Beskæftigelsesudvalget 2020-21
BEU Alm.del Bilag 444
Offentligt
2452965_0001.png
Articles
Emotional demands at work and risk of long-term sickness
absence in 1·5 million employees in Denmark: a prospective
cohort study on effect modifiers
Elisabeth Framke, Jeppe Karl Sørensen, Kristina Alexanderson, Kristin Farrants, Mika Kivimäki, Solja T Nyberg, Jacob Pedersen, Ida E H Madsen,
Reiner Rugulies
Summary
Background
High emotional demands at work can affect employees’ health and there is a need to understand whether
such an association might be modified by other working conditions. We aimed to examine emotional demands at
work as a risk factor for long-term sickness absence and analyse whether influence, possibilities for development, role
conflicts, and physical demands at work might modify this risk.
Methods
We did a nationwide, population-based, prospective cohort study in Denmark and included employed
individuals who were residing in Denmark in 2000, aged 30–59 years, who had complete data on age, sex, and
migration background, with information on emotional demands and possible effect modifiers from job exposure
matrices, and covariates and outcome (sickness absence) from population registers. Individuals with long-term
sickness absence (≥6 weeks of consecutive sickness absence) between Jan 1, 1998, and Dec 31, 2000, and self-employed
individuals were excluded. We assessed long-term sickness absence during a 10-year period from Jan 1, 2001, to
Dec 31, 2010. Using Cox regression, we estimated hazard ratios (HRs) and 95% CIs and tested interaction as departure
from additivity, estimating relative excess risk due to interaction (RERI). Multivariable adjusted models included sex,
age, cohabitation, migration background, and income.
Findings
1 521 352 employed individuals were included and contributed data between Jan 1, 2000, and Dec 31, 2010.
During 11 919 021 person-years (mean follow-up 7·8 years), we identified 480 685 new cases of long-term sickness
absence. High emotional demands were associated with increased risk of long-term sickness absence compared with
low emotional demands, after adjusting for age, sex, cohabitation, migration background, income, and the
four possible effect modifiers (adjusted HR 1·55 [95% CI 1·53–1·56]). The association between high emotional
demands and risk of long-term sickness absence was stronger in a synergistic way when individuals were also exposed
to low possibilities for development (RERI 0·35 [95% CI 0·22–0·47]; 28·9 additional cases per 1000 person-years) and
high role conflicts (0·13 [0·11–0·15]; 22·0 additional cases per 1000 person-years). No synergy was observed for
influence and physical demands at work.
Interpretation
People in occupations with high emotional demands were at increased risk of long-term sickness
absence. Our findings on synergistic interactions suggest that, in emotionally demanding occupations, increasing
possibilities for development and reducing work-related role conflicts might reduce long-term sickness absence.
Further interventional studies are needed to confirm or refute this hypothesis.
Funding
Danish Work Environment Research Fund, NordForsk.
Copyright
© 2021 The Author(s). Published by Elsevier Ltd. This is an Open Access article under CC BY-NC-ND 4.0
license.
Lancet Public Health
2021;
6: e752–59
National Research Centre for
the Working Environment,
Copenhagen, Denmark
(E Framke PhD, J K Sørensen MSc,
J Pedersen PhD,
I E H Madsen PhD,
Prof R Rugulies PhD);
Division of
Insurance Medicine,
Department of Clinical
Neuroscience, Karolinska
Institutet, Stockholm, Sweden
(Prof K Alexanderson PhD,
K Farrants PhD);
Department of
Epidemiology and Public
Health, University College
London, London, UK
(Prof M Kivimäki FMedSci);
Department of Public Health,
Clinicum, Faculty of Medicine,
University of Helsinki, Helsinki,
Finland
(Prof M Kivimäki,
S T Nyberg PhD);
Finnish
Institute of Occupational
Health, Helsinki, Finland
(Prof M Kivimäki);
Department
of Public Health and
Department of Psychology,
University of Copenhagen,
Copenhagen, Denmark
(Prof R Rugulies)
Correspondence to:
Dr Elisabeth Framke,
National Research Centre for
the Working Environment,
Copenhagen DK-2100, Denmark
[email protected]
Introduction
Emotional demands at work concern aspects of work that
require sustained emotional effort from employees.
1,2
High levels of emotional demands characterise some
professions, such as health care, education, and social
work.
3,4
Clients in such professions often include
individuals in difficult or distressing situations; for
example, patients with severe illnesses, older patients in
dementia units, or children living in socially disad­
vantaged circumstances. Public service work requires
empathy and some emotional involvement with the
clients,
5,6
and the emotional demands of such work has
www.thelancet.com/public-health
Vol 6 October 2021
been linked with poor mental health.
7
Emotional
demands at work have also been associated with an
increased risk of long­term sickness absence.
4,8–10
Concerns have been raised that these associations
might be inflated by reporting bias in the assessment of
emotional demands.
4
One previous study reported a
higher risk of long­term sickness absence in those with
high levels of emotional demands at work when
addressing this bias by comparing perceived and
content­related measurements of emotional demands,
adjusting estimates for baseline psychological state, and
aggregating individual­level emotional demands to the
e752
 BEU, Alm.del - 2020-21 - Bilag 444: Orientering om resultater fra en videnskabelig artikel fra NFA om afbødning af sammenhænge mellem høje følelsesmæssige krav i arbejdet og langvarigt sygefravær, fra beskæftigelsesministeren
2452965_0002.png
Articles
Research in context
Evidence before this study
Emotional demands at work concern aspects of work that
require sustained emotional effort from employees and are
associated with an increased risk of long-term sickness absence.
In several occupations, such as health care, education, and social
work, a high level of emotional demand is unavoidable.
We searched PubMed and Web of Science on May 13, 2021,
for articles published in English with titles or abstracts
(on PubMed in the past 10 years) or titles (on Web of Science
from Jan 1, 2011, to May 13, 2021) that included the search
terms (“emotional demands” OR “emotion*”) AND (“sickness
absence” OR “sick leave” OR “absenteeism”) AND (“influence”
OR “decision authority” OR “decision latitude” OR “job control”
OR “possibilities for development” OR “skill discretion” OR
“developmental opportunities” OR “role conflicts” OR “physical
demands” OR “physical workload” OR “ergonomic demands”).
To our knowledge, this study is the first to examine whether
influence, possibilities for development, role conflicts,
and physical demands at work modify the risk of long-term
sickness absence in relation to high emotional demands at work.
Added value of this study
In this population-based prospective cohort study, which
included 1 521 352 employees in Denmark, we used national
registers and job exposure matrices to investigate possible
effect modifiers for the association between emotional
demands at work and long-term sickness absence during a
mean follow-up of 7·8 years. Emotional demands at work were
associated with a 1·5-fold increased risk of long-term sickness
absence. This association was exacerbated when individuals
were also exposed to low possibilities for development or high
role conflicts at work. Low influence and high physical demands
at work did not exacerbate the association.
Implications of all the available evidence
Employees in occupations with high emotional demands are at
increased risk of long-term sickness absence. Our observational
findings suggest that increasing possibilities for development
and preventing or reducing high work-related role conflicts in
emotionally demanding occupations might be a successful
approach for reducing long-term sickness absence.
The association between emotional demands at work and
long-term sickness absence does not appear to be affected by
low influence or high physical demands at work. Further
interventional studies, including randomised controlled trials,
are needed on these and other psychosocial factors at work that
might modify the risk of long-term sickness absence in relation
to high emotional demands, to evaluate the effectiveness of
such interventions in real-life settings.
occupational level.
4
However, few large­scale studies
have examined the association between emotional
demands at work and long­term sickness absence with
non­self­reported data.
A further limitation of previous studies on emotional
demands at work relates to understanding of effect
modifiers. In several occupations, such as health care,
education, and social work, a high level of emotional
demand is often a part of the job and therefore it might
be difficult to reduce exposure to emotional demands.
Consequently, it is important to identify other work
factors, which are amenable to change, that might modify
the potential harmful effects of high emotional demands.
Possible effect modifiers might include influence,
possibilities for development, role conflicts, and physical
demands at work. In the psychosocial work environment
literature, the most widely used concept is the concept of
job strain (demand–control model), which assumes that
work is particularly hazardous if psychological demands
are high and job control is low.
11
Psychological demands
might partly overlap emotional demands, whereas job
control is a composite variable consisting of decision
authority (resembling influence at work) and skill
discretion (resembling possibilities for development).
When influence is low, individuals have fewer
possibilities for handling emotional demands (eg, by
prioritising or delegating certain work tasks). Low
possibilities for development inhibit individuals from
e753
learning and acquiring and using skills, and therefore
deprive them of positive experiences that might partly
compensate for the strain of high emotional demands.
Furthermore, a high level of role conflicts adds an
additional burden of complexity to work that might
overload workers who also have to handle high emotional
demands; and a high level of physical demand depletes
workers of energy that might be needed for dealing with
high emotional demands. However, we are not aware of
any previous studies that have tested these hypotheses.
We aimed to test two hypotheses: first, whether people
in occupations with high emotional demands at work are
at increased risk of long­term sickness absence and
second, whether high emotional demands are associated
with risk of long­term sickness absence in a synergistic
way in people in occupations where influence and
possibilities for development are low and role conflicts
and physical demands are high.
Methods
Study design and participants
We did a nationwide, population­based, prospective
cohort study in Denmark and included employed
individuals with information on emotional demands and
effect modifiers from job exposure matrices, and on long­
term sickness absence from national health registers.
We used data from the Job Exposure Matrix Analyses of
Psychosocial Factors and Healthy Ageing in Denmark
www.thelancet.com/public-health
Vol 6 October 2021
 BEU, Alm.del - 2020-21 - Bilag 444: Orientering om resultater fra en videnskabelig artikel fra NFA om afbødning af sammenhænge mellem høje følelsesmæssige krav i arbejdet og langvarigt sygefravær, fra beskæftigelsesministeren
2452965_0003.png
Articles
(JEMPAD) cohort—a nationwide, register­based cohort
with information on employment, psychosocial factors
at work, health, labour market attachment, and socio­
demographics.
12
JEMPAD included all employed indi­
viduals (excluding self­employed individuals) who were
residing in Denmark in 2000, aged 30–59 years, who had
complete data on age, sex, and migration background.
Individuals with long­term sickness absence (≥6 weeks of
consecutive sickness absence) between Jan 1, 1998, and
Dec 31, 2000, were excluded.
As JEMPAD did not include individual­level data on
working conditions, we ascertained data on emotional
demands, influence, possibilities for development, role
conflicts, and physical demands at work with job
exposure matrices that we created in the Danish Work
Environment Cohort Study (DWECS).
13,14
Outcome and
covariates were ascertained with individual­level register­
based data. All variables were ascertained at baseline
in 2000. Outcome was ascertained from Jan 1, 2001, to
Dec 31, 2010.
The study was approved by the Danish Data Protection
Agency through the joint notification of the National
Research Centre for the Working Environment
(2015­57­0074). All data are stored in a protected server
environment hosted by Statistics Denmark. In Denmark,
studies that are based on questionnaire and register
data only do not require approval from the National
Committee on Health Research Ethics. In DWECS,
informed consent was provided by the participants when
filling in the questionnaire. In JEMPAD, which was
register­based, informed consent was not required.
Procedures
We measured emotional demands at work at baseline
in 2000 in the JEMPAD cohort with a job exposure matrix
that we created in DWECS (appendix pp 2–5). DWECS is
a survey on working conditions and health in a randomly
selected sample of the workforce in Denmark from 1990
to 2010.
13,14
Pooling data from the 2000 and 2005 waves of
DWECS, we included the three items on emotional
demands at work and constructed a scale by calculating
the mean of the items (appendix pp 4–5). Using
multilevel modelling, we constructed the job exposure
matrix based on the mean values of emotional demands
at work as the predicted level of emotional demands
for the given occupational group (coded according to
DISCO­88, the Danish version of the International
Standard Classification of Occupations [ISCO]­88
system),
15
sex, age, and DWECS year (2000 or 2005).
Using the year 2000­specific job exposure matrix, we
assigned the predicted level of emotional demands to all
individuals in the JEMPAD study population at baseline
in 2000, according to occupational group, sex, and age.
We categorised individuals into four exposure groups on
the basis of the quartiles of the distribution of the
exposure. We defined individuals in the highest quartile
as exposed to high emotional demands and individuals
www.thelancet.com/public-health
Vol 6 October 2021
in the three remaining quartiles as exposed to low
emotional demands. More information on DWECS
items, scale, and job exposure matrix construction is
provided in the appendix (pp 2–5).
We measured influence, possibilities for development,
and physical demands at work as effect modifiers in
JEMPAD in 2000 with job exposure matrices, based on
information from DWECS and constructed using the
same approach as was used for the job exposure matrix
on emotional demands. We constructed the job exposure
matrix for role conflicts on the basis of one DWECS item
that was dichotomised and then modelled as the
predicted probability of role conflicts for the given
occupational group, sex, age, and DWECS year. For each
variable, we dichotomised individuals as either no
exposure or as adversely exposed, meaning exposed to
low influence (lowest quartile), low possibilities for
development (lowest quartile), high role conflicts
(highest quartile), and high physical demands (highest
quartile; appendix p 3).
As measures for interaction on an additive scale
16
were
developed for risk factors rather than for preventive
factors, we defined low values of the two preventive
factors (influence and possibilities for development) as
risk factors.
17
Thus, the stratum with the lowest risk
became the reference category when considering
two factors jointly
17
—ie, low emotional demands com­
bined with high influence, high possibilities for
development, low role conflicts, or low physical demands,
respectively (double­unexposed groups). By contrast, the
double­exposed groups were high emotional demands
combined with low influence, low possibilities for
development, high role conflicts, or high physical
demands, respectively.
To assess long­term sickness absence, we linked
JEMPAD to the Danish Register for Evaluation of
Marginalization,
18
which includes weekly information on
all social transfer payments in Denmark since 1991. We
defined long­term sickness absence as 6 weeks or longer
of consecutive sickness absence. We assessed long­term
sickness absence during a 10­year period from Jan 1, 2001,
to Dec 31, 2010.
Potential confounders were measured in JEMPAD at
baseline in 2000 and included sex, age, cohabitation,
migration background, and income, retrieved from
population­based registers,
19,20
as these variables have
shown associations with long­term sickness absence in
previous studies.
21–23
Cohabitation was measured as
single versus married or cohabiting. For migration
background, we used the classification of Statistics
Denmark’s population register, distinguishing between
Danish origin (the whole population except immigrants
and descendants of immigrants), immigrants (born
abroad and none of the parents were either Danish
citizens or born in Denmark), and descendants of
immigrants (born in Denmark and none of the parents
were either Danish citizens or born in Denmark). We
See
Online
for appendix
e754
 BEU, Alm.del - 2020-21 - Bilag 444: Orientering om resultater fra en videnskabelig artikel fra NFA om afbødning af sammenhænge mellem høje følelsesmæssige krav i arbejdet og langvarigt sygefravær, fra beskæftigelsesministeren
2452965_0004.png
Articles
Participants (n=1 521 352)
Sex
Female
Male
Age, years
Cohabitation
Yes
No
Missing
Migration background
Danish origin
Immigrant
Descendant of immigrants
Household income, €
Household income category, €
Low, <20 000
Medium–low, 20 000–39 999
Medium–high, 40 000–59 999
High, ≥60 000
Missing
Data are n (%) or mean (SD).
731 540 (48·1%)
789 812 (51·9%)
43·7 (8·4)
1 138 049 (74·8%)
379 795 (25·0%)
3508 (0·2%)
1 453 371 (95·5%)
65 721 (4·3%)
2260 (0·1%)
43 565 (37 998)
102 231 (6·7%)
527 615 (34·7%)
649 932 (47·7%)
153 084 (10·1%)
88 490 (5·8%)
four effect modifiers, we analysed the separate effects of
emotional demands and the effect modifier and their
joint effects, each relative to no exposure, and calculated
the relative excess risk due to interaction (RERI).
16
To
inform decision makers about the number of additional
cases for the combinations of emotional demands with
each of the effect modifiers, we multiplied the estimated
HR by the number of cases per 1000 person­years in the
reference category and calculated the differences in
number of cases in the reference group versus the
exposure groups.
As proposed by Knol and VanderWeele,
25
we also
present results for high emotional demands and risk of
long­term sickness absence across strata of each of the
four effect modifiers. In sensitivity analyses we repeated
all analyses with follow­up until Dec 31, 2001 (ie, using a
1­year follow­up), similar to the follow­up duration in
previous studies on long­term sickness absence in
Denmark.
4,26
We further analysed effect modification
without adjusting for other working conditions and with
dichotomising working conditions by median split
instead of by the highest or lowest quartile.
All analyses were done using SAS version 9.4.
Table 1:
Baseline characteristics in the year 2000
Role of the funding source
measured income as the annual disposable household
income in euros (after tax), categorised into deciles based
on the distribution within the study population.
The funders of the study had no role in study design,
data collection, data analysis, data interpretation, or
writing of the report.
Results
Statistical analysis
After confirming by visual inspection of survival plots
that proportional hazard assumptions were not violated,
we used Cox regression to estimate hazard ratios (HRs)
and 95% CIs for the association between high emotional
demands at work at baseline in 2000 and first episode of
long­term sickness absence during follow­up from
Jan 1, 2001, to Dec 31, 2010. We censored due to
retirement (early, disability, or statutory), emigration,
death, or end of study (Dec 31, 2010), whichever came
first. We fitted three models. Estimates were adjusted
for sex and age (model one), and further adjusted
for cohabitation, migration background, and income
(model two). In model three, we further adjusted for the
four other working conditions: influence, possibilities
for development, role conflicts, and physical demands at
work. We also analysed the associations between low
influence, low possibilities for development, high role
conflicts, or high physical demands, respectively, and
risk of future long­term sickness absence by fitting the
three models.
To quantify if the association between high emotional
demands and risk of long­term sickness absence was
modified by influence, possibilities for development, role
conflicts, or physical demands, we tested the inter­
action effects defined as departure from additivity as
recommended in STROBE guidelines.
24
For each of the
e755
1 680 214 individuals with complete data were identified
from the JEMPAD cohort. 158 862 individuals had
long­term sickness absence between Jan 1, 1998, and
Dec 31, 2000, and were excluded. 1 521 352 employed
individuals (789 812 [51·9%] men and 731 540 [48·1%]
women) were included and contributed data between
Jan 1, 2000, and Dec 31, 2010 (table 1). The mean age at
baseline was 43·7 years (SD 8·4). During 11 919 021 person­
years (mean follow­up 7·8 years) we identified 480 685 new
cases of long­term sickness absence. We censored
147 863 individuals due to retirement, 27 915 due to
emigration, and 13 151 due to death.
After adjustment for age and sex, high emotional
demands were associated with an increased risk of
long­term sickness absence compared with low emo­
tional demands (HR 1·23 [95% CI 1·23–1·24]; table 2).
After further adjustment for cohabitation, migration
background, income, and the four other working
conditions (effect modifiers), the estimate increased
(1·55 [1·53–1·56]). Estimates were similar for men and
women (appendix p 8).
The RERI was greater than 0 for possibilities for
development (RERI 0·35 [95% CI 0·22 to 0·47],
28·9 additional cases per 1000 person­years) and role
conflicts (0·13 [0·11 to 0·15], 22·0 additional cases
per 1000 person­years) at work, indicating more than
additive interaction (ie, synergism) in the double­exposed
www.thelancet.com/public-health
Vol 6 October 2021
 BEU, Alm.del - 2020-21 - Bilag 444: Orientering om resultater fra en videnskabelig artikel fra NFA om afbødning af sammenhænge mellem høje følelsesmæssige krav i arbejdet og langvarigt sygefravær, fra beskæftigelsesministeren
2452965_0005.png
Articles
groups. The RERI was less than 0 for influence (–0·25
[–0·28 to –0·23) and physical demands (–0·41
[–0·45 to –0·36]), indicating less than additive interaction
(ie, antagonism) in the double­exposed groups (table 3).
When estimates were stratified by high and low values
of the other four work environment factors, high
emotional demands were associated with a higher risk of
long­term sickness absence in all strata (appendix p 9).
In sensitivity analyses, the synergistic interaction
between high emotional demands and low possibilities
for development remained robust when follow­up was
for 1 year instead of 10 years (appendix p 11) and when
the estimates were not adjusted for other working
conditions (appendix p 12). When working conditions
were dichotomised by median split, the synergistic
interaction between high emotional demands and low
possibilities for development disappeared (appendix p 13).
The synergistic interaction between high emotional
demands and high role conflicts remained robust when
working conditions were dichotomised by median split
(appendix p 13), but not when follow­up was limited to
1 year (appendix p 11) and when estimates were not
adjusted for other working conditions (appendix p 12).
Person-years Cases
Cases per
1000 person-
years
Model one,
HR (95% CI)
Model two, Model three,
HR (95% CI) HR (95% CI)
Emotional demands
Low emotional
demands
High emotional
demands
Influence
High influence
Low influence
9 080 914
2 838 107
335 304
145 381
36·9
51·2
1 (ref)
1·27
(1·26–1·28)
1 (ref)
1·28
(1·27–1·28)
1 (ref)
1·08
(1·07–1·09)
1 (ref)
1·55
(1·54–1·56)
1 (ref)
1·18
(1·17–1·19)
1 (ref)
1·17
(1·16–1·17)
1 (ref)
1·09
(1·09–1·10)
1 (ref)
1·41
(1·40–1·42)
1 (ref)
1·02
(1·02–1·03)
1 (ref)
1·14
(1·13–1·15)
1 (ref)
1·01
(1·00–1·02)
1 (ref)
1·36
(1·35–1·37)
9 049 046
2 869 975
333 083
147 602
36·8
51·4
1 (ref)
1·23
(1·23–1·24)
1 (ref)
1·26
(1·26–1·27)
1 (ref)
1·55
(1·53–1·56)
Possibilities for development
High possibilities
for development
Low possibilities
for development
Role conflicts
Low role conflicts
High role
conflicts
Physical demands
Low physical
demands
High physical
demands
9 075 392
2 843 628
338 486
142 199
37·3
50·0
8 907 033
3 011 988
362 482
118 203
40·7
39·2
9 126 177
2 792 844
341 473
139 212
37·4
49·8
Discussion
In this study based on a cohort of 1 521 352 employed
individuals in Denmark, men and women working in
emotionally demanding occupations were found to be at
increased risk of long­term sickness absence compared
with those in occupations with low emotional demands.
Additive interaction analyses indicated synergism (ie, a
higher than additive risk), for double exposure to high
emotional demands and low possibilities for development
(that remained robust in two of three sensitivity analyses)
and double exposure to high emotional demands and
high role conflicts (that remained robust in one of
three sensitivity analyses). With regard to high emotional
demands and low influence and high emotional demands
and high physical demands, we identified antagonism
(ie, a lower than additive risk associated with double
exposure).
These findings support smaller­scale studies that have
reported that high emotional demands at work are a
risk factor for long­term sickness absence.
4
As we used
a job exposure matrix­based approach, our findings
are unlikely to be affected by reporting bias. To our
knowledge, with 1·5 million participants and more than
11·9 million person­years of follow­up, this study is by
far the largest study so far on emotional demands and
work absence. The next largest study was based on
26 410 participants, with 1002 incident long­term sickness
absence cases during 22 466 person­years of follow­up.
4
To our knowledge, this study is the first to examine
whether influence, possibilities for development, role
conflicts, and physical demands at work modify the
higher risk of long­term sickness absence in relation to
high emotional demands at work. Previous observational
www.thelancet.com/public-health
Vol 6 October 2021
Data are n unless otherwise stated. Model one: adjusted for sex and age. Model two: further adjusted for cohabitation,
migration background, and income. Model three: further mutually adjusted for all work environment factors.
HR=hazard ratio.
Table 2:
Risk of long-term sickness absence during 10 years of follow-up by levels of emotional demands,
influence, possibilities for development, role conflicts, and physical demands at work
studies have shown that leadership quality did not
substantially buffer the adverse effect of high emotional
demands on levels of antidepressant treatment
27
and
sickness absence.
8
However, work­related goal attainment
seemed to moderate the negative effect of high emotional
demands on employee wellbeing,
28
and finding work
emotionally enriching and meaningful seemed to buffer
the adverse effect of high emotional demands on
exhaustion.
29
Our results indicate that high emotional demands at
work might be particularly hazardous in the presence
of low possibilities for development and high role
conflicts. Low possibilities for development might
inhibit employees from learning and acquiring and
using new skills that could help with their handling of
high emotional demands. Acquiring and using new
skills might also increase job satisfaction and partly
compensate for the strain of high emotional demands.
Role conflicts might elicit negative emotions and add
another type of emotional strain to a work situation
already characterised by high emotional demands.
Furthermore, role conflicts might arise from having too
little time or resources to do the job properly, which
might add further strain and exacerbate the negative
effects of high emotional demands.
e756
 BEU, Alm.del - 2020-21 - Bilag 444: Orientering om resultater fra en videnskabelig artikel fra NFA om afbødning af sammenhænge mellem høje følelsesmæssige krav i arbejdet og langvarigt sygefravær, fra beskæftigelsesministeren
2452965_0006.png
Articles
Person-
years
Cases
Adjusted HR
Cases per
(95% CI)*
1000
person-years
Expected number of
cases per
1000 person-years
based on adjusted HR
··
32·3
34·9
51·7
Expected number RERI (95% CI)
of additional cases
per 1000 person-
years based on
adjusted HR
··
0
2·6
19·4
13·6
··
0
4·0
14·0
28·9
··
0
–0·7
17·9
22·0
··
0
12·9
23·9
23·9
–0·25
(–0·28 to –0·23)
··
··
··
··
0·35
(0·22 to 0·47)
··
··
··
··
0·13
(0·11 to 0·15)
··
··
··
··
–0·41
(–0·45 to –0·36)
··
··
··
··
Influence
Low emotional demands,
high influence
Low emotional demands,
low influence
High emotional demands,
high influence
High emotional demands,
low influence
Possibilities for development
Low emotional demands,
high possibilities for development
Low emotional demands,
low possibilities for development
High emotional demands,
high possibilities for development
High emotional demands,
low possibilities for development
Role conflicts
Low emotional demands, low role
conflicts
Low emotional demands, high role
conflicts
High emotional demands, low role
conflicts
··
6 686 851
2 362 195
2 394 063
475 912
··
6 271 571
2 777 474
2 854 606
15 369
··
7 325 420
1 723 625
1 581 613
··
216 461
116 622
118 843
28 759
··
194 925
138 158
146 548
1054
··
273 109
59 974
89 373
58 229
··
200 264
132 819
138 222
9380
··
32·3
49·4
49·6
60·4
··
31·1
49·7
51·3
68·6
··
37·3
34·8
56·5
45·8
··
31·4
49·5
51·1
57·7
··
1 (ref)
1·08
(1·07 to 1·09)
1·60
(1·58 to 1·61)
1·42
45·9
(1·40 to 1·44)
··
1 (ref)
1·13
(1·12 to 1·14)
··
31·1
35·1
1·45
45·1
(1·44 to 1·46)
1·93
60·0
(1·81 to 2·06)
··
1 (ref)
··
37·3
0·98
36·6
(0·97 to 0·99)
1·48
55·2
(1·46 to 1·49)
1·59
(1·57 to 1·61)
··
1 (ref)
1·41
(1·40 to 1·42)
1·76
(1·74 to 1·78)
1·76
(1·72 to 1·80)
59·3
··
31·4
44·3
55·3
55·3
High emotional demands, high role 1 288 362
conflicts
Physical demands
Low emotional demands,
low physical demands
Low emotional demands,
high physical demands
High emotional demands,
low physical demands
High emotional demands,
high physical demands
··
6 368 034
2 681 011
2 707 358
162 617
Data are n unless otherwise stated. HR=hazard ratio. RERI=relative excess risk due to interaction. *Adjusted for sex, age, cohabitation, migration background, income, and the
other three work environment factors.
Table 3:
Risk of long-term sickness absence during 10 years of follow-up by levels of emotional demands at work, with effect modification by influence,
possibilities for development, role conflicts, and physical demands at work
High emotional demands are part of the job in several
occupations, such as teaching and health care,
and reducing emotional demands might be difficult.
Our findings suggest that workplace interventions in
emotionally demanding occupations that aim to increase
possibilities for development of the employees and to
prevent or reduce role conflicts might be a successful
approach for reducing long­term sickness absence.
For the combinations of high emotional demands and
low influence, and high emotional demands and high
e757
physical demands, the RERI was less than 0, indicating
antagonism instead of synergism. However, it is
important to note that both low influence (model two;
table 2) and high physical demands (models two and
three; table 2) were associated with an increased risk of
long­term sickness absence compared with high
influence and low physical demands, respectively. Thus,
our analyses do not indicate that low influence and high
physical demands are not of importance in risk of
long­term sickness absence, but instead indicate that low
www.thelancet.com/public-health
Vol 6 October 2021
 BEU, Alm.del - 2020-21 - Bilag 444: Orientering om resultater fra en videnskabelig artikel fra NFA om afbødning af sammenhænge mellem høje følelsesmæssige krav i arbejdet og langvarigt sygefravær, fra beskæftigelsesministeren
2452965_0007.png
Articles
influence and high physical demands at work do not
appear to exacerbate the association between high
emotional demands and risk of long­term sickness
absence.
The measures of influence and of possibilities for
development in our study are similar to the concepts of
decision authority and skill discretion, respectively, in the
demand–control model.
11
Decision authority and skill
discretion are usually combined into the variable of job
control in the demand–control model. Our results
showed that the interactions of emotional demands with
influence and with possibilities for development,
respectively, yielded different results, which might
suggest that it could be fruitful in the demand–control
model to analyse decision authority and skill discretion
not only in combination but also separately.
The strengths of this study include the use of a large
nationwide cohort of 1 521 352 employed individuals and
the register­based outcome and covariates measurement.
The job exposure matrix­based approach enabled us to
measure emotional demands, influence, possibilities for
development, role conflicts, and physical demands at
work for all employees in the study population.
Furthermore, the job exposure matrix­based approach
reduced the risk of reporting bias as exposure was not
measured at the individual­level but was estimated by the
occupational groups of the participants.
This study has several limitations. First, the job
exposure matrix­based approach means that we did not
analyse directly the association between emotional
demands, the effect modifiers, and the risk of long­term
sickness absence, but analysed the association between
working in occupations that on average had a certain
level of exposure (with regard to emotional demands and
the effect modifiers) and risk of long­term sickness
absence. This method of analysis might lead to exposure
misclassification, as some employees might have had
low individual­level emotional demands while working
in occupations with high average emotional demands
and vice versa, which would most likely lead to
underestimation of the observed associations.
30
Second,
the Danish registers on sickness absence do not include
information on diagnoses, which meant that we could
only measure all­cause long­term sickness absence and
not cause­specific long­term sickness absence. For
example, it might be that the risk of long­term sickness
absence in relation to emotional demands is higher for
long­term sickness absence due to common mental
disorders than for long­term sickness absence due to
musculoskeletal disorders. Third, time of follow­up was
long, with a mean of 7·8 years, and level of emotional
demands might have changed during this time for some
participants, which could have led to imprecise exposure
measurement, as individuals could have changed job
groups, and underestimation of the association. How­
ever, a sensitivity analysis with a 1­year follow­up
produced estimates that were similar to the estimates of
www.thelancet.com/public-health
Vol 6 October 2021
the main analysis. Fourth, we did not incorporate
possible period effects in emotional demands, to reflect
changes in the labour market or in society. Effects were,
however, constant over time, with proportional hazards
throughout the complete follow­up period. Fifth, the
analyses did not account statistically for clustering of
data, as no such standard approach is available when
estimating confidence limits for RERI. Sixth, the cohort
included all employed individuals residing in Denmark
who were aged 30–59 years in 2000. The results,
therefore, apply to this age group and to employees in
Denmark. As sickness absence systems differ between
countries, generalisations to other countries should be
made with caution. Seventh, our analyses were limited to
incident long­term sickness absence among individuals
who had no history of long­term sickness absence in the
last 3 years before the start of follow­up, and the results
might not apply to people who have recurrent long­term
sickness absence.
In conclusion, in this large, job exposure matrix­based
cohort study in Denmark, employees in occupations with
high emotional demands had a higher risk of long­term
sickness absence than those in occupations with low
emotional demands over 7·8 years of follow­up. Additive
interaction analyses indicated that the association
between high emotional demands and risk of long­term
sickness absence was exacerbated in a synergistic way by
the presence of low possibilities for development and
high role conflicts at work. In three sensitivity analyses,
the synergism remained robust for low possibilities for
development in two analyses and for role conflicts in
one analysis. Contrary to our hypothesis, low influence
and high physical demands at work did not exacerbate
the association between high emotional demands and
long­term sickness absence. Our findings raise the
hypothesis that in emotionally demanding occupations,
increasing possibilities for development at work and
preventing or reducing high work­related role conflicts
might reduce long­term sickness absence. Our results
suggest that this hypothesis does not apply to low
influence or high physical demands at work. Further
interventional research, including randomised controlled
trials, is needed on these effect modifiers and other
psychosocial factors at work that might modify the high
risk of long­term sickness absence in relation to high
emotional demands, to evaluate the effectiveness of such
interventions in real­life settings.
Contributors
EF, JKS, IEHM, and RR designed and planned the study. EF and JKS
accessed and verified the data and did the statistical analyses. EF wrote
the first draft of the report. All authors contributed to interpretation of
data and revised the report. EF, JKS, IEHM, and RR jointly decided to
submit the report and all other authors approved the decision.
All authors had full access to all the data in the study, approved the
final version of the report, and agree to be accountable for all aspects
of the work.
Declaration of interests
We declare no competing interests.
e758
 BEU, Alm.del - 2020-21 - Bilag 444: Orientering om resultater fra en videnskabelig artikel fra NFA om afbødning af sammenhænge mellem høje følelsesmæssige krav i arbejdet og langvarigt sygefravær, fra beskæftigelsesministeren
2452965_0008.png
Articles
Data sharing
All data are stored in a protected server environment hosted by Statistics
Denmark and can be accessed by researchers registered with Statistics
Denmark. For further information, please contact Prof Reiner Rugulies
([email protected]).
Acknowledgments
We thank the funders of the study. This study was funded by grants from
the Danish Work Environment Research Fund (grant 27­2017­03) and from
the NordForsk Nordic Programme on Health and Welfare (grant 75021).
References
1
Zapf D. Emotion work and psychological well­being: a review of the
literature and some conceptual considerations.
Hum Resour Manage Rev
2002;
12:
237–68.
2
Vegchel N, Jonge J, Söderfeldt M, Dormann C, Schaufeli W.
Quantitative versus emotional demands among Swedish human
service employees: moderating effects of job control and social
support.
Int J Stress Manag
2004;
11:
21–40.
3
Madsen IEH, Diderichsen F, Burr H, Rugulies R. Person­related
work and incident use of antidepressants: relations and mediating
factors from the Danish Work Environment Cohort Study.
Scand J Work Environ Health
2010;
36:
435–44.
4
Framke E, Sørensen JK, Nordentoft M, et al. Perceived and content­
related emotional demands at work and risk of long­term sickness
absence in the Danish workforce: a cohort study of 26 410 Danish
employees.
Occup Environ Med
2019;
76:
895–900.
5
Sinclair S, Raffin­Bouchal S, Venturato L, Mijovic­Kondejewski J,
Smith­MacDonald L. Compassion fatigue: a meta­narrative review
of the healthcare literature.
Int J Nurs Stud
2017;
69:
9–24.
6
Eriksson K, Englander M. Empathy in social work.
J Soc Work Educ
2017;
53:
607–21.
7
Aronsson G, Theorell T, Grape T, et al. A systematic review
including meta­analysis of work environment and burnout
symptoms.
BMC Public Health
2017;
17:
264.
8
Rugulies R, Sørensen JK, Madsen IEH, Nordentoft M, Sørensen K,
Framke E. Can leadership quality buffer the association between
emotionally demanding work and risk of long­term sickness
absence?
Eur J Public Health
2021; ckab090.
9
Aagestad C, Johannessen HA, Tynes T, Gravseth HM, Sterud T.
Work­related psychosocial risk factors for long­term sick leave:
a prospective study of the general working population in Norway.
J Occup Environ Med
2014;
56:
787–93.
10 Slany C, Schütte S, Chastang JF, Parent­Thirion A, Vermeylen G,
Niedhammer I. Psychosocial work factors and long sickness
absence in Europe.
Int J Occup Environ Health
2014;
20:
16–25.
11 Theorell T, Karasek RA. Current issues relating to psychosocial job
strain and cardiovascular disease research.
J Occup Health Psychol
1996;
1:
9–26.
12 Rugulies R, Framke E, Sørensen JK, et al. Persistent and changing
job strain and risk of coronary heart disease: a population­based
cohort study of 1.6 million employees in Denmark.
Scand J Work Environ Health
2020;
46:
498–507.
13 Burr H, Bjorner JB, Kristensen TS, Tüchsen F, Bach E. Trends in
the Danish work environment in 1990–2000 and their associations
with labor­force changes.
Scand J Work Environ Health
2003;
29:
270–79.
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Feveile H, Olsen O, Burr H, Bach E. Danish Work Environment
Cohort Study 2005: from idea to sampling design.
Stat Transit
2007;
8:
441–58.
International Labour Organization. ISCO­88. 2004. https://www.ilo.
org/public/english/bureau/stat/isco/isco88/ (accessed
May 18, 2021).
Andersson T, Alfredsson L, Källberg H, Zdravkovic S, Ahlbom A.
Calculating measures of biological interaction.
Eur J Epidemiol
2005;
20:
575–79.
Knol MJ, VanderWeele TJ, Groenwold RHH, Klungel OH,
Rovers MM, Grobbee DE. Estimating measures of interaction on
an additive scale for preventive exposures.
Eur J Epidemiol
2011;
26:
433–38.
Burr H, Pedersen J, Hansen JV. Work environment as predictor of
long­term sickness absence: linkage of self­reported DWECS data
with the DREAM register.
Scand J Public Health
2011;
39
(suppl 7): 147–52.
Baadsgaard M, Quitzau J. Danish registers on personal income and
transfer payments.
Scand J Public Health
2011;
39
(suppl): 103–05.
Petersson F, Baadsgaard M, Thygesen LC. Danish registers on
personal labour market affiliation.
Scand J Public Health
2011;
39
(suppl): 95–98.
Nielsen ML, Rugulies R, Christensen KB, Smith­Hansen L,
Kristensen TS. Psychosocial work environment predictors of short
and long spells of registered sickness absence during a 2­year
follow up.
J Occup Environ Med
2006;
48:
591–98.
Niederkrotenthaler T, Tinghög P, Goldman­Mellor S, Wilcox HC,
Gould M, Mittendorfer­Rutz E. Medical and social determinants of
subsequent labour market marginalization in young hospitalized
suicide attempters.
PLoS One
2016;
11:
e0146130.
Sumanen H, Pietiläinen O, Lahti J, Lahelma E, Rahkonen O.
Interrelationships between education, occupational class and income
as determinants of sickness absence among young employees in
2002–2007 and 2008–2013.
BMC Public Health
2015;
15:
332.
Vandenbroucke JP, von Elm E, Altman DG, et al. Strengthening the
reporting of observational studies in epidemiology (STROBE):
explanation and elaboration.
Int J Surg
2014;
12:
1500–24.
Knol MJ, VanderWeele TJ. Recommendations for presenting
analyses of effect modification and interaction.
Int J Epidemiol
2012;
41:
514–20.
Sørensen JK, Framke E, Clausen T, et al. Leadership quality and risk
of long­term sickness absence among 53,157 employees of the
Danish workforce.
J Occup Environ Med
2020;
62:
557–65.
Madsen IEH, Hanson LL, Rugulies R, et al. Does good leadership
buffer effects of high emotional demands at work on risk of
antidepressant treatment? A prospective study from two Nordic
countries.
Soc Psychiatry Psychiatr Epidemiol
2014;
49:
1209–18.
Wong E, Tschan F, Semmer NK. Effort in emotion work and
well­being: the role of goal attainment.
Emotion
2017;
17:
67–77.
Vammen MA, Mikkelsen S, Forman JL, et al. Emotional demands
and exhaustion: cross­sectional and longitudinal associations in a
cohort of Danish public sector employees.
Int Arch Occup Environ Health
2019;
92:
639–50.
Peters S. Although a valuable method in occupational epidemiology,
job­exposure ­matrices are no magic fix.
Scand J Work Environ Health
2020;
46:
231–34.
e759
www.thelancet.com/public-health
Vol 6 October 2021