life factors at work as more important in the competition than financial arguments.

Lack of knowledge of the many possible reasons for financial losses due to a poor psychosocial environment is one of the reasons employers avoid active participation in work environment interventions. Another frequently informally reported reason is that managers may be afraid of becoming openly criticized in group meetings with staff. Employees may also fear such situations, when several parts are involved in a conflict, open confrontations may be feared by all of them. Avoidance may be the result. Techniques for solving this kind of problem are available. One such technique has been discussed by the Swedish researcher Casten von Otter. His argument is that examinations of and interventions on psychosocial risk factors are mostly based on group interviews. If the manager of the work unit participates in the group discussion, all talking in the group will be inhibited. If the process goes on in a large workplace the groups to be interviewed could be constituted by employees from different units and the manager should belong to a group with no representation in the room. Although this may not be possible in many worksites the principle could be useful in many situations. Above all, good psychological preparation of all participants is important for successful group meetings about psychosocial risks.

Scientifically established psychosocial risk factors and their impact

Research on health-related psychosocial risk factors at work started on a large scale in the late 1970s. The area has gone through many methodological and theoretical discussions. The first model for studies, Person-Environment Fit (PEF) model was introduced by researchers at the Institute for Social Research in Ann Arbor. Building upon this work, Karasek introduced the demand control model in 1976. This has been an important theoretical model since it generated a large number of epidemiological and biological mechanistic studies. From the late 1970s, researchers went from case-control studies to prospective studies controlling for such accepted risk factors as excessive smoking, high “bad” (LDL) cholesterol and high blood pressure. The psychosocial job factors in the model are high psychological demands, low decision latitude (control) and poor social support. The main risk factors that have been identified are job strain (high psychological demands and at the same time low decision latitude) and job iso-strain (strain and at the same time poor support).

The assessment of job strain and job iso-strain has been thoroughly discussed. The most widely accepted assessment tool is the JCQ (job content questionnaire) which measures demands, decision latitude, social support, and physical load. A new extended version measuring collective functioning, JCQ 2.0, is underway from Karasek´s group. A shorter version, DCQ is also widely used. Both these instruments have adequate psychometric properties although JCQ may provide more precise assessments. There has been a debate regarding the job strain combination – is the combination of demands and control crucial? The conclusion seems to be that in some conditions, decision latitude, and in others, psychological demand is the most important component. In most studies, there is an additive effect, the effect of high demand adds to the effect of low control. A similar conclusion can be drawn for the interaction between job strain and poor support. Thus, in practical work, it is recommended that all three dimensions are used together.

Other research discussions surrounding the demand control support model are of relevance to all psychosocial risk factors and could be regarded as a strategic model.

Objective versus subjective

It was argued early during the development of measures of psychosocial risk factors that self-reported data based upon self-administered questionnaires may be influenced by genetic and other variables not related to the work environment. The demand-control-support questionnaires are constructed for the exploration of “how work is”, not whether the conditions are perceived as good or bad. This gives us some control over the work environment focus in the assessments but still, the description could be flavored by the person´s way of reacting to the environment. A study^[12] based on the Swedish Twin Registry showed that there is indeed a genetic contribution to the studied self-reported work environment scores, ranging from 18% to 30% for the three dimensions of decision authority and skill discretion (two parts of decision latitude) and psychological demands respectively. However, the substantial genetic contribution to the associations between those scores and depressive symptoms was not observed in the study. This means that it is meaningful to use them in occupational health care work as a basis for interventions and preventive work.

A way of circumventing the individual bias in establishing the environmental component in the psychosocial job environment assessments is to use so-called job environment matrices (JEM). International occupational codes usually define approximately 400 different occupations in a country. Based upon working population samples, the means calculated from all participants belonging to a given occupation are used as representing all subjects of a given age group and gender in a specified occupation – a so-called aggregated measure. This allows the researcher to use scores that are not influenced by individual bias unrelated to the work environment as such. Findings in such studies have shown somewhat divergent findings due to differences between the questionnaires used in the population studies but findings for job control (decision latitude) are similar to those based upon individual self-reports. Findings for demands have been less consistent than those for job control. This may not be surprising in view of the fact that occupational title explained a large part of the variance in decision latitude (45%) and a much smaller part of the variance (7%) in psychological demands, according to results in the American Employment Survey in 1968, 1974 and 1977. An additional way of assessing demands, control and support is to use experts who visit the work sites for standardized ratings based on what they see and hear. This should be an important possibility in practical job improvement work.

Another important component in discussions about established risk factors for diseases such as socioeconomic status, smoking, overweight and alcohol consumption, all of the potential importance particularly for cardiovascular disease. There is consensus that in prospective studies of general working populations, there is an excess risk in the order of 30% of developing myocardial infarction during follow-up (Relative Risk 1.3) even after adjustment for accepted cardiovascular risk factors. This is important for prevention in the workplace.

Still, one question remains: Is it more important to spend efforts on the improvement of the employees´ physical exercise, eating habits and smoking than to reduce job strain? Part of the answer could be found in the results of a large European collaborative study^[13]. This was a prospective study of 102,000 Europeans. Among those 14,000 participants who had at least two indices of an unhealthy lifestyle (for instance cigarette smoking and being overweight) job strain added very little to excess risk. But among those in the vast majority with no or only one of the unhealthy lifestyle factors, working in a job strain situation increased the risk of developing a myocardial infarction as a follow-up by 30 to 40 per cent. Accordingly, theoretically, it should be possible to reduce the number of new myocardial infarctions among men and women of working age by 30% by eliminating job strain, but among those with an unhealthy lifestyle, it would be cost-effective to emphasize the lifestyle. However, in the latter group, it is of relevance that a poor psychosocial work environment could aggravate an unhealthy lifestyle^[14].

As pointed out above, job strain is only one of the psychosocial risk factors at work. And myocardial infarction is only one of several illness outcomes that have been associated with job strain. Both these comments are addressed below.

Another important psychosocial theoretical job stress model was introduced by Siegrist in 1996[^15]. The general theory underlying this model is that high efforts at work should be rewarded. When rewards are small in relation to the input of effort, illness risks increase. This model has two parts; one component, over-commitment, or intrinsic effort, relates to the individual´s propensity to make efforts at work, and the other component relates to the job situation as it is perceived by the individual, extrinsic effort. The extrinsic effort is related to rewards in the form of money (material), esteem (appreciation and feedback; superiors are important for this) and finally status control (status in the organization and in society). Accordingly, this model in contrast to the demand-control-support model deliberately considers the individual´s subjective assessment of the work situation as well as propensity to the effort. It is reasoned that extreme over-commitment may develop into a personal trait in environments with a chronic emphasis on effort, resulting in “immersion” (drowning in one´s work). In most of the epidemiological studies, the over-commitment part has not been included. The demand control support model and the effort-reward model have been used together and it has been shown that they supplement each other in predictions of illness risk.

Another scientific model departing from the demand-control-support is the demand resources model^[16]. This model is more flexible. Resources are defined in a broad way, and both environmental/organizational and personal resources are taken into account. In addition, users can choose to supplement with local batteries of questions to the employees. This is of course recommended also in the use of the standardized demand-control-support and effort-reward questionnaires. The standardized questionnaires provide a framework, but local descriptions and specifications are needed.

In the traditions that have developed in the use of the three models that were described, epidemiologists have used the demand-control-support model more than the others. This may have to do with the fact that the theory is focused on the objective aspects of the environment. The demand resources model has been more popular among psychologists since it is closer to individual perception. The effort reward model has also been used extensively in epidemiological studies. It occupies an intermediate position between the demand-control-support and demand-resources models. For individual prediction, particularly in relation to depression as an outcome the effort-reward model has been more successful in predictions than the demand-control model. However, it is an important possibility for interventionists who are examining a workplace to use both models since they relate to partly overlapping but partly also different aspects of the work environment.

It has been important for the plausibility of the two most frequently used psychosocial models that biological mechanisms have also been linking the theoretical models to disease mechanisms. For instance, in longitudinal studies, it can be shown that subjects who experience increasing demands in relation to their own possibility to issue control over the situation are likely to experience worsened sleep quality, increased blood pressure level during working hours, suppressed activity in the parasympathetic system (as shown in studies of the heart rate variability) and decreased levels of regenerative (repairing) hormones that protect the body from adverse effects of stress. Links with the immune system have also been shown. These links have been described elsewhere^[14]. This is important in discussions with employers and trade unions. The associations that are observed should be possible to explain, otherwise, they are not trustworthy.

Other psychosocial work variables

Apart from demand-control-support, effort-reward, and demand-resources a number of other factors have been identified as potentially causing illness risks. These are:

• Long working hours
• Shift work (night work in particular)
• Bullying
• Employment insecurity
• Lack of justice
• Lack of skill discretion

Several reviews have been published in recent years. A few of them have applied the Cochrane Institute´s principles for systematic reviews. In a review from 2015, the evidence was examined on a four-graded scale for several psychosocial job factors in relation to the development of depressive symptoms^[17]. Fifty-nine articles of high or medium-high scientific quality were included. Moderately strong evidence (grade three out of four) was found for job strain (high psychological demands and low decision latitude), low decision latitude per se and bullying in relation to the development of depressive symptoms. Limited evidence (grade two) was shown for psychological demands per se, effort-reward imbalance, low support, unfavorable social climate, lack of work justice, conflicts, limited skill discretion, job insecurity and long working hours. It should be mentioned that the highest degree of scientific evidence, grade four, cannot be achieved by studies other than those designed as blinded randomized controlled trials. Therefore, grade three should be regarded as sufficient evidence for action. The evidence for some of the psychosocial risk factors may have been underestimated since a few studies of sufficient quality had been published. This was definitely the case for effort-reward imbalance which was included in a new review after several new published studies^[18] and should now be upgraded to grade three. The same conclusion should probably be made for long working hours and employment insecurity, as shown in later reviews.

In the next step, similar reviews were published covering many different psychosocial job factors in relation to the development of ischemic heart disease^[7] and burnout^[19] with results similar to those for depressive symptoms. Niedhammer et al^[20] later published a review similar to the one from 2015. Niedhammer’s review was based on other high-quality reviews. It was concluded that there are significant associations between high job strain and long working hours as exposures and ischemic heart disease, stroke, and depression as outcomes.

The psychosocial risk factors can be described more directly, according to Hassard and Cox^[21], whose summary is quoted here:

Alternative ways of describing the societal health impact of a poor psychosocial work environment

So far, the focus was put on associations between different work environment risk factors and specified illness outcomes such as depression, ischemic heart disease and stroke. However, this may give the reader a misleading impression that the psychosocial work environment may not be so important after all. The perspective can be widened by doing two things, first, relate the relative illness risk for a given illness not only to the individual but also to the whole population and second, include all illness outcomes at the same time in order to estimate the impact on “total” health. The first calculation is labelled an etiological fraction or an attributable fraction. Two numbers are needed in order to be able to calculate the attributable fraction, first, the relative risk (or odds ratio) associated with the illness and second, the prevalence of the risk factor (how common it is among employees). For instance, job strain is mostly estimated to be experienced by 22% of the working population. The relative risk of developing depressive symptoms when the employee experiences job strain is 1.74, judging from a summary of several publications. The attributable fraction resulting from this is 14%. Theoretically, this means that 14% of all cases of depression in the working population could be prevented if job strain could be eliminated. This is of course overly optimistic, but the number gives an idea of the importance of this association on a population level. If instead, the onset of ischemic heart disease is used as the illness outcome, it has been accepted to arrive at an attributable fraction of 6% (again 22% of employees are assumed to experience job strain, but the relative risk after adjustments for other risk factors is lower for ischemic heart disease, 1.3). In this case, the conclusion would be that 6% of new cases of ischemic heart disease could theoretically be prevented if job strain could be eliminated. Although there is some overlap between these two different illness outcomes since a few people may suffer from both depression and ischemic heart disease they are mostly happening in different individuals. Therefore, it would be important to combine them if the aim is a total estimation of the impact of poor psychosocial work environments. There are indeed other methods for exploring the total societal impact.

In a prospective study of the German population^[22] it was found that job strain was associated with a 28% increase (hazard ratio) in long-term near-future sickness absence from work. Passive work (low demands and low control) was also associated with a significant excess risk for long-term near-future sickness absence but on a lower level (14%). Long-term sickness absence is very important for the societal economy. It is however difficult to compare different countries using this parameter since they differ dramatically with regard to certification and remuneration for sickness absence. Still, it gives the reader an idea of the enormous financial impact that the psychosocial work environment has on society.

Another method for describing the total effect of illnesses arising in connection with job strain is to calculate the loss in years or days of healthy life. A longitudinal study with repeated observations based upon four cohorts in Finland, Sweden, France, and Great Britain including 64,000 subjects was performed with repeated observations. Healthy (alive and in good self-rated health) life expectancy (HLE) was calculated from age 50 to 75. After adjustments for other relevant factors, it was shown that men who had been working with job strain conditions had a shortened healthy life expectancy compared to those without job strain (2.0 years). The corresponding comparison for women indicated a slightly smaller effect on the number of healthy life years (1.5 years) among those with job strain compared to the others. Similar findings were made for life expectancy without chronic diseases (cardiovascular disease, cancer, respiratory disease, and diabetes) (CDFLE), but these findings were significant only for men, not for women.  

Ervasti et al^[23] used a large European cohort for a comparison between participants who reported long working hours (at least 55 hours per week) and compared them to participants who worked 35-40 hours per week. The comparisons were made with regard to 50 different health conditions and mortality outcomes. The result indicates that long working hours constitute a risk factor associated with an elevated risk of early cardiovascular death, diabetes, injuries, musculoskeletal disorders, and hospital-treated infections before age 65.

Sultan-Taieb et al^[24] recently published results from the study of the EU’s multi-country cohort. Five psychosocial exposures - job strain, effort-reward imbalance, job insecurity, long working hours and workplace bullying – were studied in relation to five different health outcomes, coronary/ischemic heart diseases (CHD), stroke, atrial fibrillation, peripheral artery disease and depression. Disability-Adjusted Life Years (DALY) rate per 100,000 workers was calculated in each country for each outcome attributable to each exposure. In addition, the overall burden of depression was calculated. It was estimated to be 528,549 DALYs for men and 344,151 for women (respectively 7,862 and 1,823 deaths). The three highest burdens in DALYs in EU28 in 2015 were found for depression attributable to job strain (546,502 DALYs), job insecurity (294,680 DALYs) and workplace bullying (276,337 DALYs). Significant differences between countries were observed for DALY rates per 100,000 workers.

It should also be pointed out that psychosocial factors interact with physical, chemical, or other risk factors. Accordingly, it has been shown, for instance, that exposure to noise (at home or at work) amplifies the effect of a poor psychosocial work situation on mental health^[25]. Carréon et al^[26] showed that workers who had been exposed for four years to carbon disulphide (which is an established chemical risk factor for heart disease) and at the same time worked in shifts (including night shifts) had a three-fold risk of developing coronary heart disease as a consequence as compared to workers who had neither had carbon disulphide exposure nor worked in shifts in the same period. Those who had been exposed to carbon disulphide and not to shift work had no significant excess risk. This is a clear example of an interaction between a chemical and a psychosocial risk factor.

Finally, in times of financial turbulence, employment insecurity is an important psychosocial risk factor. It should be emphasized that employment insecurity is related to depressive states in two ways, forward and backwards. This double directionality was studied^[27] during the year of financial crisis, in 2008. The Swedish longitudinal work population study was used to analyze the mental effects of downsizing. It was shown that job loss consistently predicted subsequent major depression among men and women, with a somewhat greater effect size in men. “Surviving” (staying in the same job) a layoff was significantly associated with subsequent major depression in women but not in men. Women with major depression have increased risks of exclusion from employment when organizations downsized, whereas job loss in men was not significantly influenced by their mental health. Thus, there were clear gender differences in the associations between downsizing and depression. Another study of the mental effects of downsizing on employees^[28] based on cohorts in Hungary, Sweden, France, and the United Kingdom showed that in the overall sample, the risk of developing depressive symptoms was significantly reduced if participants perceived the process as transparent and understandable, fair and unbiased, well planned and democratic. It helped if they trusted the employer's veracity and agreed with the necessity for downsizing. 

Additional risk factors that were not discussed here are gender-related worksite problems, shift work (in particular at night) and exposure to noise. An important growing research field is the one dealing with interactions between physical and psychosocial risk factors, such as noise and job strain as well as high temperature and long working hours.

How to make the psychosocial assessment at the worksite?

Assessments of psychosocial factors at the worksite should be well prepared and, as pointed out previously in this review, management that starts the assessment has the responsibility to plan for a follow-up with interventions if needed. It is very important the management is engaged and interested in the process. All parts of the assessment should be prepared with discussions between all social partners. Such discussions should include methods for assessment, schedules, participants, and information about results. Not least, the information about the results should be prepared: If individuals with an important role have been criticized in the assessment, they have to be informed in advance.

The most frequently used assessment method is to use a standardized questionnaire. The most widely distributed standardized questionnaires have been constructed by researchers but are widely used in practice in European countries. There are for instance questionnaires for the assessment of psychological demands, decision latitude and workplace support, effort-reward imbalance (including commitment), job insecurity, work-home imbalance, work injustice, demands and resources. There are also standardized questionnaires that assess the mental and physical condition of the employees. Which questionnaire to choose depends upon the scientific support that can be obtained for those doing the assessment, country, branch, and several other conditions. Several of the standardized questionnaires, for instance, the COPSOQ constructed in Denmark, and translated to many European languages, have different levels of specificity, one for researchers, one for occupational health care units and one less detailed for screening. Many of the standardized questionnaires are constructed with different modules, one for examining a general perspective (with dimensions such as job strain, effort-reward imbalance, demand-resource imbalance, work injustice, job insecurity etc.), but the general level is supplemented with a questionnaire representing a level that is branch specific. Often, such modules should be supplemented with questionnaires that are specific to a given workplace.

A general rule is that the number of questions should not overwhelm the employees. Another point to be raised is that it has been customary, particularly in large companies, to distribute large questionnaires at regular intervals, for instance once every year. This could be counterproductive, because the employees may be burdened by the questionnaires and answer them non-seriously. Another risk when a company performs frequent questionnaire assessments is that the employees could start manipulating their responses.

Questionnaires do not represent the only available method for assessment. Carefully planned and executed group interviews with preformulated questions based upon scientifically founded theoretical dimensions constitute an important alternative. If the interviewers are well-trained, they can obtain information that becomes available in the group situation and is based not only on the formulations in the answers but also on group processes and body language during the group interview.

In many situations, a combination of anonymous questionnaires to the employees and interviews with carefully selected individuals and groups is the optimal solution.

Finally, there are methods for assessing the awareness, knowledge, and activity level in the management with regard to psychosocial safety work. Such methods include standardized questionnaires distributed at the company level as well as interview guides. This will allow a higher-level bench-marking process.

Conclusions

Updated scientific information is of paramount importance in the EU’s efforts to reduce psychosocial risk factors at work. For all OSH stakeholders, it is important to know what the evidence is and in what ways bad jobs influence employee health and work quality. In the review, hidden costs for companies with a bad psychosocial work environment are discussed. This should be an important point for discussion in all work sites. The review describes the present state of knowledge. Both, mental and somatic health are influenced by a poor psychosocial environment. Quantitative aspects are discussed, and it is emphasized that the total damage caused by a poor psychosocial work environment may be underestimated if each risk factor and each illness outcome are regarded separately. Therefore, this document contains several studies in which combinations have been studied and the total impact can be estimated. Large-scale studies illustrate this point well. Finally, the mental effects of downsizing on employees can be dampened by well-prepared planning which helps the employees find new jobs.