Download as pdf or txt
Download as pdf or txt
You are on page 1of 11

International Journal of Occupational Safety and

Ergonomics

ISSN: 1080-3548 (Print) 2376-9130 (Online) Journal homepage: https://www.tandfonline.com/loi/tose20

The role of physical exercise as a personal


resource against job stress

Silke Heuse, Babette Gekeler & Daniel Fodor

To cite this article: Silke Heuse, Babette Gekeler & Daniel Fodor (2020): The role of physical
exercise as a personal resource against job stress, International Journal of Occupational Safety
and Ergonomics, DOI: 10.1080/10803548.2020.1732648

To link to this article: https://doi.org/10.1080/10803548.2020.1732648

Published online: 24 Apr 2020.

Submit your article to this journal

Article views: 2

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at


https://www.tandfonline.com/action/journalInformation?journalCode=tose20
International Journal of Occupational Safety and Ergonomics (JOSE), 2020
https://doi.org/10.1080/10803548.2020.1732648

The role of physical exercise as a personal resource against job stress


a∗
Silke Heuse , Babette Gekelerb and Daniel Fodorb
a University of Applied Sciences Europe, Germany; b Charité – Universitätsmedizin Berlin and Berlin Institute of Health, Germany

Purpose. The prevention of burnout symptoms is an essential goal in occupational health promotion. Physical exercise
provides health-promotion benefits. This study aimed to verify physical exercise and its planned preparation as additional
predictors of employees’ burnout symptoms next to job demands and resources. We used the job demands–resources model
as a theoretical framework. Method. In this longitudinal online study, 342 employees completed two questionnaires at an
interval of 4 weeks. Results. Moderation and moderated mediation analyses confirmed vigorous physical exercise as a
relevant personal resource, revealing that it buffers the detrimental effects of job demands on burnout symptoms. Planning
strategies strongly predicted physical exercise and supported the debilitating effects of job resources on burnout symptoms.
Especially in employees with medium levels of job stress, coping planning supported the enactment of vigorous physical
exercise. Physical exercise did not mediate the association between job demands, job resources and burnout symptoms.
Conclusions. This study enriches our knowledge about occupational health prevention. It points to a pressing topic for the
near future, i.e., how work conditions (e.g., job stress) and leisure time (e.g., physical exercise) can be successfully combined
to keep individuals’ job stress to a minimum and to prevent burnout symptoms.
Keywords: job stress; burnout; physical exercise; action planning; coping planning; job demands–resources model

1. Introduction
physical, psychological, social or organizational factors are
Prevention of physical and mental disorders is the essen- associated with job stress [5]. These factors which help
tial goal in occupational health promotion. Stress affects to achieve personal and work goals directly or indirectly
our everyday life at work or in leisure time and is known (e.g., control, autonomy or social support) are referred to
as an antecedent of burnout [1] and physical and men- as ‘job resources’ [5]. Those factors which are associ-
tal disorders [2]. In line with the transactional idea [3], ated with certain physiological and/or psychological costs
stress emerges if events are appraised as exceeding one’s (e.g., time pressure, emotional demands or work–home
personal resources. In the context of job stress, this idea conflict) are referred to as ‘job demands’ [5]. Job demands
was taken up in the job demands–resources (JD-R) model and resources directly affect burnout or work engagement,
and was transferred to the interplay of job demands and respectively. In a transactional sense, an imbalance of job
job resources. Knowing that not only job-specific but also demands and job resources increases the likelihood of
personal resources influence employees’ stress experiences experiencing job stress, which in turn is a risk factor for
[4], the purpose of this study was to identify additional burnout. Next to direct effects on burnout, job resources
personal resources which can be used preventively to also have a buffering effect which attenuates detrimental
reduce employees’ stress and burnout symptoms. There- effects of job demands on the development of job stress [7]
fore, we asked whether self-regulatory resources (i.e., plan- (for a review, also see [8]).
ning strategies and health behavior) usually investigated in The JD-R model has been extended by including per-
the preventive context will also impact in the context of sonal resources in addition to job resources [9]. Mean-
work. while, a long list of personal resources are investi-
gated such as beliefs (e.g., self-efficacy [10]), personality
traits [11] and stress-reducing competences (e.g., mindful-
1.1. The JD-R model ness [12]). To our knowledge, only a few studies have
The JD-R model was originally developed to understand investigated whether health behaviors such as healthy
the impact of work stress, especially on burnout [5] and nutrition or physical exercise interact with job stress
work engagement [6]. The model proposes that specific [13,14]

*Corresponding author. Email: silke.heuse@ue-germany.com

© 2020 Central Institute for Labour Protection – National Research Institute (CIOP-PIB)
2 S. Heuse et al.

1.2. Physical exercise and stress burnout symptoms. Referring to the JD-R model, personal
Physical exercise can be understood as an individual cop- resources can be considered in five ways: as an addi-
ing strategy [15] which reduces stress [16,17] and protects tional predictor next to job demands and resources; as a
from adverse consequences of stress [18–20]. The World moderator of the job characteristics–burnout/work engage-
Health Organization (WHO) [21] recommends that adults ment association; as a mediator of the job characteristics–
should do at least 75 min of vigorous-intensity aerobic burnout/work engagement association; as an antecedent
physical exercise (VPE) or alternatively at least 150 min influencing the perception of job characteristics; as a ‘third
of moderate-intensity aerobic physical exercise (MPE) variable’ explaining the relationship of job characteristics
throughout the week, or an equivalent combination of both and burnout/work engagement [9]. We assumed that per-
activities. Since moderate (e.g., hiking) and vigorous (e.g., sonal resources might serve as additional moderators in
running) physical exercise differ in intensity and resource the two-way interaction with job demands or in the three-
consumption, in the present study we distinguish between way interaction with both job demands and job resources,
these two kinds of physical exercise. or as mediators in the association of job demands, job
resources and burnout symptoms. Since physical exer-
cise may either strengthen the robustness against stress or,
1.3. Planning strategies to enhance health behavior as a recreational activity, allows coping with stress, we
tested MPE and VPE as either a moderator which buffers
Physical exercise helps to reduce stress and to support
detrimental effects of job demands or unfavorable combi-
mental health [22]. But job stress also affects the interface
nations of job demands and job resources, or as a mediator
between work and home, e.g., whether employees succeed
between potentially stressing job conditions and placed
in combining work and leisure time demands [23]. This
burnout symptoms as a distal outcome. Because plan-
is supported by findings that high job demands [24,25] or
ning strategies are antecedents of health behavior, action
unfavorable combinations of high job demands and low job
planning and coping planning were considered as addi-
resources [26] are associated with reduced physical exer-
tional resources and predictors of physical exercise in the
cise. Planning strategies are additional personal resources.
mediation process:
As proposed by the health action process approach [27] or
other health behavior theories, these self-regulatory strate- H 1a : MPE moderates the association between job demands
gies mediate between intention and enactment of health and burnout symptoms.
behavior and, in this, counteract detrimental stress effects. H 1b : VPE moderates the association between job demands
In the context of stress, planning strategies can also be and burnout symptoms.
understood as proactive coping strategies which shield goal
pursuit from barriers. They can be subdivided into action H 2a : MPE moderates the association between the job
demands–job resources interaction and burnout symptoms.
planning and coping planning [28]. Action planning is a
goal-oriented strategy which links a good opportunity for H 2b : VPE moderates the association between the job
behavior enactment with details of the intended behav- demands–job resources interaction and burnout symptoms.
ior (e.g., ‘If I return home after work, then I will do
H 3a : MPE mediates the association between the job
my exercises on the ergometer for 30 minutes’ [28]; orig- demands–job resources interaction and burnout symptoms.
inal emphases). Complementarily, coping planning is a
barrier-focused strategy which specifies a suitable coping H 3b : VPE mediates the association between the job
response for anticipated barriers (e.g., ‘If we have guests demands–job resources interaction and burnout symptoms.
in the evening, then I will do my exercises already in H 4a : Action planning moderates the mediation effects of
the morning’ [28]; original emphases). Planning strategies physical exercise in the relationship of the job demands–
are proposed to enhance perceptual accessibility of situa- job resources interaction with burnout symptoms.
tional cues specified in ‘if’ parts as well as to automatize H 4b : Coping planning moderates the mediation effects of
goal-directed responses specified in ‘then’ parts [29,30]. physical exercise in the relationship of the job demands–
Consequently, action initiation is assumed to be under job resources interaction with burnout symptoms.
direct control of the specified situational cue rather than
conscious control [29]. Planning has been established as
a successful strategy to enhance physical exercise [31,32], 2. Methods
even under stress [13,33]. 2.1. Participants and procedure
A total of 1483 German employees followed the invi-
tation for study participation through the SoSci Panel,1
1.4. Aims of the study an open scientific questionnaire panel. The initial invi-
Focusing on occupational health prevention, we aimed to tation email informed employees about the study’s aim
identify personal resources in the context of job stress and procedure, and was followed by an internet link
which might help to reduce employees’ risk of developing leading to the baseline questionnaire (first measurement
International Journal of Occupational Safety and Ergonomics (JOSE) 3

point [t1]). Four weeks later, participants received a sec- 2.2.2. Burnout symptoms
ond questionnaire (second measurement point [t2]). The Burnout symptoms were assessed using two subscales
study was checked and approved by the national ethics of the Maslach Burnout Inventory (MBI) [35]: both
commission. ‘emotional exhaustion’ (Cronbach’s α = 0.877 at t1 and
A total of 871 participants completed the study. Of α = 0.892 at t2) and ‘cynicism’ (Cronbach’s α = 0.881 at
these, 342 met the inclusion criteria (no restrictions for t1 and α = 0.91 at t2) consist of five items and range from
physical activity within the 8 weeks of retrospective assess- 1 = never to 6 = very often. A mean score was computed,
ment, employment with at least a 20-h average working with higher values representing more severe symptoms of
week as well as no missing data on moderate or vigor- burnout (Cronbach’s α = 0.893 at t1 and α = 0.909 at t2).
ous physical exercise and less than 10% of overall missing
data). This met the required sample size calculated in an
a priori power analysis with an assumed power of 0.80 2.2.3. Physical exercise
and an α value of 0.05. Participants’ age ranged from 22 At t1 and t2, moderate (MPE) and vigorous (VPE) physical
to 67 years with an average of 41.370 years (SD 10.769), exercise were assessed by an open-response item indicat-
and 63.50% were female. The participants worked an aver- ing the average time per week spent on physical exercise.
age of 42.342 h per week (SD 14.601; range 20–70) at Participants were asked ‘How many minutes have you
t1 and 42.651 h per week (SD 12.933; range 20–70) at been physically active on average per week within the
t2; variables were truncated, limiting working hours to no last 4 weeks?’ accompanied by a definition of and exam-
more than 70 h per week. The sample covered a broad ples for the two different levels of physical exercise (e.g.,
range of job types, including employees from the health- for VPE: ‘rapid heartbeat, e.g., intense swimming, run-
care sector, science, service industry, banking sector and ning, cycling’). This assessment is an adaption of the
many more. Analyses revealed that employees meeting widely used Godin and Shephard leisure-time physical
the inclusion criteria were younger in age (p < 0.001) and activity questionnaire (GSLTPAQ) [36]. Participants were
reported higher job resources (p < 0.05) and less burnout instructed to report exercises which were performed for at
symptoms (p < 0.05) than those who did not meet the least 10 min per bout.
criteria.
2.2.4. Action planning and coping planning
Planning strategies were assessed at t1 by widely used
2.2. Measures
action planning and coping planning scales [28]. For action
2.2.1. Job demands and job resources planning, participants were asked to indicate on six items
Job demands were assessed at t1 in terms of ‘time pressure’ whether they have planned what, when, where, how often,
using five items (e.g., ‘How often did you work under time for how long and with whom they will perform physical
pressure?’) from the validated instrument for stress-related exercise (e.g., ‘For the upcoming four weeks I made a spe-
job analysis (Instrument zur stressbezogenen Taetigkeits- cific plan where I will perform physical exercise.’). The
analyse [ISTA] [34]). Items were framed to the previous response formats ranged from 1 = completely disagree
week. Responses could be rated from 1 = rarely/never to 5 = completely agree. A mean score was computed
to 5 = very often. A mean score was calculated, with (Cronbach’s α = 0.944). Participants were also asked to
higher values representing more job demands (Cronbach’s indicate on five items the extent to which they have used
α = 0.867). coping planning strategies (e.g., ‘For the upcoming four
Job resources were measured as a combination of two weeks I made a specific plan what to do when some-
5-item scales from the ISTA, ‘time control’ (e.g., ‘How thing unexpected comes up.’). A mean score was computed
much were you able to determine your pace of work (Cronbach’s α = 0.923).
on your own?’) and ‘control at work’ (e.g., ‘How much The means, standard deviations and correlations of all
influence did you have on the type of work that was variables are reported in Table 1.
assigned to you?’). Again, items were framed to the pre-
vious week. The response formats ranged from 1 = not
at all to 6 = more than one hour and 1 = less than five 2.3. Analytical procedure
minutes a day to 6 = more than two hours a day, respec- Regression analyses were conducted to test our assump-
tively, or from 1 = not at all to 5 = all results of work. We tions using the PROCESS macro for SPSS version 23
transformed 6-point scales into 5-point scales [34] in order [37]. Models aimed at the prediction of burnout symp-
to compute a mean score with higher values represent- toms at t2, controlled for burnout symptoms at t1. To
ing more job resources (Cronbach’s α = 0.851). Because test hypotheses H 1a and H 1b , moderation models consid-
the ISTA scales are comparable to JD-R scales, we chose ering two-way interactions of job demands and physical
them for reasons of comparability with our own previous exercise at t1 in the prediction of burnout symptoms at
studies. t2 (Models 1 and 2) were calculated using PROCESS
4
Table 1. Means, standard deviations and correlations for all study variables.

Variable M SD 1 2 3 4 5 6 7 8 9 10 11 12 13

1. JD t1 3.118 0.925
2. JR t1 3.692 0.706 −0.300***
3. MBI t1 2.969 1.043 0.331*** −0.408***
4. MBI t2 2.942 1.089 0.362*** −0.406*** 0.843***
5. MPE t1 82.652 111.504 −0.018 0.086 −0.073 −0.056

S. Heuse et al.
6. VPE t1 50.389 74.006 0.012 0.040 −0.029 −0.034 0.339***
7. MPE t2 86.754 96.726 0.020 0.070 −0.197† −0.092† 0.529*** 0.294***
8. VPE t2 43.591 65.625 0.007 0.082 −0.031 −0.036 0.344*** 0.599*** 0.265***
9. AP t1 3.525 1.307 0.052 0.054 −0.126* −0.088 0.298*** 0.345*** 0.319*** 0.359***
10. CP t1 2.656 1.273 −0.006 0.102† −0.112* −0.075 0.282*** 0.353*** 0.249*** 0.385*** 0.666***
11. WH t1 42.342 14.601 0.291*** 0.080 −0.014 −0.014 0.199† 0.155** 0.006 0.146** 0.082 0.101†
12. WH t2 42.651 12.933 0.313*** 0.041 0.049 0.085 0.050 −0.014 0.024 0.063 0.046 0.042 0.535**
13. Gendera – – −0.082 0.091† −0.032 −0.090 −0.129* −0.040 −0.087 −0.014 −0.139* −0.067 0.114* 0.113*
14. Age 41.370 10.769 −0.046 0.093† −0.083 −0.041 0.045 −0.038 0.027 −0.039 0.062 0.124* 0.017 −0.027 0.132*

*p < 0.05; **p < 0.01; ***p < 0.001; † p < 0.10; a 1 = male; 2 = female.
Note: AP = action planning; CP = coping planning; JD = job demands; JR = job resources; MBI = burnout symptoms; MPE = moderate physical exercise; t1 = first measurement
point, t2 = second measurement point; VPE = vigorous physical exercise; WH = working hours.
International Journal of Occupational Safety and Ergonomics (JOSE) 5

model 1. To test hypotheses H 2a and H 2b , moderation mod- Table 2. Moderated models predicting burnout symptoms
(PROCESS model 1 [37]).
els considering three-way interactions of job demands, job
resources and physical exercise at t1 in the prediction of Model 1 Model 2
burnout symptoms at t2 (Models 3 and 4) were calculated (DV = burnout t2) (DV = burnout t2)
using PROCESS model 3. To test hypotheses H 3a and H 3b , Predictor B SE B SE
moderated mediation models with job demands and job
resources at t1 as well as their interaction as predictors Gender −0.067* 0.030 −0.069* 0.030
and physical exercise at t1 as a mediator in the prediction Age 0.040 0.029 0.040 0.029
WH (average) 0.018 0.032 0.022 0.032
of burnout symptoms at t2 (Models 5 and 6) were calcu- MBI t1 0.810*** 0.031 0.817*** 0.031
lated using PROCESS model 8. To test hypotheses H 4a JD t1 0.081* 0.033 0.070* 0.034
and H 4b , planning strategies at t1 as additional moderators MPE t1 −0.017 0.029 – –
and their respective two-way and three-way interactions VPE t1 – – −0.017 0.031
with job demands and job resources at t1 (Models 7–10) JD × MPE t1 −0.034 0.032 – –
JD × VPE t1 – – −0.060† 0.033
were included in these moderated mediation models, using R² 0.724*** – 0.726*** –
PROCESS model 12. All models were controlled for age,
gender, average working hours and burnout at t1. Z scores *p < 0.05; ***p < 0.001; † p < 0.10.
were used to allow a comparison of effect sizes. Signifi- Note: Z-standardized variables. DV = dependent variable;
cant interactions were qualified by estimating the size of JD = job demands; MBI = burnout symptoms;
MPE = moderate physical exercise; t1 = first
the conditional effects of the predictor at combinations of measurement point, t2 = second measurement point;
high and low values of the moderators (M ± 1SD) in terms VPE = vigorous physical exercise; WH = working hours.
of simple slope analyses [38].

3. Results (hypothesis H 2a ), job demands were again positively


associated with more burnout symptoms and addition-
Because physical exercise is known to prevent burnout
ally a marginal significant interaction of job demands
symptoms, we considered physical exercise as a personal
and job resources (β = −0.52, p = 0.066) was found
resource in the prediction of burnout symptoms. Bivari-
(see Table 3, Model 3). Simple slope analyses revealed
ately, MPE but not VPE was associated with less burnout
that in individuals with a low amount of job resources,
symptoms. Action planning and coping planning are com-
job demands were positively associated with burnout
monly known as predictors of health behavior. Here, they
(β = 0.121, p = 0.007), whereas for individuals report-
were bivariately associated with MPE and VPE, but not
ing a high amount of job resources, job demands were not
bivariately related to job demands and job resources. All
related to burnout symptoms (β = 0.015, ns). The same
correlations are presented in Table 1.
marginal interaction of job demands and job resources
(β = −0.051, p = 0.071) was also found in Model 4
3.1. Extended JD-R model prediction of burnout: (hypothesis H 2b ), i.e., the prediction of burnout by job
physical exercise as a moderator demands, job resources, VPE and the respective interac-
Testing Model 1 (hypothesis H 1a ), i.e., the prediction of tion terms (see Table 3, Model 4). In addition, job demands
burnout by job demands, MPE and the respective inter- and VPE again interacted significantly in the same way
action term, only job demands were positively associated as already described. Models 3 and 4 did not show
(β = 0.081, p = 0.016) with more burnout symptoms at three-way interactions of job characteristics and physical
t2 (see Table 2, Model 1). Testing Model 2 (hypothesis exercise. Therefore, hypotheses H 2a and H 2b had to be
H 1b ), i.e., the prediction of burnout by job demands, VPE rejected.
and the respective interaction term, a marginal significant
two-way interaction (β = −0.060, p = 0.071) was found
(see Table 2, Model 2). Simple slope analyses revealed 3.2. Extended JD-R model prediction of burnout:
that in individuals with less VPE, job demands were pos- physical exercise as a mediator
itively associated with burnout (β = 0.111, p = 0.003), Moderated mediation Models 5 and 6 considered physical
whereas for individuals reporting higher amounts of exercise as a mediator of the association of job demands
VPE, job demands were not related to burnout symp- and job resources as well as their interaction with burnout
toms (β = 0.015, ns). Thus, only VPE rather than MPE symptoms. Neither job demands, job resources nor their
buffers the detrimental effects of job demands. Hypothe- interaction predicted physical exercise, and the latter also
sis H 1a had to be rejected, but hypothesis H 1b could be did not predict burnout symptoms. Consequently, neither
retained. MPE nor VPE functioned as a mediator (see Table 4, Mod-
Testing Model 3 with the interaction of MPE and both els 5 and 6). Therefore, hypotheses H 3a and H 3b were
job demands and job resources in the prediction of burnout rejected.
6 S. Heuse et al.

Table 3. Moderated models predicting burnout symptoms Table 4. Moderated mediation models predicting
(PROCESS model 3 [37]). burnout symptoms (PROCESS model 8 [37]).

Model 3 Model 4 Model 5 Model 6


(DV = burnout t2) (DV = burnout t2)
Predictor B SE B SE
Predictor B SE B SE
DV = MPE t1 DV = VPE t1
Gender −0.070* 0.030 −0.072* 0.030 Gender −0.120* 0.050 −0.045 0.054
Age 0.038 0.029 0.033 0.029 Age 0.030 0.049 −0.034 0.052
WH (average) 0.034 0.033 0.035 0.032 WH (average) 0.008 0.054 0.043 0.058
MBI t1 0.789*** 0.033 0.797*** 0.032 MBI t1 −0.058 0.054 −0.015 0.058
JD t1 0.067* 0.034 0.056 0.035 JD t1 0.011 0.057 0.002 0.061
JR t1 −0.044 0.034 −0.047 0.033 JR t1 0.065 0.056 0.022 0.060
MPE t1 −0.003 0.033 – – JD × JR t1 0.023 0.047 0.035 0.050
VPE t1 – – −0.009 0.032 R² 0.032 0.008
JD × JR t1 −0.052† 0.028 −0.051† 0.028
JD × MPE t1 −0.034 0.032 – – DV = burnout t2 DV = burnout t2
JR × MPE t1 −0.055 0.034 – – Gender −0.069* 0.030 −0.069* 0.030
JD × JR × MPE t1 0.035 0.032 – – Age 0.045 0.029 0.044 0.029
JD × VPE t1 – – −0.067* 0.034 WH (average) 0.030 0.032 0.030 0.032
JR × VPE t1 – – −0.039 0.037 MBI t1 0.797*** 0.032 0.797*** 0.032
JD × JR × VPE t1 – – −0.005 0.033 JD t1 0.067† 0.034 0.067† 0.034
R² 0.732*** – 0.733*** – JR t1 −0.047 0.034 −0.047 0.034
MPE t1 −0.007 0.033 – –
*p < 0.05; **p < 0.01; ***p < 0.001; † p < 0.10. VPE t1 – – −0.014 0.030
Note: Z-standardized variables. DV = dependent variable; JD × JR t1 −0.050† 0.028 −0.049† 0.028
JD = job demands; JR = job resources; MBI = burnout R² 0.729*** 0.728***
symptoms; MPE = moderate physical exercise; t1 = first
measurement point; t2 = second measurement point; *p < 0.05; **p < 0.01; ***p < 0.001; † p < 0.10.
VPE = vigorous physical exercise; WH = working hours. Note: Z-standardized variables. DV = dependent
variable; JD = job demands; JR = job resources;
MBI = burnout symptoms; MPE = moderate physical
exercise; t1 = first measurement point; t2 = second
measurement point; VPE = vigorous physical exercise;
3.3. Extended JD-R model prediction of burnout: WH = working hours.
planning strategies as an additional moderator
In the following Models 7–10, planning strategies were
included as an additional moderator. In Models 7–10, Interestingly, in Models 7–10, planning strategies sig-
action planning or coping planning were positively related nificantly interacted with job resources, predicting burnout
to increased MPE or VPE. In Model 10, the three-way symptoms at t2 (e.g., see Figure 1, Model 10). Simple slope
interaction of coping planning with job demands and job analyses revealed that in individuals with less use of plan-
resources was found in the prediction of VPE (β = 0.144, ning strategies, job resources were not related to burnout
p = 0.004). Simple slope analyses revealed that cop- symptoms (β = 0.046, ns), whereas for individuals report-
ing planning was associated with increased VPE for all ing a higher use of planning strategies, job resources
combinations of job demands and job resources, with were negatively associated with burnout (β = −0.168,
the strongest associations for combinations of high job p < 0.001).
demands and high job resources (β = 0.495, p = 0.001)
or low job demands and low job resources (β = 0.585,
4. Discussion
p = 0.001). Additionally, job demands were positively
associated with VPE in the presence of both high coping 4.1. Explaining the results
planning and high job resources (β = 0.197, p = 0.073), Stress affects our everyday life at work or in leisure time.
and were negatively associated if coping planning was The present longitudinal study was designed to exam-
not complemented by high job resources (β = −0.214, ine whether physical activity is able to reduce stress and
p = 0.064). However, neither MPE nor VPE predicted therefore protect against burnout symptoms. In addition,
burnout symptoms, i.e., no mediation could be detected. we assessed planning strategies known to support physi-
Figure 1 shows Model 10 to represent all of the moder- cal exercise performance. For this purpose, we used the
ated mediation models. Given the lack of both mediation JD-R model as a theoretical framework, including use
and the moderating role of action planning, hypothesis of planning strategies and physical exercise as additional
H 4a was rejected and hypothesis H 4b was only partially resources for the prevention of burnout symptoms next to
accepted with regard to the moderating role of coping job demands and job resources. The findings partially sup-
planning. port our assumption that physical exercise and planning
International Journal of Occupational Safety and Ergonomics (JOSE) 7

Figure 1. Moderated mediation models predicting burnout symptoms, considering vigorous physical exercise as a mediator
(PROCESS model 12 [37]).
**p < 0.01; ***p < 0.001; † p < 0.10.
Note: Z-standardized variables, controlled for gender, age, working hours, VPE t1 and MBI t1. CP = coping planning; JD = job
demands; JR = job resources; MBI = burnout symptoms; t1 = first measurement point; t2 = second measurement point;
VPE = vigorous physical exercise.

strategies function as additional moderators, whereas we exercise. This is in line with previous studies introduc-
failed to confirm physical exercise as a mediator between ing the use of action planning and coping planning to
job demands and job resources and burnout symptoms. support health behavior change [43]. Additionally, cop-
Incorporating physical exercise as a personal resource ing planning rather than action planning also interacted
into the JD-R model, VPE rather than MPE was identi- with job characteristics. The use of coping planning was
fied as an important personal resource. While VPE did even associated with increased VPE in individuals who
not protect against adverse effects of stressful job condi- reported balanced combinations of job demands and job
tions stemming from an unfavorable combination of job stress, resulting in medium amounts of stress [8]. This dif-
demands and job resources in general, it was able to buffer ferential effect might be explained by the fact that high
detrimental effects of job demands specifically. Such dose– stress conditions consume many resources. Under these
response relations are still known in the context of physical conditions, employees are less likely to engage in physical
exercise [39], e.g., regarding third variables like cogni- exercise [24], or even plan to exercise, because integrat-
tive functioning [40] or personality [41]. In summary, VPE ing VPE into everyday life is time-consuming and also
might not only function as a personal resource in the work affords self-regulatory effort [44]. On the contrary, the least
context, but also strengthens further influencing factors to stressful job conditions might not afford coping planning
buffer job demands. in order to support physical exercise. Consequently, espe-
Although other personal resources like work ability cially individuals experiencing medium levels of job stress,
have been established as a mediator [42], contradictory to i.e., who are still capable and willing to use job-external
our mediation assumptions, physical exercise did not medi- resources to reduce stress, might benefit from coping plan-
ate the relationship between job conditions and burnout ning, which helps to anticipate and deal with barriers and
symptoms. An explanation might be the present selective difficulties.
sample that was characterized by a low to moderate sever- Planning strategies were also shown to strengthen the
ity of burnout symptoms. Understanding physical exercise debilitating effect of job resources on burnout symptoms.
as a recreational strategy used to decrease detrimental This underlines the understanding of spontaneously used
effects of potentially harmful job conditions and the devel- action planning and coping planning as an additional per-
opment of burnout symptoms, individuals did not use it sonal resource. Spontaneous use of planning strategies is
to cope with job stress in the present sample or, in other already known as an important self-regulatory strategy
words, did not feel obliged to counteract the development [45], which also supports coping with stress. The find-
of more severe burnout symptoms. ings highlight the importance of encompassing a holistic
Regarding the use of planning strategies, our study con- perspective (e.g., job-specific and unspecific resources) to
firmed their important role in the enactment of physical effectively prevent burnout symptoms [4].
8 S. Heuse et al.

4.2. Limitations especially because planning strategies can easily be taught


There are a number of strengths and shortcomings with to employees in short planning interventions [49].
the present study that require additional commentary. It Using health-psychological and occupational approa-
can be assumed that study participation was influenced ches for the clinical context, the present study provides
by individuals’ job stress. Although this online assessment insight into action fields which are equally important for
was easily applicable and required less time and resources, employees, health professionals and decision-makers in
especially if time pressure was high, study participation companies and management. Shedding light on the inter-
might be burdensome and would likely influence individu- face of job conditions (e.g., job stress) and leisure-time
als’ decisions to participate. As a consequence, the sample activities (e.g., recreation in terms of physical exercise), the
comprised employees working under lower or medium study points to a pressing topic for the near future [50], i.e.,
job stress conditions. Dropout analysis points to a selec- how work and family or leisure time can be successfully
tive sample regarding age and job stress. Thus, beneficial combined in order to keep individuals’ stress to a minimum
effects of planning might be overestimated. This represents and to promote mental health.
a central weakness of this study and affords future studies
to replicate these findings in more stressed and overworked Note
samples.
1. The SoSci Panel is an open online scientific questionnaire
The present sample was very heterogeneous, covering panel of 90,317 individuals (as of January 2015). Depending
jobs from a number of different sectors. However, in con- on the aimed sample size, an invitation is sent to individuals
trast to the assumptions of the JD-R model [8], we did not of the panel meeting the inclusion criteria.
test the specific job demands and job resources of each job
type, although time pressure and control are considered
Disclosure statement
as general factors influencing job stress across all indus-
tries. Previous findings suggest that different job demands No potential conflict of interest was reported by the authors.
and resources influence stress experiences and health in
specific ways [46], and therefore future research should ORCID
take different job-specific demands and resources [23] as Silke Heuse http://orcid.org/0000-0003-0673-5692
well as expert ratings [5] or external assessments [47] into
account.
References
Follow-up assessments of the present longitudinal
[1] Sun J-W, Lin P-Z, Zhang H-H, et al. A non-linear
study were limited. Covering a 4-week period will hardly
relationship between the cumulative exposure to occu-
suffice to detect clinically significant changes in burnout pational stressors and nurses’ burnout and the poten-
symptoms. Therefore, future studies should cover a longer tially emotion regulation factors. J Ment Health;27(5):1–7.
period of time, including further follow-up assessments. doi:10.1080/09638237.2017.1385740
[2] Hannigan B, Edwards D, Burnard P. Stress and stress
management in clinical psychology: findings from a
systematic review. J Ment Health. 2017;13:235–245.
4.3. Conclusion doi:10.1080/09638230410001700871
Despite these limitations, the present study introduces [3] Lazarus RS. Psychological stress and the coping process.
New York (NY): McGraw-Hill; 1966.
physical exercise and use of planning strategies as personal [4] Viotti S, Martini M, Converso D. Are there any job
resources, which can be considered worthy candidates for resources capable of moderating the effect of physi-
enhancing employees’ stress management [19] and for the cal demands on work ability? A study among kinder-
prediction of health-related outcomes [48]. Physical exer- garten teachers. Int J Occup Saf Ergon. 2017;23:544–552.
cise is known as a recreational strategy to improve overall doi:10.1080/10803548.2016.1267976
[5] Demerouti E, Bakker AB, Nachreiner F, et al. The job
physical and mental health in general, and to reduce job demands–resources model of burnout. J Appl Psych.
stress, i.e., to support coping with job demands, specifi- 2001;86:499–512. doi:10.1037/0021-9010.86.3.499
cally. However, depending on the type of physical exercise, [6] Schaufeli WB, Bakker AB. Job demands, job resources,
the enactment is more or less challenging [44]. The prepa- and their relationship with burnout and engagement: a
ration of such behavior is recommendable, especially if multi-sample study. J Organ Behav. 2004;25:293–315.
doi:10.1002/job.248
medium to high job stress is experienced, which in turn [7] Bakker AB, Demerouti E, Euwema MC. Job resources
might be accompanied by limited resources, e.g., time buffer the impact of job demands on burnout. J Occup
control, social support or cognitive functioning. Planning Health Psych. 2005;10:170–180. doi:10.1037/1076-8998.
strategies have been proven successful as a resource- 10.2.170
consuming means to support goal achievement even under [8] Bakker AB, Demerouti E. The job demands–resources
model: state of the art. J Manag Psychol. 2007;22:309–328.
stress due to their automaticity in behavioral enactment doi:10.1108/02683940710733115
[28,47]. This qualifies the use of planning strategies as an [9] Schaufeli WB, Taris TW. A critical review of the job
effective tool to cope with the negative effects of job stress; demands–resources model: implications for improving
International Journal of Occupational Safety and Ergonomics (JOSE) 9

work and health. In: Bauer GF, Hämmig O, editors. Bridg- [25] Morassaei S, Smith PM. Examining the relationship
ing occupational, organizational and public health. Dor- between psychosocial working conditions, physical work
drecht: Springer; 2014. p. 43–68. demands, and leisure time physical activity in Canada.
[10] Xanthopoulou D, Heuven E, Demerouti E, et al. Work- J Occup Environ Med. 2011;53:1099–1105. doi:10.1097/
ing in the sky: a diary study on work engagement among JOM.0b013e3182307863
flight attendants. J Occup Health Psych. 2008;13:345–356. [26] Payne N, Jones F, Harris P. The impact of working life on
doi:10.1037/1076-8998.13.4.345 health behavior: the effect of job strain on the cognitive pre-
[11] Bakker AB, Boyd CM, Dollard M, et al. The role of person- dictors of exercise. J Occup Health Psych. 2002;7:342–353.
ality in the job demands–resources model. Career Dev Int. doi:10.1037/1076-8998.7.4.342
2010;15:622–636. doi:10.1108/13620431011094050 [27] Schwarzer R, Luszczynska A. Health Action Process
[12] Grover SL, Teo ST, Pick D, et al. Mindfulness as Approach. In: Conner M, Norman P, editors. Predicting
a personal resource to reduce work stress in the job health behaviours. Maidenhead: McGraw Hill Open Univer-
demands–resources model. Stress Health. 2016;33:426– sity Press; 2015. p. 252–278.
436. doi:10.1002/smi.2726 [28] Scholz U, Sniehotta FF, Burkert S, et al. Increasing physical
[13] Fodor D, Wiedemann AU, Antoni C, et al. Healthy eat- exercise levels: age-specific benefits of planning. J Aging
ing at different risk levels for job stress: testing a moder- Health. 2007;19:851–866. doi:10.1177/089826430730
ated mediation. J Occup Health Psych. 2014;19:259–267. 5207
doi:10.1037/a0036267 [29] Gollwitzer PM, Schaal B. Metacognition in action:
[14] Payne N, Jones F, Harris P. A daily diary investiga- the importance of implementation intentions. Pers Soc
tion of the impact of work stress on exercise inten- Psychol Rev. 1998;2:124–136. doi:10.1207/s15327957
tion realization: can planning overcome the disruptive pspr0202_5
influence of work? Psychol Health. 2010;25:111–129. [30] Hagger MS, Luszczynska A. Implementation intention and
doi:10.1080/08870440903337622 action planning interventions in health contexts: state of the
[15] Wickramasinghe V. Work-related dimensions and job stress: research and proposals for the way forward. Appl Psychol-
the moderating effect of coping strategies. Stress Health. Hlth We. 2014;6:1–47.
2010;26:417–429. doi:10.1002/smi.1314 [31] Bélanger-Gravel A, Godin G, Amireault S. A meta-
[16] Clark MM, Jenkins SM, Hagen PT, et al. High stress and analytic review of the effect of implementation intentions
negative health behaviors: a five-year wellness center mem- on physical activity. Health Psychol Rev. 2013;7:23–54.
ber cohort study. J Occup Environ Med. 2016;58:868–873. doi:10.1080/17437199.2011.560095
doi:10.1097/JOM.0000000000000826 [32] Carraro N, Gaudreau P. Spontaneous and experimen-
[17] Nguyen-Michel ST, Unger JB, Hamilton J, et al. Associa- tally induced action planning and coping planning for
tions between physical activity and perceived stress/hassles physical activity: a meta-analysis. Psychol Sport Exerc.
in college students. Stress Health. 2006;22:179–188. 2013;14:228–248. doi:10.1016/j.psychsport.2012.10.004
doi:10.1002/smi.1094 [33] Gerber M, Lindwall M, Brand S, et al. Longitu-
[18] Fang YY, Huang CH, Hsu MC. Effectiveness of a phys- dinal relationships between perceived stress, exercise
ical activity program on weight, physical fitness, occupa- self-regulation and exercise involvement among physi-
tional stress, job satisfaction and quality of life of over- cally active adolescents. J Sport Sci. 2015;33:369–380.
weight employees in high-tech industries: a randomized doi:10.1080/02640414.2014.946072
controlled study. Int J Occup Saf Ergon. 2019;25(4):621– [34] Semmer NK, Zapf D, Dunckel H. Instrument zur stressbezo-
629. doi:10.1080/10803548.2018.1438839 genen Taetigkeitsanalyse [Instrument for stress-related job
[19] Moreira-Silva I, Santos R, Abreu S, et al. The effect of a analysis; ISTA]. In: Dunckel H, editor. Handbuch psychol-
physical activity program on decreasing physical disabil- ogischer Arbeitsanalyseverfahren [Analytical instrumenta-
ity indicated by musculoskeletal pain and related symptoms tion handbook]. Zürich: VdF Hochschulverlag; 1999. p.
among workers: a pilot study. Int J Occup Saf Ergon. 179–204. German.
2014;20:55–64. doi:10.1080/10803548.2014.11077028 [35] Maslach C, Jackson SE. The Maslach Burnout Inventory
[20] Toker S, Biron M. Job burnout and depression: unrav- (manual). 2nd ed. Palo Alto (CA): Consulting Psychologists
eling their temporal relationship and considering the Press; 1986.
role of physical activity. J Appl Psych. 2012;7:699–710. [36] Godin G, Shephard RJ. A simple method to assess exer-
doi:10.1037/a0026914 cise behavior in the community. Can J Appl Sport Sci.
[21] World Health Organization. Global recommendations on 1985;10:141–146.
physical activity for health [Internet]; 2010 [cited 2016 Oct [37] Hayes AF. Introduction to mediation, moderation, and con-
12]. Available from: http://www.who.int/dietphysicalactivity ditional process analysis: a regression-based approach. New
/publications/9789241599979/en/. York (NY): Guilford Press; 2013.
[22] Tyson P, Wilson K, Crone D, et al. Physical activity and [38] Aiken LS, West SG. Multiple regression: testing and
mental health in a student population. J Ment Health. interpreting interactions. Newbury Park (CA): SAGE;
2010;19:492–499. doi:10.3109/09638230902968308 1991.
[23] Bakker AB, ten Brummelhuis LL, Prins JT, et al. [39] Proper KI, van den Heuvel SG, De Vroome EM,
Applying the job demands–resources model to the et al. Dose–response relation between physical activ-
work–home interface: a study among medical residents ity and sick leave. Br J Sports Med. 2006;40:173–178.
and their partners. J Vocat Behav. 2001;79:170–180. doi:10.1136/bjsm.2005.022327
doi:10.1016/j.jvb.2010.12.004 [40] Esteban-Cornejo I, Tejero-Gonzalez CM, Sallis JF, et al.
[24] Fransson EI, Heikkilä K, Nyberg ST, et al. Job strain Physical activity and cognition in adolescents: a sys-
as a risk factor for leisure-time physical inactivity: an tematic review. J Sci Med Sport. 2015;18:534–539.
individual-participant meta-analysis of up to 170,000 men doi:10.1016/j.jsams.2014.07.007
and women. The IPD-Work Consortium. Am J Epidemiol. [41] de Bruijn G-J, de Groot R, van den Putte B, et al. Con-
2012;176:1078–1089. doi:10.1093/aje/kws336 scientiousness, extroversion, and action control: comparing
10 S. Heuse et al.

moderate and vigorous physical activity. J Sport Exercise disability related to arterial hypertension and its com-
Psy. 2009;31:724–742. doi:10.1123/jsep.31.6.724 plications. Int J Occup Saf Ergon. 2017;23:259–266.
[42] Viotti S, Guidetti G, Loera B, et al. Stress, work abil- doi:10.1080/10803548.2016.1179524
ity, and an aging workforce: a study among women [47] Brandstätter V, Lengfelder A, Gollwitzer PM. Implemen-
aged 50 and over. Int J Stress Manage. 2017;24:98–121. tation intentions and efficient action initiation. J Pers
doi:10.1037/str0000031 Soc Psychol. 2001;81:946–960. doi:10.1037/0022-3514.81.
[43] Sheeran P, Milne SE, Webb TL, et al. Implementation inten- 5.946
tions. In: Conner M, Norman P, editors. Predicting health [48] Airila A, Hakanen JJ, Schaufeli WB, et al. Are job and
behaviour. Buckingham: Open University Press; 2005. p. personal resources associated with work ability 10 years
276–323. later? The mediating role of work engagement. Work Stress.
[44] Winters ER, Petrosa RL, Charlon TE. Using social cogni- 2014;28:87–105. doi:10.1080/02678373.2013.872208
tive theory to explain discretionary, ‘leisure-time’ physical [49] Hagger MS, Luszczynska A, de Wit J, et al. Implementa-
exercise among high school students. J Adolescent Health. tion intention and planning interventions in health psychol-
2003;32:436–442. doi:10.1016/S1054-139X(03)00046-6 ogy: recommendations from the synergy expert group for
[45] Burkert S, Knoll N, Scholz U, et al. Self-regulation fol- research and practice. Psychol Health. 2016;31:814–839.
lowing prostatectomy: phase-specific self-efficacy beliefs doi:10.1080/08870446.2016.1146719
for pelvic-floor exercise. Brit J Health Psychology. [50] Guha M. The changing nature of work, leisure and involve-
2012;17:273–293. doi:10.1111/j.2044-8287.2011.02037.x ment in society: some possible ideas from an old philoso-
[46] Lazaridis K, Jovanović J, Jovanović J, et al. The pher. J Ment Health. 2017;27:1–3. doi:10.1080/09638237.
impact of occupational stress factors on temporary work 2017.1417572

You might also like