1134

Sick Leave and Factors at Work and Outside Work

•

Voss et al

CME Available for this Article at ACOEM.org

How Do Job Characteristics, Family Situation, Domestic Work, and Lifestyle Factors Relate to Sickness Absence? A Study Based on Sweden Post Margaretha Voss, PhD Birgitta Floderus, PhD, Finn Diderichsen, MD, PhD Learning Objectives

T

• Recall how the risk of sickness absence, adjusted for age, was associated with work-related symptoms or complaints in this study population with work tasks ranging from clerical to heavy manual work. • Describe the observed associations of sickness absence with family matters, domestic work, and financial status. • Explain how sickness absence is a function of life style factors including substance use in men and women.

Abstract Objective: We sought to determine how do working life and private life characteristics relate to sickness absence. Methods: Questionnaire data for 1557 female and 1913 male employees were related to registered sickness absence by multivariate analyses. Results: Apart from health problems, clear associations with sickness absence were observed for complaints attributed to heavy, arduous work, and sickness presenteeism. The use of tranquilizers, occurrence of bullying, and the existence of a high total workload from paid and unpaid work were associated with sickness absence in women. In men, the use of alcohol as sedative, anxiety of reorganization, not holding a supervisor position, adverse life events, and divorce were related with sickness absence. Conclusions: Our study identified different areas at work and outside work for women and men that could be targets for actions aiming at preventing high sickness absence. (J Occup Environ Med. 2004;46:1134 –1143) From the Institute of Environmental Medicine, Karolinska Institutet, Stockholm, Sweden (Drs Voss, Floderus); National Institute for Working Life, Stockholm, Sweden (Drs Voss, Floderus); Department of Clinical Neuroscience, Karolinska Institutet, Stockholm, Sweden (Dr Voss); Department of Public Health Sciences, Karolinska Institutet, Stockholm, Sweden (Drs Floderus, Diderichsen); and Institute of Public Health, University of Copenhagen, Copenhagen, Denmark (Diderichsen). Margaretha Voss has no commercial interests related to this article. Address correspondence to: Margaretha Voss, Section of Personal Injury Prevention, Department of Clinical Neuroscience, Karolinska Institutet, Box 12718, SE-112 94 Stockholm, Sweden; E-mail: [email protected]. Copyright © by American College of Occupational and Environmental Medicine DOI: 10.1097/01.jom.0000145433.65697.8d

here has been a marked increase in sickness absence in Sweden since 1997. The public expenditure for sick-leave compensation regarding leaves of more than 14 days has increased from 1.5 billion Euros in 1997 to 4.5 billion Euros in 2003. During the same period, the number of sick leave days more than doubled for both women and men—to 68 million days for women and 40 million days for men.1 Sickness absence is determined not only by health conditions but also by general regulations and other societal conditions, the labor market, the work environment, the social situation, and individual characteristics.2–7 Several studies have shown that occupational factors, such as heavy, physically arduous, or monotonous work, may influence the worker’s health and absenteeism.2,8 –11 The importance of the work organization and the psychosocial work environment, including, for example, bullying at the workplace, also have been described.12–17 To understand the mechanisms behind absenteeism, it is important to take into account not only workrelated factors but also the individual’s situation outside work, such as family life, that is, marital status and children.2,18,19 Combining a gainful employment and parenthood may sometimes lead to a high total strain, especially in women, which may cause role conflicts and work overload, leading to an increased risk of

JOEM

•

Volume 46, Number 11, November 2004

1135

TABLE 1 Age Distribution and Incidence of Sickness Absence for Women and Men of the Total Study Population and for Respondents of the Questionnaire Study population

Mean age Median age Sickness incidence†

Respondents*

Women (No.)

Men (No.)

Women (No./%)

Men (No./%)

1557 42.8 44 52.2

1913 39.5 39 46.1

1226/78.7 42.9 45 49.8

1410/73.7 39.5 39 41.4

* 8 individuals lacking information about days at risk were excluded. † Crude sickness incidence per 10 000 days at risk.

sonal identification numbers. The questionnaire covered conditions during 1993, including the physical, psychosocial, and organizational environment at the workplace as well as family situation, financial situation, lifestyle factors, health and well-being, and leisure time activities. When available, we used items or scales developed and evaluated in previous studies.34,35

Independent Variables absenteeism.20 Multiple roles may, however, also imply a richer life, promoting health and well-being, that is, role enhancement.18 –22 In some studies, it has been pointed out that both these effects probably exist and that they tend to counterbalance each other.18,21 Experiences of adverse life events, including divorce, and the individual’s health-related behavior, including smoking and use of alcohol and other drugs, also may be followed by increased sickness absence.2,11,23–26 Sickness absence is to some extent a conscious choice of the employee, a coping strategy to handle reduced work ability caused by illness, adverse situations at work, or difficulties attributed to private life.27 Sickness absence may occasionally cover absence without diagnosed or selfperceived illness,28 but people may also be working instead of taking justified sick leave (sickness presenteeism).29,30 The aim of the present study was to give a multifactorial background to sickness absence. The impact of different features of the physical and psychosocial work environment as well as work organization was studied previously.30 These results have been integrated in the present study of factors outside work, including family situation, children, domestic workload, economy, and lifestyle. The study population includes a large variety of work tasks, clerical jobs as well as heavy manual work tasks, and a comparable representation of women and men. The knowl-

edge about the relations between work-related factors, conditions outside work, and sickness absence in women and men is limited and sometimes restricted to specific occupational groups such as nurses.5,11,31–33

Materials and Methods Participants The study population consisted of all employees in three of Sweden Post’s regional organizations in the northern part of Sweden: Letter Division Sundsvall, Letter Division Umeå, and Sales Division Sundsvall. All persons on regular employment contracts (full-time or part-time) since July 1, 1992, who were employed for at least 6 months during 1993 and who were still employed by the same division in September 1994 were included. The subjects, 1557 women and 1913 men, represented a number of different work tasks: administrators, cashiers, mail handling staff, rural mail carriers, office personnel, office-cleaning staff, computer personnel, and technicians. A questionnaire was mailed to all individuals in the autumn of 1994. The response rate was 76%, with a higher proportion among women (79%) than among men (74%). The results of the study are based on data obtained from 1219 women and 1409 men. The study population and the respondents of the questionnaire are described in Table 1. The questionnaire data were linked to information on sickness absence by per-

Health and well-being. Health variables comprised long-lasting, serious illness, and sick leave on more than two occasions in 1993 because of cold. Work-related symptoms or complaints included asthma, migraine, headache, eye discomfort, skin disorders, tiredness/exhaustion, uneasiness at work, sleeping problems, and musculoskeletal disorders. Work-Related Factors. Workrelated factors included information about the workplace (eg, size, proportion of women and men), environmental physical factors and work load, working hours, work-mates and managers, and work organization. The separate results for work-related factors, as well as definitions, have been presented elsewhere,30 and are not described in the present report. Family, domestic work, and financial situation. These included marital status (ie, unmarried/single or divorced vs. married/cohabitant); responsibility for the child/children (ie, staying at home in case of sick children, taking children to day-care/school and to leisure time activities, help with homework, possibility to get help from others taking care of sick children); participation in domestic work (ie, regular care of relatives/close friends; everyday domestic work, that is, shopping, cooking, washing, cleaning; household economy; gardening; and maintenance of the car and the house/ apartment); work situation for spouse/ cohabitant (ie, employment, reason for no employment, full-time or part-time work and salary); adverse life events (ie, relative’s serious disease, accident or death, miscarriage); and financial

1136

situation (ie, difficulties in making ends meet, inability to raise 1100 Euros within a week). Lifestyle Factors. These included medication and use of drugs (ie, use of vitamins and medicines [analgesics, tranquilizers, and soporifics] and use of alcohol and tobacco); and leisure time activities, sleeping and eating habits (ie, indoor and outdoor activities, exercise, hours of sleep, and consumption of fruits, vegetables, and cooked food).

Outcome Measure Incidence of Sickness Absence. Information about sickness absence during 1993 was obtained from Sweden Post’s register of absenteeism. For each individual the number of days at risk of sickness absence was computed. Days at risk were determined by taking the number of calendar days in the year and subtracting the number of days absent from work (days without employment, leaves of absence, and sickness absence). Incidence of sickness absence was expressed as the number of spells of sick leave per 10,000 days at risk. Recurrent sickness absence within 5 days from the end of the previous spell was handled as one and the same spell. Low and High Sickness Absence. In the analyses, we distinguished between “low” and a “high” sickness absence. The low sickness absence group (40%) comprised people with no sickness absence (30% among women, 34% among men) or less than 55.25 sickness spells per 10 000 days at risk corresponding to, approximately, less than two spells per year. The high sickness absence group (60%) comprised people with more than 55.25 sickness spells per 10 000 days at risk.

Statistical Analyses Each variable was dichotomized into exposed/unexposed. Sickness absence was analyzed in relation to all variables mentioned above adjusting for age. As a second step, multivariate analyses were performed sep-

Sick Leave and Factors at Work and Outside Work

arately for variables within the categories of: (1) family, domestic work, and financial situation and (2) lifestyle factors. As a third step, all variables showing a statistically significant association (or boarder line significance, P 聿 0.05) with high sickness absence or an odds ratio ⱖ2.0 were included in a final model. In the final model the work-related factors previously identified as potential predictors for high sickness absence according to a similar procedure were included as well. In addition, we extended the analyses of the importance of domestic workload on sickness absence among married/cohabitant women, by stratification on prevalence of children at home, strong/weak financial situation, and full-time/part-time work. High domestic workload meant “total” responsibility for the everyday domestic work. The Statistical Analysis System SAS 6.1236 was used throughout and the multivariate analyses were performed by logistic regression (LOGISTIC), yielding odds ratios (ORs) with 95% confidence intervals (95% CIs). For work-related symptoms/disorders, we estimated the attributable fraction (AF) as a measure of the proportion of cases of high sickness absence that could (partly) be explained by these symptoms given a causal association.37

Results Incidence of Sickness Absence The incidence of sickness absence in the study population during 1993 was 52 cases per 10,000 days at risk in women and 46 cases per 10,000 days at risk in men. The incidence was slightly higher in the total study population compared to those who answered the questionnaire (Table 1). High Sickness Absence. In the multivariate analyses, men older than 30 years of age showed a decreased risk of high sickness absence compared with the youngest age group (⬍31 years of age; bottom part of Table 5). For women the multivariate analyses

•

Voss et al

of the age groups did not show any association with sickness absence.

Work-Related Symptoms or Complaints Individuals reporting work-related symptoms and disorders showed, in general, an increased risk of sickness absence. Adjusting for age, particularly strong associations were found for staying at home attributed to unspecified uneasiness at work, with an OR⫽ 3.9 (95% CI ⫽ 1.6 –9.4) for women and 5.9 (2.6 –13.2) for men (Table 2). Asthma, perceived as work related, also showed a strong relation to sickness absence for both genders. For both these variables, however, the number of individuals reporting the trait was small, and in terms of the AF, taking the prevalence into account, other variables were of greater significance. These were physically exhausted after work (AF 26% for women and 28% for men), too tired to do anything after work (AF 24% for women and 27% for men), neck–shoulder complaints (AF 25% for women and 17% for men), and back complaints (AF 16% for women and men). The proportions of the study population reporting these problems were between 53% and 72%. Work-related headache showed an increased risk of high sickness absence for both women and men. Taking into account the prevalence of headache, the AF was estimated at 18% in women compared with 8% in men. The same tendency of a gender difference was also found for workrelated hand–arm– elbow complaints and psychologic exhaustion.

Family, Domestic Work, and Financial Situation: Results Adjusted for Resembling Variables and Age A weak financial situation implied a 30% increased occurrence of high sickness absence in women and a 50% increase in men after adjustment for other variables pertaining to family and domestic work (Table 3).

JOEM

•

Volume 46, Number 11, November 2004

1137

TABLE 2 Odds Ratio and Attributable Fraction for Sickness Absence Among Subjects Reporting Certain Work Related Symptoms and Disorders Women*

Variables Often/sometimes asthma Often/sometimes migraine Often/sometimes headache Often/sometimes abnormal tiredness Often/sometimes sleeping problems Often/sometimes eye discomfort Often/sometimes skin problems Now and then neck-shoulder complaints Now and then hand-arm-elbow complaints Now and then back complaints Now and then feet-knee-hip complaints Often/sometimes psychologically exhausted after work Often/sometimes physically exhausted after work Often/sometimes tired—not able to do anything after work Stayed at home due to uneasiness at work

Proportion (%) of exposed

No. of exposed cases

Age adjusted OR (95% CI)†

2 9 36 30

21 50 224 186

19

Men*

AF**

Proportion (%) of exposed

No. of exposed cases

Age adjusted OR (95% CI)†

AF**

3.2 (1.5– 6.9) 1.1 (0.8 –1.7) 1.7 (1.4 –2.2) 1.6 (1.3–2.1)

0.03 0.01 0.18 0.13

3 6 25 28

23 39 155 175

2.4 (1.3– 4.6) 1.3 (0.8 –2.1) 1.4 (1.1–1.9) 1.4 (1.1–1.8)

0.03 0.02 0.08 0.09

110

1.4 (1.0 –1.8)

0.06

20

116

1.3 (1.0 –1.7)

0.05

24 19 61

145 119 347

1.5 (1.1–1.9) 1.5 (1.1–2.0) 1.6 (1.2–2.0)

0.09 0.08 0.25

15 16 53

100 100 315

1.6 (1.2–2.1) 1.4 (1.0 –1.8) 1.4 (1.1–1.7)

0.07 0.05 0.17

29

175

1.5 (1.2–1.9)

0.11

21

122

1.3 (1.0 –1.7)

0.05

35 20

218 120

1.6 (1.3–2.1) 1.4 (1.0 –1.8)

0.16 0.07

42 28

262 169

1.5 (1.2–1.8) 1.3 (1.0 –1.6)

0.16 0.07

61

331

1.2 (0.9 –1.5)

0.11

55

306

1.1 (0.9 –1.4)

0.05

72

401

1.5 (1.1–1.9)

0.26

68

401

1.6 (1.3–2.1)

0.28

68

382

1.5 (1.2–1.9)

0.24

57

352

1.7 (1.4 –2.1)

0.27

2

18

3.9 (1.6 –9.4)

0.03

2

23

5.9 (2.6 –13.2)

0.04

* 1219 women of which 522 are cases; 1409 men of which 536 are cases. ** Attributable fraction, if the relation reflects causality. † Odds ratio with 95% confidence interval.

For men, experience of a relative’s serious sickness, accident, or death during 1993 was associated with a doubled probability of being in the high sickness absence group. A 30% reduced risk of high sickness absence was observed for both women and men who had never been at home taking care of sick children during the year. Possibilities of getting help from others taking care of sick children also implied a reduced absence among women. Furthermore, being responsible for the largest part of the everyday domestic work showed a reduced risk of sickness absence among women.

Lifestyle Factors: Results Adjusted for Resembling Variables and Age Among lifestyle variables, a substantially higher risk of sickness ab-

sence was observed for both women and men using tranquilizers with a more than doubled OR (Table 4). Other variables associated with an increased risk of high sickness absence in both genders were use of tobacco and use of analgesics. Among men, vitamins and use of alcohol as sedatives showed increased OR. A protective effect was suggested for women reporting outdoor leisure time activities at least twice a week.

Factors at Work and Outside Work: Final Multivariate Model Table 5 shows the variables of highest significance according to the gender-specific final models. Both women and men who had stayed at home more than twice during 1993 because of cold had a fourfold higher probability of being in the high sick-

ness absence group when adjusting for all other variables included in the model. Long-lasting or serious illness more than doubled the probability of being in the high sickness absence group in women and also in men. Other factors associated with high sickness absence in both women and men were complaints attributed to heavy lifting at work and working instead of taking sick leave. More than 50 hours of overtime work during 1993 showed a decreased probability of high sickness absence regardless of gender. In women, the strongest association with high sickness absence, when controlling for all other variables, was found for use of tranquilizers, with a fourfold increase in risk (Table 5). Complaints attributable to work in a forward-bent position and

1138

Sick Leave and Factors at Work and Outside Work

•

Voss et al

TABLE 3 Family, Domestic Work and Financial Situation Relative to Sickness Absence Women

Variables Family Spouse/cohabitant is not gainfully employed Spouse/cohabitant works 40 hours or more a week Spouse/cohabitant earns more money Spouse/cohabitant has a more arduous work Some in the family run a farm or a company Regularly taking care of relatives/close friends Experience of a relative’s serious sickness, accident or death Staying at home in case of sick children: all/the greater part Taking care of the children: all/ almost all for 3 out of 4 duties Can get help from others in case of sick children Own children sicker than other children Stayed at home at least 3 times with sick children in 1993 Stayed at home at least 6 days with sick children in 1993 Became under stress when taking care of sick children Became unaffected when taking care of sick children Needed the recovery from staying at home with sick children I was never at home taking care of sick child/children Domestic work Everyday domestic work: at least 3 out of 5 duties Other domestic work: at least 3 out of 4 duties Financial situation Difficulties in making ends meet Can not raise 10,000 Swedish Crowns within a week

Men

Total No. of Total No. of no. of % exp. Age adjusted Model I* OR no. of % exp. Age adjusted Model I* OR exp. exp. cases OR (95% CI)† (95% CI)† exp. exp. cases OR (95% CI)† (95% CI)† 195

16

80

0.9 (0.6 –1.2)

199

14

85

1.1 (0.8 –1.5)

708

58

298

1.0 (0.8 –1.2)

454

32

164

1.0 (0.8 –1.3)

553

45

217

0.8 (0.6 –1.0)

69

5

29

1.4 (0.9 –2.3)

233

19

88

0.8 (0.6 –1.1)

162

11

58

1.0 (0.7–1.4)

165

14

74

1.1 (0.8 –1.6)

125

9

45

1.0 (0.7–1.5)

161

13

60

0.8 (0.6 –1.1)

144

10

60

1.3 (0.9 –1.9)

55

5

26

1.2 (0.7–2.1)

56

4

30

2.2 (1.3–3.6)

174

14

84

1.3 (0.9 –1.8)

39

3

16

1.1 (0.6 –2.1)

105

9

49

1.2 (0.8 –1.8)

14

1

6

1.2 (0.4 –3.5)

191

16

69

0.7 (0.5– 0.9)

273

19

99

0.8 (0.6 –1.1)

27

2

16

1.8 (0.9 – 4.0)

30

2

14

1.4 (0.7–2.8)

123

10

60

1.3 (0.9 –1.9)

141

10

67

1.4 (1.0 –2.0)

106

9

57

1.6 (1.1–2.5)

110

8

48

1.1 (0.8 –1.7)

122

10

49

0.8 (0.6 –1.3)

114

8

56

1.5 (1.0 –2.2)

147

12

57

0.8 (0.5–1.1)

272

19

111

1.1 (0.8 –1.4)

38

3

25

2.8 (1.4 –5.5)

2.1 (1.0 – 4.3)

51

4

26

1.7 (1.0 –3.0)

1.4 (0.8 –2.6)

195

16

61

0.6 (0.4 – 0.8)

0.7 (0.5– 0.9)

236

17

68

0.7 (0.5–1.0)

0.7 (0.5–1.0)

817

67

323

0.7 (0.5– 0.9)

0.7 (0.6 –1.0)

90

6

35

1.0 (0.6 –1.5)

66

5

29

1.1 (0.6 –1.8)

659

47

238

0.9 (0.7–1.2)

346 222

28 18

172 114

1.5 (1.2–1.9) 1.5 (1.1–2.0)

492 261

35 19

226 121

1.6 (1.3–2.1) 1.5 (1.1–1.9)

0.9 (0.7–1.2)

0.7 (0.5–1.0)

1.4 (0.9 –2.2)

1.3 (1.0 –1.8) 1.2 (0.8 –1.7)

2.1 (1.2–3.6)

1.1 (0.8 –1.7)

1.2 (0.8 –1.8)

1.5 (1.2–1.9) 1.2 (0.9 –1.6)

* Bold numbers indicate variables included in the subsequent final model. † Odds ratio with 95% confidence interval.

need for recovery from staying at home with sick children were variables that more than doubled the risk. Other variables of significance were occurrence of bullying, a workplace with more than 50 employees, few or no discussions with the supervisor, social contacts through active partic-

ipation in trade union work, and being a regular smoker or snuffer. Reduced risks of high sickness absence were found for women who had never been at home taking care of sick children during the year, women reporting few or no information meetings, and for women with

work-related contacts outside of the workplace. For men, anxiety about reorganization of the workplace, adverse life events, use of alcohol as sedatives, and divorce doubled the risk of being classified with high sickness absence (Table 5). Increased risks were also

JOEM

•

Volume 46, Number 11, November 2004

1139

TABLE 4 Life Style Factors Relative to Sickness Absence Women

Variables Medication and use of drugs Often or sometimes use of vitamins Often or sometimes use of restoratives Often or sometimes use of analgesics Often, sometimes or seldom use of soporifics Often, sometimes or seldom use of tranquilizers Often or sometimes use of alcohol as sedative Regular use of tobacco during 1990 –94 Leisure time activities, sleep and food habits Less than 1 hour per day for own activities Watch TV more than 8 hours a week Often/sometimes 2 or more outdoor leisure time activities Often/sometimes 3 or more indoor leisure time activities No regularly exercise 4 – 6 hours sleep per night Never or seldom enough sleep Consume fruits and vegetables ⬎ 3 times a week

Total No. of no. of % exp Age adjusted exp. exp. cases OR (95% CI)†

Men Total No. of Model I* OR no. of % exp Age adjusted (95% CI)† exp. exp. cases OR (95% CI)†

Model I* OR (95% CI)†

660

54

300

1.3 (1.0 –1.6)

1.2 (0.9 –1.5)

491

35

215

1.5 (1.2–1.8)

1.3 (1.0 –1.7)

208

17

101

1.4 (1.0 –1.9)

1.2 (0.8 –1.6)

86

6

39

1.5 (1.0 –2.4)

1.2 (0.8 –2.0)

686

56

323

1.5 (1.2–1.9)

1.4 (1.1–1.8)

459

33

209

1.6 (1.3–2.1)

1.5 (1.2–1.9)

44

4

29

2.8 (1.5–5.1)

1.8 (0.9 –3.6)

36

3

22

3.2 (1.7– 6.3)

1.7 (0.8 –3.7)

25

2

18

3.6 (1.6 – 8.2)

2.4 (0.9 – 6.1)

32

2

20

3.3 (1.7– 6.5)

2.2 (1.0 –5.1)

22

2

10

1.2 (0.5–2.8)

81

6

46

2.4 (1.5–3.8)

1.8 (1.1–2.9)

370

30

174

1.3 (1.0 –1.7)

572

41

837

1.3 (1.0 –1.6)

1.3 (1.0 –1.6)

250

21

110

1.1 (0.8 –1.4)

176

12

72

1.2 (0.8 –1.6)

632

52

276

1.1 (0.9 –1.4)

941

67

361

1.0 (0.8 –1.3)

504

41

196

0.7 (0.6 –1.0)

668

47

242

1.0 (0.8 –1.2)

470

39

211

1.2 (0.9 –1.5)

363

26

136

0.9 (0.7–1.2)

360 349 99 195

30 29 8 16

156 164 53 87

1.0 (0.8 –1.3) 1.2 (1.0 –1.6) 1.6 (1.0 –2.4) 1.1 (0.8 –1.5)

448 572 156 467

32 41 11 33

160 242 76 187

0.9 (0.7–1.1) 1.3 (1.0 –1.6) 1.5 (1.1–2.1) 1.0 (0.8 –1.3)

1.3 (1.0 –1.7)

0.8 (0.6 –1.0)

1.1 (0.8 –1.4) 1.3 (0.8 –2.0)

1.2 (0.9 –1.5) 1.3 (0.9 –1.8)

* Bold numbers indicate variables included in the subsequent final model. † Odds ratio with 95% confidence interval.

found for men who did not hold a supervisor position, with fellowworkers or temporary employees taking care of the duties in case of sickness absence, with complaints about noise, and men who desired flexible working hours. Difficulties in making ends meet showed a weak association with sickness absence. Men reporting anxiety about downsizing had a reduced risk of sickness absence.

High Domestic Workload The relation between a high domestic workload and sickness absence was analyzed within specific subgroups among married/cohabitant women (Table 6). Overall, a high domestic workload was not associ-

ated with high sickness absence. Among married/cohabitant women with children at home, a high domestic workload increased the risk by 60%. When also stratifying for working hours, high domestic workload in women also working 40 hours or more per week showed a 130% increase in risk of high sickness absence. Furthermore, a high domestic workload implied an increased risk among married/cohabitant women with children at home and with no financial difficulties. Among women with no children at home, a high domestic work load was associated with low sickness absence, and for the subgroup of women working less than 40 hours/wk the odds ratio was equally low.

Discussion The results showed that apart from health problems both work-related and nonwork-related variables were associated with high sickness absence. Some results pertained to both women and men, and some were gender specific. In women, complaints attributed to work in a forward bent position, use of tranquilizers, and need for recovery from staying at home with sick children showed the strongest associations. In men, anxiety about reorganization of the workplace, use of alcohol as sedative, experience of a relative’s serious sickness, accident or death, and being divorced were associated with high sickness absence.

1140

Sick Leave and Factors at Work and Outside Work

•

Voss et al

TABLE 5 Work Related and Nonwork Related Factors Relative to Sickness Absence Variables Family, domestic work and financial situation Needed the recovery from staying at home with sick children I was never at home taking care of sick child/children Experience of a relative’s serious sickness, accident or death Difficulties in making ends meet Life style factors Regular use of tobacco during 1990 –1994 Often, sometimes or seldom use of tranquilizer Often or sometimes use of alcohol as sedative The workplace and work organization Workplace ⬎50 employees ⬎50 hours overtime-work during 1993 Desire to have flexible working hours Sick-leave: duties are carried out by fellow-workers Sick-leave: duties are carried out by temporary employees Often anxiety about reorganization of the workplace Often anxiety about downsizing at the workplace Physical work load and noise Complaints due to heavy lifting at work Complaints due to work in a forward-bent position Complaints due to monotonous movements Noise Interactions with work-mates and managers Seldom/never discussion with the supervisor Seldom/never information meetings Social contacts by active participation in trade union work Work related contacts outside the workplace Occurrence of bullying at the workplace No supervisor position Health and wellbeing Long-lasting or serious illness Stayed at home more than twice during 1993 due to cold Have been working instead of taking sick-leave Civil status Unmarried/single Divorced Age 31– 40 years old 41–50 years old ⬎50 years old

Women

Men

Final model OR (95% CI)*

Final model OR (95% CI)*

2.3 (1.0 –5.3) 0.6 (0.4 – 0.9) 2.1 (1.1– 4.0) 1.3 (1.0 –1.6) 1.4 (1.0 –1.8) 4.1 (1.3–13.3) 2.1 (1.3–3.7) 1.5 (1.1–2.2) 0.6 (0.5– 0.9)

1.8 (1.3–2.5) 2.4 (1.5– 4.0)

0.7 (0.5– 0.9) 1.4 (1.1–1.9) 1.5 (1.1–2.0) 1.6 (1.2–2.2) 2.1 (1.4 –3.0) 0.6 (0.5– 0.9) 1.6 (1.1–2.3) 1.3 (1.0 –1.8) 1.5 (1.1–1.9)

1.3 (1.0 –1.7) 0.7 (0.5– 0.9) 1.5 (1.0 –2.4) 0.7 (0.5– 0.9) 1.6 (1.1–2.4) 1.6 (1.2–2.2) 2.5 (1.8 –3.5) 4.0 (2.1–7.6) 1.6 (1.2–2.2)

2.3 (1.6 –3.2) 4.3 (2.7– 6.9) 1.6 (1.2–2.1) 0.9 (0.7–1.3) 1.9 (1.1–3.5)

0.9 (0.6 –1.5) 1.0 (0.6 –1.5) 1.2 (0.7–1.9)

0.7 (0.5–1.0) 0.5 (0.3– 0.7) 0.6 (0.3– 0.9)

* Odds ratio with 95% confidence interval.

Objective measurements on sickness absence were obtained from administrative files on an individual level and information about the situation at work and outside work was collected by a mail questionnaire. There is a wide scope of different measures of sickness absence in the literature.38 We used the incidence, which takes both frequency and duration of the sickness episodes, into account. This means that instead of focusing on the employee’s total amount of days on sick leave, we

preferred a measure reflecting the rate at which the spells occurred. The measure of sickness incidence was developed and used in previous studies from Sweden Post.7,30,39 This measure also has been used in other studies.4,5,6,31 Sickness absence is in general defined as the sick role a person takes in a situation with illness (subjective ill health) or disease (medical diagnosis) or in situations not involving illness or disease. We had no access to medical diagnoses for those clas-

sified with high sickness absence. To provide an “illness profile” of the individuals with a high sickness absence, we analyzed the prevalence of subjective symptoms perceived as work related. Important symptoms among both women and men with high sickness absence were physical exhaustion after work, tiredness with inability to engage in any activities after work, and musculoskeletal complaints. In addition, women reported in excess, headache, and “abnormal” tiredness (Table 2). Staying

JOEM

•

Volume 46, Number 11, November 2004

1141

TABLE 6 Sickness Absence Among Married/cohabitant Women With High Domestic Work Load Compared to Married/cohabitant Women with Lower Domestic Work Load OR (95% CI)* for sickness absence No. of women

High sickness absence

High domestic workload

Low domestic workload

High domestic workload**

983 501 240 229 169 332 482 148 303 72 410

407 196 84 98 81 115 211 66 127 37 174

174 82 36 37 33 49 92 32 52 13 79

1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0

0.9 (0.6 –1.2) 1.6 (1.0 –2.6) 1.1 (0.5–2.2) 2.3 (1.1– 4.8) 1.2 (0.5–2.6) 1.8 (1.0 –3.4) 0.5 (0.3– 0.8) 0.3 (0.1– 0.7) 0.7 (0.4 –1.3) 0.3 (0.1–1.1) 0.5 (0.3– 0.9)

Married/cohabitant Children at home Worked ⬍ 40 hours/week Worked 肁 40 hours/week Economic difficulties No economic difficulties No children at home Worked ⬍ 40 hours/week Worked 肁 40 hours/week Economic difficulties No economic difficulties

* Odds ratio with 95% confidence interval adjusted for age. ** Domestic workload ⫽ all domestic duties vs sharing the duties with someone.

at home because of uneasiness at work may be a marker of unjustified sickness absence. Only 2% of both men and women agreed that this had happened to them. The prevalence was clearly increased among those with high sickness incidence, but as indicated by the AF, this is far from an important explanation of sickness absence. All variables included in the questionnaire were selected on the basis of a priori hypotheses. It cannot be ruled out, however, that some of the findings were caused by chance because of the extensive number of variables analyzed. The crosssectional study design with information on exposures collected in retrospect (the exposure situation in 1993 was assessed during the end of 1994) may have introduced recall bias why the results should be interpreted with caution. Difficulties to remember should only have a diluting effect on the associations if these difficulties did not differ between the sickness absence groups. The cross-sectional nature of the study also means that the report of the exposure situation may pertain to times both preceding and following the sickness absence. Some exposure variables may therefore be influenced by sickness absence. For example, a weak financial

situation may have been caused by sickness absence, but it may be a risk factor as well. Despite the extensive control of confounding, residual confounding could be attributed to, for example, depressive symptoms or negative attitudes in general, influencing a person’s pattern of sickness absence as well as the responses to the questions on exposure. Most of the questions should be unrelated to attitudes however, and unmeasured confounding should not heavily influence the results. Nonrespondents had a higher incidence of sickness absence than respondents (Table 1). We do not know whether or not inclusion of this group would have changed the results. The selection of individuals into the study, excluding employees at work less than 6 months in 1993, might restrain the generalizability. Healthy workers were probably overrepresented in our study group, mainly because of some deficiency of individuals on long-term sick leave. As expected, long-lasting/serious illness and recurring common colds showed the strongest associations with a high incidence of sickness absence for both women and men. In the final analyses these variables were included, which means that the

results for other variables in the model were, to some extent, adjusted for health conditions including such illness and susceptibility to infectious disease. Sickness absence is influenced by the individual’s psychosocial status. The fourfold increase in risk of high sickness absence for women using tranquilizers may reflect a stressful situation or some psychiatric disorder. Psychiatric diagnoses are increasingly important causes of sickness absence and explain some of the difference in sickness absence between women and men.40 – 42 Men reporting use of alcohol as sedative also showed an increased risk of high sickness absence. The doubled risk of high sickness absence among men reporting adverse life events or divorce could be explained by increased psychosocial vulnerability and reduced ability to cope with such situations due to weaker social networks compared to women.25,26 Women who needed recovery from staying at home taking care of sick children had a doubled risk of high sickness absence, possibly caused by both physical and psychologic strain, and infections acquired by the children. However, children at home did not show any association with sickness absence for women or

1142

men (results not showed). We did neither find an association between high domestic workload and sickness absence. An explanation for the latter result could be that women with health problems and higher rates of sick leave take less responsibility for the domestic work. On the other hand, when analyzing subgroups of married/cohabitant women, a high domestic workload was associated with high sickness absence among women with children at home. This result was strengthened among women with children who also worked 40 hours or more per week. Our interpretation is that the total workload from paid and unpaid work may reach a threshold entailing an increased risk of sickness absence, which supports the theories of role overload.19 –21 The association between high domestic workload and high sickness absence among married/cohabitant women with children at home and without financial difficulties could be attributed to the fact that these women had better economic possibilities to take sickleave—possibly as a coping strategy to reduce a high total workload. The relatively low sickness absence in women with a high domestic work load and no children at home could mean that their domestic work was less heavy compared with women with children and that they did not reach the critical “high-strain level.” A difficult financial situation showed a similar association with high sickness absence for both women and men in the age adjusted analyses. In the final multivariate analyses an association (of borderline significance) remained only in men. One reason could be that financial issues per se are of higher significance for men than for women. Possibly, men take more responsibility in that respect, and weak conditions may cause anxiety and sickness absence further on. However, financial difficulties may also be caused by sickness absence, and the effect may be more significant in men.

Sick Leave and Factors at Work and Outside Work

In this study 37% in women and 56% in men reported sickness presenteeism, they had been working instead of taking sick leave, when ill. They had a 60% increase in risk of high sickness absence in the final multivariate analysis (Table 5). A common denominator for sickness absence and sickness presenteeism is sickness. High sickness absence may force the individual to sickness presenteeism, to avoid the income reduction. It could also be that sickness presenteeism causes sickness absence. We do not know the long-term effect of working when sick leave is warranted, but it seems likely that sickness presenteeism may lead to more serious health problems with longer periods of sick leave and rehabilitation. Our results are largely coherent with those of Aronsson et al29, who showed a prevalence of sickness presenteeism of 50% for both women and men, and an association with high sickness absenteeism. Most associations for work-related variables that were observed previously 30 were largely unaffected when adjusting for nonwork-related variables in the final multivariate model. Occurrence of bullying, which may be a marker of a detrimental psychosocial climate at the workplace, also remained as an important potential risk factor for high sickness absence in women.15–17 However, sickness absence might also cause bullying.16 The results for women of a low sickness absence among those who seldom or never had information meetings probably meant that recurrent meetings implied “negative information,” for example about reorganization. Most of the potential determinants pointed out in this study could be an objective for actions with the aim to prevent high sickness absence. We wanted to capture the most important potential determinants within separate target areas, starting with the age-adjusted analyses followed by the multivariate analyses. The final analysis combining work-related fac-

•

Voss et al

tors, family, domestic work, financial situation, and lifestyle factors as well as illness such as common colds complete the multifactorial picture of the most important factors that may cause sickness absence in women and men. These complex issues need to be further explored in prospective studies on the individual level, covering both working life and private life characteristics, and taking into account potential synergistic effects from both areas. Sickness absence is an important indicator of public health and has far-reaching economic consequences for the society.

Acknowledgment This study was supported by three of Sweden Post’s regional organizations: Letter Division Sundsvall, Letter Division Umeå and Sales Division Sundsvall. We thank Hannu Kiviranta, Mariedal Konsult AB, for help with data entry and programming.

References 1. Riksfo¨rsa¨kringsverket. Socialfo¨rsa¨kringens utgifter fo¨r sjukpenning. [The National Social Insurance Board, Social insurance expenditure for sickness cash benefit, in Swedish]. Available at www.rfv.se, Internet; accessed September 14, 2004. 2. Blank N, Diderichsen F. Short-term and long-term sick-leave in Sweden: relationships with social circumstances, working conditions and gender. Scand J Soc Med. 1995;23:265–272. 3. Fuhrer R, Shipley MJ, Chastang, JF, et al. Socioeconomic position, health and possible explanations: a tale of two cohorts. Am J Public Health. 2002;92:1290 – 1294. 4. North F, Syme SL, Feeney A, Head J, Shipley MJ, Marmot MG. Explaining socioeconomic differences in sickness absence: the Whitehall II study. BMJ. 1993;306:361–366. 5. Niedhammer I, Bugel I, Goldberg M, Leclerc A, Gue´guen A. Psychosocial factors at work and sickness absence in the Gazel cohort: a prospective study. Occup Environ Med. 1998;55:735–741. 6. Kivima¨ki M, Vahtera J, Thomson L, Griffiths A, Cox T, Pentti J. Psychosocial factors predicting employee sickness absence during economic decline. J Appl Psychol. 1997;82:858 – 872. 7. Voss M, Floderus B, Diderichsen F. Changes in sickness absenteeism follow-

JOEM

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

•

Volume 46, Number 11, November 2004

ing the introduction of a qualifying day for sickness benefit: findings from Sweden Post. Scand J Public Health. 2001; 29:166 –174. Hoogendoorn WE, van Poppel MN, Bongers PM, Koes BW, Bouter LM. Physical load during work and leisure time as risk factors for back pain. Scand J Work Environ Health. 1999;25:387– 403. Josephson M, Vingård E, and the MUSIC-Norrta¨lje Study Group. Workplace factors and care seeking for low-back pain among female nursing personnel. Scand J Work Environ Health. 1998;24: 465– 472. Sobti A, Cooper C, Inskip H, Searle S, Coggon D. Occupational physical activity and long-term risk of musculoskeletal symptoms: a national survey of post office pensioners. Am J Ind Med. 1997;32: 76 – 83. Trinkoff AM, Storr CL, Lipscomb JA. Physically demanding work and inadequate sleep, pain medication use, and absenteeism in registered nurses. J Occup Environ Med. 2001;43:355–363. Kivima¨ki M, Sutinen R, Elovainio M, et al. Sickness absence in hospital physicians: 2 year follow up study on determinants. Occup Environ Med. 2001;58: 361–366. Kristensen TS. Sickness absence and work strain among Danish slaughterhouse workers: an analysis of absence from work regarded as coping behaviour. Soc Sci Med. 1991;32:15–27. North FM, Syme SL, Feeney A, Shipley MJ, Marmot M. Psychosocial work environment and sickness absence among British civil servants: the Whitehall II study. Am J Public Health. 1996;86:332– 340. Kivima¨ki M, Elovainio M, Vahtera J. Workplace bullying and sickness absence in hospital staff. Occup Environ Med. 2000;57:656 – 660. Vartia ML. Consequences of workplace bullying with respect to the well-being of its targets and the observers of bullying. Scand J Work Environ Health. 2001;27: 63– 69. Godin IM. Bullying, workers’ health, and labour instability [letter]. J Epidemiol Community Health. 2004;58:258 –259. Martikainen P. Women’s employment, marriage, motherhood and mortality: a test of the multiple role and role accumu-

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

lation hypotheses. Soc Sci Med. 1995;40: 199 –212. Mastekaasa A. Parenthood, gender and sickness absence. Soc Sci Med. 2000;50: 1827–1842. Bratberg E, Dahl SÅ, Risa AE. “The double burden.” Do combinations of career and family obligations increase sickness absence among women? Eur Sociol Rev. 2002;18:233–249. Khlat M, Sermet C, Le Pape A. Women’s health in relation with their family and work roles: France in the early 1990s. Soc Sci Med. 2000;50:1807–1825. Waldron I, Weiss CC, Hughes ME. Interacting effects of multiple roles on women’s health. J Health Soc Behav. 1998;39:216 –236. Hensing G, Spak F. Psychiatric disorders as a factor behind sick-leave in other diagnoses. A general population based study. Br J Psychiatry. 1998;172:250 – 256. Upmark M, Mo¨ller J, Romelsjo¨ A. Longitudinal, population-based study of self reported alcohol habits, high levels of sickness absence, and disability pensions. J Epidemiol Community Health. 1999;53: 223–229. Eriksen W, Natvig B, Bruusgaard D. Marital disruption and long-term work disability. A four-year prospective study. Scand J Public Health. 1999;27:196 – 202. Kivima¨ki M, Vahtera J, Elovainio M, Lillrank B, Kevin M. Death or illness of a family member, violence, interpersonal conflict, and financial difficulties as predictors of sickness absence: longitudinal cohort study on psychological and behavioral links. Psychosom Med. 2002;64: 817– 825. Kristenssen TS. Sygefravær som coping. Sickness absence as a means of coping [in Danish]. Socialmedicinsk tidskrift. 1995;(1):12–17. Ba¨ckman O. Longitudinal Studies on Sickness Absence in Sweden [Thesis]. Stockholm, Sweden: Swedish Institute for Social Research, Stockholm University, 1998. Aronsson G, Gustafsson K, Dallner M. Sick but yet at work. An empirical study of sickness presenteeism. J Epidemiol Community Health. 2000;54:502–509. Voss M, Floderus B, Diderichsen F. Physical, psychosocial, and organisational factors relative to sickness ab-

1143

31.

32.

33.

34.

35.

36. 37.

38.

39.

40.

41.

42.

sence: a study based on Sweden Post. Occup Environ Med. 2001;58:178 –184. Bourbonnaise R, Mondor M. Job strain and sickness absence among nurses in the province of Que´bec. Am J Ind Med. 2001;39:194 –200. Eriksen W, Bruusgaard D, Knardahl S. Work factors as predictors of sickness absence: a three month prospective study of nurses’ aides. Occup Environ Med. 2003;60:271–78. Moreau M, Valente F, Mak R, et al. Occupational stress and incidence of sick leave in the Belgian workforce: the Belstress study. J Epidemiol Community Health. 2004;58:507–516. Statistiska Centralbyrån. Underso¨kningen av levnadsfo¨rhållanden, [Statistics Sweden, Investigations of living conditions, in Swedish]. Stockholm: ULF; 89:1. 1989. Bergqvist U, Knave B, Voss M, Wibom R. A longitudinal study of VDT work and health. Int J Human-Computer Interact. 1992;4:197–219. SAS/STAT User’s Guide, Version 6. 4 ed. Cary, NC: SAS Institute Inc.; 1994. Greenland S. Applications of stratified analysis methods. In: Rothman KJ, Greenland S, eds. Modern Epidemiology. Philadelphia: Lippincott-Raven Publishers; 1998:281–300. Hensing G, Alexanderson K, Allebeck P, Bjurulf P. How to measure sickness absence? Literature review and suggestion of five basic measures. Scand J Soc Med. 1998;26:133–144. Knave B, Paulsson H, Floderus B et al. Incidence of work-related disorders and absenteeism as tools in the implementation of work environment improvements: the Sweden Post strategy. Ergonomics. 1991;34:841– 848. Savikko A, Alexandersson K, Hensing G. Do mental health problems increase sickness absence due to other diseases? Soc Psychiatry Psychiatr Epidemiol. 2001; 36:310 –316. Stansfeld S, Feeney A, Head J, Canner R, North F, Marmot M. Sickness absence for psychiatric illness: the Whitehall II Study. Soc Sci Med. 1995;40:189 –197. Nystuen P, Hagen KB, Herrin J. Mental health problems as a cause of long-term sick leave in the Norwegian workforce. Scand J Public Health. 2001;29:175– 182.