Workplace discrimination as risk factor for long-term sickness absence: Longitudinal analyses of onset and changes in workplace adversity

Alice Clark; Sari Stenholm; Jaana Pentti; Paula Salo; Theis Lange; Eszter Török; Tianwei Xu; Jesper Fabricius; Tuula Oksanen; Mika Kivimäki; Jussi Vahtera; Naja Hulvej Rod

doi:10.1371/journal.pone.0255697

. 2021 Aug 5;16(8):e0255697. doi: 10.1371/journal.pone.0255697

Workplace discrimination as risk factor for long-term sickness absence: Longitudinal analyses of onset and changes in workplace adversity

Alice Clark ^1,^*,^#, Sari Stenholm ^2,^3,^*,^#, Jaana Pentti ^2,^3,⁴, Paula Salo ^5,⁶, Theis Lange ⁷, Eszter Török ¹, Tianwei Xu ^1,⁸, Jesper Fabricius ^1,⁹, Tuula Oksanen ^6,¹⁰, Mika Kivimäki ^4,^6,¹¹, Jussi Vahtera ^2,^3,^‡, Naja Hulvej Rod ^1,^‡

Editor: Ali B Mahmoud¹²

¹Department of Public Health, Section of Epidemiology, University of Copenhagen, Copenhagen, Denmark

²Department of Public Health, University of Turku and Turku University Hospital, Turku, Finland

³Centre for Population Health Research, University of Turku and Turku University Hospital, Turku, Finland

⁴Faculty of Medicine, Clinicum, University of Helsinki, Helsinki, Finland

⁵Department of Psychology and Speech-Language Pathology, University of Turku, Turku, Finland

⁶Finnish Institute of Occupational Health, Helsinki, Finland

⁷Department of Public Health, Section of Biostatistics, University of Copenhagen, Copenhagen, Denmark

⁸Stress Research Institute, Stockholm University, Stockholm, Sweden

⁹Hammel Neurorehabilitation Centre and University Research Clinic, Hammel, Denmark

¹⁰Institute of Public Health and Clinical Nutrition, University of Eastern Finland, Kuopio, Finland

¹¹Department of Epidemiology and Public Health, University College London, London, United Kingdom

¹²St John’s University, UNITED KINGDOM

Competing Interests: The authors have declared that no competing interests exist.

‡ These authors are joint senior authors on this work.

^✉

* E-mail: alcl@sund.ku.dk (AC); sari.stenholm@utu.fi (SS)

Contributed equally.

Roles

Alice Clark: Conceptualization, Formal analysis, Writing – original draft, Writing – review & editing

Sari Stenholm: Conceptualization, Writing – original draft, Writing – review & editing

Jaana Pentti: Conceptualization, Data curation, Formal analysis, Writing – review & editing

Paula Salo: Conceptualization, Writing – review & editing

Theis Lange: Conceptualization, Writing – review & editing

Eszter Török: Conceptualization, Writing – review & editing

Tianwei Xu: Conceptualization, Writing – review & editing

Jesper Fabricius: Conceptualization, Writing – review & editing

Tuula Oksanen: Conceptualization, Writing – review & editing

Mika Kivimäki: Conceptualization, Funding acquisition, Writing – review & editing

Jussi Vahtera: Conceptualization, Formal analysis, Funding acquisition, Writing – original draft, Writing – review & editing

Naja Hulvej Rod: Conceptualization, Formal analysis, Funding acquisition, Supervision, Writing – original draft, Writing – review & editing

Ali B Mahmoud: Editor

PMCID: PMC8341535 PMID: 34351965

Abstract

Workplace discrimination may affect the health of the exposed employees, but it is not known whether workplace discrimination is also associated with an increased risk of long-term sickness absence. The aim of this study was to examine the longitudinal associations of changes in and onset of workplace discrimination with the risk of long-term sickness absence. Data on workplace discrimination were obtained from 29,597 employees participating in survey waves 2004, 2006, 2008 and/or 2010 of the Finnish Public Sector Study. Four-year changes in long-term sickness absence (≥10 days of medically certified absence with a mental or non-mental diagnosis) were assessed. This covered successive study waves in analyses of onset of workplace discrimination as well as fixed effect analyses of change in workplace discrimination (concurrent i.e. during the exposure year and 1-year lagged i.e. within one year following exposure), by using each employee as his/her own control. The risk of long-term sickness absence due to mental disorders was greater for employees with vs. without onset of workplace discrimination throughout the 4-year period, reaching a peak at the year when the onset of discrimination was reported (adjusted risk ratio 2.13; 95% confidence interval (CI) 1.80–2.52). The fixed effects analyses showed that workplace discrimination was associated with higher odds of concurrent, but not 1-year lagged, long-term sickness absence due to mental disorders (adjusted odds ratio 1.61; 95% CI 1.33–1.96 and adjusted odds ratio 1.02; 95% CI 0.83–1.25, respectively). Long-term sickness absence due to non-mental conditions was not associated with workplace discrimination. In conclusion, these findings suggest that workplace discrimination is associated with an elevated risk of long-term sickness absence due to mental disorders. Supporting an acute effect, the excess risk was confined to the year when workplace discrimination occurred.

Introduction

A large body of literature has linked various adverse psychosocial work characteristics, including job strain, effort-reward imbalance, job insecurity, and long working hours, to increased risk of physical and mental chronic conditions [1–5]. These factors characterize unhealthy features of work tasks and organization of work. In addition, negative interpersonal relations at work have been hypothesized to predispose employees to ill health.

Workplace discrimination (WD) is a specific social stressor, being typically both unpredictable and uncontrollable. Among the general working population in Finland, 11% of women and 6% of men reported experiencing discrimination at their workplace [6]. As such, discrimination has been associated with mental and physical health complaints including psychological distress, depression and anxiety, cardiometabolic diseases [7–12] as well as lower job satisfaction, organizational commitment and work efficiency [12–16], thus yielding a great cost for the individual and society as a whole. Further, prior studies point to a potential link between workplace discrimination and the risk of sickness absence [12,13,17–22]. Sickness absence, especially when long-term, is considered as a measure employee health and well-being as well as a correlate of workplace productivity [23–26]. However, studies of workplace discrimination and sickness absence are few, and suffer from a number of methodological problems. Current evidence relies almost exclusively on cross-sectional data and self-reported information on sickness absence [12,13,18–22]. We are aware of no longitudinal studies on this topic.

The majority of studies on the importance of psychosocial work environment for employee health and well-being, including those on workplace discrimination, are based on exposure assessment at one, often arbitrary, time-point, and longitudinal analyses assessing the importance of the timing and duration of exposure remain scarce. The ‘one measurement approach’ fails to account for at least two common methodological concerns within psychosocial work environment research.

First, the important fact that employees who experience adverse psychosocial circumstances at work may quit their job as a means of coping [13]. Therefore, the effect of workplace adversities is likely to differ between employees remaining at the workplace and those who quit–possibly because of those adversities–leading to biased associations between WD and health.

Second, stable individual characteristics such as personality traits are important in determining how psychosocial exposures are experienced and reacted to by the individual employee, and research has shown that employees who report negative social relations at work generally differ from other employees in terms of psychological characteristics such as neuroticism, aggressiveness and negative affectivity [27–29]. With such traits also being predictive of employee health and well-being [30–33], failing to control for this may be an important source of bias. Given the intangible concept of person-stable characteristics, even studies adjusting for personality traits may not adequately address this potential source of bias.

To address these two methodological shortcomings, we combined longitudinal survey with register data from national registers from Finland and applied robust methodological approaches in assessing in the associations of onset of WD and changes in WD with risk of subsequent long-term sickness absence (LTSA). First, to reduce issues of selection bias, we analyzed longitudinal data to compare changes in LTSA between employees with and without onset of workplace discrimination in a population in which all were free of that exposure at the first assessment. Second, we compared within-person changes in LTSA before and after a change in exposure to workplace discrimination to account for stable factors, such as personality, affecting the exposure to (confounding) and the reporting of (misclassification) workplace discrimination using each employee as his/her own control in fixed effects models.

Materials and methods

Study population

Participants were from the Ten town study, a part of the Finnish Public Sector Study (FPS) survey cohort, initiated in 1997. The survey cohort is nested within the FPS register cohort consisting of all Finnish public sector employees in ten towns and five hospital districts who had at least six months of employment between 1991 and 2005 [34,35]. The Ethics committee of Helsinki and Uusimaa Hospital District approved the study (registration number HUS/1210/2016). Participants were provided information about the study and response to the survey questionnaire was considered as a consent to participate. From 2004 the public sector employees working in the ten towns have been invited to biannual surveys including questions about work related factors.

Repeated information on workplace discrimination from waves 2004 (n = 32,197, response rate 65%), 2006 (n = 34,418, response rate 69%), 2008 (n = 38,838, response rate 70%) and 2010 (n = 37,651, response rate 69%) was used to assess the associations of onset of and changes in workplace discrimination with the risk of LTSA. To be eligible for the study, participants had to respond to at least two survey waves (successive waves were required for the assessment of onset) during which they were actively working for a minimum of 50% of the time and provide information on workplace discrimination in both surveys. In addition, linkage to the national health register on LTSA was required. These criteria were fulfilled by 32,519 employees. Since some participants had participated in all four study waves, they contributed observations to two separate two-wave data cycles (S1 Table). Given this, the 32,519 participants contributed to a total of 44,704 observations.

The selection of the study samples for the individual analyses of onset of and change in workplace discrimination is illustrated in S1 Fig.

For analyses of onset of workplace discrimination, 2811 observations with missing information on covariates (6% of the observations) were excluded. Among the 41,893 remaining observations, 2480 were excluded due to exposure to workplace discrimination at baseline (2004 and/or 2008) leaving 39,413 observations (in 29,597 employees) eligible for the analyses of onset of workplace discrimination. The structure of the data-cycles and follow-up is illustrated in S1 Table.

For analyses of change in workplace discrimination, 1125 observations with missing information on time-varying covariates (1% of the observations) were excluded. Due to the application of the fixed effect models, only observations with change in outcomes around the four waves where discrimination was measured were included in the analyses. The final number of employees and observations for analyses is illustrated in S1 Fig.

Workplace discrimination

Perceived WD was measured with a single item: “I am exposed to discrimination at my workplace” with responses on a 5-point Likert scale: “Completely agree” (1), “Somewhat agree” (2), “Do not agree or disagree” (3), “Somewhat disagree” (4) and “Completely disagree” (5) [36]. Workplace discrimination was dichotomized as “yes” (response options 1–2) and “no” (response options 3–5). In 2008 and 2010, the surveys included information regarding potential causes of workplace discrimination with the following categories: discrimination due to age, sex, ethnicity, education, opinion, and position.

Long-term sickness absence

Using employees’ unique personal identification numbers, we linked participants to the records of medically-certified LTSA from the Finnish Social Insurance Institution register, including dates and diagnoses for LTSA spells lasting 10 or days or more. This register covers the entire working population including public and private sector workers, but does not include sickness absence due to maternity leave and caring for a sick child. We distinguished between LTSA with an underlying mental (ICD-10 F01-F99) and non-mental diagnosis (other ICD-10 codes), and only primary diagnosis of LTSA was considered. In the text, terms mental LTSA and non-mental LTSA are used.

In the analyses with onset of WD as the exposure, the annual risk of LTSA (one or more episodes) for the groups with and without onset of WD was assessed. The risk of LTSA was calculated from the year pertaining to the first reporting of WD i.e. year 0 (where no employees were exposed) to year 3 (one year after the second study wave when WD was assessed). The study design is illustrated in the S1 Table.

In analyses with change in WD as the exposure, two different LTSA outcomes were defined: Concurrent LTSA was defined as sickness absence spells within the year pertaining to the reporting of exposure, and 1-year lagged LTSA was defined as sickness absence spells initiated during the year that followed the assessment of exposure. The study design is illustrated in the S1 Table.

Covariates

Information on individual and work unit characteristics was drawn from surveys, employers’ records, and national registers. Potential confounders were identified based on associated factors found in research literature and the methods of directed acyclic graphs [37] (See S2 Fig).

Individual level covariates included age, sex, chronic disease (diabetes, angina, acute myocardial infarction, chronic obstructive pulmonary disease, asthma, or cancer; no/yes), type of job contract (permanent/temporary), shift work (yes/no) and occupational grade (upper-grade non-manual, lower-grade non-manual, and manual workers) based on the ISCO-88 and the Occupational Title Classification of Statistics Finland. Information on alcohol consumption (abstainer, 1–16 units/week for women and 1–24 units/week for men, >16 units/week for women and >24 units/week for men) and body mass index (BMI) (<25, 25–30, ≥30 kg/m²) was obtained from the surveys. Psychological distress was assessed using the 12-item General Health Questionnaire (<4 vs ≥4 symptoms) [38]. The six-item Trait Anxiety Inventory [39] was used to assess liability to anxiety and was analyzed on a continuous scale based on the average of responses 1–4.

Work unit level covariates included work unit size (<20/≥20 employees), work unit gender distribution (female dominated defined as <33% men, male dominated defined as <33% women, and balanced defined as ≥33% women and men), and work unit proportion of temporary employees (above or below the median of 5%). Work unit level covariates were constructed by using information on the characteristics of the eligible population for each work unit.

Statistical analysis

Two different analytic approaches were used to determine the relationship between 1) onset of WD and LTSA; and 2) changes in WD and LTSA, as outlined below.

Analyses of onset of workplace discrimination

The analyses were based on a four-year observation period, ranging from year 0 (where none was exposed to discrimination) to year 3 (one year after reported onset of WD). This time frame was chosen, because the FPS survey did not include question regarding the specific date of the onset of discrimination. It is therefore only known that the onset took place between two successive study waves. The annual average prevalence of LTSA and the corresponding 95% CIs were estimated for employees with and without onset of WD using a repeated-measures log-binomial regression and a generalized estimating equation (GEE) model to account for the intra-individual correlation between measurements and across data-cycles [40]. If this model did not converge, we used a Poisson distribution and a log link to assess the risk of sickness absence. The estimates were pooled across the two data-cycles, which contributed to an increased statistical efficiency of the analyses. Results are presented as risk ratios (RR) and 95% confidence intervals (CI).

The analyses were adjusted for individual and work-unit level confounders at baseline: age, sex, chronic disease, job contract, normal working hours, occupational grade, work unit size, gender distribution, proportion of temporary employees, alcohol consumption, BMI, psychological distress, and trait anxiety.

Analyses of changes in workplace discrimination

To control for person-stable characteristics (measured as well as unmeasured) a fixed-effect approach was used to assess the average within-person change in LTSA related to WD across the four waves.

Odds ratios (OR) of LTSA were assessed as the risk of having one or more episodes of LTSA occurring concurrently (within the exposure year) and with one-year lag (within one year after exposure) in conditional logistic regression models with the corresponding 95% CIs. Due to the within-subject comparisons, the covariates include only the potential time-varying individual and work-unit level confounders: age, job contract, work unit size, gender distribution, and proportion of temporary employees. Results are presented as ORs and 95% CI’s.

Analyses of cause-specific workplace discrimination

Cause specific WD (age, sex, ethnicity, education, opinion, or position), was analyzed using data from the 2008 and 2010 study waves. The cause-specific onset and changes in WD and LTSA of mental and non-mental origin were examined separately as described earlier.

Supplementary analyses

To assess the robustness of findings, several supplementary analyses were carried out. First, previous LTSA may affect both subsequent LTSA and be a cause of discrimination. Therefore, the occurrence of concurrent LTSA was included as a covariate in analyses of one-year lagged LTSA in the fixed effect analyses of change in workplace discrimination, and as an additional cofounder in the analyses of onset.

As the time-varying health-related factors alcohol consumption, BMI and psychological distress may be either a cause or a consequence of discrimination, these were included as additional potential confounders in the fixed effects analyses available in the 2004 and 2008 study waves.

To explore the potential risk of reverse causation of the association between WD and LSTA, an additional analysis of the effect of LTSA on onset of WD was performed.

To examine the proposed source of bias from discriminated employees leaving their job, an additional analysis of the association between onset of discrimination and turnover was performed. Turnover was defined as working less than 50% of the time following onset of WD (year 3).

Finally, due to missing data on covariates, 6% of the observations for the onset analysis and 1% of the observations for the change analysis were excluded from the analysis. Sensitivity analysis for the main analyses were conducted by using mean imputation method in the full dataset in which missing covariates were replaced with their overall estimated mean. Statistical software SAS version 9.4 (SAS Institute Inc., Cary, NC, USA) was used for all the analyses. The level of statistical significance was set at 5%.

Results

Baseline characteristics of the study population are presented in Table 1. Employees were between 18 and 66 years of age at baseline (mean 47 years). The vast majority were women (78%) reflecting the gender-distribution of the Finnish public sector employees. Among the 39,413 employee-observations from the group of 29,597 participants not exposed to WD at the baseline, 1,973 (5%) employee-observations among 1,920 participants indicated WD at the second time-point. The most common cause of WD were opinion, followed by education and position.

Table 1. Baseline characteristics of the Finnish Public Sector Study sample (in the onset analysis).

	All observations	Not discriminated	Onset of discrimination
	(n = 39,413)	(n = 37,440)	(n = 1,973)
*Causes of discrimination*^a
Age-discrimination	552 (3)	-	177 (25)
Sex-discrimination	281 (1)	-	99 (16)
Discrimination because of ethnicity	41 (0.2)	-	19 (3)
Discrimination because of education	792 (4)	-	222 (29)
Discrimination because of position	674 (3)	-	225 (30)
Discrimination because of opinion	1081 (5)	-	375 (41)
Employee characteristics
Women, n (%)	30,870 (78)	29,362 (78)	1,508 (76)
Mean age (SD)	47 (9)	47 (9)	48 (8)
Psychological distress, n (%)	8,759 (22)	8104 (22)	655 (33)
Mean trait anxiety (SD)	1.92 (0.54)	1.91 (0.54)	2.06 (0.62)
Chronic medical conditions^b, n (%)	7,308 (19)	6843 (18)	465 (24)
Obese, n (%)	5,936 (15)	5,559 (15)	377 (19)
High alcohol consumption, n (%)	4,016 (10)	3,816 (10)	200 (10)
Occupational grade, n (%)
High	15,299 (39)	14,608 (39)	691 (35)
Intermediate	9,828 (25)	9,273 (25)	555 (28)
Low	14,286 (36)	13,559 (36)	727 (37)
Temporary employment, n (%)	2,685 (7)	2534 (7)	151 (8)
Shift-work, n (%)	8,495 (22)	8038 (21)	457 (23)
Work-unit characteristics
≥5% work unit temporary employment, n (%)	20,184 (51)	19,147 (51)	1037 (53)
Work units with <20 employees, n (%)	15,045 (38)	14,315 (38)	730 (37)
Gender distribution, n (%)
Female dominated; <33% men	30,005 (76)	28,640 (77)	1365 (69)
Balanced; ≥33% women and men	4,740 (12)	4434 (12)	306 (16)
Male dominated; <33% women	4,668 (12)	4366 (12)	302 (15)

Open in a new tab

^a based on information from 2008 and 2010 where the information was available.

^b ≥1 chronic disorders.

Relative to employees not being discriminated, those experiencing onset of WD were more likely to be male, have an intermediate occupational grade, work in shifts, have more chronic medical disorders, be obese and have more psychological distress at baseline, i.e. before onset of discrimination. The distribution of work unit size was comparable between those experiencing onset of discrimination and those not. Meanwhile, a higher proportion with onset of workplace discrimination worked at male dominated or gender balanced work units, and a lower proportion worked in work units with ≥5% temporary employees.

Workplace discrimination and long-term sickness absence due to mental disorders

Analyses of onset of workplace discrimination

Fig 1 illustrates the prevalence of LTSA due to mental disorders for employees with and without onset of WD from year 0 to year 3. At year 0, where no one was exposed to WD, the prevalence of LTSA were comparable between the two exposure groups, with a prevalence of 2.7% and 2.3% for those with and without later onset of WD, respectively.

The prevalence of LTSA differed markedly between the groups during the 4-year follow-up (Fig 1). Among employees not experiencing WD, the prevalence of LTSA due to a mental disorders was almost constant over the 4-year period. Meanwhile, in employees with onset of WD there was an increase in LTSA due to mental disorders already at year 1. This difference in LTSA levels between the two exposure groups reached its peak at year 2, the time point when the onset of WD was reported (RR = 2.13; 95% CI 1.80–2.52).

Analyses of changes in workplace discrimination

As seen in Table 2, a short-term higher risk associated with WD was also observed in the fixed effect analyses. The odds of concurrent LTSA due to mental disorders were higher (OR = 1.61; 95% CI 1.33–1.96) in the year when WD was reported compared to the year(s) when WD was not reported. This finding did not persist in analyses of 1-year lagged LTSA (OR = 1.02; 95% CI 0.83–1.25).

Table 2. Changes in workplace discrimination for concurrent and 1-year lagged long-term sickness absence.

		Concurrent sickness absence^a		1-year lagged sickness absence^b
	Cause of sickness absence	Observations	OR ^c (95% CI)	Observations	OR ^c (95% CI)
Discrimination vs. no discrimination	Mental	6821	1.61 (1.33–1.96)	6621	1.02 (0.83–1.25)
Discrimination vs. No discrimination	Non-mental	26,784	1.06 (0.95–1.19)	26,906	1.11 (0.99–1.24)

Open in a new tab

^a One or more spells of absence during the exposure year.

^b One or more spells within one year after exposure.

^c Conditional logistic regression analyses adjusted for age, job contract, work-unit gender distribution, work-unit size, work-unit temporary employment.

Workplace discrimination and long-term sickness absence due to non-mental conditions

Analyses of onset of workplace discrimination

As seen in Fig 1, the group with and without onset of WD differed slightly in the prevalence of non-mental LTSA at year 0 (RR = 1.13; 95% CI 1.03–1.25), i.e. before onset of WD. This difference persisted throughout the follow-up as illustrated by the parallel lines for those with and without onset of WD, respectively.

Analyses of changes in workplace discrimination

The results showed comparable occurrence of non-mental LTSA when employees reported exposure to WD and when they did not, as seen in Table 2.

Causes of workplace discrimination and long-term sickness absence

Analyses on the causes of WD showed a similar pattern as the main analyses. No difference in LTSA due to mental disorders were observed at year 0 among employees with and without later onset of WD for all cause, except for discrimination due to education and opinion, where the groups differed already at year 0 (Table 3). The exposure groups diverged at the following years with a higher occurrence of mental LTSA among those experiencing an onset of WD for all cause except sex-discrimination. Risk of LTSA due to mental disorders at year 3 was most pronounced for WD due to ethnicity (RR = 3.20; 95% CI 1.21–8.48).

Table 3. Onset of cause-specific workplace discrimination and annual risk of mental and non-mental diagnosed long-term sickness absence.

		Mental diagnosed sickness absence				Non-mental diagnosed sickness absence
		Year 0	Year 1	Year 2	Year 3	Year 0	Year 1	Year 2	Year 3
	Observations	RR ^a (95% CI)	RR a (95% CI)	RR a (95% CI)	RR a (95% CI)	RR a (95% CI)	RR a (95% CI)	RR a (95% CI)	RR a (95% CI)
Cause-specific workplace discrimination
Age-discrimination
Not discriminated	18756	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)
Onset of discrimination	552	1.18 (0.78–1.80)	1.32 (0.87–2.00)	1.65 (1.13–2.42)	1.68 (1.12–2.51)	1.19 (1.00–1.42)	1.29 (1.09–1.52)	1.43 (1.23–1.67)	1.31 (1.10–1.56)
Sex-discrimination
Not discriminated	18735	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)
Onset of discrimination	281	1.18 (0.66–2.13)	1.45 (0.83–2.51)	1.23 (0.67–2.26)	1.48 (0.83–2.64)	0.90 (0.65–1.25)	1.30 (0.99–1.70)	1.12 (0.85–1.48)	1.15 (0.85–1.55)
Discrimination because of ethnicity
Not discriminated	18691	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)
Onset of discrimination	41	1.27 (0.31–5.23)	0.72 (0.11–4.65)	2.93 (1.13–7.56)	3.20 (1.21–8.48)	0.99 (0.52–1.89)	0.98 (0.50–1.92)	1.33 (0.78–2.29)	2.12 (1.43–3.13)
Discrimination because of education
Not discriminated	18719	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)
Onset of discrimination	792	1.40 (1.02–1.90)	1.49 (1.08–2.04)	2.13 (1.62–2.79)	1.44 (1.02–2.04)	0.98 (0.83–1.16)	1.12 (0.96–1.31)	1.15 (0.99–1.33)	1.18 (1.01–1.38)
Discrimination because of position
Not discriminated	18668	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)
Onset of discrimination	674	1.15 (0.79–1.68)	1.24 (0.84–1.82)	2.04 (1.50–2.78)	1.70 (1.19–2.42)	1.19 (1.01–1.40)	1.17 (0.99–1.38)	1.26 (1.10–1.47)	1.21 (1.02–1.44)
Discrimination because of opinion
Not discriminated	18701	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)
Onset of discrimination	1081	1.38 (1.05–1.83)	1.98 (1.54–2.54)	2.38 (1.88–3.01)	1.65 (1.24–2.21)	1.25 (1.10–1.42)	1.20 (1.05–1.36)	1.22 (1.07–1.38)	1.26 (1.10–1.45)
Additional adjustment for prior long-term sickness absence
Not discriminated	37440	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)	1 (ref.)
Onset of discrimination	1973	1.08 (0.81–1.45)	1.25 (0.93–1.68)	2.03 (1.60–2.57)	1.80 (1.38–2.36)	1.13 (1.00–1.28)	1.15 (1.02–1.31)	1.25 (1.11–1.41)	1.17 (1.03–1.34)

Open in a new tab

^a Log-binomial regression analyses adjusted for the following variables at year 0: Age, sex, chronic disease, psychological distress, anxiety, BMI, alcohol consumption, shift work, employment contract, occupational grade, work-unit size, work-unit temporary employment, work-unit gender distribution as well as data-cycle no.

Results of the fixed effect analyses supported these findings with a higher occurrence of concurrent, but not 1-yr lagged mental LTSA associated with all causes of discrimination apart from sex-discrimination. Noticeably, the occurrence of concurrent mental LTSA was tenfold associated with discrimination due to ethnicity, but with an extreme 95% CI due to the low prevalence of this exposure (OR = 10.56; 95% CI 1.36–82.14) (Table 4).

Table 4. Changes in cause-specific workplace discrimination and odds for concurrent and 1-year lagged long-term sickness absence.

	Mental diagnosed					Non-mental diagnosed
	Concurrent sickness absence^a		1-year lagged sickness absence^b			Concurrent sickness absence^a		1-year lagged sickness absence^b
	Observations	OR ^c (95% CI)	No of observations	OR ^c (95% CI)		No of observations	OR ^c (95% CI)	N of observations	OR ^c (95% CI)
Cause-specific workplace discrimination
Age-discrimination	1766	2.30 (1.19–4.43)	1578	0.88 (0.49–1.58)		7896	1.05 (0.78-1-40)	7588	1.22 (0.95–1.65)
Sex-discrimination	1694	1.49 (0.67–3.33)	1522	0.62 (0.27–1.45)		7566	0.91 (0.58–1.43)	7318	1.33 (0.84–2.10)
Discrimination because of ethnicity	1636	10.56 (1.36–82.14)	1454	1.17 (0.23–5.85)		7380	0.97 (0.45–2.09)	7130	3.09 (1.13–8.44)
Discrimination because of education	1830	1.91 (1.18–3.11)	1640	1.09 (0.66–1.81)		7990	1.00 (0.77-1-31)	7738	1.04 (0.81–1.35)
Discrimination because of position	1818	1.77 (1.13–2.76)	1600	1.09 (0.66–1.81)		7916	0.99 (0.76–1.29)	7668	1.24 (0.95–1.61)
Discrimination because of opinion	2026	1.95 (1.38–2.74)	1798	0.67 (0.47–0.96)		8294	1.00 (0.81–1.23)	8082	1.12 (0.91–1.38)
Additional adjustment for long-term sickness absence one year prior
Discrimination	-	-	6621	0.97 (0.79–1.19)		-	-	26,906	1.11 (0.99–1.24)
Additional adjustment for alcohol consumption, BMI and psychological distress
Discrimination	6576	1.62 (1.33–1.98)	6351		0.98 (0.79–1.20)	25,655	1.05 (0.94–1.18)	25,767	1.09 (0.97–1.23)

Open in a new tab

^a Conditional logistic regression analyses for the following variables at year 0: Age, sex, chronic disease, psychological distress, anxiety, BMI, alcohol consumption, shift work, employment contract, occupational grade, work-unit size, work-unit temporary employment, work-unit gender distribution as well as data-cycle no.

In terms of LTSA due to non-mental conditions, the risk of LTSA was consistently slightly higher among employees with onset of WD due to age, position and opinion compared to those without onset of WD (Table 3). Results pointed to a higher prevalence of LTSA following onset of WD due to ethnicity, with a twofold risk of LTSA at year 3 (RR = 2.12; 95% CI:1.43–3.13). The higher prevalence of LTSA associated with WD due to ethnicity, was also apparent in the fixed effect analysis in the year following exposure (1-year lagged OR = 3.09; 95% CI: 1.13–8.44) (Table 4). Similar to the main analyses, there was no association between WD for any other cause and LTSA due to non-mental conditions.

Supplementary analyses

Additional adjustment for prior LTSA, or the potential time-varying confounders alcohol, BMI and psychological distress, did not change the findings of either the relations with onset of or change in WD, as seen in Tables 2 and 3.

Additional analyses of the potential reverse causation between LTSA and onset of WD supported a likely circular association with mental LTSA. Here LTSA due to mental disorders (but not due to non-mental condition) was associated with higher odds of subsequent exposure to WD (OR = 1.46; 95% CI 1.09–1.97) (S2 Table).

In addressing the potential effect of discrimination on turnover, results pointed to a higher risk of employees working less than 50% of the time in the year following exposure to discrimination (OR = 1.34; 95% CI 1.15–1.56) indicating that employees have left work sometime during that year (S3 Table).

Sensitivity analyses with the full dataset with imputed values for missing covariates replicated the main findings (S3 Fig and S4 Table).

Discussion

In robust models determining both between and within subject differences in LTSA associated with WD, we found that perceived WD was associated with a higher risk of LTSA due to mental, but not non-mental conditions in a large-scale cohort study of Finnish public sector employees. This excess risk was highest for concurrent LTSA during the year when WD was reported, and was particularly strong for discrimination due to ethnicity.

Our results are in accordance with the theoretical hypothesis that workplace discrimination may act as an uncontrollable social stressor triggering stress reactions [9], which may manifest in mental disorders causing spells of LTSA–effects which are evident already within a year of exposure in our analyses. The adverse effect of WD on mental LTSA found in our study is in line with findings from previous cross-sectional studies on self-reported sickness absence [12,13,18–22]. Our findings are also in line with those of a longitudinal study of WD and sickness absence, which addressed age-discrimination only and showed higher rates of sickness absence after a 3-year follow-up among age-discriminated employees [17] and support the increasingly popular view that stressors often act as disease triggers [41].

We add to this literature by looking into all cause as well as specific types of discrimination and by distinguishing between LTSA due to diagnosed mental as well as non-mental conditions. This distinction seem important for understanding the relationship between WD and LTSA. It is for example interesting that we, in a subsample of the population, found that the potential effects are the strongest for discrimination due to ethnicity with higher rates of concurrent mental LTSA, and also higher rates of non-mental LTSA in the year following exposure. Our results also support a general detrimental effect of discrimination on LTSA due to mental disorders, irrespective the cause of discrimination with one exception. We did not observe associations with sex-discrimination which may be due to composition of our study population comprising primarily women.

According to a large-scale review, there is substantial support for an adverse effect of discrimination (not addressing WD per se) on self-reported mental health outcomes such as depressive symptoms, distress and low well-being, although few studies have determined these association using objective measures of diagnosed mental illness [9]. Thus, we add to current evidence by examining timing and changes in workplace discrimination and by showing a strong short-term association between workplace discrimination and medically certified spells of LTSA with an underlying mental diagnosis. The more pronounced associations with mental than non-mental diagnoses of sickness absence support previous studies on discrimination both within [21,42] and outside the workplace setting, pointing to acute or triggering health effects, which are stronger for mental than physical health outcomes [9].

This study has several strengths. The study was based on a large cohort of Finnish public sector employees from a wide range of occupations. The prospective design, use of repeated measures of workplace discrimination and objective and universally covering measure of LTSA including the underlying diagnoses confer other strengths of the study. The use of fixed effect analyses reduced concerns of bias from person-stable characteristics, while the assessment of the effect of onset of discrimination reduced problems of selection. With our longitudinal study design, we were able to minimize reverse causality as a source of bias.

Whereas previous studies on workplace discrimination and health outcomes have primarily been based on self-reported measures of both exposure and outcome, conferring risk of common method bias, we have used a universal and objective measure of LTSA from national registers. The Finnish Social Insurance Institution keeps records on all sickness allowances paid for medically certified LTSA spells of nine or more days for the entire population, thus ensuring a valid record of LTSA independent of the reporting of workplace exposures.

Further, the majority of previous studies have examined the association with exposure at an arbitrary time-point not taking into account that some employees exposed to workplace discrimination will leave their workplace as a means of coping. A mechanism also supported by our findings. To reduce this issue of selection, we have determined the risk of LTSA in association with the onset of discrimination in a study sample free of this exposure at baseline. However, information on workplace discrimination was only available in two year intervals and we have no data on when between two waves the workplace discrimination began. Therefore, to better explicate potential trends in LTSA, we assessed the association with the full trajectory of LTSA covering the period from baseline assessment to one year after reported onset of workplace discrimination. Concurrent and 1-year lagged LTSA was assessed along with the full trajectory of LTSA in the four-year period covering each of the study-cycles, thus providing information on potential acute/triggering and prolonged effects of workplace discrimination. Further, to better accommodate common concerns of person-stable characteristics confounding the relation between workplace exposures and health outcomes, fixed-effects analyses were carried out using each employee as his or her own control.

Some limitations need to be considered when interpreting our findings. First, perceived WD was measured using a single question rather than a multi-item scale. In addition, we had only limited information on the characteristics of WD; for example, no data were available about the source (leader, organization, co-worker), frequency or perceived severity of discrimination. However, the cause of WD (age, sex, ethnicity, education, opinion, or position) was requested in a sub-sample, allowing analyses examining this factor as a potential source for heterogeneity in associations. This adds new insights as previous studies have typically collapsed multiple causes of discrimination into single measure or focused on only a single cause of discrimination, such as age, race or sex [12,17,42,43]. Further studies are warranted to elaborate how different types and sources of WD may be associated with LTSA.

Second, we focused on LTSA in the current study, which have proven a good measure of employee health as well as workplace productivity and commitment. However, workplace discrimination may in addition affect the occurrence of shorter-term absences and/or the total duration of absence. To assess the extent to which workplace discrimination affects exit from the labour market, we carried out a sensitivity analysis of the risk of working less than 50% of the time in the year following onset of discrimination. These results support turnover as a means of coping, thus highlighting an important pit falls of the majority of current evidence on the relation between discrimination and LTSA, which includes only one baseline measure of discrimination neglecting the importance of the timing i.e. duration of exposure.

Lastly, the population comprised public sector employees, primarily female, and the results may not extrapolate directly to other settings such as the private sector in which the context may be different.

In conclusion, we found that workplace discrimination was associated with a higher risk of long-term sickness absence with an underlying mental diagnosis. This excess risk was primarily restricted to spells of sickness absences occurring during the year where workplace discrimination was reported, pointing to potential acute or triggering effects. This study highlights the importance of identifying and combating discriminatory practices at the workplace. We found that self-reported workplace discrimination predisposed employees to health problems and long-term sickness absence, thus being a health risk factor for the employee and a source of burden for employers and organizations in terms of decreased productivity. In Finland the government has introduced a new Non-Discrimination Act in 2014 [44], emphasizing the importance of prevention of discrimination at work. Future studies investigating the impact of this and similar law enforcement as an effective means of reducing long-term sickness absence are warranted.

Supporting information

S1 Fig. Flowchart showing the selection of eligible FPS employees for the analyses of onset of and changes in workplace discrimination and long-term sickness absence.

N; number of employees, n; number of employee-observations, LTSA; long-term sickness absence.

(PDF)

Click here for additional data file.^{(605.7KB, pdf)}

S2 Fig. Directed acyclic graph of the assumed causal network between work-place discrimination and long-term sickness absence.

(DOCX)

Click here for additional data file.^{(221.2KB, docx)}

S3 Fig. Sensitivity analysis by using imputed data on missing covariates.

Prevalence of long-term sickness absence among employees with and without onset of workplace discrimination as well as the risk ratio (RR) and 95% CI’s associated with onset of workplace discrimination.

(DOCX)

Click here for additional data file.^{(4.2MB, docx)}

S1 Table. Measurement time-points for the analyses of onset of and changes in workplace discrimination from the Finnish Public Sector Study.

(PDF)

Click here for additional data file.^{(510.4KB, pdf)}

S2 Table. Supplementary analysis.

Prior long-term sickness absence and odds for onset of discrimination.

(DOCX)

Click here for additional data file.^{(12.9KB, docx)}

S3 Table. Supplementary analysis.

Onset of workplace discrimination and odds for turnover following exposure.

(DOCX)

Click here for additional data file.^{(12.7KB, docx)}

S4 Table. Sensitivity analysis by using imputed data on missing covariates.

Changes in workplace discrimination and odds for concurrent and 1-year lagged long-term sickness absence.

(DOCX)

Click here for additional data file.^{(16.2KB, docx)}

Data Availability

The outcome of this study, long-term sickness absence, is obtained from the Finnish Social Insurance Institution’s register and linked to survey data by using the unique personal identification number. According to the Finnish law, we are not allowed to share identified sensitive health data to other researchers. We are allowed to share anonymised questionnaire data of the Finnish Public Sector Study by application for with bona fide researchers with an established scientific record and bona fide organisations. For information about the Finnish Public Sector Study contact Prof. Mika Kivimaki mika.kivimaki[at]helsinki.fi / PI of the Finnish Public Sector study Dr. Jenni Ervasti jenni.ervasti[at]ttl.fi.

Funding Statement

This work was supported by NordForsk, the Nordic Research Program on Health and Welfare [grant no. 75021 to NHR], Project on Psychosocial Work Environment and Healthy Ageing and the Danish Working Environment Foundation [grant no. 13-2015-09 to NHR]. MK was supported by grants from NordForsk, the UK Medical Research Council [grant no. K013351], the Academy of Finland [grant no. 311492], the Finnish Work Environment Foundation and a Helsinki Institute of Life Science (HILIFE) fellowship. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

References

1.Kivimäki M, Nyberg ST, Batty GD, Fransson EI, Heikkilä K, Alfredsson L, et al. Job strain as a risk factor for coronary heart disease: a collaborative meta-analysis of individual participant data. Lancet (London, England). 2012. Oct;380(9852):1491–7. doi: 10.1016/S0140-6736(12)60994-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Madsen IEH, Nyberg ST, Magnusson Hanson LL, Ferrie JE, Ahola K, Alfredsson L, et al. Job strain as a risk factor for clinical depression: systematic review and meta-analysis with additional individual participant data. Psychol Med. 2017. Jun;47(8):1342–56. doi: 10.1017/S003329171600355X [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Kivimäki M, Jokela M, Nyberg ST, Singh-Manoux A, Fransson EI, Alfredsson L, et al. Long working hours and risk of coronary heart disease and stroke: a systematic review and meta-analysis of published and unpublished data for 603,838 individuals. Lancet (London, England). 2015. Oct;386(10005):1739–46. doi: 10.1016/S0140-6736(15)60295-1 [DOI] [PubMed] [Google Scholar]
4.Dragano N, Siegrist J, Nyberg ST, Lunau T, Fransson EI, Alfredsson L, et al. Effort-Reward Imbalance at Work and Incident Coronary Heart Disease: A Multicohort Study of 90,164 Individuals. Epidemiology. 2017. Jul;28(4):619–26. doi: 10.1097/EDE.0000000000000666 [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Ferrie JE, Virtanen M, Jokela M, Madsen IEH, Heikkilä K, Alfredsson L, et al. Job insecurity and risk of diabetes: a meta-analysis of individual participant data. C Can Med Assoc J = J l’Association medicale Can. 2016. Dec;188(17–18):E447–55. doi: 10.1503/cmaj.150942 [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Official Statistics of Finland (OSF). Quality of work life [e-publication]. [Internet]. ISSN = 2342–2890. 02 2013, 1. Discrimination at workplaces. Helsinki: Statistics Finland; [referred: 9.6.2021]. Available from: http://www.stat.fi/til/tyoolot/2013/02/tyoolot_2013_02_2014-05-15_kat_001_en.html. [Google Scholar]
7.O’Brien KR, McAbee ST, Hebl MR, Rodgers JR. The impact of interpersonal discrimination and stress on health and performance for early career STEM academicians. Front Psychol. 2016;7(APR):1–11. doi: 10.3389/fpsyg.2016.00615 [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Pavalko EK, Mossakowski KN, Hamilton VJ. Does perceived discrimination affect health? Longitudinal relationships between work discrimination and women’s physical and emotional health. J Health Soc Behav [Internet]. 2003;44(1):18–33. Available from: http://www.ncbi.nlm.nih.gov/pubmed/12751308. [PubMed] [Google Scholar]
9.Pascoe EA, Smart-Richman L. Perceived discrimination and health: a meta-analytic review. Psychol Bull [Internet]. 2009;135(4):531–54. Available from: http://psycnet.apa.org/journals/bul/135/4/531/. doi: 10.1037/a0016059 [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Whitaker KM, Everson-Rose SA, Pankow JS, Rodriguez CJ, Lewis TT, Kershaw KN, et al. Experiences of discrimination and incident type 2 diabetes mellitus: The Multi-Ethnic Study of Atherosclerosis (MESA). Am J Epidemiol. 2017;186(4):445–55. doi: 10.1093/aje/kwx047 [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Yuan ASV. Perceived Age Discrimination and Mental Health. Soc Forces [Internet]. 2007;86(1):291–311. Available from: https://academic.oup.com/sf/article-lookup/doi/10.1353/sof.2007.0113. [Google Scholar]
12.Furunes T, Mykletun RJ. Age discrimination in the workplace: Validation of the Nordic Age Discrimination Scale (NADS). Scand J Psychol. 2010;51(1):23–30. doi: 10.1111/j.1467-9450.2009.00738.x [DOI] [PubMed] [Google Scholar]
13.Volpone SD, Avery DR. “It’s self defense: How perceived discrimination promotes employee withdrawal”: Correction to Volpone and Avery (2013). J Occup Health Psychol [Internet]. 2014;19(1):45. Available from: http://doi.apa.org/getdoi.cfm?doi=10.1037/a0035163. [DOI] [PubMed] [Google Scholar]
14.Di Marco D, López-Cabrera R, Arenas A, Giorgi G, Arcangeli G, Mucci N. Approaching the discriminatory work environment as stressor: The protective role of job satisfaction on health. Front Psychol. 2016;7(AUG):1–10. doi: 10.3389/fpsyg.2016.01313 [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Cortina; Magley Vicki J.; Williams Jill Hunter; Langhout Regina Day 1 LM 1. Incivility in the Workplace: Incidence and Impact. J Occup Health Psychol. 2001;6(1):64–80. [PubMed] [Google Scholar]
16.Ensher E, Grant-Vallone E, Donaldson S. Effects of perceived discrimination on job satisfaction, organizational commitment, organizational citizenship behavior, and grievances. Hum Resour Dev Q [Internet]. 2001;12(1):53–72. Available from: . [DOI] [Google Scholar]
17.Viitasalo N, Nätti J. Perceived Age Discrimination at Work and Subsequent Long-Term Sickness Absence Among Finnish Employees. J Occup Environ Med [Internet]. 2015;57(7):801–5. Available from: http://content.wkhealth.com/linkback/openurl?sid=WKPTLP:landingpage&an=00043764-201507000-00014. doi: 10.1097/JOM.0000000000000468 [DOI] [PubMed] [Google Scholar]
18.Svedberg P, Alexanderson K. Associations between sickness absence and harassment, threats, violence, or discrimination: A cross-sectional study of the Swedish Police. Work. 2012;42(1):83–92. doi: 10.3233/WOR-2012-1333 [DOI] [PubMed] [Google Scholar]
19.Slany C, Schütte S, Chastang JF, Parent-Thirion A, Vermeylen G, Niedhammer I. Psychosocial work factors and long sickness absence in Europe. Int J Occup Environ Health. 2014;20(1):16–25. doi: 10.1179/2049396713Y.0000000048 [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Niedhammer I, Chastang JF, Sultan-Taïeb H, Vermeylen G, Parent-Thirion A. Psychosocial work factors and sickness absence in 31 countries in Europe. Eur J Public Health. 2013;23(4):622–9. doi: 10.1093/eurpub/cks124 [DOI] [PubMed] [Google Scholar]
21.Min JY, Park SG, Kim SS, Min KB. Workplace injustice and self-reported disease and absenteeism in South Korea. Am J Ind Med. 2014;57(1):87–96. doi: 10.1002/ajim.22233 [DOI] [PubMed] [Google Scholar]
22.D’Errico A, Costa G. Socio-demographic and work-related risk factors for medium- and long-term sickness absence among Italian workers. Eur J Public Health. 2012;22(5):683–8. doi: 10.1093/eurpub/ckr140 [DOI] [PubMed] [Google Scholar]
23.Marmot M, Feeney A, Shipley M. Sickness absence as a measure of health status and functioning: from the UK Whitehall II study. Epidemiol Community Heal Epidemiol Community Heal [Internet]. 1995;4949:124–30. Available from: http://jech.bmj.com/content/jech/49/2/124.full.pdf. doi: 10.1136/jech.49.2.124 [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Kivimäki M, Head J, Ferrie JE, Shipley MJ, Vahtera J, Marmot MG. Sickness absence as a global measure of health: evidence from mortality in the Whitehall II prospective cohort study. Br Med J [Internet]. 2003;327(7411). Available from: http://www.bmj.com/content/bmj/327/7411/364.full.pdf. doi: 10.1136/bmj.327.7411.364 [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Kivimäki M, Forma P, Wikström J, Halmeenmäki T, Pentti J, Elovainio M, et al. Sickness absence as a risk marker of future disability pension: the 10-town study. J Epidemiol Community Heal [Internet]. 2004;58:710–1. Available from: http://jech.bmj.com/content/jech/58/8/710.full.pdf. doi: 10.1136/jech.2003.015842 [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Whitaker SC. The management of sickness absence. Occup Environ Med [Internet]. 2001;58(6):420–4. Available from: http://oem.bmj.com/content/oemed/58/6/420.full.pdf. doi: 10.1136/oem.58.6.420 [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Persson R, Høgh A, Grynderup MB, Willert MV, Gullander M, Hansen ÅM, et al. Relationship Between Changes in Workplace Bullying Status and the Reporting of Personality Characteristics. J Occup Environ Med [Internet]. 2016;58(9):902–10. Available from: http://insights.ovid.com/crossref?an=00043764-201609000-00009. doi: 10.1097/JOM.0000000000000822 [DOI] [PubMed] [Google Scholar]
28.Aquino K, Bradfield M. Perceived Victimization in the Workplace: The Role of Situational Factors and Victim. Organ Sci [Internet]. 2000;11(5):525–37. Available from: http://www.jstor.org/stable/2640343. [Google Scholar]
29.Einarsen S. The nature, causes and consequences of bullying at work: The Norwegian experience. Perspect Interdiscip sur le Trav la santé [Internet]. 2005;(7–3):0–19. Available from: http://pistes.revues.org/3156. [Google Scholar]
30.Vlasveld MC, Van Der Feltz-Cornelis CM, Anema JR, Van Mechelen W, Beekman ATF, Van Marwijk HWJ, et al. The associations between personality characteristics and absenteeism: A cross-sectional study in workers with and without depressive and anxiety disorders. J Occup Rehabil. 2013;23(3):309–17. doi: 10.1007/s10926-012-9406-9 [DOI] [PubMed] [Google Scholar]
31.Kivimäki M, Vahtera J, Koskenvuo M, Uutela A, Pentti J. Response of hostile individuals to stressful change in their working lives: Test of a psychosocial vulnerability model. Psychol Med. 1998;28(4):903–13. doi: 10.1017/s003329179800659x [DOI] [PubMed] [Google Scholar]
32.Räikkönen K, Matthews KA, Flory JD, Owens JF, Gump BB. Effects of optimism, pessimism, and trait anxiety on ambulatory blood pressure and mood during everyday life. J Pers Soc Psychol [Internet]. 1999;76(1):104–13. Available from: http://doi.apa.org/getdoi.cfm?doi=10.1037/0022-3514.76.1.104. [DOI] [PubMed] [Google Scholar]
33.Gunthert KC, Cohen LH, Armeli S. The role of neuroticism in daily stress and coping. J Pers Soc Psychol [Internet]. 1999;77(5):1087–100. Available from: http://www.ncbi.nlm.nih.gov/pubmed/10573882. doi: 10.1037//0022-3514.77.5.1087 [DOI] [PubMed] [Google Scholar]
34.Salo P, Vahtera J, Hall M, Rod NH, Virtanen M, Pentti J, et al. Using Repeated Measures of Sleep Disturbances to Predict Future Diagnosis-Specific Work Disability: A Cohort Study. Sleep. 2012. doi: 10.5665/sleep.1746 [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Halonen JI, Kivimäki M, Kouvonen A, Pentti J, Kawachi I, Subramanian S V, et al. Proximity to a tobacco store and smoking cessation: a cohort study. Tob Control [Internet]. 2014;23:146–51. Available from: doi: 10.1136/tobaccocontrol-2012-050726 [DOI] [PubMed] [Google Scholar]
36.Lehto A-M. Statistics Finland scale on workplace climate. Quality of working life and equity. 1991. [Google Scholar]
37.Greenland S, Pearl J, Robins JM. Causal diagrams for epidemiologic research. Epidemiology [Internet]. 1999;10(1):37–48. Available from: http://www.ncbi.nlm.nih.gov/pubmed/9888278. [PubMed] [Google Scholar]
38.Holi MM, Marttunen M, Aalberg V. Comparison of the GHQ-36, the GHQ-12 and the SCL-90 as psychiatric screening instruments in the Finnish population. Nord J Psychiatry [Internet]. 2003;57(3):233–8. Available from: http://www.tandfonline.com/doi/full/10.1080/08039480310001418. [DOI] [PubMed] [Google Scholar]
39.Marteau TM, Bekker H. The development of a six‐item short‐form of the state scale of the Spielberger State—Trait Anxiety Inventory (STAI). Br J Clin Psychol. 1992;31(3):301–6. doi: 10.1111/j.2044-8260.1992.tb00997.x [DOI] [PubMed] [Google Scholar]
40.Lipsitz SR, Kim K, Zhao L. Analysis of repeated categorical data using generalized estimating equations. Stat Med [Internet]. 1994;13(11):1149–63. Available from: http://www.ncbi.nlm.nih.gov/pubmed/8091041. doi: 10.1002/sim.4780131106 [DOI] [PubMed] [Google Scholar]
41.Kivimäki M, Steptoe A. Effects of stress on the development and progression of cardiovascular disease. Nat Rev Cardiol [Internet]. 2018;15(4):215–29. Available from: doi: 10.1038/nrcardio.2017.189 [DOI] [PubMed] [Google Scholar]
42.Triana M del C, Jayasinghe M, Pieper JR. Perceived workplace racial discrimination and its correlates: A meta-analysis. J Organ Behav. 2015;36(4):491–513. [Google Scholar]
43.Dovidio JF, Gaertner SL, Kawakami K, Hodson G. Why can’t we just get along? Interpersonal biases and interracial distrust. Cult Divers Ethn Minor Psychol. 2002;8(2):88–102. doi: 10.1037/1099-9809.8.2.88 [DOI] [PubMed] [Google Scholar]
44.Non-discrimination Act (1325/2014) [Internet]. Ministry of Justice, Finland; Available from: https://www.finlex.fi/sv/laki/kaannokset/2014/en20141325.pdf.

PLoS One. doi: 10.1371/journal.pone.0255697.r001

Decision Letter 0

Ali B Mahmoud

30 Apr 2021

PONE-D-21-07363

Workplace discrimination as risk factor for long-term sickness absence: Longitudinal analyses of onset and changes in workplace adversity

PLOS ONE

Dear Dr. Stenholm,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by Jun 14 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Ali B. Mahmoud, Ph.D.

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please provide additional details regarding participant consent. In the ethics statement in the Methods and online submission information, please ensure that you have specified what type you obtained (for instance, written or verbal, and if verbal, how it was documented and witnessed). If your study included minors, state whether you obtained consent from parents or guardians. If the need for consent was waived by the ethics committee, please include this information.

Once you have amended this/these statement(s) in the Methods section of the manuscript, please add the same text to the “Ethics Statement” field of the submission form (via “Edit Submission”).

For additional information about PLOS ONE ethical requirements for human subjects research, please refer to http://journals.plos.org/plosone/s/submission-guidelines#loc-human-subjects-research.

3. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially identifying or sensitive patient information) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. Please see http://www.bmj.com/content/340/bmj.c181.long for guidelines on how to de-identify and prepare clinical data for publication. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

4. Please amend the manuscript submission data (via Edit Submission) to include author Tianwei Xu.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: No

Reviewer #2: Yes

Reviewer #3: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: I Don't Know

Reviewer #2: Yes

Reviewer #3: No

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: No

Reviewer #2: Yes

Reviewer #3: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Thanks for the opportunity to review this paper. I have several queries/clarifications set out below.

Methods

Overall I found the methods and results sections very difficult to read. Many of your paragraphs explaining your methods were just not clear. The problem is compounded possibly because of your use of run on sentences - sentences which run to 5/6/7 lines - and which become increasingly incomprehensible the longer they are. You could try splitting these long sentences into shorter sentences which might improve comprehension. See for example lines 332-337.

Line 126 - you suddenly start discussing number of observations instead of number of participants - and this is displayed in some of your tables. Why have you done this - I could see no explanation provided in your manuscript?

Line 139 - has the workplace discrimination question been used/validated elsewhere? What type of discrimination is it measuring?

Paragraph commencing line 155 is not clear - perhaps you can show this in a diagram? And how do you know your participants didn't experience discrimination in year 0 - was this because the data wasn't collected?

Similarly the paragraph commencing on line 169 is unclear.

Line 172 - what was your cut off for psychological distress using the GHQ? Ditto the TAI?

Results

Line 274. These sentences are not clear - they don't seem to relate to figure 1 - which has more than you are describing in this paragraph - and appears to be set out differently to your depiction?

Reviewer #2: Thank you for the opportunity to review the work entitled “Workplace discrimination as risk factor for long-term sickness absence: Longitudinal analyses of onset and changes in workplace adversity” This paper focuses on a relevant and important topic. In particular, I appreciate the aim to examine the onset of workplace discrimination (WD) and changes of WD on long-term sickness absence (LTSA). I think the methods are convincing. However, the paper has some areas for improvement and clarification. Please see my comments below.

1.The Introduction is well written. The authors have emphasized the lack of longitudinal studies on this topic and the methodological issues of previous studies, however, I believe that the authors need to explain more the rationale behind the study in the introduction section. In the introduction, it’s unclear how workplace discrimination is different from other working environment factors, such as job stress, job demand, or job support. Why workplace discrimination is a practical factor that the authors want to look at? In the discussion, the authors mentioned that workplace discrimination is an uncontrollable social stressor triggering stress reactions, but this ‘social stressor’ theory was not mentioned in the Introduction. Also, I think the introduction can benefit from explaining why focusing on LTSA? What is the government policy for LTSA in Finland, and whether the policy for LTSA is the same for those working in the public sector and those who work in other sectors?

2. Table 1 needs some clarification. For example, the authors said the study population is N=32519, but this number is not shown in Table 1. Also, as described “Of the 29597 participants who were not discriminated at baseline (39,413 employee-observations), 5.0% reported WD at the second time-point.” But this information is not clearly presented in Table 1.

Reviewer #3: Thank you for the opportunity to review this interesting manuscript. Overall the details were presented coherently to approach the aim of the proposed study. I have a few minor comments for the authors to consider:

• Abstract:

o Please clarify if the RR or OR (line 48, line 51) were adjusted or crude ones. It would be good to clarify so by adding ‘adjusted’ or ‘crude’ to the RR or OR.

• Introduction:

o It would be good to clarify the setting of the study, e.g. Finnish evidence around this research topic is lacking, to link the research gap to the research aim. This clarification could be added to the beginning of the last paragraph (line 90), to set the scope of this study so that the readers are not expected worldwide data.

• Methods:

o Line 132: The authors mentioned that “observations with missing information on time-varying covariates (n=1125) were excluded”. Please justify a reason to not perform analyses using the “missing by random” (or other) technique.

o Line 175: The authors mentioned that “directed acyclic graphs” method was used to identify potential confounders. Please include another supplementary graph of the DAG.

o Sensitivity analyses (line 221 onwards): This may be renamed as ‘subgroup analyses’ or ‘supplementary analyses’. In saying that, it would be good to have a sensitivity analyses done on the full dataset and on the missing data dataset (or simulated for the missing variables) to assure the readers if there is any difference in the finding.

o Please clarify if the level of significance was set at 5% (or 1% or lower due to the multiple comparisons).

• Results:

o It would be good to also specify the analyses done (e.g. logistic regression) in the footnotes of the tables (e.g., added to footnote c of Table 2, footnote a of Table 3, etc)

o Sensitivity analyses (line 343 onwards): please update as per suggestions under the Methods section.

o Figure 1: Please add analyses done as a footnote too.

• Discussion

o The limitations of the study may be grouped into a long paragraph, after the paragraph around the strengths of the study.

o The last sentence in the Discussion section (line 445-446: “With our longitudinal study design, we were able to minimize reverse causality as a source of bias.”) could be placed elsewhere (e.g. in line 397) so that the ‘strengths’ of the study go together in a paragraph.

o It would be good to add a sentence around the implication of this research finding (or a few sentences around suggestions for future studies) as the concluding remark.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

PLoS One. 2021 Aug 5;16(8):e0255697. doi: 10.1371/journal.pone.0255697.r002

Author response to Decision Letter 0

16 Jun 2021

Response to Reviewers’ comments:

Reviewer 1

COMMENT 1:

Thanks for the opportunity to review this paper. I have several queries/clarifications set out below.

RESPONSE: We thank the Reviewer for useful comments. Please see our point-by-point responses below.

COMMENT 2:

RESPONSE: Thank you for this suggestion. We have now revised the Methods and Results section by splitting long sentences into shorter ones and by clarifying the sentences. For example, the sentence highlighted by the Reviewer is revised as follows:

“In terms of LTSA due to non-mental conditions, risk of LTSA was consistently slightly higher among employees with onset of WD due to age, position and opinion compared to those without onset of WD (Table 3). Results pointed to a higher prevalence of LTSA following onset of WD due to ethnicity, with a twofold risk of LTSA at year 3 (RR=2.12; 95% CI:1.43-3.13).” (page 16, lines 338-342)

(WD = workplace discrimination; LTSA = long-term sickness absence)

COMMENT 3:

RESPONSE: We agree that this was not clearly described in the original version. There were four data collection waves in the FPS study and participating in two or more waves was required to be included in the analysis. Since some participants had responded to all four waves, they contributed to two data cycles (one cycle includes two waves) and thus two observations as illustrated in Supplemental Table 1. Therefore, the inclusion criteria were fulfilled by 32,519 employees. Some of these participants contributed with several data cycles, totalling of 44,704 observations. This is now illustrated in the Supplemental Figure 1.

We have modified the description of the study population in the Methods section as follows:

“Since some participants had participated in all four study waves, they contributed observations to two separate two-wave data cycles (S1 Table). Given this, the 32,519 participants contributed with a total of 44,704 observations.” (page 5, lines 126-128)

COMMENT 4:

Line 139 - has the workplace discrimination question been used/validated elsewhere? What type of discrimination is it measuring?

RESPONSE: Thank you for this comment. The question about workplace discrimination has been included in the Finnish Public Sector survey since 1997. It originates from the Statistics Finland’s workplace climate survey, which was conducted for the first time in 1991 (Statistics Finland 1991). The question measures participants’ general experiences on discrimination at their workplace. Since 2008 , information regarding potential causes of workplace discrimination has also been inquired (age, sex, ethnicity, education, opinion, and position). However, detailed characteristics of workplace discrimination is unknown, such as the source (leader, organization, co-worker), frequency or perceived severity of discrimination.

We have added a reference to the workplace discrimination measure in the Methods section (page 6) and commented it in the Discussion as follows:

“First, perceived WD was measured using a single question rather than a multi-item scale. In addition, we had only limited information on the characteristics of WD; for example, no data were available from the source (leader, organization, co-worker), frequency, or perceived severity of discrimination. However, the cause of WD (age, sex, ethnicity, education, opinion, or position) was requested in a sub-sample, allowing for analyses examining this factor as a potential source for the heterogeneity in associations. This adds new insights as previous studies have typically collapsed multiple causes of discrimination into a single measure or focused on only a single cause of discrimination, such as sex or race (12, 17, 42, 43). Further studies are warranted to elaborate on how different types and sources of WD may be associated with LTSA.” (page 23, lines 438-448)

Reference:

Statistics Finland scale on workplace climate. Lehto A-M. Quality of working life and equity. Helsinki, Finland: Statistics Finland, 1991.

COMMENT 5:

RESPONSE: Thank you for these comments, we are happy to clarify. In fact, we show the study design and respective measurement points in the S1 Table.

Regarding the analyses of the onset of WD, the inclusion criteria was that the participant at baseline (year 0, i.e. Tx) reported not to be exposed to discrimination. According to the pseudo trial design, some participants reported onset of discrimination and some participants remained free of discrimination at year 2 (i.e. Tx+1). The outcome, prevalence of sickness absence, was assessed annually from year 0 to year 3 (i.e. Tx+2).

Regarding the analyses of change in WD, two different LTSA outcomes were defined. First, a concurrent LTSA was assessed within the year pertaining to the reporting of exposure, that is WD. Second, a 1-year lagged LTSA was assessed as spells initiated during the year following the assessment of exposure.

We now provide a more detail description for both analytical approaches in the Methods section:

“In the analyses with onset of WD as the exposure, the annual risk of LTSA (one or more episodes) for the groups with and without onset of WD was assessed. The risk of LTSA was calculated from the year pertaining to the first reporting of WD i.e. year 0 (where no employees were exposed to discrimination) to year 3 (one year after the second study wave when WD was assessed). The study design is illustrated in the S1 Table.

In analyses with change in WD as the exposure, two different LTSA outcomes were defined: Concurrent LTSA was defined as sickness absence spells within the year pertaining to the reporting of exposure, and 1-year lagged LTSA was defined as sickness absence spells initiated during the year that followed the assessment of exposure. The study design is illustrated in the S1 Table.” (page 7, lines 161-170)

COMMENT 6:

Line 172 - what was your cut off for psychological distress using the GHQ? Ditto the TAI?

RESPONSE: As suggested by Goldberg 1972, the cut off for psychological distress scale was four or more symptoms. For the Trait Anxiety Inventory, we used a continuous scale based on the average response score from the six-item scale.

This is now clarified in the Methods section as follows:

“Psychological distress was assessed using the 12-item General Health Questionnaire and the score was dichotomized (<4 vs ≥4 symptoms) (38). The six-item Trait Anxiety Inventory (39) was used to assess liability to anxiety and was analyzed on a continuous scale based on the average of responses 1-4.“ (page 8, lines 184-187)

References:

Goldberg DP. The detection of psychiatric illness by questionnaire; a technique for the identification and assessment of non-psychotic psychiatric illness. London: Oxford University Press 1972.

Marteau TM, Bekker H. The development of a six‐item short‐form of the state scale of the Spielberger State—Trait Anxiety Inventory (STAI). Br J Clin Psychol. 1992;31(3):301–6.

COMMENT 7:

Results: Line 274. These sentences are not clear - they don't seem to relate to figure 1 - which has more than you are describing in this paragraph - and appears to be set out differently to your depiction?

RESPONSE: We apologise that the description was unclear. We have now revised the text in the Results section to improve clarity:

“The prevalence of LTSA differed markedly between the groups during the 4-year follow-up (Figure 1). Among employees not experiencing WD, the prevalence of LTSA due to a mental disorders was almost constant over the 4-year period. Meanwhile, in employees with onset of WD there was an increase in LTSA due to a mental health disorders already at year 1. This difference in LTSA levels between the two exposure groups reached its peak at year 2, the time point when the onset of WD was reported (RR=2.13; 95% CI 1.80-2.52).” (page 13, lines 293-298)

Reviewer 2

COMMENT 1:

Thank you for the opportunity to review the work entitled “Workplace discrimination as risk factor for long-term sickness absence: Longitudinal analyses of onset and changes in workplace adversity” This paper focuses on a relevant and important topic. In particular, I appreciate the aim to examine the onset of workplace discrimination (WD) and changes of WD on long-term sickness absence (LTSA). I think the methods are convincing. However, the paper has some areas for improvement and clarification. Please see my comments below.

RESPONSE: Thank you for the encouraging comments and useful suggestions. Please see our point-by-point responses below.

COMMENT 2:

The Introduction is well written. The authors have emphasized the lack of longitudinal studies on this topic and the methodological issues of previous studies, however, I believe that the authors need to explain more the rationale behind the study in the introduction section. In the introduction, it’s unclear how workplace discrimination is different from other working environment factors, such as job stress, job demand, or job support. Why workplace discrimination is a practical factor that the authors want to look at? In the discussion, the authors mentioned that workplace discrimination is an uncontrollable social stressor triggering stress reactions, but this ‘social stressor’ theory was not mentioned in the Introduction. Also, I think the introduction can benefit from explaining why focusing on LTSA? What is the government policy for LTSA in Finland, and whether the policy for LTSA is the same for those working in the public sector and those who work in other sectors?

RESPONSE: Thank for these important comments. As suggested by the Reviewer, we have provided a more comprehensive description about the conceptual differences between workplace discrimination and other psychosocial work characteristics in the Introduction as follows:

“A large body of research has linked various adverse psychosocial work characteristics, including job strain, effort-reward imbalance, job insecurity and long working hours, to increased risk of physical and mental chronic conditions (1-5). These factors characterize unhealthy features of work tasks and organization of work. In addition, negative interpersonal relations at work have been hypothesized to predispose employees to ill health.” (page 3, lines 56-60)

We have also emphasized the importance to investigate workplace discrimination, as it is a common social stressor at work:

“Workplace discrimination is a specific social stressors, being typically both unpredictable and uncontrollable. Among the general working population in Finland, 11% of women and 6% of men reported experiencing discrimination at their workplace (6).” (page 3, lines 61-63)

We have also added a justification for LTSA as an outcome as follows:

“Further, prior studies point to a potential link between workplace discrimination and the risk of sickness absence (12, 13, 17-22). Sickness absence, especially when long-term, is considered as a measure of employee health and well-being as well as a correlate of workplace productivity (23-26).’’ (page 3, lines 67-70)

We have also specified the policies related to LTSA in the Methods as follows:

“Using employees’ unique personal identification numbers, we linked participants to the records of medically-certified LTSA from the Finnish Social Insurance Institution register, including dates and diagnoses for LTSA spells lasting 10 days or more. This register covers the entire working population including public and private sector, but does not include sickness absence due to maternity leave and caring for a sick child.” (page 7, lines 153-157)

COMMENT 3:

Table 1 needs some clarification. For example, the authors said the study population is N=32519, but this number is not shown in Table 1. Also, as described “Of the 29597 participants who were not discriminated at baseline (39,413 employee-observations), 5.0% reported WD at the second time-point.” But this information is not clearly presented in Table 1.

RESPONSE: Thank you for these important comments. For the sake of clarity, we have now presented baseline characteristics for the study sample included in the “Onset of workplace discrimination” analysis. There were a total of 39,413 employee-observations obtained from 29,597 participants. We have revised the text as follows:

“Among the 39,413 employee-observations from the group of 29,597 participants not exposed to WD at the baseline, 1,973 (5%) employee-observations among 1,920 participants indicated WD at the second time-point.” (page 12, lines 257-260)

Reviewer 3

COMMENT 1: Thank you for the opportunity to review this interesting manuscript. Overall the details were presented coherently to approach the aim of the proposed study. I have a few minor comments for the authors to consider:

RESPONSE: Thank you for the positive comments and good suggestions. Please see our point-by-point responses below.

COMMENT 2: Abstract: Please clarify if the RR or OR (line 48, line 51) were adjusted or crude ones. It would be good to clarify so by adding ‘adjusted’ or ‘crude’ to the RR or OR.

RESPONSE: As suggested by the Reviewer, we have clarified in the abstract that RRs and ORs are based on adjusted models.

COMMENT 3: Introduction: It would be good to clarify the setting of the study, e.g. Finnish evidence around this research topic is lacking, to link the research gap to the research aim. This clarification could be added to the beginning of the last paragraph (line 90), to set the scope of this study so that the readers are not expected worldwide data.

RESPONSE: Thank you for this suggestion. We have supplemented the last paragraph in the Introduction to address also context of the study as follows:

“To address these two methodological shortcomings, we combined longitudinal survey with register data from national registries from Finland and applied robust methodological approaches in assessing the associations of onset of WD and changes in WD with risk of subsequent long-term sickness absence (LTSA).” (page 4, lines 95-98)

COMMENT 4: Methods: Line 132: The authors mentioned that “observations with missing information on time-varying covariates (n=1125) were excluded”. Please justify a reason to not perform analyses using the “missing by random” (or other) technique.

RESPONSE: Thank you for this comment. Indeed, we excluded observations with missing information on covariates. However, 1125 represents only 1.3% of the entire number of observations (n=89,408) included in the “Change in workplace discrimination” analysis. This is also shown in the flowchart S1 Fig. Due to the very small proportion of observations, excluding them did not influence the results of this study.

To illustrate the role of missing observations in the main findings, we imputed values for missing covariates and repeated the main analysis. These results are now shown in the S2 Figure and S4 Table indicating that the main results were replicated and only minor differences in the second decimal of the RRs or ORs and their 95% CIs were observed.

COMMENT 5: Line 175: The authors mentioned that “directed acyclic graphs” method was used to identify potential confounders. Please include another supplementary graph of the DAG.

RESPONSE: We have now included the DAG in the supplementary material (S2 Figure).

COMMENT 6: Sensitivity analyses (line 221 onwards): This may be renamed as ‘subgroup analyses’ or ‘supplementary analyses’. In saying that, it would be good to have a sensitivity analyses done on the full dataset and on the missing data dataset (or simulated for the missing variables) to assure the readers if there is any difference in the finding.

RESPONSE: As suggested by the Reviewer, we have replaced term “Sensitivity analyses” by “Supplementary analyses” in the Statistical Analysis and Results sections. In addition, as suggested we have conducted sensitivity analysis with the full dataset by imputing values for missing covariates. This is now described in the Statistical Analysis section as follows:

“Finally, due to missing data on covariates, 6% of the observations for the onset analysis and 1% of the observations for the change analysis were excluded from these analyses. Sensitivity analysis for the main analyses were conducted by using mean imputation method in the full dataset in which missing covariates were replaced with their overall estimated mean.” (page 11, lines 246-249)

The results are shown in the S2 Figure and S4 Table indicating that the main results were replicated and only minor differences in the second decimal of the RRs or ORs and their 95% CIs were observed. This is also mentioned in the Results section as follows:

“Sensitivity analyses with the full dataset with imputed values for missing covariates replicated the main findings (S2 Figure and S4 Table).” (page 20, lines 368-369)

COMMENT 7: Please clarify if the level of significance was set at 5% (or 1% or lower due to the multiple comparisons).

RESPONSE: We have clarified this in the Statistical Analysis section as follows:

“The level of statistical significance was set at 5%.” (page 11, line 251)

COMMENT 8: Results: It would be good to also specify the analyses done (e.g. logistic regression) in the footnotes of the tables (e.g., added to footnote c of Table 2, footnote a of Table 3, etc).

RESPONSE: As suggested by the Reviewer, we have specified the analytical methods in the footnotes of the tables (Table 2, Table 3, Table 4, S2 Table, S3 Table and S4 Table).

COMMENT 9: Sensitivity analyses (line 343 onwards): please update as per suggestions under the Methods section.

RESPONSE: As suggested by the Reviewer, we have replaced term “Sensitivity analyses” by “Supplementary analyses” in the Statistical Analysis and Results sections.

COMMENT 10: Figure 1: Please add analyses done as a footnote too.

RESPONSE: As suggested by the Reviewer, we have specified the analytical methods in the Figure 1 footnote.

COMMENT 11: Discussion: The limitations of the study may be grouped into a long paragraph, after the paragraph around the strengths of the study.

RESPONSE: Thank you for this suggestion. We have grouped the limitations of the study into longer section towards the end of discussion section.

COMMENT 12: The last sentence in the Discussion section (line 445-446: “With our longitudinal study design, we were able to minimize reverse causality as a source of bias.”) could be placed elsewhere (e.g. in line 397) so that the ‘strengths’ of the study go together in a paragraph.

RESPONSE: As suggested, we have moved the sentence earlier in the Discussion (page 22, lines 413-414).

COMMENT 13: It would be good to add a sentence around the implication of this research finding (or a few sentences around suggestions for future studies) as the concluding remark.

RESPONSE: As suggested by the Reviewer, we have elaborated on the implications of the study by the end of the manuscript as follows:

“This study highlights the importance of identifying and combating discriminatory practices at the workplace. We found that self-reported workplace discrimination predisposed employees to health problems and long-term sickness absence, thus being a health risk factor for the employee and a source of burden for employers and organizations in terms of lowered productivity. In Finland the government has introduced a new Non-Discrimination Act in 2014 (44), emphasizing the importance of prevention of discrimination at work. Future studies investigating the impact of this and similar law enforcement as an effective means of reducing long-term sickness absence are warranted.” (page 24, lines 464-472)

Journal Requirements

POINT 1: Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming.

RESPONSE: We have modified the manuscript and it now meets the journal’s style requirements.

POINT 2: Please provide additional details regarding participant consent. In the ethics statement in the Methods and online submission information, please ensure that you have specified what type you obtained (for instance, written or verbal, and if verbal, how it was documented and witnessed). If your study included minors, state whether you obtained consent from parents or guardians. If the need for consent was waived by the ethics committee, please include this information.

RESPONSE: In the questionnaire participants were provided information about the study and response to the questionnaire was considered as a consent to participate in the study.

We have specified the participant consent in the Methods section as follows:

“Participants were provided information about the study and response to the survey questionnaire was considered as a consent to participate.” (page 5, lines 115-116)

POINT 3:. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

We will update your Data Availability statement on your behalf to reflect the information you provide.

RESPONSE: The outcome of this study, long-term sickness absence, is obtained from the Finnish Social Insurance Institution’s register and linked to survey data by using the unique personal identification number. According to the Finnish law, we are not allowed to share identified sensitive health data to other researchers. We have also explained this in the Cover letter.

POINT 4: Please amend the manuscript submission data (via Edit Submission) to include author Tianwei Xu.

RESPONSE: Thank you for noticing this mistake. We have now included Tianwei Xu in the submission system.

Attachment

Submitted filename: Response PlosOne R1_submitted 09062021.docx

Click here for additional data file.^{(48.8KB, docx)}

PLoS One. doi: 10.1371/journal.pone.0255697.r003

Decision Letter 1

Ali B Mahmoud

22 Jul 2021

Workplace discrimination as risk factor for long-term sickness absence: Longitudinal analyses of onset and changes in workplace adversity

PONE-D-21-07363R1

Dear Dr. Stenholm,

I'm pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Ali B. Mahmoud, Ph.D.

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

Reviewer #3: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

6. Review Comments to the Author

Reviewer #1: all my earlier comments were addressed appropriately..........................................................

Reviewer #2: The manuscript has been thoroughly revised and my concerns and suggestions have been addressed. In particular, the introduction and method sections are now much clearer.

Reviewer #3: Thank you for addressing my previous comments and improving the quality of this revised manuscript.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

PLoS One. doi: 10.1371/journal.pone.0255697.r004

Acceptance letter

Ali B Mahmoud

27 Jul 2021

PONE-D-21-07363R1

Workplace discrimination as risk factor for long-term sickness absence: Longitudinal analyses of onset and changes in workplace adversity

Dear Dr. Stenholm:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Ali B. Mahmoud

Academic Editor

PLOS ONE

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

S1 Fig. Flowchart showing the selection of eligible FPS employees for the analyses of onset of and changes in workplace discrimination and long-term sickness absence.

N; number of employees, n; number of employee-observations, LTSA; long-term sickness absence.

(PDF)

Click here for additional data file.^{(605.7KB, pdf)}

S2 Fig. Directed acyclic graph of the assumed causal network between work-place discrimination and long-term sickness absence.

(DOCX)

Click here for additional data file.^{(221.2KB, docx)}

S3 Fig. Sensitivity analysis by using imputed data on missing covariates.

(DOCX)

Click here for additional data file.^{(4.2MB, docx)}

S1 Table. Measurement time-points for the analyses of onset of and changes in workplace discrimination from the Finnish Public Sector Study.

(PDF)

Click here for additional data file.^{(510.4KB, pdf)}

S2 Table. Supplementary analysis.

Prior long-term sickness absence and odds for onset of discrimination.

(DOCX)

Click here for additional data file.^{(12.9KB, docx)}

S3 Table. Supplementary analysis.

Onset of workplace discrimination and odds for turnover following exposure.

(DOCX)

Click here for additional data file.^{(12.7KB, docx)}

S4 Table. Sensitivity analysis by using imputed data on missing covariates.

Changes in workplace discrimination and odds for concurrent and 1-year lagged long-term sickness absence.

(DOCX)

Click here for additional data file.^{(16.2KB, docx)}

Attachment

Submitted filename: Response PlosOne R1_submitted 09062021.docx

Click here for additional data file.^{(48.8KB, docx)}

Data Availability Statement

[pone.0255697.ref001] 1.Kivimäki M, Nyberg ST, Batty GD, Fransson EI, Heikkilä K, Alfredsson L, et al. Job strain as a risk factor for coronary heart disease: a collaborative meta-analysis of individual participant data. Lancet (London, England). 2012. Oct;380(9852):1491–7. doi: 10.1016/S0140-6736(12)60994-5 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0255697.ref002] 2.Madsen IEH, Nyberg ST, Magnusson Hanson LL, Ferrie JE, Ahola K, Alfredsson L, et al. Job strain as a risk factor for clinical depression: systematic review and meta-analysis with additional individual participant data. Psychol Med. 2017. Jun;47(8):1342–56. doi: 10.1017/S003329171600355X [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0255697.ref003] 3.Kivimäki M, Jokela M, Nyberg ST, Singh-Manoux A, Fransson EI, Alfredsson L, et al. Long working hours and risk of coronary heart disease and stroke: a systematic review and meta-analysis of published and unpublished data for 603,838 individuals. Lancet (London, England). 2015. Oct;386(10005):1739–46. doi: 10.1016/S0140-6736(15)60295-1 [DOI] [PubMed] [Google Scholar]

[pone.0255697.ref004] 4.Dragano N, Siegrist J, Nyberg ST, Lunau T, Fransson EI, Alfredsson L, et al. Effort-Reward Imbalance at Work and Incident Coronary Heart Disease: A Multicohort Study of 90,164 Individuals. Epidemiology. 2017. Jul;28(4):619–26. doi: 10.1097/EDE.0000000000000666 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0255697.ref005] 5.Ferrie JE, Virtanen M, Jokela M, Madsen IEH, Heikkilä K, Alfredsson L, et al. Job insecurity and risk of diabetes: a meta-analysis of individual participant data. C Can Med Assoc J = J l’Association medicale Can. 2016. Dec;188(17–18):E447–55. doi: 10.1503/cmaj.150942 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0255697.ref006] 6.Official Statistics of Finland (OSF). Quality of work life [e-publication]. [Internet]. ISSN = 2342–2890. 02 2013, 1. Discrimination at workplaces. Helsinki: Statistics Finland; [referred: 9.6.2021]. Available from: http://www.stat.fi/til/tyoolot/2013/02/tyoolot_2013_02_2014-05-15_kat_001_en.html. [Google Scholar]

[pone.0255697.ref007] 7.O’Brien KR, McAbee ST, Hebl MR, Rodgers JR. The impact of interpersonal discrimination and stress on health and performance for early career STEM academicians. Front Psychol. 2016;7(APR):1–11. doi: 10.3389/fpsyg.2016.00615 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0255697.ref008] 8.Pavalko EK, Mossakowski KN, Hamilton VJ. Does perceived discrimination affect health? Longitudinal relationships between work discrimination and women’s physical and emotional health. J Health Soc Behav [Internet]. 2003;44(1):18–33. Available from: http://www.ncbi.nlm.nih.gov/pubmed/12751308. [PubMed] [Google Scholar]

[pone.0255697.ref009] 9.Pascoe EA, Smart-Richman L. Perceived discrimination and health: a meta-analytic review. Psychol Bull [Internet]. 2009;135(4):531–54. Available from: http://psycnet.apa.org/journals/bul/135/4/531/. doi: 10.1037/a0016059 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0255697.ref010] 10.Whitaker KM, Everson-Rose SA, Pankow JS, Rodriguez CJ, Lewis TT, Kershaw KN, et al. Experiences of discrimination and incident type 2 diabetes mellitus: The Multi-Ethnic Study of Atherosclerosis (MESA). Am J Epidemiol. 2017;186(4):445–55. doi: 10.1093/aje/kwx047 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0255697.ref011] 11.Yuan ASV. Perceived Age Discrimination and Mental Health. Soc Forces [Internet]. 2007;86(1):291–311. Available from: https://academic.oup.com/sf/article-lookup/doi/10.1353/sof.2007.0113. [Google Scholar]

[pone.0255697.ref012] 12.Furunes T, Mykletun RJ. Age discrimination in the workplace: Validation of the Nordic Age Discrimination Scale (NADS). Scand J Psychol. 2010;51(1):23–30. doi: 10.1111/j.1467-9450.2009.00738.x [DOI] [PubMed] [Google Scholar]

[pone.0255697.ref013] 13.Volpone SD, Avery DR. “It’s self defense: How perceived discrimination promotes employee withdrawal”: Correction to Volpone and Avery (2013). J Occup Health Psychol [Internet]. 2014;19(1):45. Available from: http://doi.apa.org/getdoi.cfm?doi=10.1037/a0035163. [DOI] [PubMed] [Google Scholar]

[pone.0255697.ref014] 14.Di Marco D, López-Cabrera R, Arenas A, Giorgi G, Arcangeli G, Mucci N. Approaching the discriminatory work environment as stressor: The protective role of job satisfaction on health. Front Psychol. 2016;7(AUG):1–10. doi: 10.3389/fpsyg.2016.01313 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0255697.ref015] 15.Cortina; Magley Vicki J.; Williams Jill Hunter; Langhout Regina Day 1 LM 1. Incivility in the Workplace: Incidence and Impact. J Occup Health Psychol. 2001;6(1):64–80. [PubMed] [Google Scholar]

[pone.0255697.ref016] 16.Ensher E, Grant-Vallone E, Donaldson S. Effects of perceived discrimination on job satisfaction, organizational commitment, organizational citizenship behavior, and grievances. Hum Resour Dev Q [Internet]. 2001;12(1):53–72. Available from: . [DOI] [Google Scholar]

[pone.0255697.ref017] 17.Viitasalo N, Nätti J. Perceived Age Discrimination at Work and Subsequent Long-Term Sickness Absence Among Finnish Employees. J Occup Environ Med [Internet]. 2015;57(7):801–5. Available from: http://content.wkhealth.com/linkback/openurl?sid=WKPTLP:landingpage&an=00043764-201507000-00014. doi: 10.1097/JOM.0000000000000468 [DOI] [PubMed] [Google Scholar]

[pone.0255697.ref018] 18.Svedberg P, Alexanderson K. Associations between sickness absence and harassment, threats, violence, or discrimination: A cross-sectional study of the Swedish Police. Work. 2012;42(1):83–92. doi: 10.3233/WOR-2012-1333 [DOI] [PubMed] [Google Scholar]

[pone.0255697.ref019] 19.Slany C, Schütte S, Chastang JF, Parent-Thirion A, Vermeylen G, Niedhammer I. Psychosocial work factors and long sickness absence in Europe. Int J Occup Environ Health. 2014;20(1):16–25. doi: 10.1179/2049396713Y.0000000048 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0255697.ref020] 20.Niedhammer I, Chastang JF, Sultan-Taïeb H, Vermeylen G, Parent-Thirion A. Psychosocial work factors and sickness absence in 31 countries in Europe. Eur J Public Health. 2013;23(4):622–9. doi: 10.1093/eurpub/cks124 [DOI] [PubMed] [Google Scholar]

[pone.0255697.ref021] 21.Min JY, Park SG, Kim SS, Min KB. Workplace injustice and self-reported disease and absenteeism in South Korea. Am J Ind Med. 2014;57(1):87–96. doi: 10.1002/ajim.22233 [DOI] [PubMed] [Google Scholar]

[pone.0255697.ref022] 22.D’Errico A, Costa G. Socio-demographic and work-related risk factors for medium- and long-term sickness absence among Italian workers. Eur J Public Health. 2012;22(5):683–8. doi: 10.1093/eurpub/ckr140 [DOI] [PubMed] [Google Scholar]

[pone.0255697.ref023] 23.Marmot M, Feeney A, Shipley M. Sickness absence as a measure of health status and functioning: from the UK Whitehall II study. Epidemiol Community Heal Epidemiol Community Heal [Internet]. 1995;4949:124–30. Available from: http://jech.bmj.com/content/jech/49/2/124.full.pdf. doi: 10.1136/jech.49.2.124 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0255697.ref024] 24.Kivimäki M, Head J, Ferrie JE, Shipley MJ, Vahtera J, Marmot MG. Sickness absence as a global measure of health: evidence from mortality in the Whitehall II prospective cohort study. Br Med J [Internet]. 2003;327(7411). Available from: http://www.bmj.com/content/bmj/327/7411/364.full.pdf. doi: 10.1136/bmj.327.7411.364 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0255697.ref025] 25.Kivimäki M, Forma P, Wikström J, Halmeenmäki T, Pentti J, Elovainio M, et al. Sickness absence as a risk marker of future disability pension: the 10-town study. J Epidemiol Community Heal [Internet]. 2004;58:710–1. Available from: http://jech.bmj.com/content/jech/58/8/710.full.pdf. doi: 10.1136/jech.2003.015842 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0255697.ref026] 26.Whitaker SC. The management of sickness absence. Occup Environ Med [Internet]. 2001;58(6):420–4. Available from: http://oem.bmj.com/content/oemed/58/6/420.full.pdf. doi: 10.1136/oem.58.6.420 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0255697.ref027] 27.Persson R, Høgh A, Grynderup MB, Willert MV, Gullander M, Hansen ÅM, et al. Relationship Between Changes in Workplace Bullying Status and the Reporting of Personality Characteristics. J Occup Environ Med [Internet]. 2016;58(9):902–10. Available from: http://insights.ovid.com/crossref?an=00043764-201609000-00009. doi: 10.1097/JOM.0000000000000822 [DOI] [PubMed] [Google Scholar]

[pone.0255697.ref028] 28.Aquino K, Bradfield M. Perceived Victimization in the Workplace: The Role of Situational Factors and Victim. Organ Sci [Internet]. 2000;11(5):525–37. Available from: http://www.jstor.org/stable/2640343. [Google Scholar]

[pone.0255697.ref029] 29.Einarsen S. The nature, causes and consequences of bullying at work: The Norwegian experience. Perspect Interdiscip sur le Trav la santé [Internet]. 2005;(7–3):0–19. Available from: http://pistes.revues.org/3156. [Google Scholar]

[pone.0255697.ref030] 30.Vlasveld MC, Van Der Feltz-Cornelis CM, Anema JR, Van Mechelen W, Beekman ATF, Van Marwijk HWJ, et al. The associations between personality characteristics and absenteeism: A cross-sectional study in workers with and without depressive and anxiety disorders. J Occup Rehabil. 2013;23(3):309–17. doi: 10.1007/s10926-012-9406-9 [DOI] [PubMed] [Google Scholar]

[pone.0255697.ref031] 31.Kivimäki M, Vahtera J, Koskenvuo M, Uutela A, Pentti J. Response of hostile individuals to stressful change in their working lives: Test of a psychosocial vulnerability model. Psychol Med. 1998;28(4):903–13. doi: 10.1017/s003329179800659x [DOI] [PubMed] [Google Scholar]

[pone.0255697.ref032] 32.Räikkönen K, Matthews KA, Flory JD, Owens JF, Gump BB. Effects of optimism, pessimism, and trait anxiety on ambulatory blood pressure and mood during everyday life. J Pers Soc Psychol [Internet]. 1999;76(1):104–13. Available from: http://doi.apa.org/getdoi.cfm?doi=10.1037/0022-3514.76.1.104. [DOI] [PubMed] [Google Scholar]

[pone.0255697.ref033] 33.Gunthert KC, Cohen LH, Armeli S. The role of neuroticism in daily stress and coping. J Pers Soc Psychol [Internet]. 1999;77(5):1087–100. Available from: http://www.ncbi.nlm.nih.gov/pubmed/10573882. doi: 10.1037//0022-3514.77.5.1087 [DOI] [PubMed] [Google Scholar]

[pone.0255697.ref034] 34.Salo P, Vahtera J, Hall M, Rod NH, Virtanen M, Pentti J, et al. Using Repeated Measures of Sleep Disturbances to Predict Future Diagnosis-Specific Work Disability: A Cohort Study. Sleep. 2012. doi: 10.5665/sleep.1746 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0255697.ref035] 35.Halonen JI, Kivimäki M, Kouvonen A, Pentti J, Kawachi I, Subramanian S V, et al. Proximity to a tobacco store and smoking cessation: a cohort study. Tob Control [Internet]. 2014;23:146–51. Available from: doi: 10.1136/tobaccocontrol-2012-050726 [DOI] [PubMed] [Google Scholar]

[pone.0255697.ref036] 36.Lehto A-M. Statistics Finland scale on workplace climate. Quality of working life and equity. 1991. [Google Scholar]

[pone.0255697.ref037] 37.Greenland S, Pearl J, Robins JM. Causal diagrams for epidemiologic research. Epidemiology [Internet]. 1999;10(1):37–48. Available from: http://www.ncbi.nlm.nih.gov/pubmed/9888278. [PubMed] [Google Scholar]

[pone.0255697.ref038] 38.Holi MM, Marttunen M, Aalberg V. Comparison of the GHQ-36, the GHQ-12 and the SCL-90 as psychiatric screening instruments in the Finnish population. Nord J Psychiatry [Internet]. 2003;57(3):233–8. Available from: http://www.tandfonline.com/doi/full/10.1080/08039480310001418. [DOI] [PubMed] [Google Scholar]

[pone.0255697.ref039] 39.Marteau TM, Bekker H. The development of a six‐item short‐form of the state scale of the Spielberger State—Trait Anxiety Inventory (STAI). Br J Clin Psychol. 1992;31(3):301–6. doi: 10.1111/j.2044-8260.1992.tb00997.x [DOI] [PubMed] [Google Scholar]

[pone.0255697.ref040] 40.Lipsitz SR, Kim K, Zhao L. Analysis of repeated categorical data using generalized estimating equations. Stat Med [Internet]. 1994;13(11):1149–63. Available from: http://www.ncbi.nlm.nih.gov/pubmed/8091041. doi: 10.1002/sim.4780131106 [DOI] [PubMed] [Google Scholar]

[pone.0255697.ref041] 41.Kivimäki M, Steptoe A. Effects of stress on the development and progression of cardiovascular disease. Nat Rev Cardiol [Internet]. 2018;15(4):215–29. Available from: doi: 10.1038/nrcardio.2017.189 [DOI] [PubMed] [Google Scholar]

[pone.0255697.ref042] 42.Triana M del C, Jayasinghe M, Pieper JR. Perceived workplace racial discrimination and its correlates: A meta-analysis. J Organ Behav. 2015;36(4):491–513. [Google Scholar]

[pone.0255697.ref043] 43.Dovidio JF, Gaertner SL, Kawakami K, Hodson G. Why can’t we just get along? Interpersonal biases and interracial distrust. Cult Divers Ethn Minor Psychol. 2002;8(2):88–102. doi: 10.1037/1099-9809.8.2.88 [DOI] [PubMed] [Google Scholar]

[pone.0255697.ref044] 44.Non-discrimination Act (1325/2014) [Internet]. Ministry of Justice, Finland; Available from: https://www.finlex.fi/sv/laki/kaannokset/2014/en20141325.pdf.

PERMALINK

Workplace discrimination as risk factor for long-term sickness absence: Longitudinal analyses of onset and changes in workplace adversity

Alice Clark

Sari Stenholm

Jaana Pentti

Paula Salo

Theis Lange

Eszter Török

Tianwei Xu

Jesper Fabricius

Tuula Oksanen

Mika Kivimäki

Jussi Vahtera

Naja Hulvej Rod

Roles

Abstract

Introduction

Materials and methods

Study population

Workplace discrimination

Long-term sickness absence

Covariates

Statistical analysis

Analyses of onset of workplace discrimination

Analyses of changes in workplace discrimination

Analyses of cause-specific workplace discrimination

Supplementary analyses

Results

Table 1. Baseline characteristics of the Finnish Public Sector Study sample (in the onset analysis).

Workplace discrimination and long-term sickness absence due to mental disorders

Analyses of onset of workplace discrimination

Fig 1. Prevalence of long-term sickness absence among employees with and without onset of workplace discrimination as well as the risk ratio (RR) and 95% CI’s associated with onset of workplace discrimination.

Analyses of changes in workplace discrimination

Table 2. Changes in workplace discrimination for concurrent and 1-year lagged long-term sickness absence.

Workplace discrimination and long-term sickness absence due to non-mental conditions

Analyses of onset of workplace discrimination

Analyses of changes in workplace discrimination

Causes of workplace discrimination and long-term sickness absence

Table 3. Onset of cause-specific workplace discrimination and annual risk of mental and non-mental diagnosed long-term sickness absence.

Table 4. Changes in cause-specific workplace discrimination and odds for concurrent and 1-year lagged long-term sickness absence.

Supplementary analyses

Discussion

Supporting information

Data Availability

Funding Statement

References

Decision Letter 0

Ali B Mahmoud

Roles

Author response to Decision Letter 0

Decision Letter 1

Ali B Mahmoud

Roles

Acceptance letter

Ali B Mahmoud

Roles

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases