Elsevier

Social Science & Medicine

Volume 57, Issue 5, September 2003, Pages 807-824
Social Science & Medicine

Job characteristics, physical and psychological symptoms, and social support as antecedents of sickness absence among men and women in the private industrial sector

https://doi.org/10.1016/S0277-9536(02)00450-1Get rights and content

Abstract

Most longitudinal studies on the relationship between psychosocial health resources and risks, and the employees’ subsequent sickness absences have been conducted in the public sector. The purpose of this study was to find out psychosocial antecedents of sickness absenteeism in the private industrial sector. The effects of job characteristics (job autonomy and job complexity), physical and psychological symptoms, and social support (from coworkers and supervisors) on sickness absenteeism were investigated. The number of long (4–21 days) and very long (>21 days) sickness absence episodes of 3895 persons (76% men and 24% women, mean age 44 years) was obtained from the health registers of a multinational forest industry corporation in 1995–1998. A questionnaire survey on the working conditions and health of the workers was carried out in 1996. The follow-up time of the sickness absences was 1-year 9-month. Job autonomy was found to be associated with long and very long episodes in men (rate ratio (RR) in the lowest autonomy group approximately 2 times higher than the highest autonomy group), and with very long episodes of absence in women (2–3 times higher RR between the low vs. the high category). Low job complexity predicted men's very long absences (RR 1.4). Long and very long episodes were associated with physical and psychological symptoms (RR 1.2–1.7) among men and women. Lack of coworkers’ support increased the frequency of very long sickness absence among men (RR 1.4), and lack of supervisor's support among women (RR 1.6). Also, some interaction effects of social support variables were observed among both genders. We conclude that the studied psychosocial factors are associated with subsequent sickness absence, and that the associations are partly gender-specific. The results showing which variables are related to employees’ sickness absenteeism in the private industrial sector can be applied in human resource management and health service planning.

Introduction

Sickness absences have increasingly been a focus of research since the 1970s. The psychosocial literature on absenteeism involves a commonly accepted view that sickness absences are closely related to stressful characteristics of work (e.g. Schechter, Green, Olsen, Kruse, & Cargo, 1997). Some researchers have even estimated that about 60–70% of all sickness absences are associated with stress-related illness (e.g. Kearns, 1986). According to other estimates, an average employee is on sick leave for about 18 months during his/her life solely because of stress-related illnesses (Long, 1991). Furthermore, it has been purported that the work environment can be a better predictor of absenteeism than demographic factors or the psychological characteristics of a person (Farrell & Stamm, 1988). Several studies have concluded that stressful work increases the risk of diseases, e.g., of cardiovascular diseases (Belcastro, Gold, & Grand, 1982; Karasek & Theorell, 1990).

The relation between psychosocial factors of work, physical and psychological distress, and sickness absenteeism (used as an objective measurement of health) has been investigated in large-scale longitudinal studies (e.g. Kivimäki et al., 1997; Kivimäki, Vahtera, Koskenvuo, Uutela, & Pentti, 1998; Kivimäki, Vahtera, Pentti, & Ferrie, 2000; Niedhammer, Bugel, Goldberg, Leclerc, Guéguen, 1998; North et al., 1993; North, Syme, Feeney, Shipley, & Marmot, 1996; Stansfeld, Fuhrer, Head, Ferrie, & Shipley, 1997; Smulders & Nijhuis, 1999; Vahtera, Kivimäki & Pentti, 1997; Vahtera, Kivimäki, Pentti, & Theorell, 2000). The majority of the psychosocially oriented studies on sickness absenteeism have, however, been carried out in the public sector, and only a few have been conducted in the private sector (e.g. Donaldson, Sussman, Dent, Severson, & Stoddard, 1999; Kristensen, 1991). Furthermore, none of the large-scale prospective studies have focused on the private industrial sector, even though it is estimated that stress-related health disorders cause huge individual, social and economic losses in all sectors of work life (Kearns, 1997; Martocchio, 1992). Recent cross-sectional findings indicate that the psychosocial work environment affects absenteeism more strongly than, for example, traditional indicators of health behavior (sleep pattern, regular exercise, etc.) among the private industrial employees (Donaldson et al., 1999). Maes, Verhoeven, Kittel, and Scholten (1998) found in their worksite health program that when decision latitude increased, absenteeism decreased simultaneously in a private industrial enterprise.

The private industrial sector may have unique psychosocial determinants of employee health. For instance, unlike public organizations, private firms more actively encourage their clients to become consumers and more clearly seek a profit. This can heighten employees’ work stress. Furthermore, the psychosocial work environments may differ in that private enterprises place greater emphasis on participation in decision making and obligation towards others while working, and their human resource management practices utilize more aspects such as autonomy, task significance and supervisory control. Moreover, the service sectors and other sectors may differ concerning the psychosocial predictors of objectively measured health (Luz & Green, 1997; North et al., 1996; Smulders & Nijhuis, 1999). For instance, social support did not predict sickness absence among employees working mostly in the service sector (Kivimäki et al., 1997), but it did predict sickness absence among employees mainly in technical and administrative occupations of an electricity company (Niedhammer et al., 1998). Finally, recent theoretical models for health promotion in industry have started to stress the role of the psychosocial work environment in the reduction of absenteeism (Donaldson & Klien, 1997), but large-scale prospective studies are lacking.

In the present study, we first assumed that there are stress-producing risks and stress-protecting resources at work, as well as antecedent indicators of deteriorating health, which are associated with sickness in the course of time. Secondly, we assumed that medically certified absences reflect adequately the health of the employees (Jenkins, 1985; Marmot, Feeney, Shipley, North, & Syme, 1995). Our aim was to investigate whether psychosocial work variables and potential antecedent symptoms of ill health predict subsequent sickness absenteeism in the private industrial sector. In the following we review what is known about the impact of job characteristics, symptoms, work-related social support, and gender based on studies mostly from the public sector, due to lack of information on private industrial enterprises. We will present separately findings from the literature for each psychosocial predictor to corroborate research questions and hypotheses (Fig. 1).

Acute unpleasant life events and chronic psychosocial difficulties have usually been linked with stress (Jenkins, 1991). Different empirical results and theories about occupational stress have regarded job autonomy to be crucial for the health of employees, mainly because greater autonomy is associated with more opportunities to cope with stressful situations (see Jenkins, 1991; Lazarus, 1966). In our study job autonomy means, on the one hand, independence from other workers while carrying out tasks, and on the other, decision latitude concerning one's work pace and phases. In the recent studies on absenteeism, this dimension of work has often been called job control due to the strong influence of Karasek's (1979) work stress model.

Prior research has shown that weak control/autonomy over one's work associates with psychological stress and, if long-lasting, causes health problems (see Ganster & Schaubroeck, 1991; Karasek & Theorell, 1990) and relates to such health indicators as long medically certified sick leaves (North et al., 1993; see also Houtman, Bongers, Smulders, & Kompier, 1994). Recently, many prospective absenteeism studies in the public sector have brought out the importance of job control/autonomy/decision latitude on future health (Kivimäki et al., 1997; Peiró, González-Romá, Lloret, Bravo, & Zurriaga, 1999; Smulders & Nijhuis, 1999, Vahtera et al., 2000).

If we assume that work is one of the central sources of stress and well-being in one's life (Buunk & Ybema, 1997), it is important to study how demanding tasks occurring in today's work life are associated with health. Therefore, job complexity, indicating the level of challenges and variety of tasks at work, was chosen as the second job-related characteristic. Job complexity has some similarities with concepts such as skill requirements and intellectual discretion that have been used to capture the nature of cognitively demanding work (see Houtman et al., 1994; Stansfeld et al., 1997). It can mean interesting opportunities for the worker, and can lead to increasing commitment, higher job satisfaction and fewer intentions to leave the organization (see Björvell & Brodin, 1992).

Job complexity has not been studied in previous studies of absenteeism. However, a positive connection has been found between job complexity and type A behavior related to future cardiovascular health (Schaubroeck, Ganster, & Kemmerer, 1994). In the study made by Melamed, Ben Avi, Luz, and Green (1995) psychological stress was related to task monotony and increased absenteeism in the industrial sector. A high level of, and increase in, skill discretion protected against various types of sickness absences among public service employees (Kivimäki, Vahtera, Ferrie, Hemingway, & Pentti, 2001; Stansfeld et al., 1997). Thus the first research question addressed job characteristics: What is the role of job autonomy and job complexity in sickness absenteeism in the private industrial sector? Our hypothesis was that low job autonomy and low job complexity predict sickness absenteeism.

For a long time, studies on stress have shown the relationship between a stressful life situation and ill health (e.g. Kessler, 1997). The response model of stress suggests that stress first causes various minor physical and psychological symptoms, later more serious physiological and mental responses, and finally illnesses (Skelton & Pennebaker, 1982). Still, usually only subjective reports of symptoms have been used as indicators of health status rather than objective indicators such as sickness absence. Some studies have shown that mental symptoms and emotional distress are related to ill health, indicated by sickness absenteeism (Borgquist, Hansson, Nettelbladt, Nordstrom, & Lindelöw, 1993; Grossi, Soares, Angesleva, & Perski, 1999). Leaning on these theories and results, we expected that physical and psychological symptoms would be associated with sickness absence (see also Rees & Smith, 1991).

In the present study, we investigated the predictive power of psychological and physical symptoms on sickness absences longitudinally. We also wanted to study prospectively the potential deteriorating effect of these symptoms on subsequent health, because in welfare societies, the number of people permitted to retire because of diagnoses associated with psychological and physical symptoms is increasing (Social Insurance Institution, Finland, Statistics, 1997). To our knowledge, only in Whitehall Study II (e.g. North et al., 1996) among civil servants in London, and in a Finnish study (e.g. Kivimäki et al., 1997; Vahtera, Pentti, & Uutela, 1996) among municipal employees, has the relation between these symptoms and sickness absence been investigated prospectively using absence record data as an indicator of more serious illness. These studies have shown positive relations between psychological and physical symptoms and sickness absence. Thus our second research question was: Do psychological and physical symptoms predict sickness absenteeism? The hypothesis was that physical and psychological symptoms increase sickness absences.

Previous investigations of sickness absence point out that studies should focus on social interaction within the work group and the organization (see Anderson, 1991). Normally this aspect of work life has been studied using the concept of social support. Social support predicted absenteeism in the Whitehall Study II (North et al (1993), North, Syme (1996)), in a Swedish Study (Unden, 1996), and in a Finnish Study (Vahtera, Pentti (1996), Vahtera, Kivimäki (2000)). Stansfeld et al. (1997) found that social support inversely predicted short and longer psychiatric absence episodes. However, social support may sometimes be associated with increased absenteeism (Rael, Stansfeld, Shipley, & Head, 1995).

Although the research results on social support are partly contradictory, they mostly show negative relations with sickness absence. It is striking, however, that the operationalization of social support normally includes all sources of support under the same concept, although the type of support (Unden, 1996) and from whom it comes (Dean, Kolody, & Wood, 1990) can be a critical factor in absenteeism. In the present study, we included belonging support, instrumental support and a supportive work atmosphere (see House, 1981) in our measure of social support, because of their potential relevance to health (Unden, 1996). We used two social support sources, coworkers’ support and supervisor's support, as separate predictors of sickness absence.

We treated the support factors also as moderators, because several studies suggest that the effect of social support on health is not direct, but is interactive with stress and job characteristics (e.g. Karasek & Theorell, 1990; Vahtera et al., 1996). The stress-buffering hypothesis states that social support protects employees from the pathological consequences of stressful experiences (Cohen & Wills, 1985). Numerous studies have supported this view (House, Umberson, & Landis, 1988; Parkes, Mendham, & Von Rabenau, 1994; Ulleberg & Rundmo, 1997). Roberts, Cox, Shannon, and Wells (1994) have suggested that different sources of support can moderate the effects of stress on well-being in a different manner. However, only a few combined effects have so far been studied in prospective absence studies. They have shown that the highest risk of sick leave is found with the combined effects of poor job control, and negative changes in job control and low social support (Vahtera et al., 2000). Thus the third research question was: Does support from the supervisor and coworkers have a direct effect on sickness absence? Our hypothesis was that low levels of social support are associated with sickness absence. A related research question was: Do sources of support moderate the effect of job characteristics and of symptoms on sickness absenteeism? Our hypothesis was that high levels of social support buffer the negative effects of job characteristics and symptoms on sickness absenteeism.

During the past two decades, the interest in factors related to women's health has grown. Studies have shown that women usually have poorer self-rated health than men, as measured by the number of reported symptoms (Krantz & Östergren, 2000). Interestingly, the sources of stress seem to vary between women and men (Hendrix, Spencer, & Gibson, 1994). Women seem to have less coping resources in stressful situations due to their various role demands (see Jenkins, 1991; Kushnir & Kasan, 1992/93), and this seems to affect their health (Dixon, Dixon, & Spinner, 1991). Some researchers have suggested that there are basic structural differences underlying health differences between the genders, and that they are partly psychosocial (Van den Heuvel & Wooden, 1995). Separate models for both genders should therefore be created, or the genders should be investigated separately, in order to study the psychosocial construction of health adequately (Hendrix et al., 1994; Luz & Green, 1997).

The gender-specific approach has been adopted when investigating the effects of psychosocial factors on absenteeism in a few large-scale follow-up studies in the public sector (e.g., Kivimäki et al., 1997; Niedhammer et al., 1998; North et al (1993), North, Syme (1996); Vahtera et al., 1996). The findings on gender differences have nevertheless been partly contradictory (North et al., 1993; Vahtera et al., 1996). In the Whitehall Study II among civil servants in London, a low level of work control, work demands and social support was associated with higher rates of long sick leaves, especially among men (North et al., 1993; Rael et al., 1995). In line with these results, a French study showed that a low level of coworkers’ support increased the number of sick leaves only in men (Niedhammer et al., 1998). It has also been reported that supervisor's support is a strong protector mainly against women's absenteeism in some service sector occupations (Schokking-Siegrist, 1981). The results of the Finnish study (Vahtera et al., 1996) among municipal employees indicated that the connections were more complex, depending on the demand-control interaction, gender, and other psychosocial factors at work. Finally, many cross-sectional studies have pointed out that different sources of social support may moderate the association between stress and self-rated health differently in men and in women (Abdalla, 1991; Wolgemuth & Betz, 1991). These previous results have been contradictory to some extent, but particularly interesting from the point of view of our study. We therefore separated male and female employees in the analyses in order to find out common and gender-specific features in their predictors of sickness absence. Thus, the research question on gender was: Do the variables of interest affect sickness absence in men and in women? Our hypotheses was that psychosocial factors predict sickness absence in both genders, but some work-related predictors have a stronger effect on male employees, and the sources of social support may function differently in men and women.

Section snippets

Study design and collection of data

This was a prospective follow-up study on psychosocial predictors of sickness absence in an industrial corporation. It is based on a questionnaire survey on health and the work environment, and on sickness absence records.

Questionnaires were sent to the work units during the spring and summer of 1996, and distributed by the supervisors. All employees, domestic and non-domestic, were allowed to fill out the questionnaire at that time. At the same time, permission for collecting and using the

Descriptive results

The total number of long sick leaves (4–21days) among men was 997, and of very long sick leaves (21< days) 293, and among women 388 and 141, respectively. On average, men had 25 and women had 32 long and very long sick leaves per 100 person-years. On average male employees had 8 days of absence, and female employees 12 days of absence per year. Overall 60% of participants did not have sick leaves at all (63% of men and 54% of women) and only 11% had more than two sick leaves during the 1-year

Discussion

In this study the predictive effects of psychosocial work characteristics, social support resources, and physical and psychological symptoms on subsequent sickness absenteeism were studied in a private industrial company. Comparison of women and men showed that women had more sick leaves as noted in previous studies (Alexanderson, Leijon, Åkerlind, Rydh, & Bjurulf, 1994; Hendrix et al., 1994; Jacobson, Aldana, Goetzel, & Vardell, 1996; North et al., 1996). Our results regarding the psychosocial

Acknowledgements

This study was financially supported by the University of Tampere, the Academy of Finland (grant #36281, and project #77560), the Finnish Work Environmental Fund (grant #10014), the Finnish Work Health Developmental Foundation, the FIOH and Enso Oyj.

References (85)

  • P.A. Belcastro et al.

    Stress and burnoutPhysiologic effects on correctional teachers

    Criminal Justice and Behaviour

    (1982)
  • H. Björvell et al.

    Hospital staff members are satisfied with their jobs

    Scandinavian Journal of Caring Sciences

    (1992)
  • L. Borgquist et al.

    Perceived health and high consumers of careA study of mental health problems in a Swedish primary health care district

    Psychological Medicine

    (1993)
  • Buunk, B. P., & Ybema, J. F. (1997). Social comparisons and occupational stress: The identification-contrast model. In...
  • S. Cohen et al.

    Stress, social support, and the buffering hypothesis

    Psychological Bulletin

    (1985)
  • A. Dean et al.

    Effects of social support from various sources on depression in elderly persons

    Journal of Health and Social Behavior

    (1990)
  • J.P. Dixon et al.

    Tensions between career and interpersonal commitments as a risk factor for cardiovascular disease among women

    Women's Health

    (1991)
  • S.I. Donaldson et al.

    Creating healthful work environments for ethnically diverse employees working in small and medium-sized businessesA non-profit industry/community/university collaboration model

    Employee Assistance Quarterly

    (1997)
  • S.I. Donaldson et al.

    Health behavior, quality of work life, and organizational effectiveness in the lumber industry

    Health Education Behavior

    (1999)
  • Elo, A. -L. (1994). Assessment of mental stress factors at work. In: C. Zent, O. B. Dickerson, & E. P. Howarth (Eds.)...
  • Elo, A. -L., Leppänen, A., Lindström, K., & Roponen, T. (1992). Occupational stress questionnaire: User's instructions....
  • D. Farrell et al.

    Meta-analysis of the correlates of employee absence

    Human Relations

    (1988)
  • D.C. Ganster et al.

    Work stress and employee health

    Journal of Management

    (1991)
  • S.A. Geurts et al.

    Social comparisons and absenteeismA structural modeling approach

    Journal of Applied Social Psychology

    (1994)
  • D.P. Goldberg et al.

    A scaled version of a General Health Questionnaire

    Psychological Medicine

    (1979)
  • G. Grossi et al.

    Psychosocial correlates of long-term sick-leave among patients with musculoskeletal pain

    Pain

    (1999)
  • R.D. Hackett et al.

    An evaluation of employee absenteeism as coping mechanism among hospital nurses

    Journal of Occupational and Organizational Psychology

    (1996)
  • W.H. Hendrix et al.

    Organizational and extraorganizational factors affecting stress, employee well-being, and absenteeism for males and females

    Journal of Business and Psychology

    (1994)
  • W.A. Hochwarter et al.

    Gender differences in perceptions of stress-related variablesDo the people make the place or does the place make the people?

    Journal of Managerial Issues

    (1995)
  • House, J. S. (1981). Work, stress, and social support. Reading, MA:...
  • J.S. House et al.

    Structures and processes of social support

    Annual Review of Sociology

    (1988)
  • I.L.D. Houtman et al.

    Psychosocial stressors at work and musculoskeletal problems

    Scandinavian Journal of Work, Environment and Health

    (1994)
  • P. Huuhtanen et al.

    Ageing workers and changing working life

    (1999)
  • B.H. Jacobson et al.

    The relationship between perceived stress and self-reported illness-related absenteeism

    American Journal of Health Promotion

    (1996)
  • R. Jenkins

    Minor psychiatric morbidity in employed young men and women and its contribution to sickness absence

    British Journal of Industrial Medicine

    (1985)
  • Jenkins, R. (1991). Demographic aspects of stress. In C. L. Cooper, & R. Payne (Eds.), Personality and stress:...
  • Kalimo, R., & Toppinen, S. (1999). Finland: Organizational well-being: Ten years of research and development: in a...
  • R. Karasek

    Job demands, job decision latitude and mental strainImplications for job redesign

    Administrative Science Quarterly

    (1979)
  • R.A. Karasek et al.

    Healthy work, stress, productivity and the reconstruction of work life

    (1990)
  • J. Kearns

    Stress at workThe challenge of change

    (1986)
  • W.L. Kearns

    Cash benefits for short-term sickness, 1970-94

    Social Security Bulletin

    (1997)
  • R. Kessler

    The effects of stressful life events on depression

    Annual Review of Psychology

    (1997)
  • Cited by (0)

    View full text