Long-term sickness absence
Job mismatching, unequal opportunities and long-term sickness absence in female white-collar workers in Sweden
Hélène Sandmark a, b
a School of Health and Medical Sciences, Public Health Science, Örebro University, Sweden
b Department of Medical Sciences, Occupational and Environmental Medicine, Uppsala University, Sweden
Corresponding author:
Dr Hélène Sandmark,
School of Health and Medical Sciences, Public Health Science
Örebro University
SE-701 82 Örebro
Sweden
Phone +46 (0)19 303676
Abstract
Aim: To investigate associations between long-term sick-listing, and factors at work and in family life.
Methods: Associations were investigated in a cross-sectional case-referent study. The study base included women in white-collar jobs, aged 30 to 55 years, living in three urban areas in Sweden between February 2004 and October 2004. A postal questionnaire was constructed with questions on occupational and family circumstances, and sent to 510 randomly selected female white-collar workers, of whom 190 had an ongoing sick-leave of 90 days or more. The response rate was 81% (n=413).
Results: In the bivariate associations, sick-listed women with children showed the highest estimates regarding reported long working hours, bullying, high mental strain, low control and low influence at work, and work–family imbalance. In a regression model, the strongest associations were: experiencing too high mental strain in work tasks (OR=2.57 95 % CI 2.09-3.15) and low control and influence at work (OR=2.17 95 % CI 1.60-2.94). Sick-listed women rated an overall higher dissatisfaction with their workplace and working life.
Conclusions: There seemed to be a greater tendency for sick-listed women to live in traditional family relationships with unequal opportunities. All the women in this study were primarily in managerial positions and in work groups previously dominated by men, and this could induce job strain and stress. The women who were sick-listed were probably less able to cope with these exposures. Policies in society have not yet succeeded in overcoming health hazards and facilitating occupational life for all working women.
Key words: work–family imbalance; mental strain; long working hours; bullying; case-referent study; job autonomy
Background
During the past century, one of the major social changes in western societies has been the change of women’s social status and economic behaviour, and their participation in paid work. Apart from their high rate of participation in the workforce today, women have made advances in management, which used to be a male domain [1]. However, the labour market and social conditions vary from one country to another [2]. Those who are at risk of long-term sickness absence may vary, and without a good knowledge of the basic associations between demographic factors and workplace characteristics it is difficult to target interventions aimed at reducing sickness absence.
In Sweden the average number of working days lost per year due to sickness is the highest in the OECD [3]. This is due to a high rate of sick leave, and a long average length of each spell. Sickness benefits can last for several years in Sweden, or be changed into early pension, and both have increased since the end of the last decade. Since the end of the 1990s, women in the Swedish population take more than twice as much long-term sick leave as men [3,4]. From 1998 until 2002, sick-listing for more than 90 days increased by 230% among Swedish white-collar workers, and for white-collar women with higher wages the increase was 335% during the same period [5].
Exposures to increased job demands and work intensity have been suggested as possible reasons for the increase in long-term sickness absence. Organizational restructuring in companies, globalization of enterprises, and change in work contracts, especially during the last decade, have also taken place. These changes probably affect the health and well-being of the workforce. The increase of female participation in the workforce and dual earners in families, which have taken place during the last decades could also affect the health of the workers [6, 7, 8].
Sickness absence is not identical with health status, and the mechanism behind it is complex. New definitions of sickness and changed attitudes could underlie at least part of the increase since the end of the 1990s. It has been suggested that sickness absence could be an indicator of health in a population if it includes social, psychological and physiological dimensions, as well as medical dysfunctions [9,10].
Aims
The aim of this study was to investigate associations between long spells of sick-listing and factors at work and in private life, in female white-collar workers in high positions in the private sector in Sweden. It was hypothesized that there are factors beyond those which are strictly connected to a medical diagnosis that could be associated with long-term sick-listing in Swedish women.
Methods
The Swedish social insurance system
Social insurance, which is uniform throughout the country and funded by the Swedish state, is administrated by the Swedish Social Insurance Agency. Everyone who lives or works in Sweden is covered by social insurance. Sickness compensation is possible in the case of a significant reduction in work capacity caused by sickness or injury. The insurance distinguishes between personal sickness absence and staying at home to take care of sick children, which is covered by a special insurance within the Social Insurance Agency [4].
For employees in the private sector in Sweden, a mutual insurance company, Alecta, operated at the time of this investigation, on behalf of the Confederation of Swedish Enterprise and the Federation of Salaried Employees in Industry and Services, to provide additional compensation in the case of long-term sickness lasting more than 90 days [11].
Study population and design
The association between long-term sick-listing, and factors related to work and private life were investigated in a cross-sectional case-control study. The study base included women in white-collar jobs, aged 30 to 55 years, living in three urban areas in Sweden between February 2004 and October 2004. In order to investigate women in higher positions with present full-time positions, only those who had a monthly salary of at least SEK 25,000 (EUR 2750) were included.
From a register at Alecta, with a total of 1,170,000 policyholders, women which on the days of the data selection had been continuously sick listed for more than 90 days, and living in three urban areas in Sweden, were randomly selected (cases). Women on sick leave due to stress reactions, musculoskeletal dysfunctions, medical problems, minor psychiatric dysfuntions, burnout syndrome, occupational fatigue and depression, were included. Those with pregnancy complications or severe medical diagnoses, such as cancer, severe coronary dysfuntions, stroke, major psychiatric diagnoses or complications after major accidents, were not included. Female referents, who were not on long-term sick leave, but of the same age as the cases, in the same salary band, and living in the same geographical areas, were randomly selected from the same register.
Of the 233 randomly selected women who fulfilled the criteria to be included in the study as cases, 190 (82%) chose to participate. Of those 280 referents who fulfilled the criteria for inclusion, 223 (80%) participated.
Questionnaire and data collection
In order to find contextual factors related to long-term sick-listing and develop a questionnaire for this study, open-ended interviews were initially conducted with 16 women from the study population of this investigation [12]. A questionnaire was subsequently constructed with background questions on age, civil status, number of children, present and earlier occupational positions, and estimated hours spent at work per week.
Questions were included on self-rated physical and mental demands at work, responsibility at work and in work tasks, sense of stimulus from work, and competence for work tasks, the occurrence of bullying, and feedback from supervisors and workmates [13].
Other questions concerned motivation and satisfaction in relation to occupational work, and also sense of demand, influence and control at work, as well as questions on perceived stress at work and in private life [14,15,16,17].
Regarding responsibility for children and taking care of children, and responsibility for domestic work and carrying out domestic work during the years, the following items were used: “Who takes the main responsibility for the children/domestic work?” with 4 response alternatives and “Who carries out the domestic work/childcare at home?” with 4 response alternatives [15].
The questionnaire and a covering letter were sent by post to the study participants and in the case of missing answers the questionnaire was sent once again or a supplementary telephone interview was conducted. The ethical review board at Karolinska Institutet approved the study.
Data analysis
In this study an exposure is a factor that might possibly influence the risk of sick-listing for 90 days or more. Data collection was based on self-reports, and each variable was dichotomized into exposed or low/unexposed in the analysis. The work variables reported by the sick-listed women represent the situation before sick-listing.
Occupational titles were classified into four broadly similar categories according to a national standard for classification of skill levels in occupational life [18], in order to confirm skill levels and make sure cases and controls were comparable. Most of the cases (81%) and controls (79%) were at the highest skill level.
Odds ratios were used to estimate the strength of the association between variables in work and private life and sick-listing ≥90 days. The study population was divided into two age groups, 30 to 42 and 43 to 55 years of age respectively, to investigate potential confounding due to age, and to reflect the impact of age. Analyses were also performed for women with and without children, respectively.
Seven of the significant variables with the highest estimates in the bivarite analyses were then entered into a logistic regression model with long-term sick-listing of 90 days or more as outcome. Age was also included in the model for adjustment purposes. The statistical analyses were performed using the statistical package, SPSS for Windows, version 13.0.
Results
Table I shows that the personal characteristics and occupational positions of the cases and referents were similar. Most of the women were employed in private companies, where most of them had a managerial position, and the second most common was a position as a specialist or researcher in industry. However, a few more of the cases compared with the referents had a non-managerial position, such as expert or researcher. The groups did not differ with regard to earlier occupational situations (Table I).
Sick-listed women with children were more likely to have been working regularly more than 50 hours per week, and reported difficulty in combining occupational work with household work and care of children. They also experienced high demands from family and friends (Table II).
Associations were also found with reports of too high mental strain in work tasks, low job control and low influence at work, and perceiving lack of appreciation from superiors, workmates and clients or patients. Long-term sick-listing was associated with reported bullying, and for women with children this association seemed to be even stronger. Association with a less stimulating job and increasing demands on work performance during the past two years was found, and also an overall dissatisfaction with the workplace. Sick-listed women reported to a greater extent that they would rather have another job (Table II).
The results of the logistic regression show that the strongest factors regarding the present occupational situation were: experiencing too high mental strain in work tasks, and low control and influence at work. There was a significant association with regard to a lack of work–family balance (Table III).
Discussion
The findings in this study indicate that long-term sick-listed female white-collar workers with children tend to live in traditional family relationships, with unequal opportunities. Associations found were: too high work-related mental strain, lack of influence and control at work, and too long working hours. Reported bullying at work, lack of appreciation, and dissatisfaction with working life were also related to long-term sick-listing. This dissatisfaction with working life probably explains why the sick-listed women more often stated they would rather have another job.
The strong association with high mental strain in work tasks demonstrates a failure in the balance between sick-listed women’s resources and perceived strain, probably leading to impaired health and later sickness absence. The association with low control and low influence at work is in concordance with earlier studies showing that low job control, especially in women, is associated with negative stress and long-term sick leave [19, 20, 21].
These results address the classical demand–control model, which states that health and the way in which the individual rates his/her health, are dependent on the interaction between experienced demands and perceived control and decision latitude over one’s work [22]. Deterioration in general health and high demands in private life and at work may underlie the associations found in this study. But it remains to be investigated what actually comes first, as the design of this study is cross-sectional.
The women in this study were exposed to traditionally male-dominated working environments that might induce sex discrimination and role conflicts [23, 24]. Women who break into former male domains might not be promoted, but could rather be opposed. The reported bullying is serious and could be linked with women entering these high-level positions, which until now have been strictly male-dominated. Bullying deeply affects the individual and reflects a poor psychosocial climate at the workplace, as well as insufficient leadership. Apart from being in the minority, which could be stressful enough and induce mental strain, the woman could also be consciously discriminated due to entering these traditionally male worksites and challenging male positions. This could also explain the lack of appreciation which was also more frequently reported by the sick-listed women.
The association with reports of long working hours in paid work could be interpreted in different ways. This might depend on difficulty in performing work tasks due to lack of correct competence, but also to an overambitious approach to work performance. An adverse health status and dysfunction could contribute to less efficient work performance, and the necessity to spend more hours at work in order to cope with it. Sick-listed women with children show an even stronger association, despite the fact that there are family policies with generous and extensive opportunities for childcare and leave in Sweden, which might facilitate combining family life and professional ambitions. To date, this has obviously not enabled Swedish woman to devote themselves to a career without health hazards. The women in the reference group seemed to be in more equal family conditions. Multiple demands from family and work may well entail increased stress and be a challenge to women’s health and well-being, as well as a determinant for long-term sickness absence. This has been indicated in our study, and the results are in concordance with the role stress theory, which has been shown to be more pronounced for women [25, 26, 27].