Article Text

Download PDFPDF

Resource-enhancing group intervention against depression at workplace: who benefits? A randomised controlled study with a 7-month follow-up
  1. Kirsi Ahola,
  2. Jukka Vuori,
  3. Salla Toppinen-Tanner,
  4. Pertti Mutanen,
  5. Teija Honkonen
  1. Work Organizations, Finnish Institute of Occupational Health, Helsinki, Finland
  1. Correspondence to Dr Kirsi Ahola, Finnish Institute of Occupational Health, Topeliuksenkatu 41 a A, FI-00250 Helsinki, Finland; kirsi.ahola{at}ttl.fi

Abstract

Objectives The aim of the present study was to investigate whether participation in a structured resource-enhancing group intervention at work would act as primary prevention against depression. The authors analysed whether the intervention resulted in universal, selected or indicated prevention.

Methods A total of 566 persons participated in a prospective, within-organisation, randomly assigned field experimental study, which consisted of 34 workshops in 17 organisations. The participants filled in a questionnaire, were randomly assigned to either intervention (n=296) or comparison (n=324) groups and returned another questionnaire 7 months later. The intervention, lasting four half-day sessions, was delivered by trainers from occupational health services and human resources. The aim of the structured programme was to enhance participants' career management preparedness by strengthening self-efficacy and inoculation against setbacks. The comparison group received a literature package. The authors measured depressive symptoms using the short version of the Beck Depression Inventory. A high number of depressive symptoms (over 9 points) were used as a proxy for depression.

Results At follow-up, the odds of depression were lower in the intervention group (OR=0.40, 95% CI 0.19 to 0.85) than in the comparison group when adjusted for baseline depressive symptoms, job strain and socio-demographics. In addition, the odds of depression among those with job strain (OR=0.15, 95% CI 0.03–0.81) at baseline were lower after the intervention. The intervention had no statistically significant effect on those with depressive symptoms (over 4 points) at baseline.

Conclusion The resource-enhancing group intervention appeared to be successful as universal and selective prevention of potential depression.

  • Depression
  • group intervention
  • primary prevention
  • RCT
  • work
  • organ system
  • disease
  • disease type
  • stress
  • general expertise
  • psychology
  • methodology
  • speciality
  • burnout
  • materials
  • exposures and occupational groups
  • workload
  • epidemiology
  • statistics
  • psychiatry

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

What this paper adds

  • Although depressive symptoms can be reduced via preventive programmes, only a few programmes have been successful in preventing depression.

  • Only a small proportion of interventions aimed at preventing depression have been conducted at workplaces.

  • In this randomised controlled study, a structured, incompany resource-enhancing group intervention had a positive effect on the prevention of depression indicated by a high number of depressive symptoms.

  • The intervention was successful among all participants (universal prevention) and among those with job strain (selective prevention).

Introduction

It has been estimated that in 2030, depression will be the leading cause of disability-adjusted life-years in developed countries.1 Depressive symptoms have already been shown to predict future work disability, indicated by long sickness absences2 and disability pensions.3 Depression may also result in impaired job performance.4 The resulting loss due to reduced work capacity has been estimated to account for approximately two working days per month5 or 27 working days per year.6 All in all, the total cost of depression to society may be substantial.7 ,8

Depression is multifactorial in origin.9 ,10 Along with a variety of contributing factors, psychosocial factors at work significantly affect the onset of depression.11–13 The associations have been the strongest and most consistent for job strain defined as high demands and low decision latitude at work. In addition, evidence has been found of an association between depressive symptoms and effort-reward imbalance, low social support and high job insecurity.

Work disability caused by depression is more common among workers who are older, have a low socioeconomic position and have a severe level of depression.14 Individual factors have also been shown to affect depression-related work disability. Feelings of hopelessness15 and low self-esteem16 predict disability pensions among depressed outpatients. Work-related factors may also affect early retirement. In a Finnish population-based sample, job strain was an independent predictor of disability pension granted due to common mental disorders.17

According to the health promotion paradigm,18 primary prevention is used to counteract illness among people who do not yet meet any diagnostic criteria for a mental disorder. Such preventive activities can be further divided into three main categories according to the target population. Universal preventive interventions are targeted towards a whole population. They are provided to individuals with all levels of risk, including no risk at all. Selective preventive interventions are targeted towards individuals whose risk of developing a mental disorder has increased based on earlier evidence. Indicated preventive interventions are targeted towards high-risk individuals who already show mild but detectable symptoms of a foreshadowing mental disorder.

Earlier research has shown that the burden of depression can be reduced through preventive activities.19 ,20 Successful interventions against depression have been lengthy and provided by a combination of healthcare professionals and lay personnel. They have included several methods involving a competence enhancement component, have been conducted according to a well-defined programme and have been integrated into a high-quality research design.19 However, only a small portion of mental health interventions have been conducted at workplaces and their positive effect on depressive symptoms has been small.20–22 In a subgroup analysis, the interventions that included a component focused directly on mental health (ie, psychoeducation or training in coping skills) had a similar beneficial effect to those with an indirect focus on mental health (ie, via risk factors for depression, such as low physical activity, poor work environment, and cardiovascular disease).

Modern work life is characterised by inevitable changes, continuous challenges and ever-increasing insecurity due to various mergers, acquisitions, downsizing and off-shoring caused by globalisation and the deregulation of financial markets. Therefore, the personal resources of workers for managing their own careers in changing situations have become increasingly important for workers' health and work ability.23 ,24 One resource-enhancing group intervention25 was developed based on earlier experiences in preventive group methods aimed at empowering people and increasing their career preparedness during unemployment.26–29 Further development of the intervention was guided by analyses of successful interventions and the Michigan Prevention Research Centre principles for effective group training.19 ,30

The empowering intervention25 relied on the principles of social cognitive theories on social action, behavioural control and individual coping resilience.31–33 It found that well-being can be improved by developing individual preparedness and abilities related to goal setting and attainment. Preparedness is defined here as a cognitive–motivational construct that contains specific self-efficacy32 and preparation against setbacks.33 The intervention aimed at enhancing career management preparedness by (1) identifying in peer groups the career-related goals needed for progress, (2) defining in peer groups the solutions and tasks for carrying out these goals and (3) practising the required skills and actions in small groups.25 Social modelling, gradual exposure and practice through role-playing were used as the main methods. The modules of the structured programme were built on skills training, active learning, skilled trainers, social support and inoculation training.34 Previous studies show that the intervention increased career management preparedness, which in turn led to positive longer term effects, that is, an increase in mental resources and mental health as well as a decrease in intentions to retire, regardless of perceived job security.25 ,35

The aim of the present study was to investigate if such a resource-enhancing group intervention at work would succeed in preventing depression indicated by a high number of depressive symptoms. Increase in career management preparedness might strengthen employees' personal resources and help them to mould their psychosocial factors at work in such a way that the risk for depression would decrease. In order to get the ultimate benefit from such an intervention, we were especially interested in whether it would act most effectively as universal, selected or indicated prevention.

Methods

Procedure and participants

A total of 43 medium-sized and large-sized organisations in southern Finland were contacted and offered the opportunity to participate in a study on an intervention programme.25 Fifteen organisations from different sectors, both private and public, agreed to participate. Another two organisations chose to participate after hearing about the study from other participating organisations. Of the 17 participating organisations, nine were city administrations, five were governmental organisations (two city departments, a research institute, an employment office and an insurance office) and three were private enterprises (a banking company, a multiservice company and an occupational health service organisation). The size of the participating organisations varied from 106 to 13 871 employees, and the approximate proportion of women among their personnel was 75%.

All material and training of the trainers were free of charge to the organisations. Those participating agreed to invest through working hours of trainers and participants. Two employees from each organisation were trained as trainers for four full days at the Finnish Institute of Occupational Health, one from occupational health services and the other from human resources. The researchers provided the organisations with information leaflets regarding the programme.

The opportunity for the employees to take part in the intervention was marketed to the whole personnel via personnel magazines, the intranet and other channels used by the organisations in their daily communication, as a way to ‘get an extra buzz from your work’. The information letter introduced the study as having two experimental conditions: one would involve participation in group training and the other would involve receiving a literature package. To become a participant, the employees had to agree to a randomisation procedure and to hand in the baseline questionnaire (T1). Between the autumn of 2006 and the spring of 2008, a total of 718 individuals took part in the study. From the 732 persons who originally agreed to participate, 14 were discarded due to returning the questionnaire too late, unknown address, preference for intervention group or quitting the study.

The 718 participants were randomly assigned to either the intervention group (N=369) or comparison group (N=349). The randomisation was carried out by the researchers separately for each participating organisation. Two researchers shuffled the received sealed questionnaire envelopes and dealt them into two piles. The participants were given a research code and their questionnaires were sent to an independent company to be saved as data. The results of the randomisation were sent to the participating organisations. Those randomised into the intervention group were invited to take part in the group workshop, whereas the comparison group received a literature package on career management and health-related information during the week the intervention group received their training.

A total of 34 workshops were conducted. The groups, comprising 8–15 employees and supervisors, assembled in the participating organisations with two trained trainers from the same organisation. The participants were sent a follow-up questionnaire (T2) 7 months after the intervention. The follow-up questionnaires were returned by 86% (n=616) of the study participants: by 88% (n=324) of the intervention group and by 84% (n=292) of the comparison group. Again, the questionnaires were sent to an independent company that combined the follow-up data and the baseline data. After excluding the participants with missing values in some of the relevant variables (n=50), the final study population comprised 566 people (79%): 296 (80%) in the intervention group and 270 (77%) in the comparison group.

Intervention

The intervention was delivered by two trainers, one from occupational health services and one from human resources, strictly according to the structured programme,25 as presented in the published manual.34 The intervention consisted of four half-day sessions, which were delivered over 1 or 2 weeks. The skills training element included defining one's own strengths and career interests, introducing the principles of lifelong learning, practising organisational change management, obtaining career-related resources from social networks, solving social conflicts and managing one's career. The trainers relied on active learning methods by making use of participants' own career knowledge and career choice skills in discussions and role-plays. The skilled trainers worked together in pairs in order to build trust and facilitate group processes, which promote learning, and social support was provided by facilitating modelling and strengthening supportive behaviour in the groups. Preparation against setbacks was accomplished through inoculation training. The modules are presented in figure 1.

Figure 1

Contents of the four half-day sessions of the ‘Towards Successful Seniority’ intervention (http://www.ttl.fi/successfulseniority).

The no-show bias and the dropout effect in T2 were examined. When we compared the socio-demographic characteristics of the participants and no-shows, we found that the no-shows were more often employed by the government than were the participants (p<0.05). Of the 369 participants in the intervention group, 25 (6%) did not finish the intervention. There were no significant differences among the dropouts in any of the factors studied between the intervention and the comparison group.

Measures

Outcome

We used potential depression as the outcome variable of the intervention. A high number of depressive symptoms was used as a proxy of depression. The number of depressive symptoms at T2 was measured using the 13-item short form of the Beck Depression Inventory (BDI).36 The reliability of the measure (Cronbach's α) was 0.86. The summary score of the items was calculated and dichotomised as ‘no depression’ (0–9 points) and ‘depression’ (10–39 points).37

Job strain

In order to obtain an indicator for the risk of depression in selective prevention,18 job strain was assessed as the combination of job demands and job control, which were measured using the Job Content Questionnaire.38 Job demands comprised five items (Cronbach's α=0.79) and job control comprised nine items (Cronbach's α=0.80). We summed up job demand items and job control items and split them into two groups (low and high) at their median.38 We then formed the following categories for job strain: ‘no job strain’ (groups of high demands and high control, low demands and low control, and low demands and high control) and ‘job strain’ (high demands and low control).

Depressive symptoms

As an indicator for a high risk of depression in indicated prevention,18 we measured depressive symptoms at T1 using the 13-item short form of the BDI.36 The summary score of the items was calculated and dichotomised as ‘no depressive symptoms’ (0–4 points) and ‘depressive symptoms’ (5–39 points).36

Socio-demographic factors

Socio-demographic factors included age, gender, marital status and occupational status. Age was categorised by 5-year intervals starting at 25 years. Marital status was dichotomised as married (married or cohabiting) and unmarried (divorced, widowed or single). Occupational status was dichotomised as white collar (mostly mental work) or blue collar (physical work for at least half of the working day).

Statistical analysis

We used cross-tabulations to describe the prevalence of depression at T2 among all participants and also according to job strain and depressive symptoms at T1. We analysed the likelihood of depression at T2 using logistic regression. We then examined (model 0) the effect of group intervention on the odds of depression at T2. The model was adjusted for the depressive symptoms at T1 (model 1) and for socio-demographic factors and job strain at T1 (model 2). Next, the model was modified by including the interaction term of intervention and job strain, to obtain the intervention results separately for those without job strain (model 3a) and those with job strain (model 3b). Finally, the model (model 0) was modified by including the interaction term of intervention and depressive symptoms at T1 to obtain the intervention results separately for those without depressive symptoms and those with depressive symptoms at T1 (models 4a and 4b, respectively). These models were adjusted for socio-demographic factors and job strain.

Results

The mean age of the participants was 50 years (SD=6.5, range 31–64). A total of 89% were women and 74% were married. Most of the participants worked in the public sector: 66% in municipalities and 15% in state administration, while 18% worked in the private sector. Over half of the participants (56%) were graduates and 31% had completed high school. The majority of the participants (74%) were white-collar workers. At baseline, 10% of the participants had potential depression and 39% showed depressive symptoms. Table 1 presents the characteristics of the study participants by group. There were no statistically significant differences between the intervention and the comparison group at baseline.

Table 1

Characteristics of the study population by group at baseline

Seven months after the intervention, the prevalence of potential depression was 9% among the study participants. It was 8% (n=23) in the intervention group and 12% (n=33) in the comparison group (figure 2). The intervention had a statistically significant effect (OR=0.40, 95% CI 0.19 to 0.82, p=0.01) on depression at T2 after depressive symptoms at baseline were taken into account (model 1). The OR for depression at T2 in the intervention group compared with the comparison group was 0.40 (95% CI 0.19 to 0.85, p=0.02), after additional adjustment for socio-demographic factors and job strain (model 2).

Figure 2

Prevalence of depression at 7-month follow-up by preventive activity. Asterisk indicates a statistically significant difference between the intervention and control groups at p<0.05.

The analyses were then stratified according to job strain at baseline. Seven months after the intervention, 7% of the participants (n=18) in the intervention group and 10% (n=23) in the comparison group had potential depression among those without job strain at baseline (figure 2). The corresponding numbers among those with job strain at baseline were 10% (n=5) and 23% (n=10). Compared with the comparison groups, the adjusted ORs for depression at T2 in the intervention group were 0.52 (95% CI 0.23 to 1.19, p=0.12) among those without job strain (model 3a) and 0.15 (95% CI 0.03 to 0.81, p=0.03) among those with job strain (model 3b).

The analyses were alternatively stratified according to whether participants had depressive symptoms at baseline. Seven months after the intervention, 1% of the participants (n=1) in the intervention group and 2% (n=3) in the comparison group had potential depression among those without depressive symptoms at baseline. The corresponding numbers among those with depressive symptoms at baseline were 19% (n=22) and 29% (n=30). The intervention had no statistically significant effect among those without depressive symptoms at baseline (adjusted OR=0.30, 95% CI 0.03 to 2.96, p=0.31; model 4a) or among those with depressive symptoms at baseline (adjusted OR=0.55, 95% CI 0.29 to 1.05, p=0.07; model 4b).

Discussion

In this randomised controlled field trial, a resource-building group intervention at the workplace had a statistically significant effect against depression indicated by a high number of depressive symptoms at 7-month follow-up, after the baseline socio-demographic factors, depressive symptoms and job strain of the participants were taken into account. The effect of the intervention was significant among the heterogeneous group of all participants (universal prevention) and among the subgroup of those who suffered from job strain at baseline (selective prevention).

As the major strength of the study, the participants were randomly assigned into the intervention and comparison groups. The randomisation was successful in that there were no differences between the intervention and the comparison group in any of the study variables. The voluntary participants came from different sectors and jobs of both large-sized and medium-sized organisations. The follow-up period covered 7 months, which is a relatively long follow-up for the effects of an intervention. Finally, the theory-based intervention was delivered in a structured way strictly according to a published manual.34 A previous study had confirmed increased career management preparedness as the immediate effect of the intervention.25

A few limitations must also be considered. We could not compare the characteristics of the participating and declined organisations. The participating organisations were large-sized and medium-sized, situated in urban areas, operated predominantly in the public sector and were dominated by female workers (75%). Therefore, the results can be carefully generalised to similar settings in which the participation is voluntary. We were also unable to compare the participating employees and those who did not take part. The participants were mainly women (89%), reflecting the overall proportion of women in the participating organisations, and white-collar workers (74%). Therefore, the results need to be replicated in other kinds of samples. The final study sample comprised 79% of those who originally took part in the intervention. During the intervention, there was no statistically significant difference between the dropouts and the respondents.

The number of participants in our sample was quite high (n=566). However, the stratified analyses may have suffered from diminished statistical power. Many interesting issues worth studying as well as important factors to control for were outside the focus of the present study. For example, psychological contract, psychological capital and psychological involvement with work among the participants may have affected the success of the intervention, and the level of income of the participants may have affected the general risk of depression.

A detailed cost and benefit analysis could not be performed in the present study. The costs of the intervention in each organisation included the training of the two trainers (4 days each), and then in each intervention group, the work time of the participants and two trainers (16 h each). The benefits included a decreased level of depressive symptoms and diminished intentions to retire prematurely, and an increase in the employees' mental resources.25 The intervention may also have helped to prevent subsequent cases of depression. These consequences may lead to savings in terms of decreased sickness absence days and increased productivity, results that need to be verified with a detailed analysis of costs and benefits in any subsequent studies.

We assessed depressive symptoms using the short form of the BDI.36 A high number of depressive symptoms (over 9 points) were used as a proxy for depression.37 However, depression can never be reliably defined with a questionnaire. The prevalence of a high number of depressive symptoms was 9% after the follow-up in the present sample, which consisted mainly of women. This prevalence was equivalent to that of the prevalence of depressive disorders among women assessed using the Composite International Diagnostic Interview in a representative Finnish study.39 Furthermore, no participants were excluded due to prior illness in our study.

The results demonstrated that a resource-enhancing group intervention can be successfully applied as a form of universal preventive intervention targeted towards all employees. It helped to decrease the likelihood of potential depression. Universal preventive interventions against depression have been scarce and have so far yielded quite modest results, whereas procedures targeted towards risk groups that have also proven successful have been more frequent.19–22 Mental health intervention studies have seldom been conducted at workplaces. Therefore, the most important contribution of our study to previous knowledge is that the likelihood of depression can be decreased in the work context. These results are very promising because even mild levels of depression have shown to cause enormous losses in productivity to employers due to its high prevalence and per patient condition impact.40

Job strain appears to be a risk factor for subsequent depression11–13 and related work disability.17 In the present study, the intervention proved successful as a form of selective preventive intervention18 against depression among those suffering from job strain at baseline. This is a very important finding that demonstrates that the adverse consequences of job strain can be combated within organisations, through training and collaboration. Participants in the peer groups discussed the everyday challenges in their work and exchanged practical ideas and ways of action to battle these challenges. However, the effect of the intervention as indicated prevention of depression among the employees with depressive symptoms did not reach significance in this study. One must be careful in giving recommendations regarding the targeting of the intervention. A heterogeneous group of participants including mentally healthy employees may be necessary in order to provide role models in the peer group-based intervention as examples of positive experiences and alternative interpretations. More research is needed to clarify how the intervention effect depends on the composition of the group.

Our intervention employed many characteristics described as predictors of efficacy in depression prevention programmes.19 Our programme was based on social cognitive theory on preparedness being the mediator of the beneficial effects of the intervention on mental well-being and mental resources.28 The intervention was provided in a structured format applicable by the organisations' own employees, as a combination of healthcare professionals (occupational healthcare) and lay personnel (human resources). The programme included several active learning methods and emphasised a supportive learning environment for enhancing the effectiveness of training. The focal active ingredient of the programme, preparedness for career management, included two intertwined components of competence enhancement, that is, enhancement of self-efficacy related to career management and the enhancement of inoculation against setbacks.19 ,25 The intensive programme lasted for 16 h, and the sessions were divided across several days. The main difference compared to other programmes was that we had four long sessions (4 h each).19 Our intervention was strongly based on group training and supportive learning environment, which require longer meetings. The implementation of the programme was also integrated into a high-quality, randomly assigned, field experimental research design.19 It was not possible in the present study design to analyse which of these aspects of our intervention were the most effective in creating the positive results.

The beneficial long-term mental health effects of interventions based on enhancing preparedness of individuals have earlier been shown to be mediated by proximal increase in preparedness.25 ,28 Employees who are well prepared and motivated to manage their careers have confidence in their career management skills, as well as the knowledge and emotional readiness to deal with the setbacks frequently encountered in the modern work life. This preparedness is strengthened through concrete advice, social support and examples of new ways of action received during the intervention. Based on previous studies, it seems that the component of inoculation against setbacks is especially effective in reducing the risk of depression.29 ,30

Inoculation against setbacks is used in cognitive behavioural therapy to build a sense of mastery when confronting stressful setbacks and obstacles. The inoculation process involves the anticipation of high-risk situations and the acceptance of relapses as learning opportunities,33 which help maintaining participants' motivation to handle difficult situations. As the employee learns to better manage stressful career transitions or other situations, exposure to manageable levels of stress leads to learnt resourcefulness and prevention relapse. In this process, the employee builds up a defence composed of skills and positive expectations in order to deal effectively with stressful situations in the future. This in turn leads to better personal career outcomes, and consequently, to beneficial effects on mental health.29

In the future, more studies in different employee groups are warranted. These studies should focus on the concrete mediating mechanisms of the intervention, composition of the groups, different kinds of implementation procedures and cost–benefit calculations. Other factors, such as level of income, psychological contract, psychological capital or psychological involvement with the work of the participants, all of which affect the success of the intervention, should be included in the design. In addition, the characteristics of the declining organisations and employees should be more carefully registered and compared with the participants in the future studies. Furthermore, more information about the intervention would help wider distribution and testing.

In conclusion, the studied resource-building group intervention targeted towards a heterogeneous group of employees was successful as a form of universal prevention for decreasing depression indicated by a high number of depressive symptoms. The intervention also had a beneficial effect, as a form of selective prevention, on those with job strain at baseline.

Acknowledgments

We are very grateful to the participants and trainers for their time and effort. We also thank Mrs Alice Lehtinen for linguistic editing of our manuscript.

References

Footnotes

  • Funding This article is based on research supported by the Finnish Work Environment Fund (106093), the Finnish Ministry of Finance (31/37/2007) and the Academy of Finland (124294).

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval Ethics Committee of the Finnish Institute of Occupational Health.

  • Provenance and peer review Not commissioned; externally peer reviewed.