Beskæftigelsesudvalget 2019-20
BEU Alm.del Bilag 252
Offentligt
Persistent and changing job strain and risk of coronary heart disease. A population-based
cohort study of 1.6 million employees in Denmark
1
by Reiner Rugulies, PhD,
2
Elisabeth Framke, PhD, Jeppe Karl Sørensen, MSc, Annemette Coop Svane-Petersen, PhD, Kristina
Alexanderson, PhD, Jens Peter Bonde, MD, PhD, Kristin Farrants, PhD, Esben Meulengracht Flachs, PhD, Linda L Magnusson Hanson,
PhD, Solja T Nyberg, PhD, Mika Kivimäki, FMedSci, Ida EH Madsen, PhD
1. Supplementary material
2. Correspondence to: Reiner Rugulies, National Research Centre for the Working Environment, Lersø
Parkallé 105, DK-2100 Copenhagen, Denmark. [E-mail: [email protected]]
•
•
•
•
Appendix 1:
Ascertainment of job strain with a job exposure matrix
Appendix 2:
Overview of the three main analyses in the study
Appendix 3:
Analysis of the association between job strain, defined by quartiles, and risk of CHD
Appendix 4:
Analysis of the association between job strain and risk of CHD conducted separately for
men and women
•
Appendix 5:
Analysis of the association between job strain and risk of CHD additionally adjusted for
education
•
Appendix 6:
Analysis of the association between job strain and risk of CHD conducted separately for
educational groups
1
 BEU, Alm.del - 2019-20 - Bilag 252: Orientering om resultater fra en NFA-artikel om psykosocialt arbejdsmiljø og risiko for iskæmisk hjertesygdom, fra beskæftigelsesministeren
Appendix 1: Ascertainment of job strain with a job exposure matrix
To ascertain job strain we constructed a job exposure matrix based on information from the
Danish Work Environment Cohort Study (DWECS) and assigned the matrix to individuals in the
JEMPAD (Job Exposure Matrix Analysis of Psychosocial Factors and Healthy Ageing in Denmark)
cohort. DWECS was a survey on working conditions and health conducted in a random sample of
employed individuals in Denmark, aged 18 to 64 years (1, 2), first drawn in 1990 and followed-up
every fifth year until 2010, with inclusion of additional individuals in each wave. We included
DWECS data from the years 2000 (response rate: 75%) (1) and 2005 (response rate 63%) (2). In
accordance with previous research on job strain and cardiovascular disease using DWECS data (3-
5), we measured job strain by combining three items measuring job demands and five items
measuring job control (see
Table S1).
Several of the items originated from the Copenhagen
Psychosocial Questionnaire, version I (6).
Respondents were included if they responded to at least two of the three job demands items and
to at least three of the five job control items. We calculated scales for job demands and job
control with higher scores indicating higher job demands and higher job control, respectively. We
defined job strain in DWECS respondents as scoring simultaneously above the median on job
demands and below the median on job control, as in previous research (3-5).
We categorized DWECS respondents in job groups according to the DISCO-88 occupational
classification system (7). DISCO-88 is the Danish version of the International Standard
Classification of Occupations (ISCO) developed by the International Labour Organization (ILO) and
was used from 1991 to 2009, when it was replaced with DISCO-08 (8). We used the four-digit level
classification and required a minimum of five respondents in DWECS within each job group. Job
groups with four or less respondents were collapsed with other similar small job groups at the
three-digit or two-digit classification level in the DISCO-88 classification system. This approach was
used to avoid developing the job exposure matrix based on job groups with a very small number of
respondents, as this would lead to imprecise estimates.
2
 BEU, Alm.del - 2019-20 - Bilag 252: Orientering om resultater fra en NFA-artikel om psykosocialt arbejdsmiljø og risiko for iskæmisk hjertesygdom, fra beskæftigelsesministeren
Using the Glimmix procedure in SAS 9.4, we estimated the predicted probabilities of job strain
given job group (according to the DISCO-88 classification system), sex, age and year of data
collection (2000, 2005). Thus, we constructed a job exposure matrix that was job group-, sex-,
age-, and period-specific. After linking the predicted probability for job strain on a yearly basis to
the individuals in the JEMPAD cohort, we dichotomized the participants into “no job strain” versus
“job strain”, based on median split of the distribution of individuals within each year. We chose to
use yearly median split rather than absolute cut-off points for the categorization of job strain
because there was a rather strong effect of wave of data collection in the data used to estimate
the job exposure matrix (DWECS 2000 vs. DWECS 2005). Because exposure was assigned based on
predicted values from the year 2000 period-specific job exposure matrix during the years 1996-
2003 and predicted values from the year 2005 period-specific job exposure matrix during the years
2004-2009, we could not establish absolute cut-off values for the categorization of job strain that
would be meaningful throughout the entire period of exposure assessment.
Participants who were not employed, e.g. due to unemployment or retirement, were assigned to a
separate category of not applicable (NA) job strain during their time outside of employment.
To assess the performance of the job exposure matrix measuring job strain, we calculated the area
under the curve (AUC) for job strain. The AUC was estimated by comparing the agreement
between the job exposure matric values for job strain to the individual-level self-reported job
strain in DWECS. The AUC was 0.70, which can be regarded as a fair agreement, considering that
AUCs may range from 0.50 (agreement by chance) to 1.00 (perfect agreement). We conclude that
the job exposure matrix assessing job strain performed reasonably well in DWECS and could
therefore be used for estimating the association between job strain and coronary heart disease in
register-based population studies.
3
 BEU, Alm.del - 2019-20 - Bilag 252: Orientering om resultater fra en NFA-artikel om psykosocialt arbejdsmiljø og risiko for iskæmisk hjertesygdom, fra beskæftigelsesministeren
2195800_0004.png
Table S1. Scales, items and response options for job strain variable
Scales
Items
Job demands
Do you have to work very fast?
Response options
Always; Often; Sometimes;
Seldom; Never/hardly ever
Always; Often; Sometimes;
Seldom; Never/hardly ever
Correct; From time to time
correct; Not correct
How often do you not have time to complete all
your work tasks?
Contradictory demands are placed on you at
work?
Job control
Do you have any influence on what you do at
work?
Can you use your skills or expertise in your
work?
Always; Often; Sometimes;
Seldom; Never/hardly ever
To a very large extent; To a large
extent; Somewhat; To a small
extent; To a very small extent
To a very large extent; To a large
extent; Somewhat; To a small
extent; To a very small extent
To a very large extent; To a large
extent; Somewhat; To a small
extent; To a very small extent
To a high degree; To some
degree; Only to a lesser degree;
No, or only to a slight degree
Do you have the possibility of learning new
things through your work?
Does your work require you to take the
initiative?
Is your work varied?
4
 BEU, Alm.del - 2019-20 - Bilag 252: Orientering om resultater fra en NFA-artikel om psykosocialt arbejdsmiljø og risiko for iskæmisk hjertesygdom, fra beskæftigelsesministeren
2195800_0005.png
References
1.
Burr H, Bjorner JB, Kristensen TS, Tüchsen F, Bach E. Trends in the Danish work environment in
1990-2000 and their associations with labor-force changes. Scand J Work Environ Health.
2003;29(4):270-9.
2.
Feveile H, Olsen O, Burr H, Bach E. Danish Work Environment Cohort Study 2005: From idea to
sampling design. Statistics in Transition. 2007;8(3):441-58.
3.
Fransson EI, Nyberg ST, Heikkilä K, Alfredsson L, De Bacquer D, Batty GD, et al. Comparison of
alternative versions of the job demand-control scales in 17 European cohort studies: the IPD-Work
consortium. BMC Public Health. 2012;12(1):62.
4.
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. 2012;380(9852):1491-7.
5.
Fransson EI, Nyberg ST, Heikkilä K, Alfredsson L, Bjorner JB, Borritz M, et al. Job strain and the risk
of stroke: an individual-participant data meta-analysis. Stroke. 2015;46(2):557-9.
6.
Kristensen T, Hannerz H, Høgh A, Borg V. The Copenhagen Psychosocial Questionnaire. A tool for
the assessment and improvement of the psychosocial work environment. Scand J Work Environ
Health. 2005;31(6):438-49.
7.
International Labour Organization (ILO). ISCO-88. 2004. Available from:
http://www.ilo.org/public/english/bureau/stat/isco/isco88/.
(Accessed: 22 May 2019).
8.
International Labour Organization. ISCO-08 Structure, index correspondence with ISCO-88. 2016.
Available from:
http://www.ilo.org/public/english/bureau/stat/isco/isco08/.
(Accessed: 22 May
2019).
5
 BEU, Alm.del - 2019-20 - Bilag 252: Orientering om resultater fra en NFA-artikel om psykosocialt arbejdsmiljø og risiko for iskæmisk hjertesygdom, fra beskæftigelsesministeren
2195800_0006.png
Appendix 2: Overview of the three main analyses in the study
Figure S1: Design of the three main analyses in the study
6
 BEU, Alm.del - 2019-20 - Bilag 252: Orientering om resultater fra en NFA-artikel om psykosocialt arbejdsmiljø og risiko for iskæmisk hjertesygdom, fra beskæftigelsesministeren
2195800_0007.png
Appendix 3: Analysis of the association between job strain, defined by
quartiles, and risk of CHD
Table S2. Association between job strain, defined by quartiles, measured in 2000 and incident coronary
heart disease from 2001 to 2010 among 1,660,150 employees in Denmark
Cases
Person-
Number
per 10,000
Adjusted HR
Job strain at baseline
years
of cases person-years
(95% CI)
All
Low job strain
4,027,420
5,028
12.5
1.00
Medium-low strain
4,018,175
6,004
14.9
1.20 (1.15-1.24)
Medium-high job strain 4,031,494
6,686
16.6
1.18 (1.14-1.23)
High job strain
4,040,423
6,441
15.9
1.24 (1.19-1.29)
HR: Hazard ratio; 95% CI: 95% Confidence interval-
Adjusted for age, family type, migration background, health service use and household disposable income.
Covariates were measured in the year 2000 and were treated as time-invariant.
7
 BEU, Alm.del - 2019-20 - Bilag 252: Orientering om resultater fra en NFA-artikel om psykosocialt arbejdsmiljø og risiko for iskæmisk hjertesygdom, fra beskæftigelsesministeren
2195800_0008.png
Appendix 4: Analysis of the association between job strain and risk of
CHD conducted separately for men and women
Table S3. Association between job strain measured in 2000 and incident coronary heart disease from
2001 to 2010 among 850,999 men and 809,151 women
Cases
Person-
Number
per 10,000
Adjusted HR
Job strain at baseline
years
of cases person-years
(95% CI)
Men
No job strain
3,921,227
8,435
21.5
1.00
Job strain
4,271,418
10,444
24.5
1.10 (1.06-1.13)
Women
No job strain
4,124,368
2,597
6.3
1.00
Job strain
3,800,499
2,683
7.1
1.11 (1.05-1.17)
HR: Hazard ratio; 95% CI: 95% Confidence interval.
Adjusted for age, family type, migration background, health service use and household disposable income.
Covariates were measured in the year 2000 and were treated as time-invariant.
Table S4. Association between number of years with job strain measured from 1996 to 2000 and incident
coronary heart disease from 2001 to 2010 among 718,906 men and 634,343 women employed
throughout the years 1996 to 2000
Cases
Person-
Number
per 10,000
Adjusted HR
Five-year exposure to job strain
years
of cases
person-years
(95% CI)
Men
Zero years with job strain
2,337,612
4,900
21.0
1.00
One to two years with job strain
1,145,571
2,563
22.4
1.22 (1.16-1.28)
Three to four years with job strain
1,176,730
2,704
23.0
1.22 (1.16-1.28)
Five years with job strain
2,291,559
5,749
25.1
1.15 (1.10-1.20)
Women
Zero years with job strain
2,314,851
1,460
6.3
1.00
One to two years with job strain
657,551
367
5.6
1.15 (1.03-1.29)
Three to four years with job strain
963,434
625
6.5
1.16 (1.06-1.28)
Five years with job strain
2,288,915
1,683
7.4
1.14 (1.06-1.22)
HR: Hazard ratio; 95% CI: 95% Confidence interval.
Adjusted for age, family type, migration background, health service use and household disposable income.
Covariates were measured in the year 2000 and were treated as time-invariant.
8
 BEU, Alm.del - 2019-20 - Bilag 252: Orientering om resultater fra en NFA-artikel om psykosocialt arbejdsmiljø og risiko for iskæmisk hjertesygdom, fra beskæftigelsesministeren
2195800_0009.png
Table S5. Association between persistent, onset and removal of job strain measured from 2000 to 2009
and incident coronary heart disease from 2001 to 2010 among 850,999 men and 809,151 women with a
one-year time lag between exposure and outcome
Cases
Exposure to job strain from one
Person-
Number
per 10,000
Adjusted HR
year to the subsequent year
years
of cases person-years
(95% CI)
Men
Persistent no job strain
3,208,901
6,455
20.1
1.00
Persistent job strain
3,143,688
6,476
20.6
1.05 (1.01-1.09)
No job strain to job strain (onset)
334,377
689
20.6
1.18 (1.09-1.28)
Job strain to no job strain (removal)
353,981
797
22.5
1.18 (1.10-1.27)
Job strain to out of employment
No job strain to out of employment
Out of employment to job strain
Out of employment to no job strain
Persistent out of employment
Women
Persistent no job strain
Persistent job strain
No job strain to job strain (onset)
Job strain to no job strain (removal)
152,280
164,182
72,427
87,698
653,339
561
598
175
188
2,940
36.8
36.4
24.2
21.4
45.0
1.21 (1.11-1.33)
1.15 (1.05-1.25)
1.20 (1.03-1.40)
1.04 (0.90-1.20)
1.22 (1.16-1.28)
3,023,671
3,126,587
266,966
251,371
1,616
1,796
147
159
5.3
5.7
5.5
6.3
1.00
1.13 (1.05-1.21)
1.28 (1.08-1.51)
1.26 (1.07-1.48)
Job strain to out of employment
164,482
166
10.1
1.29 (1.09-1.51)
No job strain to out of employment
162,382
189
11.6
1.39 (1.19-1.62)
Out of employment to job strain
94,876
60
6.3
1.31 (1.01-1.69)
Out of employment to no job strain
83,214
42
5.0
1.03 (0.76-1.40)
Persistent out of employment
731,121
1,105
15.1
1.45 (1.32-1.59)
HR: Hazard ratio; 95% CI: 95% Confidence interval.
Adjusted for age, family type, migration background, health service use and household disposable income.
The covariates age, family type and income were time-varying and were measured annually concurrent
with job strain. The covariate health service use was time-varying and was measured annually one year
before job strain was measured. The covariant migration background was time-invariant and was measured
in the year 2000.
9
 BEU, Alm.del - 2019-20 - Bilag 252: Orientering om resultater fra en NFA-artikel om psykosocialt arbejdsmiljø og risiko for iskæmisk hjertesygdom, fra beskæftigelsesministeren
2195800_0010.png
Appendix 5: Analysis of the association between job strain and risk of
CHD additionally adjusted for education
Table S6. Association between job strain measured in 2000 and incident coronary heart disease from
2001 to 2010 among 1,660,150 employees in Denmark, additionally adjusted for education at baseline
Cases
per 10,000
person-years
Job strain at baseline
Person-
years
Number
of cases
Adjusted HR
(95% CI)
No job strain
8,045,595
11,032
13.7
1.00
Job strain
8,071,917
13,127
16.3
1.00 (0.98-1.03)
HR: Hazard ratio; 95% CI: 95% Confidence interval.
Adjusted for sex, age, family type, migration background, health service use, household disposable income
and education.
Covariates were measured in the year 2000 and were treated as time-invariant.
Table S7. Association between number of years with job strain measured from 1996 to 2000 and incident
coronary heart disease from 2001 to 2010 among 1,353,249 employees in Denmark employed
throughout the years 1996 to 2000, additionally adjusted for education at baseline
Cases
per 10,000
person-years
Five-year exposure to job strain
Person-
years
Number
of cases
Adjusted HR
(95% CI)
Zero years with job strain
4,652,463
6,360
13.7
1.00
One to two years with job strain
1,803,121
2,930
16.2
1.08 (1.04-1.13)
Three to four years with job strain
2,140,164
3,329
15.6
1.07 (1.03-1.12)
Five years with job strain
4,580,474
7,432
16.2
1.02 (0.98-1.06)
HR: Hazard ratio; 95% CI: 95% Confidence interval.
Adjusted for sex age, family type, migration background, health service use, household disposable income
and education.
Covariates were measured in the year 2000 and were treated as time-invariant.
10
 BEU, Alm.del - 2019-20 - Bilag 252: Orientering om resultater fra en NFA-artikel om psykosocialt arbejdsmiljø og risiko for iskæmisk hjertesygdom, fra beskæftigelsesministeren
2195800_0011.png
Table S8. Association between persistent, onset and removal of job strain measured from 2000 to 2009
and incident coronary heart disease from 2001 to 2010, among 1,660,150 employees in Denmark with a
one-year time lag between exposure and outcome, additionally adjusted for education at baseline
Cases
Exposure to job strain from one year
Person-
Number per 10,000
Adjusted HR
to the subsequent year
years
of cases person-years
(95% CI)
Persistent no job strain
Persistent job strain
No job strain to job strain (onset)
Job strain to no job strain (removal)
6,232,572
6,270,275
601,343
605,352
8,071
8,272
836
956
12.9
13.2
13.9
15.8
1.00
0.98 (0.95-1.01)
1.11 (1.03-1.19)
1.11 (1.03-1.18)
Job strain to out of employment
316,762
727
23.0
1.14 (1.05-1.23)
No job strain to out of employment
326,564
787
24.1
1.17 (1.09-1.26)
Out of employment to job strain
167,302
235
14.0
1.14 (1.00-1.30)
Out of employment to no job strain
170,912
230
13.5
1.01 (0.88-1.15)
Persistent out of employment
1,384,460
4,045
29.2
1.20 (1.15-1.26)
HR: Hazard ratio; 95% CI: 95% Confidence interval.
Adjusted for sex, age, family type, migration background, health service use, household disposable income
and education.
The covariates age, family type, income and education were time-varying and were measured annually
concurrent with job strain. The covariate health service use was time-varying and was measured annually
one year before job strain was measured. The covariates sex and migration background were time-invariant
and were measured in the year 2000.
11
 BEU, Alm.del - 2019-20 - Bilag 252: Orientering om resultater fra en NFA-artikel om psykosocialt arbejdsmiljø og risiko for iskæmisk hjertesygdom, fra beskæftigelsesministeren
2195800_0012.png
Appendix 6: Analysis of the association between job strain and risk of
CHD conducted separately for education groups
Table S9. Association between job strain measured in 2000 and incident coronary heart disease from
2001 to 2010 among 1,660,150 employees in Denmark by educational level at baseline
Cases
per 10,000
person-years
20.0
20.7
Job strain at baseline
Low education
No job strain
Job strain
Medium education
No job strain
Job strain
Person-
years
1,427,985
2,315,358
Number
of cases
2,863
4,783
Adjusted HR
(95% CI)
1.00
0.96 (0.91-1.00)
3,037,071
4,374,273
4,517
6,925
14.9
15.8
1.00
1.01 (0.97-1.05)
High education
No job strain
3,498,922
3,501
10.0
1.00
Job strain
1,268,159
1,129
8.9
1.06 (0.99-1.13)
HR: Hazard ratio; 95% CI: 95% Confidence interval.
Adjusted for sex, age, family type, migration background, health service use and household disposable
income.
Covariates were measured in the year 2000 and were treated as time-invariant.
12
 BEU, Alm.del - 2019-20 - Bilag 252: Orientering om resultater fra en NFA-artikel om psykosocialt arbejdsmiljø og risiko for iskæmisk hjertesygdom, fra beskæftigelsesministeren
2195800_0013.png
Table S10. Association between number of years with job strain measured from 1996 to 2000 and
incident coronary heart disease from 2001 to 2010 among 1,353,249 employees in Denmark employed
throughout the years 1996 to 2000 by educational level at baseline
Cases
Number per 10,000
Adjusted HR
Five-year exposure to job strain
Person-years of cases person-years
(95% CI)
Low education
Zero years with job strain
661,348
1,272
19.2
1.00
One to two years with job strain
462,174
1,061
23.0
1.02 (0.94-1.11)
Three to four years with job strain
562,946
1,188
21.1
1.03 (0.95-1.11)
Five years with job strain
1,254,953
2,549
20.3
0.99 (0.92-1.05)
Medium education
Zero years with job strain
One to two years with job strain
Three to four years with job strain
Five years with job strain
1,424,615
899,247
1,200,057
2,667,050
2,275
1,425
1,767
4,184
16.0
15.8
14.7
15.7
1.00
1.11 (1.04-1.19)
1.09 (1.03-1.16)
1.03 (0.97-1.08)
High education
Zero years with job strain
2,533,565
2,745
10.8
1.00
One to two years with job strain
425,189
389
9.1
1.13 (1.02-1.26)
Three to four years with job strain
354,198
322
9.1
1.09 (0.97-1.23)
Five years with job strain
609,077
558
9.2
1.03 (0.94-1.13)
HR: Hazard ratio; 95% CI: 95% Confidence interval.
Adjusted for sex, age, family type, migration background, health service use and household disposable
income.
Covariates were measured in the year 2000 and were treated as time-invariant.
13
 BEU, Alm.del - 2019-20 - Bilag 252: Orientering om resultater fra en NFA-artikel om psykosocialt arbejdsmiljø og risiko for iskæmisk hjertesygdom, fra beskæftigelsesministeren
2195800_0014.png
Table S11. Association between persistent, onset and removal of job strain measured from 2000 to 2009
and incident coronary heart disease from 2001 to 2010, among 1,660,150 employees in Denmark with a
one-year time lag between exposure and outcome by educational level at baseline
Cases
Exposure to job strain from one year
Person-years Number per 10,000
Adjusted HR
to the subsequent year
of cases person-
(95% CI)
years
Low education
Persistent no job strain
938,644
1,839
19.6
1.00
Persistent job strain
1,601,852
2,785
17.4
0.99 (0.93-1.05)
No job strain to job strain (onset)
157,988
293
18.5
1.08 (0.96-1.22)
Job strain to no job strain (removal)
160,117
312
19.5
1.02 (0.91-1.15)
Job strain to out of employment
No job strain to out of employment
Out of employment to job strain
Out of employment to no job strain
Persistent out of employment
Medium education
Persistent no job strain
Persistent job strain
No job strain to job strain (onset)
Job strain to no job strain (removal)
Job strain to out of employment
No job strain to out of employment
Out of employment to job strain
Out of employment to no job strain
Persistent out of employment
High education
Persistent no job strain
Persistent job strain
No job strain to job strain (onset)
Job strain to no job strain (removal)
84,367
110,492
41,076
55,669
453,118
257
287
66
79
1,592
30.5
26.0
16.1
14.2
35.1
1.18 (1.03-1.35)
1.13 (1.00-1.28)
0.93 (0.72-1.18)
0.99 (0.79-1.24)
1.24 (1.14-1.34)
2,200,467
3,570,306
301,891
302,737
130,307
161,855
68,420
82,646
616,724
3,219
4,629
404
479
329
364
99
131
1,821
14.6
13.0
13.4
15.8
25.2
22.5
14.5
15.9
29.5
1.00
0.96 (0.91-1.00)
1.07 (0.96-1.18)
1.09 (0.99-1.20)
1.16 (1.03-1.30)
1.10 (0.98-1.23)
1.00 (0.82-1.22)
1.25 (1.05-1.50)
1.19 (1.11-1.28)
3,046,515
1,031,492
134,583
135,302
2,868
697
116
146
9.4
6.8
8.6
10.8
1.00
0.94 (0.87-1.02)
1.22 (1.01-1.47)
1.36 (1.15-1.61)
Job strain to out of employment
106,796
189
17.7
1.14 (0.98-1.33)
No job strain to out of employment
38,313
61
15.9
1.31 (1.01-1.69)
Out of employment to job strain
57,946
61
10.5
1.13 (0.88-1.46)
Out of employment to no job strain
24,992
21
8.4
1.03 (0.66-1.60)
Persistent out of employment
290,758
564
19.4
1.12 (1.01-1.24)
HR: Hazard ratio; 95% CI: 95% Confidence interval.
Adjusted for sex, age, family type, migration background, health service use and household disposable
income.
The covariates age, family type and income were time-varying and were measured annually concurrent
with job strain. The covariate health service use was time-varying and was measured annually one year
before job strain was measured. The covariates sex and migration background were time-invariant and
were measured in the year 2000.
14