Abstract
Aims
Social workers report high levels of stress and have an increased risk for hospitalisation with mental diagnoses. However, it is not known whether the risk of work disability with mental diagnoses is higher among social workers compared with other human service professionals. We analysed trends in work disability (sickness absence and disability pension) with mental diagnoses and return to work (RTW) in 2005–2012 among social workers in Finland and Sweden, comparing with such trends in preschool teachers, special education teachers and psychologists.
Methods
Records of work disability (>14 days) with mental diagnoses (ICD-10 codes F00–F99) from nationwide health registers were linked to two prospective cohort projects: the Finnish Public Sector study, years 2005–2011 and the Insurance Medicine All Sweden database, years 2005–2012. The Finnish sample comprised 4849 employees and the Swedish 119 219 employees covering four occupations: social workers (Finland 1155/Sweden 23 704), preschool teachers (2419/74 785), special education teachers (832/14 004) and psychologists (443/6726). The reference occupations were comparable regarding educational level. Risk of work disability was analysed with negative binomial regression and RTW with Cox proportional hazards.
Results
Social workers in Finland and Sweden had a higher risk of work disability with mental diagnoses compared with preschool teachers and special education teachers (rate ratios (RR) 1.43–1.91), after adjustment for age and sex. In Sweden, but not in Finland, social workers also had higher work disability risk than psychologists (RR 1.52; 95% confidence interval 1.28–1.81). In Sweden, in the final model special education teachers had a 9% higher probability RTW than social workers. In Sweden, in the final model the risks for work disability with depression diagnoses and stress-related disorder diagnoses were similar to the risk with all mental diagnoses (RR 1.40–1.77), and the probability of RTW was 6% higher in preschool teachers after work disability with depression diagnoses and 9% higher in special education teachers after work disability with stress-related disorder diagnoses compared with social workers.
Conclusion
Social workers appear to be at a greater risk of work disability with mental diagnoses compared with other human service professionals in Finland and Sweden. It remains to be studied whether the higher risk is due to selection of vulnerable employees to social work or the effect of work-related stress in social work. Further studies should focus on these mechanisms and the risk of work disability with mental diagnoses among human service professionals.
Key words: Depression, epidemiology, mental disorder, occupation, sick leave, social work, work disability
Introduction
The increasing rate and costs of mental disorders in the working population are major concerns in industrialised countries (Alexanderson & Norlund, 2004; Wittchen & Jacobi, 2005; Kessler et al. 2009; OECD, 2010) and the risk for mental disorders is particularly high among human service professions such as social workers, psychologists and teachers (Wieclaw et al. 2006; Samuelsson et al. 2013; Kokkinen et al. 2014). Social workers, especially in child welfare work, report high job strain, effort-reward imbalance, emotional exhaustion, role conflicts and lack of social support at work (Mor Barak et al. 2001; DePanfilis & Zlotnik, 2008; Rugulies et al. 2009; Tham & Meagher, 2009; Saarinen et al. 2012; Aronsson et al. 2014). Work-related stressors, particularly job strain, are associated with higher risk of mental disorders (Stansfeld & Candy, 2006; Virtanen et al. 2007; Bonde, 2008; Nieuwenhuijsen et al. 2010; Stansfeld et al. 2012). Wieclaw et al. (2006) found that the risk of affective and stress-related disorders was 1.5–2.5 times higher among social workers compared with employees in non-human service occupations. Education and health professionals, including special education teachers and psychologists, were also at risk, but with lower rates than among social workers.
Most previous studies on occupational risks for mental disorders have used self-report data and compared heterogeneous occupational groups (e.g., Stansfeld et al. 2013). Only few studies have applied a longitudinal design and used register data to analyse differences between occupations regarding risk of disability pension (Samuelsson et al. 2013) or hospitalisation (Wieclaw et al. 2006) with mental diagnoses. We are not aware of studies about the relative risk of work disability with mental diagnoses or return to work (RTW) after such a work disability period among social workers compared with other human service occupations.
To address such limitations, we analysed trends in work disability (sickness absence and disability pension) with mental diagnoses in 2005–2012 among 25 000 social workers in Finland and Sweden. In Finland and Sweden, social workers have a university degree, belong to upper-grade non-manual occupations and their job is to assess and facilitate the client's life situation, operate plans to resolve their problems and provide support and guidance to the client in implementing the plan (Statistics Finland, 2001; Statistics Sweden MIS, 2001). We chose preschool teachers, special education teachers and psychologists as reference occupations, because they share job characteristics with social work, such as face-to-face interaction with people who may have social, emotional and mental problems. The occupations were mostly comparable regarding educational level, apart from a lower level in pre-school teachers in both countries, and in Sweden most social workers have a bachelor level education. In addition, we analysed the probability of RTW after work disability with mental diagnoses.
Methods
Two prospective cohort studies were conducted using data from the Insurance Medicine All-Sweden project (IMAS) (Wang et al. 2014) covering the whole working-aged (16–64 years) population in Sweden and the Finnish Public Sector study (FPS) (Kivimäki et al. 2004), a large ongoing prospective cohort study including personnel, aged 18–65 years, of 10 municipalities and 21 hospitals in Finland. The FPS cohort comprised employees who had worked at least 6 months for the participating organisations in years 1991–2005 (N = 151 901). From IMAS we used information about the 5 750 278 people aged 16–64 years who lived in Sweden in December 2004. In both cohorts, the employees were linked to nationwide records of sickness absence and disability pension via personal identification numbers assigned to each resident. The registers contain data on beginning and end dates and diagnoses of work disability covering the years 2004–2011 in Finland and 2005–2012 in Sweden.
In the FPS cohort in 2004, there were 81 707 employees who had been employed for at least 6 months in one of the organisations. We selected all employees in one of the occupations (N = 4899), and excluded those who were on disability (part-time or full) or old-age pension at baseline at 1st January 2005 or deceased before that. From IMAS, we included all employed in one of the occupations in December 2004 (N = 133 083, no data on turnover available). We excluded those on disability pension (part-time or full) or on long-term sick leave (>1 year) at baseline 1st January 2005.
The occupations were derived from the employers’ registers in Finland and Statistics Sweden in Sweden using the occupational classification codes. In Finland, we used the Classification of Occupations 2001 created by Statistics Finland; a revised version of the International Standard Classification of Occupations (ISCO-88) (Statistics Finland, 2001) here referred to as ISCO codes. In Sweden we used the Swedish Standard Classification of Occupations (Standard för svensk yrkesklassificering 1996, SSYK 96) (Statistics Sweden MIS, 2001). From both cohorts, we included social workers (ISCO code 24461; SSYK code 2492), preschool teachers (2332; 3310), special education teachers (234; 2340), and psychologists (24451; 2491). In Finland, the final sample (N = 4849) comprised 1155 (24%) social workers, 2419 (50%) preschool teachers, 832 (17%) special education teachers and 443 (9%) psychologists. In Sweden, the final sample (N = 119 219) was 23 704 (20%) social workers, 74 785 (63%) preschool teachers, 14 004 (12%) special education teachers and 6726 (5%) psychologists.
Ethics statement
The approval of the Ethics Committee of the Helsinki-Uusimaa Hospital District was obtained for FPS. The IMAS-project was approved by the Regional Ethical Review Board of Stockholm, Sweden.
Work disability with mental diagnoses
In Finland, data on sickness absence (>9 days) and disability pensions (temporary, permanent, full-time and part-time) was derived from the registers kept by Social Insurance Institution of Finland and the Centre for Pensions, including information on main diagnoses according to the International Classification of Diseases (ICD-10) (World Health Organization, 2010). The employee has to present to the employer a physician's certificate usually from the 4th day onwards after being absent from work. Sickness benefit is paid after 10 days of sickness for a maximum of 300 working days during two consecutive years. After that, the employee is to apply for temporary or permanent disability pension.
In Sweden, information on sickness absence (>14 days) and disability pensions (75% or full-time) were obtained from the Social Insurance Agency, regarding main diagnoses (ICD-10) and start and end dates. The employee has to present a physician's certificate from the day 8 of sickness. Up to 2008, there was no maximum time limit for sick leave, but the duration of sickness absence has since been set at 364 days in the period of 15 months. In cases of serious disease, extended sickness benefit can be granted additionally 550 days (in all 914 days).
To harmonise the data between Finland and Sweden, we analysed spells >14 days in both countries. We examined work disability spells with any mental diagnosis (ICD-10 codes F00–F99) and combined overlapping and consecutive spells, using the diagnosis of the first period. For comparison, we examined trends in work disability with somatic diagnosis (all non F-diagnoses, ICD-10). Specific work disability diagnoses are reported in online Appendix 1.
Covariates
In Finland, the covariates were age, sex, area of residence, chronic somatic disease and work disability with mental diagnoses (ICD-10 codes F00–F99) in 2004. Age and sex were derived from the employers’ records. Area of residence was derived from the Population Centre of Finland. These were coded into the Helsinki region (the Finnish capital) and other regions to form two adequately large categories. The presence of chronic somatic disease at baseline was identified from the national health records: (1) The incidence of diagnosed prevalent hypertension, cardiac failure, ischemic heart disease, diabetes, asthma or other chronic obstructive lung disease and rheumatoid arthritis was obtained from the Drug Reimbursement Register kept by the Social Insurance Institution of Finland; (2) information on malignant tumours diagnosed during the preceding 5 years was obtained from the Finnish Cancer Register covering all diagnosed cancer cases in Finland.
In Sweden, the covariates were age, sex, area of residence, marital status and chronic somatic diseases in 2000–2004. Chronic somatic diseases were identified in 1.1.2000–31.12.2004 from two registers: (1) in- or specialised outpatient treatment (from 2001) with diagnoses for cardiovascular diseases and prevalent hypertension (ICD-10 codes I10, 111, I15, I20, I21, I25), diabetes (E10, E11), asthma or other chronic obstructive lung disease (J41–J45) or rheumatoid arthritis (M02, M05, M06, M08, M13), obtained from the National Patient Register or (2) malignant tumours, obtained from the Swedish Cancer Register. Area of residence had three categories: big city (Stockholm, Gothenburg and Malmo), medium size city (>90 000 inhabitants within 30 km distance from the city centre), and small city or village (≤90 000 inhabitants within 30 km distance from the city centre).
Statistical analyses
First, we calculated separately the cumulative days of work disability with mental diagnoses and with somatic diagnoses per person years in 2005–2011 in Finland and 2005–2012 in Sweden among social workers, preschool teachers, special education teachers and psychologists. The cumulative work disability days comprised sickness absence and disability pension per calendar year. We then examined incident work disability with mental diagnoses and somatic diagnoses by using negative binomial regression models and estimated rate ratios (RR) and their 95% confidence intervals (CI). The risk of work disability was analysed among social workers in reference to preschool teachers, special education teachers and psychologists.
Duration of work disability periods was calculated as days from the beginning of work disability to the date when the employee RTW, was granted an old-age pension (in Sweden 31st December in the year of 65th birthday), died, or end of follow-up. We calculated the hazard ratios (HR) and their 95% CI and used the Cox proportional hazards model to analyse probability of RTW from work disability episodes during the study period. We tested the proportionality hazards assumption by including the interaction of each occupation with work disability duration (Allison, 2000). In the Finnish sample, the interaction was not statistically significant in any case. In the Swedish sample, all interactions were statistically significant (p < 0.001). However, the Kaplan-Meier curve suggested that the magnitude of the HRs remained over time, supporting the hazards assumption (online Appendix 2). For descriptive purposes, we coded the duration of work disability periods into six categories: <2 months (14–59 days), 2–3 months (60–89 days), 3–6 months (90–119 days), 6–8 months (120–239 days), 8–12 months (240–365 days) and >12 months (>365 days).
In sensitivity analyses we analysed the risk of work disability and probability of returning to work with more specific diagnosis categories. The results are reported in online Appendix 3. We used SAS statistical software, version 9.4 for all analyses.
Results
In the FPS, the proportion of women within the occupations ranged from 76.2% (special education teachers) to 96.7% (preschool teachers) (Table 1); in the IMAS-cohort from 69.7% (psychologists) to 91.7% (preschool teachers). In each occupation, proportion of men was greater in the Swedish population-based data than in the FPS data. Participants were also in general older in the Swedish than Finnish cohort. Preschool teachers were more often younger than 30 years than other employees both in Finland and Sweden. In Finland, more employees were ≥40 years in special education teachers (68.2%) and social workers (67.1%) compared with preschool teachers (52%) and psychologists (59.6%). In Sweden, 86.5% of the special education teachers and 76.1% of psychologists were ≥40 years, whereas age was more evenly distributed among social workers and preschool teachers.
Table 1.
Characteristics of the two cohorts
| FPS (Finland) | IMAS (Sweden) | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Social workers (n = 1155) | Preschool teachers (n = 2419) | Special education teachers (n = 832) | Psychologists (n = 443) | All (n = 4849) | Social workers (n = 23 704) | Preschool teachers (n = 74 785) | Special education teachers (n = 14 004) | Psychologists (n = 6726) | All (n = 119 219) | |
| Sex | ||||||||||
| Men | 124 (10.7) | 81 (3.3) | 198 (23.8) | 66 (14.9) | 469 (9.7) | 4206 (17.7) | 6200 (8.3) | 2733 (19.5) | 2037 (30.3) | 15 176 (12.7) |
| Women | 1031 (89.3) | 2338 (96.7) | 634 (76.2) | 377 (85.1) | 4380 (90.3) | 19 498 (82.3) | 68 585 (91.7) | 11 271 (80.5) | 4689 (69.7) | 104 043 (87.3) |
| Age in years | ||||||||||
| 18–29 | 141 (12.2) | 368 (15.2) | 58 (7.0) | 62 (14.0) | 629 (13.0) | 2276 (9.6) | 9984 (13.4) | 329 (2.3) | 279 (4.2) | 12 868 (10.8) |
| 30–39 | 239 (20.7) | 794 (32.8) | 206 (24.8) | 117 (26.4) | 1356 (28.0) | 6068 (25.6) | 22 530 (30.1) | 1567 (11.2) | 1326 (19.7) | 31 491 (26.4) |
| 40–49 | 375 (32.5) | 891 (36.9) | 300 (36.0) | 115 (26.0) | 1681 (34.6) | 6798 (28.7) | 22 613 (30.2) | 3291 (23.5) | 1644 (24.4) | 34 346 (28.8) |
| >50 | 400 (34.6) | 366 (15.1) | 268 (32.2) | 149 (33.6) | 1183 (24.4) | 8562 (36.1) | 19 658 (26.3) | 8817 (63.0) | 3477 (51.7) | 40 514 (34.0) |
| Marital status | ||||||||||
| Married | – | – | – | – | – | 11 708 (49.4) | 39 513 (52.8) | 8996 (64.2) | 3561 (52.9) | 63 778 (53.5) |
| Not married | – | – | – | – | – | 11 996 (50.6) | 35 272 (47.2) | 5008 (35.8) | 3165 (47.1) | 55 441 (46.5) |
| Type of area of residence | ||||||||||
| Big size citya | – | – | – | – | – | 9670 (40.8) | 25 863 (34.6) | 4781 (34.1) | 3421 (50.9) | 43 735 (36.7) |
| Medium size citya | – | – | – | – | – | 8526 (36.0) | 28 149 (37.6) | 5206 (37.2) | 2189 (32.5) | 44 070 (37.0) |
| Small size city /villagesa | – | – | – | – | – | 5508 (23.2) | 20 773 (27.8) | 4017 (28.7) | 1116 (16.6) | 31 414 (26.3) |
| Helsinki region | 418 (36.2) | 1048 (43.3) | 343 (41.2) | 126 (28.4) | 1935 (39.9) | – | – | – | – | – |
| Other region in Finland | 737 (63.8) | 1371 (56.6) | 489 (58.8) | 317 (71.6) | 2914 (60.1) | – | – | – | – | – |
| Total work disability spells during follow-upb | 1099 | 2413 | 649 | 308 | 4469 | 22 549 | 78 672 | 10 997 | 5190 | 117 408 |
| F-diagnosis | 385 (35.0) | 486 (20.1) | 160 (24.7) | 86 (27.9) | 1117 (25.0) | 8018 (35.6) | 19 229 (24.4) | 2987 (27.1) | 1531 (29.5) | 31 765 (27.1) |
| Non F-diagnosis | 714 (65.0) | 1927 (79.9) | 489 (75.3) | 222 (72.1) | 3352 (75.0) | 14 531 (64.4) | 59 443 (75.6) | 8010 (72.8) | 3659 (70.1) | 85 643 (72.9) |
| Employees with at least one work disability spell during follow upb | 687 (59.5) | 1429 (59.1) | 425 (51.1) | 210 (47.4) | 2751 (56.7) | 11 129 (47.0) | 37 523 (50.2) | 5777 (41.3) | 2661 (39.6) | 57 090 (47.9) |
| F-diagnosis | 245 (21.2) | 336 (13.9) | 116 (13.9) | 63 (14.2) | 760 (15.7) | 4981 (21.0) | 12 372 (16.5) | 1998 (14.3) | 995 (14.8) | 20 346 (17.1) |
| Non F-diagnosis | 442 (38.3) | 1093 (45.2) | 309 (37.1) | 147 (33.2) | 1991 (41.1) | 8378 (35.3) | 31 455 (42.1) | 4641 (33.1) | 2105 (31.3) | 46 579 (39.1) |
| Employees with chronic somatic diseases before the baselinec | 76 (6.6) | 84 (3.5) | 15 (1.8) | 24 (5.4) | 199 (4.3) | 1028 (4.3) | 2243 (3.0) | 611 (4.4) | 294 (4.4) | 4176 (3.6) |
FPS, Finnish Public Sector Study; IMAS, Insurance Medicine All Sweden.
Note: Values are n (%) unless stated otherwise.
Area of residence: big cities: Stockholm, Gothenburg and Malmö; medium sized cities: cities with more than 90 000 inhabitants within 30 km distance from the centre of the city; small cities/villages.
With work disability spells lasting over 14 days in 2005–in Finland and 2005–in Sweden. Diagnosis is based on International Classification of Diseases, 10th revision codes F00–F99 (mental diagnosis) and all other diagnosis (somatic diagnosis).
In Finland, chronic somatic diseases or malignant tumours in 2004. In Sweden, chronic somatic diseases in 1.1.2000–31.12.2004 from two registers: (1) in- or specialised outpatient treatment (from 2001).
During the study period, 4469 work disability spells were recorded in the Finnish dataset (years 2005–2011) and 117 408 in the Swedish dataset (years 2005–2012) (Table 1). In both cohorts, the proportion of work disability spells with mental diagnoses of all work disability spells was the highest among social workers (Finland 35.0%; Sweden 35.6%) and the lowest among preschool teachers (Finland 20.1%; Sweden 24.4%). In the Finnish sample, a total of 760 participants (15.7%) had at least one spell >14 days with mental diagnoses and in Sweden 20 346 participants (17.1%) (Table 1). Both in Finland and Sweden, about 21% of social workers had at least one work disability spell with a mental diagnosis, while the corresponding figures in other occupations were 14–17%.
In both countries, social workers had the highest rate of work disability with mental diagnoses throughout the study period (Fig. 1), unlike for work disability with somatic diagnoses. At the end of follow up, the rates in Finland and Sweden were 7.24 and 10.46 in social workers, and 3.23–4.35 and 7.08–7.73 in the reference occupations, respectively (Table 2). In Finland, the rates accumulated steadily in all occupations, but in Sweden the rates started decreasing slowly from 2007 to 2008 (Fig. 1).
Fig. 1.
Cumulative days per cumulative person-years by occupation for work disability with mental diagnoses and somatic diagnoses in FPS (a, c) and IMAS (b, d).
Table 2.
Associations between occupation and work disability due to mental diagnoses in Finland and in Sweden. Social workers are compared separately with each reference occupation
| FPS (Finland) | IMAS (Sweden) | |||||||
|---|---|---|---|---|---|---|---|---|
| Unadjusted model | Model 1a | Model 2b | Unadjusted model | Model 1a | Model 2c | |||
| Cum. disab. daysd | RR (95% CI) | RR (95% CI) | RR (95% CI) | Cum. disab. daysd | RR (95% CI) | RR (95% CI) | RR (95% CI) | |
| Social workers | 7.24 | 10.46 | ||||||
| v. preschool teachers | 3.63 | 2.37 (1.52–3.70) | 1.77 (1.04–2.60) | 1.62 (1.04–2.52) | 7.32 | 1.43 (1.30–1.56) | 1.43 (1.30–1.57) | 1.40 (1.27–1.53) |
| v. special education teachers | 3.23 | 2.73 (1.55–4.82) | 1.91 (1.08–3.41) | 1.45 (0.80–2.61) | 7.73 | 1.36 (1.19–1.55) | 1.50 (1.31–1.72) | 1.46 (1.28–1.67) |
| v. psychologists | 4.35 | 2.12 (1.06–4.26) | 1.46 (0.73–2.92) | 1.91 (0.92–3.92) | 7.08 | 1.50 (1.26–1.78) | 1.53 (1.29–1.81) | 1.52 (1.28–1.81) |
FPS, Finnish Public Sector Study; IMAS, Insurance Medicine All Sweden; RR, rate ratios; CI, confidence interval.
Model 1 adjusted for age and sex.
Model 2 adjusted as model 1 and for area of residence and presence of chronic somatic disease and work disability due to mental diagnoses (ICD-10 codes F00–F99) in 2004.
Model 2 adjusted as model 1 and area of residence, marital status, presence of chronic somatic diseases in 2000–2004.
Cumulative work disability days due to mental diagnoses in 2005–in Finland and 2005–in Sweden.
Risk of work disability with mental diagnoses
In Finland, social workers were 1.77 (95% CI 1.04–2.60) times more likely to have work disability with mental diagnoses than preschool teachers and 1.91 (95% CI 1.08–3.41) times more likely than special education teachers, after adjustment for age and sex (Table 2). After further adjustment for area of residence and either presence of chronic somatic disease or work disability with mental diagnoses in 2004, the higher risk of work disability with mental diagnoses among social workers remained statistically significant only in comparison with preschool teachers. In comparison with psychologists, social workers had a higher risk in the crude model; when adjusting for age and sex the association attenuated to non-significant (RR 1.46, 95% CI 0.73–2.92).
In Sweden among social workers, the risk of work disability with mental diagnoses was 1.43 (95% CI 1.30–1.57) times greater compared with preschool teachers, 1.50 (95% CI 1.31–1.72) compared with special education teachers and 1.53 (95% CI 1.29–1.81) compared with psychologists, after adjustment for age and sex (Table 2). Additional adjustment for area of residence, marital status and chronic somatic diseases in 2000–2004 did not change the risks markedly.
In additional analyses, men and women were analysed separately in the Swedish sample. In the fully adjusted model, male social workers had a higher risk in relation to psychologists (RR 1.52, 95% CI 1.01–2.30) than female social workers (RR 1.33, 95% CI 1.10–1.61). The risk was higher in female social workers in relation to preschool teachers (female: RR 1.49, 95% CI 1.35–1.63; male: 1.36, 95% CI 1.01–1.83) and special education teachers (female: RR 1.36, 95% CI 1.18–1.75; male: 1.29, 95% CI 0.89–1.86). Due to the low number of men in the FPS cohort, sex-stratified analyses were not conducted.
Duration of work disability with mental diagnoses
In both the Finnish and Swedish cohorts, a large proportion of the work disability spells with mental diagnoses lasted for less than 60 days (71.4% in Finland; 41.0% in Sweden; Fig. 2). Work disability periods of 1 year or longer were more common among Swedish (19.1%) than Finnish (6.3%) employees. In the Finnish sample, the percentage of spells lasting ≥3 months was 21.9% in social workers, 18.9% in preschool teachers, 16.1% in special education teachers and 20.8% in psychologists. In the Swedish sample, percentages were 48.3, 53.0, 51.7 and 49.7, respectively.
Fig. 2.
The duration of work disability spells with mental diagnoses and their proportions of all work disability spells with mental diagnoses in FPS and IMAS.
The Kaplan-Meier curve (online Appendix 2) revealed a modestly higher probability of RTW in preschool teachers compared with employees in the other occupations. However, there was a clear difference in the patterns in the two countries. In Finland, after 3 months of work disability about 80% of the employees had returned to work, in Sweden around 60%.
In the Finnish sample, the HRs for probability of RTW were non-significant in all analysed models (Table 3). In the Swedish sample in the final model, the probability of RTW after work disability with mental diagnoses was slightly higher in special education teachers compared with social workers.
Table 3.
Associations between occupation and probability of return to work after work disability due to mental diagnoses
| FPS (Finland) | IMAS (Sweden) | |||||||
|---|---|---|---|---|---|---|---|---|
| Unadjusted model | Model 1a | Model 2b | Unadjusted model | Model 1a | Model 2c | |||
| Number of spells | HR (95% CI) | HR (95% CI) | HR (95% CI) | Number of spells | HR (95% CI) | HR (95% CI) | HR (95% CI) | |
| Social workers | 385 | 1 = Referent | 1 = Referent | 1 = Referent | 8018 | 1 = Referent | 1 = Referent | 1 = Referent |
| Preschool teachers | 486 | 1.15 (0.99–1.33) | 1.07 (0.92–1.25) | 1.08 (0.94–1.25) | 19 229 | 1.04 (1.01–1.08) | 1.03 (1.00–1.06) | 1.02 (0.99–1.06) |
| Special education teachers | 160 | 1.15 (0.93–1.43) | 1.18 (0.93–1.50) | 1.17 (0.96–1.44) | 2987 | 1.03 (0.98–1.09) | 1.09 (1.03–1.15) | 1.09 (1.03–1.15) |
| Psychologists | 86 | 1.06 (0.81–1.39) | 1.03 (0.79–1.34) | 1.04 (0.81–1.34) | 1531 | 0.97 (0.91–1.04) | 1.00 (0.93–1.07) | 0.99 (0.93–1.07) |
FPS, Finnish Public Sector Study; IMAS, Insurance Medicine All Sweden; HR, hazard Ratio; CI, confidence interval
Model 1 adjusted for age and sex.
Model 2 adjusted as model 1 and for area of residence and either presence of chronic somatic disease and work disability due to mental disorder diagnoses (ICD-10 codes F00–F99) in 2004.
Model 2 adjusted as model 1 and area of residence, marital status, presence of chronic somatic diseases in 2000–2004.
In additional analysis, probability of RTW was analysed separately for men and women in the Swedish sample. In the final model, the differences in HRs were small between men and women, among men the results were all non-significant (HRs 1.00–1.02). Among female preschool teachers, special education teachers and psychologists the HRs were 1.03 (95% CI 0.99–1.06), 1.10 (95% CI 1.04–1.17), and 0.99 (95% CI 0.92–1.07). Low number of male employees prevented sex-stratified analyses in the Finnish sample.
Discussion
Our study on nearly 25 000 social workers in Finland and Sweden showed that they had higher rates of work disability with mental diagnoses compared with preschool teachers, special education teachers and psychologists during years 2005–2011/2012. In preschool teachers, special education teachers and psychologists, the rates were relatively similar to each other at the end of the study period. In Sweden, the risk of work disability with mental diagnoses was higher in social workers compared with each of these professions, whereas in Finland, the risk was higher only in comparison with preschool teachers, after full adjustments. In Sweden, social workers had a small but statistically significant risk of delayed RTW compared with preschool and special education teachers.
Our findings are supported by previous studies that have shown a high risk of mental disorders and work disability in social workers (Wieclaw et al. 2005, 2006; Lund et al. 2007; Samuelsson et al. 2013; Kokkinen et al. 2014). In two Danish studies, social workers had a higher risk of affective and stress-related disorders compared with referents matched by age and gender (Wieclaw et al. 2005) and non-human service professionals (Wieclaw et al. 2006). In the latter study, the risk for either disorder was higher in social workers than in health and education professionals in general, except for a slightly higher risk of affective disorders in female nurses. Other studies have also shown a high risk of mental disorders in nursing and other elementary personal care occupations (e.g., Stansfeld et al. 2013). In a Swedish study, higher risk of disability pension with mental diagnoses in social workers and healthcare employees was observed even after controlling for multiple confounding factors, including psychosocial working conditions (Samuelsson et al. 2013). Finally, in a recent Finnish study social work, education and health professionals had a 2-fold risk of mental disorders compared with employees in financial services (Kokkinen et al. 2014).
It remains unclear whether the observed association in this study reflects causation or selection, or both. Causation hypothesis, meaning mental disorders among social workers is due to work-related exposures, is supported by previous findings indicating that social workers report multiple work-related risk factors (Mor Barak et al. 2001; DePanfilis & Zlotnik, 2008; Rugulies et al. 2009; Tham & Meagher, 2009), which increase the risk for mental disorders and work disability (Clausen et al. 2012; Magnusson Hanson et al. 2013). Of human service professions, social workers are the most likely to be forced to give clients negative information, experience conflicts with clients, feel unease towards work and wanting to quit (Jönsson, 2005). Also, turnover is high. According to the selection hypothesis, people vulnerable to mental health issues could be more likely to become social workers. Cross-sectional studies based on self-report have suggested that adverse childhood experiences and a caretaking role assumed as a child may lead to an unconscious motivation to help others in later life (Burnett et al. 2006; Nikcevic et al. 2007). Among social workers, a caretaking role was associated with burnout, depression and difficulty of seeking help (Siebert & Siebert, 2005). In other human service professionals (eldercare workers), increased prevalence of antidepressant treatment was observed both 10 years before and 4 years after entering human service work, suggesting a selection into the profession (Madsen et al. 2012). Thus, somewhat limited results exist to support both hypotheses and they may not be mutually exclusive. Unfortunately, we had no data on mental health prior to entering work life, thus we could not examine selection into social work in our study.
There is strong evidence for older age (>50 years) and limited evidence for female sex, education, socio-economic status, negative recovery expectations and history of sickness absence as risk factors (Cornelius et al. 2011). There were some differences between the occupations in age and sex distributions in our study, however the results remained after adjusting for this. In additional analyses, we analysed the risks for men and women separately. In previous studies, male social workers had a higher risk of mental disorders than female social workers (Wieclaw et al. 2006). In our study, the risks were similar for male and female social workers: only in comparison with psychologists the risk of work disability with mental diagnoses was somewhat higher in male social workers.
In our study, special education teachers in Sweden had a slightly higher probability of RTW after work disability compared with social workers. Delayed RTW from work disability with mental diagnoses has been associated with factors such as old age, low education, health risk behaviours, symptom severity, depressive disorder diagnosis and high job stressors (Blank et al. 2008). The occupational differences in RTW were modest, but there was a clear difference between the countries with longer duration of RTW in Sweden.
The higher rate of work disability and longer duration of RTW may be explained by the datasets used or different social security systems. Sweden has had more generous and easily accessible sickness benefits than other countries, lighter restrictions on the duration of sickness absence and a higher amount of long-term sickness absences compared with other organisation for economic co-operation and development (OECD) countries (OECD, 2010). In 2005–2007 sickness absences >6 months comprised 50% of all work disability days compared with 10% in Finland and most OECD countries. Since 2003, the rate of sickness absence decreased greatly, but still in 2007, 6% of employees were absent at any time compared with 3% in Finland. In the Swedish sample, work disability rates decreased in all occupations at the turn of 2007–2008. At this time, a benefit reform was introduced with the aim to avoid long-term work disability and to encourage those who were on temporary disability pension to RTW.
Strengths and limitations
This study had multiple strengths: a longitudinal design, large sample, register data (not self-reports) and comparison between two countries and between specific human service occupations. Also, only few studies have previously analysed RTW after work disability.
The differences between the reference occupations enable hypothesising explanations for the different risks in the occupations. For example, the difference in the risk between social workers and psychologists may be due to job characteristics (e.g., responsibility or role conflicts) or resilience for emotional demands from the skills gained through education (e.g., self-care strategies). The occupations were mostly comparable regarding educational level and within occupations, the job contents descriptions were similar between countries. Although information on mental health and behavioural diagnoses may be considered sensitive, national health registers have been shown to provide a reliable measure of those (Melchior et al. 2009; Svedberg et al. 2010; Ahola et al. 2011; Ludvigsson et al. 2011; Dewa et al. 2014).
Limitations mainly concern differences in the datasets and the available variables. The Swedish sample was population-based, while the Finnish sample comprised public sector employees. Older age is associated with work disability and the average age was higher in the Swedish sample. However, in the analyses we adjusted for age. There were differences in the available variables in the two datasets, which affected the exclusion criteria and the covariates and which have to be taken into account when comparing the results from the two countries. From the Finnish sample, those employees were excluded who were on old-age pension, disability pension, or long-term sickness absence at the beginning of the follow-up, or had been employed in the occupation for less than 6 months in 2004. In the Swedish data, information was not available about old-age pension or duration of employment, otherwise exclusion criteria were equal. Use of covariates was further limited by the fact that this was a register-based study and the registers contained no data on such potential risk factors as psychosocial work characteristics or employee-related factors, such as history of mental disorders prior to entering work life.
Conclusions
This study shows that social workers are at a greater risk for work disability compared with other human service professionals. Future studies should examine whether this risk is due to selection to social work or causation by work-related stressors.
Acknowledgements
None.
Financial support
This work was financially supported by the Finnish Work Environment Fund (grant number: 113300) and the Swedish Research Council for Health, Working Life and Welfare.
Conflict of interest
None.
Ethical standards
All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. The use of a questionnaire acts as a form of written informed consent in the FPS cohort.
Availability of Data and Materials
Data were obtained from the Finnish Public Sector study (FPS), and requests for the FPS data can be made to the principal investigators of the Finnish Public Sector study (tuula.oksanen@ttl.fi; mika.kivimaki@helsinki.fi). The data from the Insurance Medicine All Sweden project are available from the Statistics Sweden, The National Social Insurance Agency and The National Board of Health and Welfare in Sweden for researchers who meet the criteria for access to confidential data.
Supplementary material
For supplementary material accompanying this paper visit http://dx.doi.org/10.1017/S2045796016000597.
click here to view supplementary material
References
- Ahola K, Virtanen M, Honkonen T, Isometsä E, Aromaa A, Lönnqvist J (2011). Common mental disorders and subsequent work disability: a population-based Health 2000 Study. Journal of Affective Disorders 134, 365–372. [DOI] [PubMed] [Google Scholar]
- Alexanderson K, Norlund A (2004). Swedish council on technology assessment in health care (SBU). Chapter 1. Aim, background, key concepts, regulations, and current statistics. Scandinavian Journal of Public Health 32 (Suppl. 63), 12–30. [DOI] [PubMed] [Google Scholar]
- Allison PD (2000). Survival Analysis using the SAS® System: A Practical Guide. SAS Institute Inc: Cary, NC. [Google Scholar]
- Aronsson G, Astvik W, Gustafsson K (2014). Work conditions, recovery and health: a study among workers within pre-school, home care and social work. British Journal of Social Work 44, 1654–1672. [Google Scholar]
- Blank L, Peters J, Pickvance S, Wilford J, MacDonald E (2008). A systematic review of the factors which predict return to work for people suffering episodes of poor mental health. Journal of Occupational Rehabilitation 18, 27–34. [DOI] [PubMed] [Google Scholar]
- Bonde JPE (2008). Psychosocial factors at work and risk of depression: a systematic review of the epidemiological evidence. Occupational and Environmental Medicine 65, 438–445. [DOI] [PubMed] [Google Scholar]
- Burnett G, Jones RA, Bliwise NG, Ross LT (2006). Family unpredictability, parental alcoholism, and the development of parentification. American Journal of Family Therapy 34, 181–189. [Google Scholar]
- Clausen T, Nielsen K, Carneiro IG, Borg V (2012). Job demands, job resources and long-term sickness absence in the Danish eldercare services: a prospective analysis of register-based outcomes. Journal of Advanced Nursing 68, 127–136. [DOI] [PubMed] [Google Scholar]
- Cornelius LR, van der Klink JJ, Groothoff JW, Brouwer S (2011). Prognostic factors of long term disability due to mental disorders: a systematic review. Journal of Occupational Rehabilitation 21, 259–274. [DOI] [PMC free article] [PubMed] [Google Scholar]
- DePanfilis D, Zlotnik JL (2008). Retention of front-line staff in child welfare: a systematic review of research. Children and Youth Services Review 30, 995–1008. [Google Scholar]
- Dewa CS, Loong D, Bonato S, Hees H (2014). Incidence rates of sickness absence related to mental disorders: a systematic literature review. BMC Public Health 14, 205. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Jönsson S (2005). Client work, job satisfaction and work environment aspects in human service organizations. Doctoral dissertation, Department of Psychology: Lund University, Sweden. [Google Scholar]
- Kessler RC, Aguilar-Gaxiola S, Alonso J, Chatterji S, Lee S, Ormel J, Ustun TB, Wang PS (2009). The global burden of mental disorders: an update from the WHO world mental health (WMH) surveys. Epidemiology and Psychiatric Sciences 18, 23–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kivimäki M, Forma P, Wikström J, Halmeenmäki T, Pentti J, Elovainio M, Vahtera J (2004). Sickness absence as a risk marker of future disability pension: the 10-town study. Journal of Epidemiology, Community and Health 58, 710. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kokkinen L, Kouvonen A, Koskinen A, Varje P, Väänänen A (2014). Differences in hospitalizations between employment industries, Finland 1976 to 2010. Annals of Epidemiology 24, 598–605. [DOI] [PubMed] [Google Scholar]
- Ludvigsson JF, Andersson E, Ekbom A, Feychting M, Kim J, Reuterwall C, Heurgren M, Olausson PO (2011). External review and validation of the Swedish national inpatient register. BMC Public Health 11, 1–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lund T, Labriola M, Villadsen E (2007). Who is at risk for long-term sickness absence? A prospective cohort study of Danish employees. Work 28, 225–230. [PubMed] [Google Scholar]
- Madsen IEH, Aust B, Burr H, Carneiro IG, Diedrichsen F, Rugulies R (2012). Paid care work and depression: a longitudinal study of antidepressant treatment in female eldercare workers before and after entering their profession. Depression and Anxiety 29, 605. [DOI] [PubMed] [Google Scholar]
- Magnusson Hanson LL, Madsen IE, Westerlund H, Theorell T, Burr H, Rugulies R (2013). Antidepressant use and associations with psychosocial work characteristics. A comparative study of Swedish and Danish gainfully employed. Journal of Affective Disorders 149, 38. [DOI] [PubMed] [Google Scholar]
- Melchior M, Ferrie JE, Alexanderson K, Goldberg M, Kivimaki K, Singh-Manoux A, Vahtera J, Westerlund H, Zins M, Head J (2009). Using sickness absence records to predict future depression in a working population: prospective findings from the gazel cohort. American Journal of Public Health 99, 1417–1422. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Mor Barak ME, Nissly JA, Levin A (2001). Antecedents to retention and turnover among child welfare, social work, and other human service employees: what can we learn from past research? A review and meta-analysis. Social Service Review 75, 625–661. [Google Scholar]
- Nieuwenhuijsen K, Bruinvels D, Frings-Dresen M (2010). Psychosocial work environment and stress-related disorders – a systematic review. Occupational Medicine (London) 60, 277–286. [DOI] [PubMed] [Google Scholar]
- Nikcevic AV, Kramolisova-Advani J, Spada MM (2007). Early childhood experiences and current emotional distress: what do they tell us about aspiring psychologists? Journal of Psychology 141, 25–34. [DOI] [PubMed] [Google Scholar]
- OECD (2010). Sickness, Disability and Work: Breaking the Barriers – Sweden: Will the Recent Reforms make it?
- Rugulies R, Aust B, Siegrist J, von dem Knesebeck O, Bultmann U, Bjorner JB, Burr H (2009). Distribution of effort-reward imbalance in Denmark and its prospective association with a decline in self-rated health. Journal of Occupational and Environmental Medicine 51, 870–878. [DOI] [PubMed] [Google Scholar]
- Saarinen A, Blomberg H, Kroll C (2012). Liikaa Vaadittu? Sosiaalityöntekijöiden Kokemukset Työnsä Kuormittavuudesta Ja Ristiriitaisuudesta Pohjoismaissa (nordic social workers’ experiences of occupational stress and contradictions in their work). Yhteiskuntapolitiikka 77, 403–418. [Google Scholar]
- Samuelsson Å, Ropponen A, Alexanderson K, Svedberg P (2013). Psychosocial working conditions, occupational groups, and risk of disability pension due to mental diagnoses: a cohort study of 43 000 Swedish Twins. Scandinavian Journal of Work, Environment and Health 39, 351–360. [DOI] [PubMed] [Google Scholar]
- Siebert DC, Siebert CF (2005). The caregiver role identity scale: a validation study. Research on Social Work Practice 15, 204–212. [Google Scholar]
- Stansfeld S, Candy B (2006). Psychosocial work environment and mental health – a meta-analytic review. Scandinavian Journal of Work, Environment and Health 32, 443–462. [DOI] [PubMed] [Google Scholar]
- Stansfeld SA, Shipley MJ, Head J, Fuhrer R (2012). Repeated job strain and the risk of depression: longitudinal analyses from the Whitehall II study. American Journal of Public Health 102, 2360–2366. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Stansfeld SA, Pike C, McManus S, Harris J, Bebbington P, Brugha T, Hassiotis A, Jenkins R, Meltzer H, Moran P, Clark C (2013). Occupations, work characteristics and common mental disorder. Psychological Medicine 43, 961–973. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Statistics Finland (2001). Classification of Occupations 2001. Statistics Finland: Helsinki. [Google Scholar]
- Statistics Sweden MIS (2001). SSYK 96 Swedish Standard Classification of Occupations 1996. Statistiska centralbyrån (Vol. 1).
- Svedberg P, Ropponen A, Lichtenstein P, Alexanderson K (2010). Are self-report of disability pension and long-term sickness absence accurate? Comparisons of self-reported interview data with national register data in a Swedish twin cohort. BMC Public Health 10, 763. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Tham P, Meagher G (2009). Working in human services: how do experiences and working conditions in child welfare social work compare? British Journal of Social Work 39, 807–827. [Google Scholar]
- Virtanen M, Honkonen T, Kivimäki M, Ahola K, Vahtera J, Aromaa A, Lönnqvist J (2007). Work stress, mental health and antidepressant medication findings from the health 2000 study. Journal of Affective Disorders 98, 189–197. [DOI] [PubMed] [Google Scholar]
- Wang M, Alexanderson K, Runeson B, Head J, Melchior M, Perski A, Mittendorfer-Rutz E (2014). Are all-cause and diagnosis-specific sickness absence, and sick-leave duration risk indicators for suicidal behaviour? a nationwide register-based cohort study of 4.9 million inhabitants of Sweden. Occupational and Environmental Medicine 71, 12–20. [DOI] [PubMed] [Google Scholar]
- Wieclaw J, Agerbo E, Mortensen PB, Bonde JP (2005). Occupational risk of affective and stress-related disorders in the Danish workforce. Scandinavian Journal of Work, Environment and Health 31, 343–351. [DOI] [PubMed] [Google Scholar]
- Wieclaw J, Agerbo E, Mortensen PB, Bonde JP (2006). Risk of affective and stress related disorders among employees in human service professions. Occupational and Environmental Medicine 63, 314–319. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wittchen HU, Jacobi F (2005). Size and burden of mental disorders in Europe–a critical review and appraisal of 27 studies. European Neuropsychopharmacology 15, 357–376. [DOI] [PubMed] [Google Scholar]
- World Health Organization (2010). International Statistical Classification of Diseases and Related Health Problems (ICD-10).
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
For supplementary material accompanying this paper visit http://dx.doi.org/10.1017/S2045796016000597.
click here to view supplementary material
Data Availability Statement
Data were obtained from the Finnish Public Sector study (FPS), and requests for the FPS data can be made to the principal investigators of the Finnish Public Sector study (tuula.oksanen@ttl.fi; mika.kivimaki@helsinki.fi). The data from the Insurance Medicine All Sweden project are available from the Statistics Sweden, The National Social Insurance Agency and The National Board of Health and Welfare in Sweden for researchers who meet the criteria for access to confidential data.