Content uploaded by Marc Van Veldhoven
Author content
All content in this area was uploaded by Marc Van Veldhoven
Content may be subject to copyright.
Int Arch Occup Environ Health (2009) 82:291–303
DOI 10.1007/s00420-008-0334-0
123
ORIGINAL ARTICLE
Emotional exhaustion and mental health problems among
employees doing “people work”: the impact of job demands,
job resources and family-to-work conXict
Geertje van Daalen · Tineke M. Willemsen ·
Karin Sanders · Marc J. P. M. van Veldhoven
Received: 17 December 2006 / Accepted: 7 May 2008 / Published online: 30 May 2008
© Springer-Verlag 2008
Abstract
Objective This study investigates the relationship
between four job characteristics and family-to-work con-
Xict on emotional exhaustion and mental health problems.
Methods Multiple regression analyses were performed
using data from 1,008 mental health care employees. Sepa-
rate regression analyses were computed for high and low
patient interaction jobs.
Results DiVerent job characteristics as well as family-to-
work conXict were associated with emotional exhaustion
and mental health problems in each job type. The relation-
ship between family-to-work conXict and emotional
exhaustion was mitigated by social support from colleagues
for those who worked in low patient interaction jobs.
Conclusion In addition to general and speciWc stressors, it
is worthwhile to include home-related stressors that inter-
fere with the work domain in stress research.
Keywords Emotional exhaustion · Mental health
problems · Job characteristics · Family-to-work conXict ·
Patient interaction · Mental health care workers
Introduction
Work stress is one of the most prevalent health problems of
these days (Paoli and Merllié 2000). Burnout is often used to
characterize a reaction to prolonged work stress (Greenglass
and Burke 2003), especially feelings of emotional exhaus-
tion, which are believed to be at the core of burnout (Cordes
and Dougherty 1993). Initially, the concept of burnout was
developed to explain the longer-term process of chronic
stressors leading to occupational stress resulting in the
inability to cope with one’s work, both psychologically and
emotionally, among employees doing “people work” of
some kind (Maslach 1982; Maslach and Schaufeli 1993).
As the concept of burnout was originally restricted to the
human service sector, doing people work, i.e., “processing”
people, rather than things or information, was considered a
prerequisite for burnout (Demerouti et al. 2001). However,
burnout occurs in occupations outside the service sector as
well (Buunk et al. 1998; de Jonge and Schaufeli 1998;
Demerouti et al. 2001). As the primary tasks of these jobs
do not involve people work, the current opinion is that also
commonly found job characteristics that are not speciWcally
related to people work may aVect burnout, especially feel-
ings of emotional exhaustion (Demerouti et al. 2001).
Job characteristics that seem important in this respect are,
on the one hand, job demands, (e.g., workload and time
G. van Daalen (&)
OSA, Institute for Labour Studies, Tilburg University,
Warandelaan 2, P.O. Box 90153,
5000 LE Tilburg, The Netherlands
e-mail: G.vanDaalen@uvt.nl
T. M. Willemsen
Emerita, Department of Psychology, Tilburg University,
Warandelaan 2, P.O. Box 90153,
5000 LE Tilburg, The Netherlands
e-mail: t.m.willemsen@uvt.nl
K. Sanders
Work and Organisation Psychology, Twente University,
P.O. Box 217, Cubicus Building,
7500 AE Enschede, The Netherlands
e-mail: k.sanders@gw.utwente.nl
M. J. P. M. van Veldhoven
Human Resource Studies, Tilburg University, Warandelaan 2,
P.O. Box 90153, 5000 LE Tilburg, The Netherlands
e-mail: M.J.P.M.vanVeldhoven@uvt.nl
292 Int Arch Occup Environ Health (2009) 82:291–303
123
pressure) and on the other hand, job resources (e.g., auton-
omy and social support from colleagues). Jobs with high
demands and few resources are found to be detrimental to
employees’ physical and mental health, especially in combi-
nation with poor rewards (Bakker et al. 2003; Calnan et al.
2000; Heuven and Bakker 2003; van Vegchel et al. 2001).
For instance, De Jonge et al. (2000), found an elevated risk
of emotional exhaustion for employees with high job
demands and low job control of about eleven times compared
with employees with low demands and high control. Also
Demerouti et al. (2001) among three diVerent occupational
groups––human services, transport, and industry—found
that job demands aVected emotional exhaustion.
In addition to these commonly found job characteristics
that may aVect employees working in all types of jobs,
emotional demands have been found to be speciWc for those
working in human services (Borritz et al. 2005; Söderfeldt
et al. 1996), as it is more complicated to work with and for
people than with inanimate objects (de Jonge and Dormann
2003). Therefore, de Jonge and Dormann (2003) proposed
that in studies of stress it is important to take the amount as
well as the nature of interaction work performed in jobs
into account, especially in the human service sector.
In the present study, we focus on mental health care
organizations, and distinguish between employees with
high and low levels of interaction with patients in their job
(referred to as high patient interaction (HPI) jobs and low
patient interaction (LPI) jobs). Employees with high levels
of interaction with patients in their jobs are those employ-
ees whose primary task is to alter patients either physically
or psychologically. Employees with low levels of interac-
tion with patients in their jobs are those who, although not
primarily involved with patients, perform person-related
tasks that require (some) interaction with patients, like e.g.,
secretaries or receptionists.
Not all stress experienced on the job originates in the
workplace; it may also be related to stress at home (Green-
haus and Parasuraman 1999; Hammer et al. 2004). A home
domain-related stressor, that many employees experience
from time to time, is family-to-work conXict (Behson
2002). Family-to-work conXict occurs if demands from the
home and work domain are incompatible such that partici-
pation in the work domain becomes more diYcult due to
the demands of participation in the home domain (Frone
et al. 1997; Greenhaus and Beutell 1985; Greenhaus and
Parasuraman 1999; Gutek et al. 1991). Not only have fam-
ily and home demands been found predictive of daily dis-
tress (Almeida and Kessler 1998), consequences of family-
to-work conXict also have been found detrimental for
employees’ mental and physical health (Grzywacz and
Bass 2003; Torkelson and Muhonen 2003). Adverse
(mental) health outcomes associated with family-to-work
conXict are psychosomatic complaints and medication use
(Burke and Greenglass 1999), depression and poor physical
health (Frone et al. 1996, 1997), and hypertension (Frone
et al. 1997). Although family-to-work conXict originates in
the home domain, generally it aVects the work domain neg-
atively, as employees either worry about concerns at home
while at work, or simply are in lack of time. Therefore, it is
important to take family-to-work conXict into account in a
study concerning job stress and employee health.
The aim of the present study is threefold. First, we
examine how four job characteristics (i.e., two job demands
and two job resources) are related to emotional exhaustion
and mental health problems of employees working in HPI
jobs and employees in LPI jobs. Second, we examine
whether family-to-work conXict is related to emotional
exhaustion and mental health problems after controlling for
workload, emotional demands, autonomy and social sup-
port from colleagues. Finally, in addition to the direct
eVects of the job characteristics and family-to-work con-
Xict, we examine whether the job characteristics and fam-
ily-to-work conXict have a multiplicative interaction eVect
on emotional exhaustion and mental health problems.
Workload, social support and autonomy generally are
found predictive of occupational burnout and health symp-
toms among employees in various occupations (Muhonen
and Torkelson 2003). Therefore, we expect these job char-
acteristics to be related to emotional exhaustion and mental
health problems of employees in both HPI and LPI jobs. In
line with previous research, we expect workload to be
harmful for employee health and well-being. Social support
and autonomy, on the other hand, are expected to improve
employees’ health and well-being. Therefore, we hypothe-
size that workload is related positively to emotional
exhaustion and mental health problems (hypothesis 1a), and
that autonomy and social support from colleagues are
related negatively to both health outcomes (hypothesis 1b).
With regard to emotional demands, we expect that they
may aVect the health of employees in both job types, as
both employees in HPI and LPI jobs are confronted with
the behavior characteristics of others (e.g., aggressive or
irritating behaviors). As emotional demands, like workload,
lead to diminished health and well-being (de Jonge et al.
1999; Peeters and Le Blanc 2001), we hypothesize that
emotional demands are positively related to emotional
exhaustion and mental health problems for employees
working in HPI jobs as well as for those in LPI jobs
(hypothesis 1c).
Previous research found family-to-work conXict detri-
mental to one’s (mental) health (Grzywacz and Bass 2003;
Torkelson and Muhonen 2003). Moreover, employees
doing people work may be more vulnerable than other
employees to stress resulting from the demands of their
work and family roles, especially when both roles involve a
high degree of emotion management (Wharton and Erick-
Int Arch Occup Environ Health (2009) 82:291–303 293
123
son 1993). So, especially for types of jobs that are charac-
terized by high levels of emotion management, not only
work-related stressors but home-related stressors should be
included as well. Hence, we hypothesize that (strain-based)
family-to-work conXict is related positively to emotional
exhaustion and mental health problems (hypothesis 2).
In job stress research, it is generally assumed that job
resources may buVer the detrimental eVect of job demands
on health-related outcomes (Bakker et al. 2004; Demerouti
et al. 2001; Karasek and Theorell 1990; Siegrist 1996).
Recently, van Vegchel et al. (2005) demonstrated the plausi-
bility of this multiplicative interaction or buVer eVect
between job demands and job resources in relation to work-
related strain. In addition to previous stress research, which
focused on job-related demands, we examine the multiplica-
tive interaction between two job resources and a home-
related demand; family-to-work conXict. This interaction
eVect moderates the relationship between the job resources
(i.e., autonomy and social support) and family-to-work con-
Xict, such that high levels of resources prevent the occur-
rence of emotional exhaustion and mental health problems
despite high levels of family-to-work conXict. So, the rela-
tionship between family-to-work conXict and feelings of
emotional exhaustion and mental health problems will be
particular strong when autonomy and social support are low.
Accordingly, we expect that the job resources (autonomy and
social support from colleagues) decrease the assumed nega-
tive relation between family-to-work conXict and emotional
exhaustion and mental health problems (hypothesis 3a).
In job stress research it is not common to examine whether
job demands strengthen each other and hence intensify their
detrimental consequences for employee health and well-
being. According to Hockey’s (1997) cognitive-energetic
framework, people use active compensatory and control pro-
cesses to cope with disruptions in order to reach their primary
task. Hockey argues that, especially in cases of high demands,
it is hard for employees to cope. Because the work environ-
ment typically encourages a direct coping style that leaves
(too) little opportunity for recovery (Hockey 1997). Extend-
ing this line of reasoning to our study, we believe that
employees who are confronted with high demands from their
work on a daily basis, such as mental health care workers, are
at risk when confronted with demands from the home
domain. It is likely that these employees can not recovery
fully, because of the demands from their job. Accordingly,
there is no room for compensation between the two domains
suggesting that the demands from both domains strengthen
each other. Therefore we expect that the job demands (work-
load and emotional demands) increase the assumed negative
relation between family-to-work conXict and emotional
exhaustion and mental health problems (hypothesis 3b).
Prior to these issues, we test the assumption that employ-
ees working in HPI jobs report more emotional demands
than employees working in LPI jobs, and examine diVer-
ences in emotional exhaustion and mental health problems
between employees in the two job types.
Method
Sample and procedures
Data were obtained from employees of ten Dutch mental
health care organizations. Questionnaires were sent to all
employees (n = 2,463) within these organizations. A total
of 1,650 employees returned the questionnaire (response
rate 67%). For the present study we only selected respon-
dents who did people work. That is, respondents had to be
working in jobs that were primarily directed at modifying
patients either physically or psychologically, (referred to as
HPI jobs), or in jobs that, although not primarily concerned
with the interaction with patients, some interaction was
involved (referred to as LPI jobs). A total of 1,223 respon-
dents met these inclusion criteria; 916 respondents worked
in HPI jobs and 307 in LPI jobs. Listwise exclusion of
missing data from all variables (dependent, independent
and background) resulted in a Wnal sample of 1,008 respon-
dents. About equal percentages of respondents were lost
from both job types, i.e., HPI jobs 17% and LPI jobs 20%.
Thus, of the 1008 respondents, 762 respondents worked in
HPI jobs (76%), and 246 in LPI jobs (24%). Respondents
working in HPI jobs were psychiatrists, psychologists, psy-
chotherapists, nurses, social workers, physicians or other jobs
involving patients. Respondents working in LPI jobs were
secretaries, receptionists or other supporting staV employees.
Most respondents of the total Wnal sample were women
(71%). The age range was 21–64 years and the mean age
was 39.5 years. On average, respondents worked 29 h per
week with a range of 8–40 h per week (SD = 6.9). Average
job tenure was 12 years (SD = 8.8). Of the respondents
working in HPI jobs 10% had an academic degree or com-
pleted some form of higher (55%) or secondary vocational
education (27%). Most of the respondents working in LPI
jobs completed either a secondary (25%) or higher voca-
tional education (28%).
1
Measures
Dependent variables
Emotional exhaustion was measured by a subscale of the
UBOS (Schaufeli and van Dierendonck 1994), a Dutch ver-
sion of the Maslach Burnout Inventory (MBI) (Maslach and
1
For an overview of the diVerences in demographic characteristics of
both job types see Table 1.
294 Int Arch Occup Environ Health (2009) 82:291–303
123
Table 1 Means, SD and correlations among study variables
N=1,008 except for the correlations with education, these are based on 959 respondents
HPI high patient interaction, LPI low patient interaction, SS colleagues social support from colleagues, FWC family-to-work conXict
* P < 0.05; ** P < 0.01 (two-tailed)
a
SigniWcant at the P < 0.001 level
b
SigniWcant at the P < 0.05 level
Variables Mean and (SD) 1234567891011
Total HPI jobs LPI jobs
1Age
a
39.48 (9.41) 38.47 (9.40) 42.59 (8.71) 1.0
2 Education 5.37 (1.22) 5.62 (.95) 4.59 (1.60) ¡0.20** 1.0
3 Working hours
c
29.16 (6.86) 29.66 (6.30) 27.63 (8.20) ¡0.05 0.13** 1.0
4 Job tenure 12.35 (8.85) 12.61(8.69) 11.55 (9.31) 0.61** ¡0.07* 0.04 1.0
5 Workload 2.30 (.44) 2.30 (.43) 2.32 (.45) 0.08* 0.08* 0.09** 0.13* 1.0
6 Autonomy
a
2.68 (.48) 2.62 (.43) 2.85 (.57) 0.11* 0.03 0.03 0.09** ¡0.17** 1.0
7 SS colleagues 3.35 (.37) 3.34 (.36) 3.39 (.40) ¡0.04 ¡0.10** ¡0.07* ¡0.13** ¡0.23** 0.18** 1.0
8 Emotional demands
a
2.27 (.47) 2.42 (.40) 1.82 (.39) ¡0.07* 0.22** 0.17** 0.12 0.23** ¡0.22** ¡0.20** 1.0
9FWC
a
1.33 (.37) 1.35 (.38) 1.25 (.35) ¡0.08* 0.05 0.06 ¡0.03 0.04 ¡0.09** ¡0.13** 0.14** 1.0
10 Emotional exhaustion
a
2.63 (1.12) 2.72 (1.11) 2.38 (1.09) ¡0.03 0.11** 0.08* 0.03 0.41** ¡0.23** ¡0.26** 0.35** 0.21** 1.0
11 Mental health problems
b
1.13 (.17) 1.14 (.17) 1.11 (.16) 0.04 0.03 0.02 0.04 0.24** ¡0.14** ¡0.23** 0.26** 0.28** 0.61** 1.0
Int Arch Occup Environ Health (2009) 82:291–303 295
123
Jackson 1986). This scale consists of Wve items, each with a
7-point rating scale ranging from 1, “never”, to 7, “always”.
A sample item is: “I feel mentally exhausted due to my
work”. Reliability and construct validity of the Dutch ver-
sion are comparable with the original American version
(Schaufeli and van Dierendonck 1994; Schaufeli and van
Dierendonck 2001). Moreover, the UBOS has been used in
many other Dutch studies (Bakker et al. 2002; Demerouti
et al. 2004; Schaufeli and Bakker 2004). In the present
study the Cronbach alpha was 0.87.
Mental health problems were measured with a scale
based on the Four Dimensional Symptom Questionnaire
(4DSQ) (Terluin 1996). The original questionnaire mea-
sures four dimensions of common mental health problems:
distress, depression, anxiety and somatisation. For the pres-
ent study the questionnaire was adapted and shortened to a
single scale. All somatisation items were dropped, so the
Wnal scale covers distress, depression and anxiety and con-
sists of 16 dichotomous items (1 = no, 2 = yes). The scale
score was obtained by calculating a mean score across all
16 items, leading to one scale ranging from 1 to 2. Sample
items are: “Did you have problems getting asleep last
week?”, “Did you feel everything was pointless last
week?”, “Did you feel anxious last month?”. The Cronbach
alpha for this scale was 0.80. The predictive validity of the
4 DSQ has been found satisfactory in Dutch studies (Ter-
luin et al. 2004).
Independent variables
All four job characteristics were measured with a Dutch
questionnaire assessing psychosocial job demands (VBBA)
(van Veldhoven and Meijman 1994). The VBBA is a
widely used instrument in research on psychological job
factors and job stress in the Netherlands (see also van Vel-
dhoven et al. 2005). In the present study, we used four
scales of the VBBA, measuring workload (Cronbach’s
alpha 0.89), emotional demands (Cronbach’s alpha 0.76),
autonomy (Cronbach’s alpha 0.87), and social support from
colleagues (Cronbach’s alpha 0.82). Items were scored on a
4-point rating scale ranging from 1, “always” to 4, “never”.
The workload scale consists of 11 items. A sample item
of this scale is “Do you have to work very fast?”. The psy-
chometric properties of this scale have been found to be
good (de Croon et al. 2004; Sluiter et al. 2003; van Yperen
and Janssen 2002). The emotional demands scale consists
of seven items. A sample item of this scale is: “Do others
call on you personally in your work?”. Good validity and
reliability of this scale have been demonstrated in other
Dutch studies (Jansen et al. 2003; Schaufeli and Bakker
2004; Sluiter et al. 2003). The autonomy scale also consists
of 11 items. A sample item of this scale is: “Do you resolve
problems arising in your work yourself?”. The psychometric
properties of this scale have been found to be good (Claes-
sens et al. 2004; de Croon et al. 2004; Sluiter et al. 2003).
Finally, the scale measuring social support from colleagues
consists of nine items. A sample item of this scale is: “Is
there a good atmosphere between you and your col-
leagues?”. The psychometric properties of this scale have
been found to be good (Bakker et al. 2004; Geurts et al.
2003; Schaufeli and Bakker 2004). For all scales scores
were reversed so that higher scores reXect more workload,
more emotional demands, more autonomy and more social
support from colleagues.
Family-to-work conXict was measured by the Wve strain-
based items of the negative home-work interference scale
of the “Survey Work–Home Interference-Nijmegen”
(SWING) (Wagena and Geurts 2000). Items were scored on
a 4-point rating scale ranging from 1, “always”, to 4,
“never”. A sample items is: “How often does it happen that
you do not fully enjoy your work because you worry about
your home situation?”. Scores were reversed so that higher
scores reXect more family-to-work con
Xict. Cronbach’s
alpha was 0.83. Good validity and reliability of this scale
have been demonstrated in other Dutch studies regarding
work–family interference (Demerouti et al. 2004; Geurts
et al. 2003; Jansen et al. 2003). For all measures (dependent
and independent variables), a scale score was obtained by
calculating a mean score across all items of the particular
scale.
The background variables measured were gender
(1 = women, 2 = men), age (continuous), educational level
(measured with one item consisting of seven response cate-
gories ranging from 1, “primary education”, to 7, “univer-
sity”), contract hours (continuous), job tenure (continuous)
and job type, which was dichotomized into 1 = LPI jobs
and 2 = HPI jobs.
Data analyses
Zero order correlations were used to examine the general
pattern of relations among the variables. Prior to regression
analyses, the assumed diVerences in emotional demands as
well as diVerences in the other study variables between
employees working in HPI and LPI jobs were analyzed by a
MANOVA based on the general linear model (GLM).
Levene’s test for equality of variances was used. When the
assumption of equal variance was violated, t-tests were
computed and group means were compared using the statis-
tics that do not assume equal variances. For each outcome
variable, i.e., emotional exhaustion and mental health
problems, a separate regression analysis was performed for
employees working in HPI jobs and LPI jobs. The independent
variables were entered as a block into the regression
equation in the following order: (1) background variables;
(2) job characteristics; (3) emotional demands, (4) family-
296 Int Arch Occup Environ Health (2009) 82:291–303
123
to-work conXict, and (5) the interaction eVects between the
job characteristics and family-to-work conXict, as well as
between emotional demands and family-to-work conXict.
In order to eliminate non-essential correlation between the
interaction terms and their component variables, all predic-
tor variables were centered (Aiken and West 1991; Tabach-
nick and Fidell 2001). To assess the model Wt in each step
the change in R
2
was tested. All analyses were performed
with the statistical program SPSS 12.01 for Windows.
Results
Means, standard deviations and correlations of all study
variables are shown in Table 1.
The job characteristics, workload and emotional
demands correlated positively with emotional exhaustion
and mental health problems, autonomy and social support
from colleagues negatively. Family-to-work conXict was
positively related with both the outcome variables.
DiVerences in study variables between the two job types
Means of both job types are also displayed in Table 1.
Employees working in HPI jobs were higher educated (t
(285) = ¡9.29, P < 0.001), worked more hours per week (t
(343) = ¡3.55, P < 0.001) and were on the average
younger (F =39.65, P < 0.001) than employees in LPI jobs.
Furthermore, employees working in HPI jobs reported
more emotional demands (F = 443.32, P < 0.001), less
autonomy (t (342) = 5.82, P < 0.001), more family-to-work
interference (t (445) = ¡3.97, P < 0.001), and both more
emotional exhaustion (F =21.05, P < 0.001) and mental
health problems (F =4.41, P <0.05).
Results of the regression analyses
In order to answer the question how the four job character-
istics are related to emotional exhaustion Model 3 of the
regression analyses had to be inspected, as this Model esti-
mates the eVect of the four job characteristics on the depen-
dent variable taking into account the background variables.
Initially, regression analyses were performed including
educational level as a background variable. As educational
level was not signiWcantly related to both dependent vari-
ables, but did lead to 49 more missing cases, regression
analyses were ran for both dependent variables without
educational level. SigniWcant relations between the other
independent variables and the two dependent variables
were the same for both the regression with and without edu-
cational level. Accordingly, regression analyses without
educational level were used in the current study. Model 4
was inspected to answer whether family-to-work conXict is
related to emotional exhaustion, after controlling for the
background variables and the job characteristics. Model 5
displays the interaction eVects between the job characteris-
tics and family-to-work conXict.
Emotional exhaustion
Results for emotional exhaustion are shown in Table 2
(HPI jobs) and 3 (LPI jobs). For employees working in HPI
jobs it was found that women reported more feelings of
emotional exhaustion ( = ¡0.07, P < 0.05) than men. For
those working in LPI jobs it was found that working more
hours lead to more feelings of emotional exhaustion
( =0.14, P <0.05).
Model 3 of Tables 2 and 3 showed that of the job charac-
teristics, workload was related positively to emotional
exhaustion for employees working in both HPI and LPI
jobs ( =0.30, P < 0.001 and =0.31, P < 0.001, respec-
tively). Autonomy was related negatively to emotional
exhaustion in both job types ( = ¡0.11, P < 0.01 for HPI
jobs and = ¡0.12, P < 0.05 for LPI jobs), and social sup-
port from colleagues was related negatively to feelings of
emotional exhaustion only of employees working in HPI
jobs ( = ¡0.14, P < 0.001). Thus, employees in both HPI
and LPI jobs who report high levels of autonomy report less
feelings of emotional exhaustion, while only employees in
HPI jobs, who report high levels of social support from col-
leagues report less feelings of emotional exhaustion. Emo-
tional demands were related positively to emotional
exhaustion of employees working in both job types
( =0.22, P < 0.001 and = 0.18, P < 0.01). That is,
employees who report high levels of emotional demands
report more feelings of emotional exhaustion.
Model 4 of Tables 2 and 3 showed that family-to-work
conXict was related positively to emotional exhaustion for
employees in both job types ( =0.11, P <0.01 for HPI
and =0.26, P < 0.001 for LPI). That is, employees who
report high levels of family-to-work conXict report more
feelings of emotional exhaustion.
Table 3 showed that only for employees working in LPI
jobs a signiWcant interaction eVect was found (see Model
5). The “chunk” test (Kleinbaum 1992) revealed a signiW-
cant increase in explained variance between the model with
all interaction terms (Model 5) and the model with none of
the interaction terms (Model 4). Evaluation of the individ-
ual interaction terms showed that the interaction eVect of
social support from colleagues by family-to-work conXict
was signiWcant ( =0.18, P < 0.01) for employees working
in LPI jobs. This interaction eVect showed that employees
who receive little social support from colleagues and at the
same time experience high levels of family-to-work con-
Xict, report more feelings of emotional exhaustion than
employees who receive much social support from their
Int Arch Occup Environ Health (2009) 82:291–303 297
123
colleagues and experience high levels of family-to-work
conXict.
As shown in Tables 2 and 3, after controlling for the
background variables, the job characteristics explained a
large portion of the variability in both job types (R
2
=0.22,
P < 0.001, for HPI jobs, and R
2
=0.14, P < 0.001 for LPI
jobs). The proportion of additional variance explained by
emotional demands, after controlling for the background
variables and the job characteristics was small, but signiW-
cant for employees in both job types (R
2
= 0.04, P < 0.001
and R
2
=0.03, P < 0.01, respectively). Family-to-work
conXict explained only a small portion of emotional exhaus-
tion after controlling for the other study variables for
employees working in HPI jobs (R
2
= 0.01, P < 0.01). For
employees in LPI jobs the additional variance explained was
somewhat larger (R
2
= 0.06, P < 0.001). The interaction
eVect of social support from colleagues by family-to-work
conXict explained a signiWcant portion of the variability
Table 2 Job characteristics and FWC as predictors of emotional exhaustion of employees working in HPI jobs (n = 762)
All predictor variables were mean centered
Gender 1 female, Gender 2 male, SS colleagues social support from colleagues, FWC family-to-work conXict, NonsigniWcant interactions are not
displayed
*P< 0.05; ** P < 0.01; *** P <0.001
Step Variable Model 1 Model 2 Model 3 Model 4
B SE B B SE B B SE B B SE B
1 Age 0.00 0.01 ¡0.00 0.00 0.01 0.01 0.00 0.01 0.01 0.00 0.01 0.01
Gender ¡0.22 0.10 ¡0.09* ¡0.15 0.08 ¡0.06 ¡0.17 0.08 ¡0.07* ¡0.16 0.08 ¡0.06
Working hours 0.01 0.01 0.05 0.00 0.01 0.02 0.00 0.01 0.01 0.00 0.01 0.00
Job tenure ¡0.00 0.01 ¡0.01 ¡0.01 0.01 ¡0.05 ¡0.01 0.01 ¡0.05 ¡0.01 0.01 ¡0.05
2 Workload 0.95 0.09 0.37*** 0.79 0.09 0.30*** 0.78 0.09 0.30***
Autonomy ¡0.32 0.09 ¡0.12*** ¡0.27 0.08 ¡0.11** ¡0.26 0.08 ¡0.10**
SS colleagues ¡0.49 0.10 ¡0.16*** ¡0.43 0.10 ¡0.14***
¡0.39 0.10 ¡0.13***
3 Emotional demands 0.61 0.09 0.22*** 0.59 0.09 0.21***
4FWC 0.31 0.09 0.11**
5(No signiWcant interactions)
R
2
0.01 0.23 0.27 0.28
F for change in R
2
1.64 72.81*** 42.39*** 11.18**
Table 3 Job characteristics and FWC as predictors of emotional exhaustion of employees working in LPI jobs (n = 246)
All predictor variables were mean centered
Gender 1 female, Gender 2 male, SS colleagues social support from colleagues, FWC family-to-work conXict, NonsigniWcant interactions are not
displayed
* P < 0.05; ** P < 0.01; *** P <0.001
Step Variable Model 1 Model 2 Model 3 Model 4 Model 5
B SE B B SE B B SE B B SE B B SE B
1Age ¡0.00 0.01 ¡0.01 ¡0.01 0.01 ¡0.05 ¡0.01 0.01 ¡0.06 ¡0.00 0.01 ¡0.03 0.00 0.01 0.00
Gender ¡0.33 0.18 ¡0.14 ¡0.30 0.17 ¡0.12 ¡0.31 0.16 ¡0.13 ¡0.23 0.16 ¡0.10 ¡0.23 0.16 ¡0.09
Working hours 0.03 0.01 0.20** 0.02 0.01 0.16* 0.02 0.01 0.14* 0.02 0.01 0.12 0.02 0.01 0.12
Job tenure 0.02 0.01 0.19* 0.01 0.01 0.10 0.01 0.01 0.06 0.01 0.01 0.05 0.01 0.01 0.06
2 Workload 0.78 0.15 0.33*** 0.74 0.15 0.31*** 0.80 0.14 0.33*** 0.92 0.16 0.38***
Autonomy ¡0.26 0.12 ¡0.14* ¡0.24 0.12 ¡0.12* ¡0.21 0.11 ¡0.11 ¡0.17 0.13 ¡0.09
SS colleagues ¡0.23 0.18 ¡0.08 ¡0.14 0.18 ¡0.05 ¡0.08 0.17 ¡0.03 0.07 0.18 0.03
3 Emotional demands 0.51 0.18 0.18** 0.49 0.17 0.17** 0.45 0.17 0.16**
4FWC 0.80 0.17 0.26*** 0.75 0.27 0.24**
5 SS colleagues* FWC 1.38 0.44 0.18**
R
2
0.07 0.21 0.24 0.30 0.33
F for change in R
2
4.41** 14.72*** 8.41** 21.26*** 9.74**
298 Int Arch Occup Environ Health (2009) 82:291–303
123
associated with emotional exhaustion for those working in
LPI jobs (R
2
= 0.03, P < 0.05). For employees working in
HPI jobs the full regression models accounted for 28% of
the variability in emotional exhaustion and for employees in
LPI jobs for 30%.
Mental health problems
Results for mental health problems are shown in Table 4
(HPI jobs) and Table 5 (LPI jobs). For employees working
in HPI jobs, it was found that women reported more mental
health problems ( = ¡0.12, P < 0.01) than men. For those
working in LPI jobs none of the background variables were
signiWcant.
Model 3 of Tables 4 and 5 showed that of the job charac-
teristics, workload was related positively to mental health
problems for employees in both job types ( =0.12,
P < 0.01 for HPI jobs and =0.15, P < 0.01 for LPI jobs).
Autonomy was not signiWcantly related to mental health
problems of employees in HPI or LPI jobs. Social support
from colleagues and emotional demands were only signiW-
cant for those working in HPI jobs ( = ¡0.18, P < 0.001
and =0.22, P < 0.001, respectively). So, employees
working in HPI jobs who report low levels of social support
from colleagues or experience high levels of emotional
demands report more mental health problems than employ-
ees in HPI jobs who report high levels of social support or
low levels of emotional demands.
Model 4 of Tables 4 and 5 showed that family-to-work
conXict was related to mental health problems of employ-
ees working in both job types ( = 0.22, P < 0.001 for HPI
jobs and =0.29, P < 0.001 for LPI jobs), indicating that
employees who experience family-to-work conXict report
more mental health problems than those who do not experi-
ence family-to-work conXict.
None of the interactions between family-to-work conXict
and the job characteristics were signiWcant, indicating that
the direct eVects of family-to-work conXict and the job
characteristics do not have a multiplicative eVect on mental
health problems for employees working in HPI or LPI jobs
(see Model 5 of Tables 4 and 5).
As shown in Tables 4 and
5, for employees in HPI jobs a
large part of the variability in mental health problems was
explained by the job characteristics (R
2
= 0.10,
P < 0.001), family-to-work conXict explained a large part
of the variability in mental health problems of those in LPI
jobs (R
2
=0.08, P <0.001).
The job characteristics and family-to-work conXict
explained a signiWcant part in the variability in mental
health problems for both employees in HPI and LPI jobs
((R
2
=0.10, P < 0.001 for HPI jobs, and R
2
= 0.04,
P < 0.05 for LPI jobs) and (R
2
=0.05, P < 0.001 for HPI
jobs, and R
2
= 0.08, P < 0.001 for LPI jobs) respectively).
The proportion of additional variance explained by emo-
tional demands, after controlling for the background vari-
ables and the job characteristics was only signiWcant for
employees working in HPI jobs (R
2
= 0.04, P <0.001).
For both job types the interaction eVects did not explained a
signiWcant portion of the variability associated with mental
health problems. For employees in HPI jobs the full regres-
sion models accounted for 21% of the variability in mental
health problems and for employees in LPI jobs for 17%.
Table 4 Job characteristics and FWC as predictors of mental health problems of employees working in HPI jobs (n = 762)
All predictor variables were mean centered
Gender 1 female, Gender 2 male, SS colleagues social support from colleagues, FWC family-to-work conXict, NonsigniWcant interactions are not
displayed
*P< 0.05; ** P < 0.01; *** P <0.001
Step Variable Model 1 Model 2 Model 3 Model 4
B SE B B SE B B SE B B SE B
1 Age 0.00 0.00 0.01 0.00 0.00 0.08 0.01 0.00 0.09 0.00 0.00 0.09*
Gender ¡0.05 0.01 ¡0.13** ¡0.04 0.01 ¡0.11** ¡0.04 0.01 ¡0.12** ¡0.04 0.01 ¡0.11**
Working hours 0.00 0.00 0.03 0.00 0.00 0.02 0.00 0.01 0.01 0.00 0.00 ¡0.01
Job tenure 0.00 0.00 ¡0.02 ¡0.00 0.00 ¡0.05 ¡0.00 0.00 ¡0.06 ¡0.00 0.00 ¡0.05
2 Workload 0.07 0.01 0.18*** 0.05 0.01 0.12** 0.05 0.01 0.12**
Autonomy ¡0.03 0.01 ¡0.09* ¡0.03 0.01 ¡0.07 ¡0.02 0.01 ¡0.06
SS colleagues ¡0.09 0.02 ¡0.19*** ¡0.08 0.02 ¡0.18*** ¡0.07 0.02 ¡
0.15***
3 Emotional demands 0.09 0.02 0.22*** 0.09 0.02 0.20***
4FWC 0.10 0.02 0.22***
5(No signiWcant interactions)
R
2
0.02 0.12 0.16 0.21
F for change in R
2
2.91* 29.74*** 36.11*** 43.03***
Int Arch Occup Environ Health (2009) 82:291–303 299
123
Considering these results for emotional exhaustion and
mental health problems, it appears that hypothesis 1a,
assuming that workload is related positively to emotional
exhaustion and mental health problems, was conWrmed for
both job types. Hypothesis 1b, that autonomy and social
support from colleagues are related negatively to emotional
exhaustion and mental health problems, was partially con-
Wrmed. That is, with regard to emotional exhaustion this
hypothesis was only supported among those working in
HPI jobs, and partially among those working in LPI jobs.
That is, both social support from colleagues and autonomy
were related negatively with emotional exhaustion among
those working in HPI jobs, whereas autonomy was related
negatively with emotional exhaustion among employees in
LPI jobs. With regard to mental health problems, only
among employees working in HPI jobs this hypothesis was
partially conWrmed. That is, social support from colleagues
was negatively related with mental health problems of HPI
employees. No signiWcant relations were found for employ-
ees in LPI jobs. Hypothesis 1c, assuming that emotional
demands were related positively to emotional exhaustion
and mental health problems, was conWrmed for emotional
exhaustion. For mental health problems, emotional
demands were only related to mental health problems of
employees in HPI jobs. Hence, this hypothesis was partially
conWrmed.
Hypothesis 2, in which we assume that family-to-work
conXict is positively related to both health outcomes, was
conWrmed in both job types.
Hypothesis 3a assumed a multiplicative interaction eVect
of job resources and family-to-work conXict on emotional
exhaustion and mental health problems, i.e., job resources
decrease the negative eVect of family-to-work conXict on
emotional exhaustion and mental health problems. We only
found support for a multiplicative eVect of social support
from colleagues and family-to-work conXict on emotional
exhaustion, and only among employees working in LPI
jobs. Hypothesis 3b, in which the job demands are assumed
to increase the assumed negative eVect of family-to-work
conXict on emotional exhaustion and mental health prob-
lems, was not at all conWrmed.
Discussion
The present study contributes to the existing occupational
health literature in two ways. First, to gain more insight into
the impact of service jobs on service employees’ health and
well-being, following de Jonge and Dormann(2003), who
stress the importance of the distinction between service
jobs based on the amount of interaction work performed by
service workers when examining stress in the service sec-
tor, our study distinguishes between diVerent levels of
interaction with patients among mental health care employ-
ees. More speciWcally, our study distinguishes between
employees working in high patient interaction (HPI) jobs,
i.e., jobs aimed at altering or modifying patients, and
employees working in low patient interaction (LPI) jobs,
i.e., jobs that involve performing person-related tasks that
require some interaction with patients. Second, in addition
to common job stressors, a home-related stressor that inter-
feres with the work domain, i.e., strain-based family-to-
Table 5 Job characteristics and FWC as predictors of mental health problems of employees working in LPI jobs (n = 246)
All predictor variables were mean centered
Gender 1 female, Gender 2 male, SS colleagues social support from colleagues, FWC family-to-work conXict, NonsigniWcant interactions are not
displayed
*P< 0.05; ** P < 0.01; *** P <0.001
Step Variable Model 1 Model 2 Model 3 Model 4
B SE B B SE B B SE B B SE B
1 Age 0.00 0.00 0.07 0.00 0.00 0.06 0.01 0.01 0.06 0.00 0.00 0.09
Gender ¡0.05 0.03 ¡0.13 ¡0.05 0.03 ¡0.13 ¡0.05 0.03 ¡0.13 ¡0.03 0.03 ¡0.10
Working hours 0.00 0.00 0.10 0.00 0.00 0.06 0.00 0.00 0.05 0.00 0.00 0.03
Job tenure 0.00 0.00 0.13 0.00 0.00 0.06 0.00 0.00 0.04 0.00 0.00 0.03
2 Workload 0.05 0.02 0.16* 0.05 0.02 0.15* 0.06 0.02 0.17**
Autonomy ¡0.00 0.02 ¡0.01 ¡0.00 0.02 ¡0.01 0.00 0.02 0.01
SS colleagues ¡0.05 0.03 ¡0.12 ¡0.04 0.03 ¡0.10 ¡0.03 0.03 ¡0.08
3 Emotional demands 0.05 0.03 0.11 0.04 0.03 0.11
4FWC 0.13 0.03 0.29***
5 (No signiWcant interactions)
R
2
0.04 0.08 0.09 0.17
F for change in R
2
2.38 3.77* 2.74 44.58***
300 Int Arch Occup Environ Health (2009) 82:291–303
123
work conXict, was incorporated in this study. This, we
believe is a contribution as most occupational health
research focuses on work–family conXict in general or on
work-to-family conXict.
The aim of the present study was to examine the impact
of three common job stressors (i.e., workload, autonomy
and social support from colleagues), a job stressor speciWc
to the health care sector (i.e., emotional demands) and a
home-related stressor (i.e., family-to-work conXict) on
mental health outcomes among employees working in HPI
jobs and LPI jobs. Furthermore, we examined whether
these job characteristics and family-to-work conXict have a
multiplicative interaction eVect on employees’ feelings of
emotional exhaustion and mental health problems.
Generally accepted assumptions concerning diVerences
between employees in HPI jobs and LPI jobs were tested.
Results were in favor of the current assumption that
employees working in HPI jobs report more emotional
demands than employees in LPI jobs. Furthermore it
appeared that employees working in HPI jobs also report
more feelings of emotional exhaustion and more mental
health problems. With regard to the level of complaints,
however, it should be noted that the level of emotional
exhaustion among the respondents of our study were mod-
erate. Compared with the norm scores (Schaufeli and van
Dierendonck 1994) of healthy employees, respondents in
our sample report higher levels of emotional exhaustion,
whereas compared with employees who report burnout
complaints respondents in our sample report lower levels of
emotional exhaustion.
2
Scores of mental health problems
could not be compared with their norm scores, as we used
an adapted shortened scale to measure mental health prob-
lems.
In accordance with previous research, both emotional
demands and workload were found to be associated with
emotional exhaustion and mental health problems. How-
ever, it appeared that emotional demands only contributed
to the explanation of mental health problems of employees
in HPI jobs, whereas mental health of employees working
in LPI jobs was not aVected by emotional demands. Results
for job autonomy and social support from colleagues were
not as expected; autonomy only decreased feelings of emo-
tional exhaustion and was not related with mental health
problems. Social support from colleagues only decreased
feelings of emotional exhaustion and mental health prob-
lems of those working in HPI jobs.
Although the level of complaints in our study were mod-
est, employees working in HPI jobs seem to be more at risk
when in comes to their health and well-being than employ-
ees in LPI jobs. That is, employees working in HPI jobs,
who encounter already a higher level of emotional demands
from their jobs than LPI employees, and for whom these
emotional demands seem to have a higher impact on their
well-being, suVer from a lack of autonomy and social sup-
port. A lack of autonomy and support adds to their stress
level in a work situation that is generally believed to be
more stressful than in LPI jobs. Employees working in LPI
jobs are confronted with less emotional demands, have
more autonomy, and receive more social support from col-
leagues. Moreover, they report less family-to-work conXict
and work less hours per week.
With respect to family-to-work conXict, support was
found for the idea that, in addition to job stressors, this
home-related stressor relates to emotional exhaustion and
mental health problems. Where previous studies showed
that work-to-family conXict has a signiWcant eVect on
employee’s health outcomes (Frone et al. 1997; Proost
et al. 2004), we extended these Wndings by showing that
family-to-work conXict is associated with emotional
exhaustion and mental health problems. Furthermore, this
study showed that the adverse eVect of family-to-work con-
Xict on feelings of emotional exhaustion is mitigated by
social support from colleagues for those working in LPI
jobs. Given these results, we consider it useful to integrate
family-to-work conXict into job stress research and suggest
to further investigate the eVect of family-to-work conXict
on health-related outcomes, both directly and combined
with other stressors.
Generally, the job stressors explained the largest part of
variance in both emotional exhaustion and mental health
problems of employees working in HPI jobs, whereas for
employees in LPI jobs the largest part of variance was
explained by family-to-work conXict, especially of mental
health problems. These Wndings, in line with Borritz et al.
(2005), de Jonge and Dormann (2003), and Sparks and
Cooper (1999), show that it is important to take account of
diVerent stressors in diVerent job types.
Several limitations of the present study and issues for
future research can be mentioned. First, as our data were
cross-sectional it is not possible to draw any causal conclu-
sions based on Wndings of the present study. The direction
of the relationships between stressors and outcome vari-
ables can only be determined theoretically. Thus, although
it is likely that job demands, job resources and family-to-
work conXict inXuence feelings of emotional exhaustion
and mental health problems, it might be the other way
around as well. That is, feelings of emotional exhaustion
and mental health complaints may inXuence the perception
of stress-related factors at work and at home. Second, we
only used self-report data, which may lead to data contami-
nation due to common method variance. Although several
studies have shown common method variance not to be as
problematic as once thought (Semmer et al. 1996; Spector
2
Detailed information about the norm scores can be obtained from the
corresponding author.
Int Arch Occup Environ Health (2009) 82:291–303 301
123
1992), common method variance would enlarge main
eVects at the cost of Wnding interaction eVects (cf. de Jonge
et al. 1999). In our case this may have resulted in not Wnd-
ing signiWcant interaction eVects apart from the social sup-
port from colleagues by family-to-work conXict interaction.
Third, we did not incorporate intrinsic personal characteris-
tics, such as for example overcommitment, in our study.
This may be a limitation, as some studies found evidence
for the impact of overcommitment on the relation between
stress and employee health (de Jonge et al. 2000; Siegrist
and Peter 1994). On the other hand, other studies did not
support these Wndings (de Jonge and Hamers 2000). Fourth,
employees in LPI jobs were lower educated and worked on
lower job levels than those in HPI jobs, which may have led
to confounded eVects of job type and job level. Finally, our
data are from a Dutch population which may limits its gen-
eralizability to other countries. An illustration of this is for
instance that working part-time is very common among
women in the Netherlands (SCP 2006), making it diYcult
to generalize our results to studies performed in countries
where working part-time is not the norm among women.
In conclusion, the present Wndings provide evidence that
diVerent job characteristics are related to emotional exhaus-
tion and mental health problems of employees working in
HPI and LPI jobs. Furthermore, it appears that the job
resources; autonomy and social support act in a diVerent
way for those working in HPI and LPI jobs. That is,
although social support did not reduce emotional exhaus-
tion and mental health problems of employees in LPI jobs
directly, it buVered the impact of family-to-work conXict on
emotional exhaustion for those in LPI jobs. Whereas Wnd-
ings for employees in HPI jobs were the other way around.
In accordance with Frone et al. (1997) and Reid Keene and
Reynolds (2005), we found that family-to-work conXict has
negative consequences for one’s emotional exhaustion and
mental health. Hence stress research as well as organiza-
tional programs designed to promote employees’ health
should take this home-related stressor into account.
The present Wndings have practical implications as well.
Although, levels of complaints and stress were moderate,
Wndings of the present study can be helpful in programs of
stress management or stress prevention within the health care
services. Especially as it provides insight into service jobs
with diVerent levels of patient interaction. With regard to
HPI jobs, where emotional demands have a high impact, and
often cannot easily be diminished due to the nature of these
jobs, it will be useful to pay careful attention to those factors
that can be improved in order to diminish the stress level of
the work situation. Giving those employees autonomy if pos-
sible, encouraging mutual social support among colleagues,
protecting them from too heavy workloads, and providing
work-family support are all measures that are within reach of
management and can probably prevent or at least diminish
health problems due to the stressful nature of HPI jobs. Nev-
ertheless, one should be careful not to overlook employees
with low levels of patient interaction in their jobs. Although
most research emphasizes the adverse health eVects of stress-
ors for those working in HPI jobs, our study demonstrates
that employees with low levels of patient interaction also
may su
Ver adverse eVects from various stressors as well. For
them also, paying attention to workload and, especially, to
diminishing family-to-work conXict by providing relevant
work-family programs may enhance their well-being and
prevent burnout. Moreover, as diVerent stressors are associ-
ated with emotional exhaustion and mental health problems
of employees working in the two job types, employees with
high and low levels of patient interaction should not be
treated in the same way in programs of stress prevention.
Acknowledgments Data collection for the present study was sup-
ported by a research grant from GGZ Nederland, an organization rep-
resenting all Mental Health Care Institutions in the Netherlands.
References
Aiken L, West S (1991) Multiple regression: testing and interpreting
interactions. Sage, Newbury Park
Almeida DM, Kessler RC (1998) Everyday stressors and gender diVer-
ences in daily distress. J Pers Soc Psychol 75:670–680.
doi:10.1037/0022-3514.75.3.670
Bakker AB, Demerouti E, Schaufeli WB (2002) Validation of the Mas-
lach Burnout Inventory––general survey: an internet study. Anxiety
Stress Coping 15:245–260. doi:10.1080/1061580021000020716
Bakker AB, Demerouti E, Schaufeli WB (2003) Dual processes at
work in a call centre: an application of the job demands-resources
model. Eur J Work Organ Psychol 12:393–417. doi:10.1080/
13594320344000165
Bakker AB, Demerouti E, Verbeke W (2004) Using the job demands-
resources model to predict burnout and performance. Hum Re-
sour Manage 43:83–104. doi:10.1002/hrm.20004
Behson SJ (2002) Coping with family-to-work conXict: the role of
informal work accommodations to family. J Occup Health Psy-
chol 7:324–341. doi:10.1037/1076-8998.7.4.324 Medline
Borritz M, Bültman U, Rugulies R, Christensen KB, Villadsen E, Kris-
tensen TS (2005) Psychosocial work characteristics as predictors
for burnout: Wndings from 3-year follow up of the PUMA study.
J Occup Environ Med 47:1015–1025. doi:10.1097/01.jom.
0000175155.50789.98
Burke RJ, Greenglass ER (1999) Work-family conXict, spouse sup-
port, and nursing staV well-being during organizational restruc-
turing. J Occup Health Psychol 4:327–336. doi:10.1037/1076-
8998.4.4.327
Buunk BP, de Jonge J, Ybema JF, de WolV CJ (1998) Psychosocial as-
pects of occupational stress. In: Drenth PJD, Thierry H, De WolV
CJ (eds) Handbook of work and organizational psychology. Psy-
chology Press, Hove, pp 145–182
Calnan M, Wainright D, Almond S (2000) Job strain, eVort-reward
imbalance and mental distress: a study of occupations in general
medical practice. Work Stress 14:297–311. doi:10.1080/
02678370110040920
Claessens BJC, van Eerde W, Rutte CG, Roe RA (2004) Planning
behavior and perceived control of time at work. J Organ Behav
25:937–950. doi:10.1002/job.292
302 Int Arch Occup Environ Health (2009) 82:291–303
123
Cordes CL, Dougherty TW (1993) A review and an integration of re-
search on job burnout. Acad Manage Rev 18:621–656.
doi:10.2307/258593
de Jonge J, Hamers JPH (2000) Inspanningen en beloningen in het
werk van verpleegkundigen en verzorgenden: Een kwestie van
balans of disbalans? (EVorts and rewards in the work of health
care workers: a matter of balance or imbalance?). Verpleegkunde
15:64–73
de Jonge J, Schaufeli WB (1998) Job characteristics and employee
well-being: a test of Warr’s vitamin model in health care workers
using structural equation modeling. J Organ Behav 19:387–407.
doi:10.1002/(SICI)1099-1379(199807)19:4<387::AID-JOB851>
3.0.CO;2-9
de Jonge J, Mulder MJGP, Nijhuis FJN (1999) The incorporation of
diVerent demand concepts in the job demand-control model:
eVects on health care professionals. Soc Sci Med 48:1149–1160.
doi:10.1016/S0277-9536(98) 00429-8
de Jonge J, Bosma H, Peter R, Siegrist J (2000) Job strain, eVort-
reward imbalance and employee well-being: a large-scale
cross-sectional study. Soc Sci Med 50:1317–1327. doi:10.1016/
S0277-9536(99)00388-3
de Jonge J, Dormann C (2003) The DISC model: demand-induced
strain compensation mechanisms in job stress. In: Dollard MF,
WineWled AH, WineWeld HR (eds) Occupational stress in the ser-
vice professions. Taylor and Francis, London, pp 43–74
de Croon EM, Sluiter JK, Blonk RWB, Broersen JPJ, Frings-Dresen
MHW (2004) Stressful work, psychological job strain, and turn-
over: a 2-year prospective cohort study of truck drivers. J Appl
Psychol 89:442–454. doi:10.1037/0021-9010.89.3.442
Demerouti E, Bakker AB, Nachreiner F, Schaufeli WB (2001) The job
demands-resources model of burnout. J Appl Psychol 80:499–
512. doi:10.1037/0021-9010.86.3.499
Demerouti E, Bakker AB, Bulters AJ (2004) The loss spiral of work
pressure, work–home interference and exhaustion: reciprocal
relations in a three-wave study. J Vocat Behav 64:131–149.
doi:10.1016/S0001-8791(03)00030-7
Frone MR, Russell M, Barnes GM (1996) Work–family conXict, gen-
der, and health-related outcomes: a study of employed parents in
two community samples. J Occup Health Psychol 1:57–69.
doi:10.1037/1076-8998.1.1.57
Frone MR, Russell M, Cooper ML (1997a) Relation of work–family
conXict to health outcomes: a four-year longitudinal study of em-
ployed parents. J Occup Organ Psychol 70:325–335
Frone MR, Yardly JK, Markel KS (1997b) Developing and testing an
integrative model of the work–family interface. J Vocat Behav
50:145–167. doi:10.1006/jvbe.1996.1577
Geurts SAE, Kompier MAJ, Roxburgh S, Houtman ILD (2003) Does
work–home interference mediate the relationship between work-
load and well-being? J Vocat Behav 63:532–559. doi:10.1016/
S0001-8791(02) 00025-8
Greenglass ER, Burke RJ (2003) Teacher stress. In: Dollard MF,
WineWeld AH, WineWeld HR (eds) Occupational stress in the ser-
vice professions. Taylor and Francis, London, pp 213–236
Greenhaus JH, Beutell NJ (1985) Sources of conXict between work and
family roles. Acad Manage Rev 10:76–88. doi:10.2307/258214
Greenhaus JH, Parasuraman S (1999) Research on work, family and
gender. In: Powell GN (ed) Handbook of gender and work. Sage
publications, Thousand Oaks
Grzywacz JG, Bass BL (2003) Work, family, and mental health: test-
ing diVerent models of work–family Wt. J Marriage Fam 65:248–
262. doi:10.1111/j.1741-3737.2003.00248.x
Gutek BA, Searle S, Klepa L (1991) Rational versus gender role expla-
nations for work–family conXict. J Appl Psychol 76:560–568.
doi:10.1037/0021-9010.76.4.560
Hammer TH, Saksvik PO, Nytro K, Torvatn H, Bayazit M (2004)
Expanding the psychosocial work environment: workplace norms
and work–family conXict as correlates of stress and health. J Oc-
cup Health Psychol 9:38–97. doi:10.1037/1076-8998.9.1.83
Heuven E, Bakker AB (2003) Emotional dissonance and burnout
among cabin attendants. Eur J Work Organ Psychol 12:81–100.
doi:10.1080/13594320344000039
Hockey GRJ (1997) Compensatory control in the regulation of human
performance under stress and high workload: a cognitive-energet-
ical framework. Biol Psychol 45:73–93. doi:10.1016/S0301-
0511(96) 05223-4
Jansen NWH, Kant I, Kristensen TS, Nijhuis FJN (2003) Antecedents
and consequences of work–family conXict: a prospective cohort
study. J Occup Environ Med 45:479–491. doi:10.1097/
01.jom.0000063626.37065.e8
Karasek RA, Theorell T (1990) Healthy work: stress, productivity, and
the reconstruction of working life. Basic Books, New York
Kleinbaum DG (1992) Logistic regression: a self learning text. Spring-
er, New York
Maslach C (1982) Burnout: the cost of caring. Englewood CliVs, Pren-
tice Hall
Maslach C, Jackson SE (1986) MBI: Maslach Burnout Inventory:
manual research edition. Consulting Psychologists Press, Palo
Alto
Maslach C, Schaufeli WB (1993) Historical and conceptual develop-
ment of burnout. In: Schaufeli WB, Maslach C, Marek T (eds)
Professional burnout: recent developments in theory and research.
Taylor & Francis, Washington, pp 1–16
Muhonen T, Torkelson E (2003) The demand-control-support model
and health among women and men in similar occupations. J Be-
hav Med 26:601–613. doi:10.1023/A:1026257903871
Paoli P, Merllié D (2000) Third European survey on working
conditions 2000. OYce for OYcial Publications of the European
Communities, Luxembourg, pp 1–72
Peeters MCW, Le Blanc PM (2001) Towards a match between job de-
mands and sources of social support: a study among oncology
care providers. Eur J Work Organ Psychol 10:53–72.
doi:10.1080/13594320042000034
Proost K, de Witte H, de Witte K, Evers G (2004) Burnout among nurs-
es: extending the job demand-control-support model with work–
home interference. Psychol Belg 44:269–288
Reid Keene J, Reynolds JR (2005) The costs of family demands: gen-
der diVerences in negative family-to-work spillover. J Fam Issues
26:275–299. doi:10.1177/0192513X04270219
Schaufeli WB, Bakker AB (2004) Job demands, job resources, and
their relationship with burnout and engagement: a multi-sample
study. J Organ Behav 25:293–315. doi:10.1002/job.248
Schaufeli WB, van Dierendonck D (1994) Burnout, een begrip geme-
ten: De Nederlandse versie van de Maslach Burnout Inventory
(MBI-NL) (Burnout––the measurement of a concept: the Dutch
version of the Maslach Burnout Inventory (MBI-NL)]. Gedrag
Organ 22:153–172
Schaufeli WB, van Dierendonck D (2001) Utrechtse Burnout
Schaal (UBOS) Psychodiagnostisch gereedschap. De psycho-
loog: maandblad van het Nederlands Instituut van Psycholo-
gen 36:9–12
SCP 2006 Emancipatiemonitor 2006 (Emancipation Monitor 2006).
Den Haag: Sociaal en Cultureel Planbureau/Centraal Bureau voor
de Statistiek
Semmer N, Zapf D, Greif S (1996) Shared job strain: a new approach
for assessing the validity of job stress measures. J Occup Organ
Psychol 69:293–310
Siegrist J (1996) Adverse health eVects of high-eVort/low-reward con-
ditions. J Occup Health Psychol 1:27–41. doi:10.1037/1076-
8998.1.1.27
Siegrist J, Peter R (1994) Job stressors and coping characteristics in
work-related disease: issues of validity. Work Stress 8:130–140.
doi:10.1080/02678379408259985
Int Arch Occup Environ Health (2009) 82:291–303 303
123
Sluiter JK, de Croon EM, Meijman TF, Frings-Dresen MHW (2003)
Need for recovery from work related fatigue and its role in the
development and prediction of subjective health complaints. Oc-
cup Environ Med 60:162–170. doi:10.1136/oem.60.5.373
Söderfeldt B, Söderfeldt M, Muntaner C, O’Campo P, Ohlson C-G,
Warg L-E (1996) Psychosocial work environment in human ser-
vice organizations: a concept analysis and development of the de-
mand-control model. Soc Sci Med 42:1217–1226. doi:10.1016/
0277-9536(95) 00231-6
Sparks K, Cooper CL (1999) Occupational diVerences in the work–
strain relationship: towards the use of situation-speciWc models.
J Occup Organ Psychol 72:219–229
Spector PE (1992) A consideration of the validity and meaning of self-
report measures of job conditions. Wiley, Chichester
Tabachnick BG, Fidell LS (2001) Using multivariate statistics. Allyn
and Bacon, Boston
Terluin B (1996) De Vierdimensionele Klachtenlijst (4DKL). Een vra-
genlijst voor het meten van distress, depressie, angst en somatisa-
tie (The four-dimensional symptom questionnaire (4DSQ). A
questionnaire to measure distress, depression, anxiety, and soma-
tization). Huisarts en Wetenschap 39:538–547
Terluin B, van Rhenen W, Schaufeli WB, de Haan M (2004) The four-
dimensional symptom questionniare (4DSO): measuring distress
and other mental health problems in a working population. Work
Stress 18:178–207. doi:10.1080/0267837042000297535
Torkelson E, Muhonen T (2003) Stress and health among women and
men in a Swedish telecom company. Eur J Work Organ Psychol
12:171–186. doi:10.1080/13594320344000066
van Vegchel N, de Jonge J, Meijer T, Hamers JP (2001) DiVerent eVort
constructs and eVort-reward imbalance: eVects on employee well-
being in ancillary health care workers. J Adv Nurs 34:128–136.
doi:10.1046/j.1365-2648.2001.3411726.x
van Vegchel N, de Jonge J, Landsbergis PA (2005) Occupational stress
in (inter) action: the interplay between job demands and job re-
sources. J Organ Behav 26:535–560. doi:10.1002/job.327
van Veldhoven M, Meijman TF (1994) Het meten van psychosociale
arbeidsbelasting met een vragenlijst: De vragenlijst beleving en
beoordeling van de arbeid (VBBA). (The measurement of psy-
chosocial job demands with a questionnaire: the questionnaire on
the experience and evaluation of work (QEEW)]. NIA, Amster-
dam
van Veldhoven M, Taris TW, de Jonge J, Broersen S (2005) The rela-
tionship between work characteristics and employee health and
well-being: how much complexity doe we really need? Int J Stress
Manag 12:3–28. doi:10.1037/1072-5245.12.1.3
van Yperen N, Janssen O (2002) Fatigues and dissatisWed or fatigued
but satisWed? Goal orientations and responses to high job de-
mands. Acad Manage J 45:1161–1171. doi:10.2307/3069431
Wagena E, Geurts S (2000) Ontwikkeling en validering van de ‘Survey
Werk-thuis Interferentie Nijmegen’ (Development and validation
of the ‘Survey Work–home Interference-Nijmegen). Gedrag Gez-
ond 28:138–158
Wharton AS, Erickson RJ (1993) Managing emotions on the job and at
home: understanding the consequences of multiple emotional
roles. Acad Manage Rev 18:457–486. doi:10.2307/258905
