Article Text
Abstract
Background Prospective studies on the relationship between organisational merger and mental health have been conducted using subjective health indicators. The objective of this prospective occupational cohort study was to examine whether a negative change during an organisational merger is an independent predictive factor of psychiatric morbidity.
Method Survey data on organisational characteristics, health and other factors were collected prior to (1996) and after the merger (2000); register data on psychiatric morbidity were collected at baseline (1/1/1994–30/9/2000) and during the follow-up (1/10/2000–31/12/2005). Participants were 6511 (77% men) industrial employees aged 21–65 years with no register-based diagnosed psychiatric events prior to the follow-up (the Still Working Study). During the follow-up, 252 participants were admitted to the hospital due to psychiatric disorders, were prescribed a psychotropic drug or attempted or committed suicide.
Results A negative self-reported change in the work organisation during the merger was associated with increased risk of postmerger psychiatric event (HR 1.60, 95% CI 1.19 to 2.14). This association was independent of mental health-related factors measured before the merger announcement, such as demographic characteristics, occupational status, personal orientation to life, self-rated health, self-reported psychiatric morbidity or chronic disease.
Conclusion A negative change in work organisation during an organisational merger may elevate the risk for postmerger psychiatric morbidity.
- Psychiatric disorders
- prospective studies
- organisational change
- mental health DI
- occupational stress
- psychiatric
- psychosocial epidemiology
- workplace
Statistics from Altmetric.com
- Psychiatric disorders
- prospective studies
- organisational change
- mental health DI
- occupational stress
- psychiatric
- psychosocial epidemiology
- workplace
Introduction
To stay competitive and efficient in the current global business environment, companies often have to merge with one another. Mergers often cause considerable changes in work organisations, work teams, roles and job statuses. From the employees' perspective, factors such as continuity,1 security2 and predictability3 may be threatened. Merger-related changes may therefore worsen working conditions,3–5 increase psychopathological symptoms6 and have adverse effects on employees' mental health.6–8
Although organisational mergers have become globally increasingly common, surprisingly little is known on the mental health effects related to them. Longitudinal studies on the health effects of organisational changes have mainly dealt with the effects of downsizing,9 changes in labour market status,10 employment instability11 and reorganisations within one workplace.6 Both the survivors of the downsized firms and the workers with an unstable job suffer from mental health problems (eg, increased use of psychotropic drugs) more often than those who have not experienced downsizing or who have a more stable job.6 11 12 Indeed, sparse prospective evidence suggests that a decline in one's status during a merger may be associated with at least a temporary decline in one's psychological well-being.13 However, no long-term prospective studies have been conducted on the impact of an organisational merger on the postmerger psychiatric morbidity of employees.
In this 5-year prospective cohort study, our objective was to examine whether a self-assessed change (improved or worsened situation) during an organisational merger (1/8/1998–30/9/2000) predicted the onset of new psychiatric events (measured by hospitalisation, drug prescription, attempted or committed suicide) after it (1/10/2000–31/12/2005) among employees who did not have a psychiatric disorder prior to or during the merger according to the health records (1/1/1994–30/9/2000). In addition, we examined whether the association was independent of various indicators of premerger health status and other baseline characteristics and whether certain sociodemographic groups were particularly vulnerable to the negative effects of a merger.
Methods
Procedure
Data were derived from the ongoing Still Working prospective cohort study examining the work-related antecedents of morbidity and mortality in a multinational forest industry corporation.14 15 In this study, we used a link between a personal identification code of the surveys and the national ID number given to all Finns at birth to merge the questionnaire data with data from national health registers. The approval of the ethics committee of the Finnish Institute of Occupational Health was obtained for the study.
Study context
The merger was performed between two companies based originally in Finland and Sweden employing >40 000 employees altogether globally. It was announced in the summer of 1998 and was carried out in 1999 and in early 2000. In Finland, where this prospective study was conducted, the company had been a stable traditional employer for several decades. The majority of the industrial plants were dispersed around the country, often being the major employer in the sparsely populated local areas. The merger did not lead to dramatic changes, such as large-scale redundancies during the study period, but evident changes in work roles, supervision, information practices and other procedures did take place.
Study design
As figures 1 and 2 show, we collected the data in four phases, twice before the merger and twice after. At time 1, the data on baseline psychiatric morbidity were collected between 1 January 1994 and 30 September 2000 from the national health registers (Hospital Discharge Register and Drug Imbursement Register). In spring 1996, prior to the merger, data on the subpopulation of employees who responded to the survey on psychosocial factors and subjective health were gathered (time 2). During autumn 2000, a postmerger survey (time 3) on work conditions, recent organisational changes and subjective health was carried out. In both survey phases, the questionnaires were sent to the work units, distributed to employees by their supervisors, and, once completed, mailed directly to the Finnish Institute of Occupational Health. Participation was voluntary, and confidentiality was assured to all employees. After the second survey assessment, data on postmerger psychiatric morbidity from 1 October 2000 to 31 December 2005 (time 4) were collected from the national mortality register and the same national health registers as at time 1.
Participants
Two study populations were formed. When generating study population I (including assessments at times 1, 3 and 4), we excluded participants with a history of hospital admissions or drug purchases related to psychiatric disorders at baseline (n=420) from the eligible study population (n=7850). The final cohort of 6511 employees consisted of employees who (1) were free from psychiatric disorders (no recorded hospital admissions or drug purchases) before the survey on 1 October 2000, (2) had worked for the company for at least 12 months before the survey in 2000, (3) responded to the scale of organisational change, (4) had no missing values for any other study variables and (5) were identified from the database of the National Population Register Centre. The mean length of the follow-up was 5 years and 2 months (range 0.0–5.3 years). At time 3, the mean age was 45 years (range 21–65 years). Women (23% vs 22%, p=0.031), non-manual workers (36% vs 28%, p<0.001) and married participants (63% vs 61%, p=0.001) were over-represented in the final study population, whereas no difference emerged according to age. The same exclusion criteria (1–5) was applied in study population II of 4096 employees who had responded to the surveys both before (time 2) and after the merger (time 3). Their baseline characteristics were more or less identical to the study population I; the participants were only somewhat older (mean age 47 years). The use of this population offered an opportunity to control the impact of several potential confounders measured prior to the merger announcement.
Assessment of change during organisational merger
At time 3, we assessed the employees' experiences of the merger period by asking what the course of events has been like during the organisational change. The study participants evaluated (1) the change in one's own standing at work (improved, unaltered and deteriorated), (2) the development within the work unit (1=very positive… very negative=5) and (3) the trends in the whole company (1=very positive… very negative=5) during the preceding 1–2 years. A summary measure of organisational change during a merger was derived by summing the responses across the three items (α=0.68). Based on summary scores, we formed three categories indicating the experienced change: improved situation (range 0–2.32), unaltered situation (range 2.33–3.32) and worsened situation (range 3.33–5).
Ascertainment of psychiatric events
Data on all persons who had been hospitalised for psychiatric disorders (ICD9 codes: 291–319; ICD10 F04–F99), who had been prescribed a psychotropic drug (ATC codes N05A, N05B, N06A) or had attempted suicide before the assessment of organisational change (1/1/1994–30/9/2000) and after (1/10/2000–31/12/2005) were obtained from the Hospital Discharge Register and from the Drug Imbursement Register, respectively. These registers reliably cover all information on the hospital admissions and drug prescription purchases for each Finnish citizen residing in Finland. Diagnosis-specific data related to depression were also used (ICD9 296.1, 300.4A; ICD10 F32, F33, D34.1; ATC N06A). In addition, data on suicides (10/2000–12/2005) from the Statistics Finland national mortality register were collected.
Ascertainment of mortality
The dates and causes of death from 1 October 2000 to 31 December 2005 were obtained from Statistics Finland. The database provides virtually complete population mortality data.
Assessment of demographic, health-related and psychosocial baseline characteristics
In study population I, data on covariates were collected at time 1 (premerger psychiatric events) and at time 3 (sociodemographic characteristics). All data on sociodemographic characteristics were obtained from the National Population Register Centre, except occupational status (employer's records). Marital status (married vs unmarried) and occupational status (manual vs non-manual employees) were used as dichotomised variables. Age was used as a continuous measure, except in the age-stratified analyses (younger than 50 years vs 50 years or older). In study population II, job characteristics were assessed at time 2 using the Occupational Stress Questionnaire,16 and sense of coherence (α=0.86)—a health-enhancing personality characteristic—was measured using a 13-item version of Antonovsky's Orientation to Life Questionnaire.14 17 Both decision authority (α=0.80) and skill discretion (α=0.83) at work were measured using five items.15 18 Summary scales were computed. For descriptive purposes, the scales were trichotomised, while in the main analyses, they were used as continuous measures. At time 2, data on premerger subjective health status were also collected using measures of self-reported psychiatric morbidity (10-item measure on anxiety, depression and other symptoms14), self-rated health (1-item measure,19 average/good vs bad) and chronic diseases (no vs yes). Self-reported psychiatric morbidity was used as a continuous variable, while other health measures were used as dichotomised variables in the analyses.
Statistical analysis
Descriptive statistics were applied to obtain the prevalence of organisational change during a merger and the prevalence of psychiatric events by covariates. The associations between organisational change levels and confounding factors were described by cross-tabulations and χ2 tests. The association between organisational change and psychiatric morbidity was assessed using Cox proportional hazards models. For each participant, person-days of follow-up were calculated from 1 October 2000 to the death of the employee, his or her hospitalisation or drug purchase or death due to psychiatric reasons, or to 31 December 2005, whichever of these three options came first. The time-dependent interaction term between predictor and logarithm of the follow-up period was not statistically significant, confirming that the proportional hazards assumption was justified (all ps>0.32). The adjusted HRs and 95% CIs for three categories of organisational change (improved, unaltered or worsened situation) were calculated, and the unaltered situation was used as a reference category. The main analyses were conducted in four steps. First, age and sex were adjusted for. Second, marital status and occupational status were also adjusted for. Third, sense of coherence and job characteristics were added to the models in the study population II, information of which could be used from time 2. Fourth, the subjective health measures from time 2 were additionally adjusted for in the same population. Additional analyses were conducted separately for other than alcohol- and drug-related diagnoses and for diagnoses related to depression. In the study population I, stratified analyses were run by sociodemographic characteristics. The analyses were conducted using the PHREG procedure of the SAS V.9.2 statistical software package.
Results
Of all 6511 participants with neither hospital admissions for psychiatric disorders nor use of psychotropic drugs at baseline, 59 died (excluding suicides) during the follow-up and were censored at the time of death. By the end of the follow-up, a total of 44 participants were admitted to the hospital due to a psychiatric disorder, 199 were prescribed psychotropic medication by a physician, 5 participants committed and 4 attempted suicide (total number of cases=252, 3.9% of the study population). The mean time before the first event was 2.7 years (range 0.0–5.2 years). As table 1 shows, employees with a psychiatric disorder were more often older employees and men. Employees with a psychiatric event after the merger had reported more symptoms of psychiatric morbidity, poorer self-rated health, chronic disease and a lower sense of coherence prior to the merger than those with no psychiatric events.
Table 2 presents variations in the prevalence of organisational change by demographic and other factors. Nearly 60% of the participants had experienced the organisational change rather neutrally, >20% saw that the situation had improved, while nearly 20% considered the changes during the merger negative. The participants who were older than 50 years, had low self-rated health or chronic disease prior to the merger, reported premerger symptoms of psychiatric morbidity, and had a weak sense of coherence and weak psychosocial work characteristics, reported negative organisational changes more often than the others. Non-manual employees experienced both negative and positive changes more frequently than manual employees.
As table 3 shows, after adjustment for age, sex, marital status and occupational status, a negative change in the organisation during the merger was associated with a 60% increased risk of psychiatric events (95% CI 1.19 to 2.14). The association slightly attenuated but remained statistically significant in the study population II, even after the work-related psychosocial risk factors, sense of coherence and various dimensions of subjective health status measured prior to the merger announcement were controlled (HR=1.53, 95% CI 1.06 to 1.99). The associations were more or less identical after alcohol-related psychiatric disorders (n=31) were excluded. The negative change in the organisation during the merger was also related to subsequent events of depression. In all models, the participants who reported a negative change in organisation showed the shortest time before the onset of psychiatric event.
Table 4 shows rather similar associations between organisational change and psychiatric events in separate groups according to age and sex. However, non-manual employees who had experienced the negative organisational change had a twofold risk of subsequent psychiatric events (95% CI 1.25 to 3.18), whereas manual employees were not significantly affected by organisational changes (HR=1.38, 95% CI 0.94 to 2.03).
Discussion
Our unique longitudinal data connecting organisational surveys and national health registers and covering a period of 11 years with four assessment phases reveal that merger-related experiences may affect employees' mental health long after the merger process is over. Evidence suggests that when initially healthy employees considered the situation after an organisational merger worse than before, it predicted the onset of a psychiatric event during the 5-year follow-up period after the merger. The association was independent of conventional mental health risk factors. The finding lends support to the earlier longitudinal findings on organisational changes and mental health using both register-based12 and subjective6 13 measures of psychiatric morbidity as outcomes.
A merger can change one's work in many ways, for example, the nature of the job itself, the human relations in one's work team and the information flow in the organisation.20 A merger may increase demands at work 4 and result in sleep disturbances.6 A negative organisational change is likely to be experienced as a potential threat, which can exceed the adaptive resources of the employee21 via various pathways such as uncertainty,2 decreased status,13 role conflicts3 5 and anxiety.22 Based on our data, it seems plausible that these types of merger-related stressful negative experiences can lead to elevated risk of psychiatric events after the most hectic period of organisational change has been passed.
Previous evidence on the association between organisational merger and mental health problems has mainly been based on subjective outcomes, and the analyses have not sufficiently controlled for the confounding effect of permanent personality characteristics, job content or baseline mental health status. Thus, chronically weak psychosocial resources and/or personal orientation to life may have confounded the associations.23 Pre-existing mental or chronic illness may also have a role in the organisational change–psychiatric morbidity relationship. We believe that this study is able to surmount these weaknesses. First, we used register data to assess the outcome. Furthermore, all those who had had psychiatric problems according to the medical records prior to the assessment of merger were excluded from the cohort. Finally, in the additional analyses (study population II), the premerger subjective measures of psychiatric morbidity, general health status, personal orientation to life and job content were used as covariates. Therefore, due to our exclusion criteria and the large set of covariates used, health selection, personal orientation to life or adverse job content was unlikely to play a major role in our results.
It is noteworthy that our stratified analyses by occupational status showed that unlike the manual employees, the non-manual employees who were often responsible for carrying out changes during the merger were at a particularly high risk of psychiatric morbidity after the merger period if they had reacted negatively to the organisational changes.
Some weaknesses may limit the generalisability of our findings. It is possible that some unexplored factors such as severe negative life events may have contributed to mental health risk and confounded the associations. However, it is likely that personal life crises have a rather independent or additional role in the impairment of mental health with regard to organisational changes. A great proportion of our study participants worked at industrial plants in rather small and often remote communities, where it is difficult to find a new job. Hence, for those employees who had experienced negative changes and with family responsibilities, future attempts to find a new job would have required major changes in their lives. This may have increased the impact of the negative organisational changes. In contrast, lengthy follow-up after the merger and exclusion of employees with psychiatric events during the merger (1999–2000) attenuated the relationship between the stressor and the outcome, and this is likely to have minimised the estimated associations between organisational change and psychiatric morbidity. Yet we acknowledge that the findings of the present study should be validated in further large-scale studies. Such a replication would be useful in determining the generalisability of our findings, that is, whether they apply to different types of merging organisations and whether differences between societies and sectors of work moderate the effects of an organisational merger on employee's psychiatric morbidity.
Practical implications
Organisational mergers and other organisational changes have become increasingly common throughout the world. The steep increase in new cases of mental ill health among working population may partly reflect the incapacity to manage organisational changes properly.24 Our longitudinal data suggest that negative experiences during a merger may lead to elevated psychiatric morbidity among employees with no earlier register-based diagnosed psychiatric disorders, especially non-manual employees seem to be vulnerable. Policymakers, employers and occupational health professionals should recognise that an organisational merger may pose a severe risk to employees' mental health. Organisational changes such as mergers may, therefore, be important to consider in mental health prevention of employed population. Procedures for managing the inevitable organisational changes in a way that would minimise negative experiences should be further studied and implemented.
What is already known on this subject
Mergers often cause considerable changes in work organisations, work teams, roles and job statuses
Negatively experienced merger-related changes seem to increase stress at work and have adverse effects on employees' subjective health
What this study adds
The results from this follow-up study covering a period of 11 years with four assessment phases indicate that merger-related negative experiences can lead to elevated risk of psychiatric events
It seems that the negative change in the organisation during the merger may represent a risk factor for psychiatric disorders
Acknowledgments
The authors thank all study participants, the Still Working members and the leader of the earlier survey phases Raija Kalimo.
References
Footnotes
Funding The research group was financially supported by the Finnish Work Environment Fund (OSH-ERA project, #109395). AV was also supported by the Academy of Finland (#128089).
Competing interests None declared.
Ethics approval This study was conducted with the approval of the ethics committee of the Finnish Institute of Occupational Health.
Provenance and peer review Not commissioned; externally peer reviewed.