Content uploaded by Ellen Melbye Langballe
Author content
All content in this area was uploaded by Ellen Melbye Langballe
Content may be subject to copyright.
Work 32 (2009) 179–188 179
DOI 10.3233/WOR-2009-0804
IOS Press
The relationship between burnout and
musculoskeletal pain in seven Norwegian
occupational groups
Ellen Melbye Langballea,∗, Siw Tone Innstrandd, Knut Arne Hagtvetc, Erik Falkuma,band
Olaf Gjerløw Aaslande,f
aDiakonhjemmet Hospital, Vinderen DPS, Oslo, Norway
bInstitute of Psychiatry, University of Oslo, Norway
cInstitute of Psychology, University of Oslo,Norway
dDepartment of Social Work and Health Sciences, Norwegian University of Science and Technology, Trondheim,
Norway
eThe Research Institute, Norwegian Medical Association, Oslo, Norway
fInstitute of Health Management and Health Economics, University of Oslo, Norway
Received 30 September 2007
Accepted 24 November 2007
Abstract. Occupational and gender differences were investigated in the relationship between burnout and musculoskeletal pain in
the head, neck, shoulders, and back. Representative samplesof lawyers, physicians, nurses, teachers, church ministers, bus drivers,
and information technology workers in Norway (N =4507) were analyzed using structural equation modeling. The exhaustion
dimension of burnout was positively associated with musculoskeletal pain inall groups, and the strength of the relationship ranged
from moderate to strong. The disengagement dimension of burnout was negatively associated with musculoskeletal pain in five
groups and only ranged from −0.15 to −0.42. Professional efficacy was slightly weaker, and inconsistently (i.e.,both positively
and negatively) associated with musculoskeletal pain in four of the groups. There were larger differences in the strength of the
relationships between the seven occupational groups than between males and females within thesame profession. Results suggest
that burnout and musculoskeletal pain are related, but the strengthof the associations varies according to gender and occupation.
Overall, occupational factors appear to be stronger predictors of the co-occurrence of burnout and musculoskeletal pain than
gender.
Keywords: Burnout, musculoskeletal pain, occupational groups, gender, structural equation modeling
1. Introduction
Burnout can be viewed as an affective response to
work stress in which one’s professional attitudes and
behaviors change [10]. It is considered a negativestate
that reflects exposure to one or more stressors over a
∗Address for correspondence: Ellen Melbye Langballe, Årrund-
kroken 5A, 0588 Oslo, Norway. Tel.: + 47 22408376; Fax: +47
23408101; E-mail: ellen.melbye.langballe@fhi.no.
period of time, and can thus be viewedas a proxy vari-
able to assess work-related stress that has depleted peo-
ple’s coping resources [39,50]. Burnout is frequent-
ly associated with somatic symptoms such as physi-
cal exhaustion, headache, nausea, dizziness, and mus-
cle pain [47], but studies on the relationship between
burnout and physical health are relatively rare [26,47,
50].
Musculoskeletalpain is generallyconsidereda major
healthproblemintheindustrialized world[43]. Where-
1051-9815/09/$17.00 2009 – IOS Press and the authors. All rights reserved
180 E.M. Langballe et al. / The relationship between burnout and musculoskeletal pain
as the underlying pathomechanisms are poorly under-
stood, there is increasing evidence that psychosocial
factors related to the job and work environment play
a role in the development of musculoskeletal disor-
ders of the upper extremity and back [3,44]. Different
psychological reactions and complex neuroendocrine
mechanisms are involved when people are exposed to
situations perceived as demanding [45]. Physiologi-
cal, emotional, behavioral and cognitive components
influence a person’s perception of pain. Psychobiolog-
ical sensations involvingboth the central nervous sys-
tem and psychological mechanisms may lead to health
problems that are difficult to measure objectively[42].
Nevertheless, the individual may perceive them as in-
tolerable [19]. A recent study showedthat among those
with chronic musculoskeletal pain, only one in four
have single-site chronic pain. The rest have pain in
multiple sites [9]. Such pain most often cannot be con-
sidered as an isolated factor as several different biome-
chanical and psychosocial factors can account for its
origin [54].
Both burnout [47] and musculoskeletal pain [34] are
thought to be reactions to perceived stressors at work.
It is reasonable to assume that increased psychological
strain over a period of time may produce muscularten-
sion, resulting in loss of muscular strength and mobili-
ty [51]. The linkage between work-relatedfactors and
health outcomes of all varieties is often complex mul-
ti factorial in nature [3,37,38,54]. A poor psychoso-
cial work environment may lead to increased reports
of injury and illness. However, in an experimental
study where subjects were asked to perform standard
lifting under psychosocial stressed and unstressed con-
ditions, the same physical challenge resulted in much
greater spine loadings in the stressed conditions; sug-
gesting that the interaction between psychosocialstress
and biomechanical responses may partly explain how
psychosocial factors can increase the risk for muscu-
loskeletal disorders [36].
Several studies have examined the relationship be-
tween musculoskeletal pain and different types of or-
ganizational stress (see the recent reviews [2,27,56]),
but only two studies on burnout and musculoskeletal
pain were located in the review for the present study.
Bothused nationwide representativesamples from Fin-
land and reported that symptoms of burnout were more
prevalent among respondents reporting shoulder pain
thanthosewithoutsuchpain[41]andthat the incidence
of musculoskeletal disorders increased with the sever-
ity of burnout, particularly among women [26]. It is
well documented in Western societies that a consider-
able share of sickness absences and disability pensions
are related to musculoskeletal pain [1], and it has also
been reported that the duration of disability caused by
musculoskeletal pain tends to be longer if it is associ-
ated with burnout [25].
A recent study of physicians’ responses to patients
with medically unexplained symptoms indicated that
potential stressors in the home or work environment
often are not sufficiently investigatedby general prac-
titioners. The study revealed that these physicians did
not examine psychosocial concerns more thoroughly
with such patients compared to a group of patients
with gastroesophangeal reflux disorders [18]. Suggest-
ing that medically unexplained health complaints (e.g.,
burnout and musculoskeletal pain) may in some cases
be insufficiently investigated by general practitioners.
1.1. The aim of the present study
Whereas the relationship between burnout and gen-
der is not clear-cut [47], previous studies have shown
that there are gender differences in self-reported pain,
with a higher prevalenceof musculoskeletal pain found
in women [55]. However, accordingto a recent study
by Bingefors and Isacson [4], the primary explanation
probably includes gender disparities at work, econom-
ically, in daily living, and in expectations. The main
research question in this investigation was whether
burnout and musculoskeletal pain are related, and if so,
how strong is the relationship for females and males in
the different occupational groups?
2. Method
2.1. Participants and procedures
This research was based on a survey on burnout in
different occupational groups that was carried out by
the Norwegian Medical Association and Statistics Nor-
way. The data analyzed in the present study was col-
lected in 2003 and included lawyers, physicians, nurs-
es, teachers, church ministers, bus drivers, and infor-
mation technologists (IT).
For each occupational group a random sample of
1000 persons was drawn from the central Norwegian
registers of employees and employment by the Statis-
tics Norway for a total of 7000. Equal numbers of men
and women were drawn from all occupations except
from the population of church ministers where the 401
women, and 599 male church ministers were asked to
E.M. Langballe et al. / The relationship between burnout and musculoskeletal pain 181
participate. Among bus drivers, employees between 50
and 59 years were somewhat overrepresented. Other-
wise, the samples were representative ofeach gender in
each occupation. The respondents were asked to com-
plete an extensive questionnaire on burnout and related
issues. The total number of respondents was 4507, and
the overall response rate was 64% with a range of 60%
among lawyers to 70% among church ministers. The
number of respondents, response rate and mean age
with standard deviation for each occupational group is
presented in Table 1.
2.2. Measures
Burnout. A modified version of The Maslach
Burnout Inventory – General Survey (MBI-GS) was
used which measures three burnout dimensions: ex-
haustion, disengagement, and professional efficacy.
The inventorywas translated into Norwegianby the au-
thors and back-translated by a bilingual native English
speaking psychologist. She compared the back trans-
lation with the original instrument and confirmed iden-
tity of contents. The original MBI-GS scale includes
seven categories that are not always mutually exclusive
(for instance “a few times a year or less” and “once
a month or less”). Therefore, we applied a five point
scale ranging from 1 =do not agree and 5 =totally
agree in the present study.
Whereas the items in the exhaustion dimension in
this study were similar to the MBI-GS originals, pro-
fessional efficacy was measured by only five of the
original six MBI-GS items. One professional efficacy
item was excluded due to its multidimensionality. The
cynicism dimension has been criticized for low reli-
ability and conceptual overlap with the two other di-
mensions in previous studies [15,46,49]. The modified
survey included one original MBI-GS item, two items
from the Oldenburg Burnout Inventory [15] and one
item formulated by authors. The new items were con-
sidered less ambiguous and added behavioral and cog-
nitive aspects to the cynicism dimension. Because this
aspect of the modified survey focused on disengage-
mentfrom work than theoriginalthisdimension was re-
labeled Disengagement. Hence, the modified MBI-GS
used in the present study included 14 items measuring
three correlated factors: exhaustion (five items), disen-
gagement (four items), and professional efficacy (five
items). This three factor burnout construct has shown
a reasonably good model fit in a previous multi-group
confirmatory factor analyses [32].
Musculoskeletalpain. Musculoskeletalpainmostof-
ten occurs in multiple body sites [9]. A critical review
of the existing studies on work-related musculoskeletal
disorders concluded that the evidence for a relation-
ship between psychosocial factors and upper extremity
disorders appears to be strong for neck/shoulder dis-
orders and musculoskeletal symptoms in general [3].
In the present study the latent variable Musculoskeletal
pain was assessed by five items from The Subjective
Health Complaint Inventory (SHC) measuring pain in
the head, neck, shoulders, upper back and lower back.
In the SHC inventory, respondents are asked to report
pain experienced in different body sites during the last
month using a four-pointscale where 1 =not at all and
4=severe [20].
2.3. Statistical analysis
The data were analyzed using structural equation
modeling (SEM) with Lisrel 8.80. The association be-
tweenthe burnout dimensionsand musculoskeletalpain
was examined in a multi-group regression model us-
ing covariance matrices. Measurement invariance [5]
was tested because invariance constraints are necessary
to allow the model to be fitted to several samples si-
multaneously. In the first model, no restrictions were
specified. In the second model, the factor loadings
were set to be invariant across the different samples.
Lack of measurement invariance indicate that items
may have different meanings in different groups, so
thatconclusionsaboutgroup differences (or lack of dif-
ferences) cannot be accurately interpreted [17,22,40].
Researchers have to consider whether a violation of the
invariance restriction interferes with the intended use
of the scale, which in this case was to investigate the re-
lationship between constructs in different occupational
groups on a common metric.
Becauseof a high correlationbetweentheexhaustion
and the disengagement dimensions in the different oc-
cupational groups (range 0.78–0.91), multicollinearity
was considered a potential problem that had to be dealt
with. Multicollinearity may occur when highly corre-
lated factors are included in the same regression mod-
el. The higher the correlation between the predictor
variables the less the chance that unique predictive in-
formation is provided by each of them. The regression
coefficients may change in magnitude and even in sign.
The consequence can be unstable coefficients that are
difficult to interpret [12]. For this reason, a regression
analysis based on Cholesky factorization was applied
182 E.M. Langballe et al. / The relationship between burnout and musculoskeletal pain
Table 1
Number of respondents, response rate, mean years of age and standard deviation in each occupational group
Bus drivers Lawyers Nurses Physicians Church ministers Teachers IT workers
Number of respondents 583 588 684 683 685 684 601
(rate in %) (60) (60) (69) (69) (70) (69) (61)
Mean years of age 44.2 38.8 41.4 44.7 46.2 45.7 38.3
(standard deviation) (10.4) (10.2) (10.0) (10.6) (10.7) (11.1) (8.9)
Exhaustion
Disengagement
Professional efficacy
Musculoskeletal pain
Y
Y
Y
Y
Fig. 1. The standardized structural regression model tested in the present study. The relationship between the three burnout dimensions and
muscloskeletal pain.
to correct for the potential multicollinearity problems.
An illustration of the model is presented in Fig. 1.
In the Cholesky factorization procedure a sequential
decomposition of variancewas introduced based on the
hypothesized causal sequence in the burnout process in
which exhaustion is followed by disengagement,which
in turn is followed by a decrease in perceived efficacy
at work. First, the regression coefficient of exhaus-
tion on musculoskeletal pain was estimated with a free
variance; thus, it was not decomposed. Each of the
subsequent burnout dimensions were then predicted by
their precursors. According to this method for com-
puting fixed-order regression [14], disengagement and
professional efficacy were regressed as phantom fac-
tors that do not have their own indicators. Hence, a
multigroup analysis with Cholesky factorization was
conducted, with 14 groups (gender x occupation), to
allow for a comparison of the strength of the relation-
ships on a common scale (common metric completely
standardized solution).
Inlargesamples therisk of type I error(i.e., rejecting
atruenullhypothesis)exists if conclusionsare made on
the basis of the χ2-tests only [28]. The Lisrel program
provides several other fit indices that are less sensitive
tosamplesize. Thefitindices used in this study was the
Non-Normed Fit Index (NNFI) [29], the Comparative
Fit Index (CFI), and the Root Mean Square Error of
Approximation (RMSEA). Generally, NNFI and CFI
scores below 0.9 indicate that the model fit can be
improved, but Hu and Bentler [30] suggest that CFI
should be close to 0.95 before it can be assumed that
there is a good fit between the hypothesized model and
the observed data. The RMSEA estimates the overall
amount of error per degree of freedom. A RMSEA
of 0.05 or less indicates a close model fit, 0.05–0.08
indicates a good fit, and 0.08–0.10indicates a moderate
fit. A RMSEA exceeding 0.10 indicates a poor fit to
the observed data [6,16].
3. Results
3.1. Frequency distribution
Mean scores, standard deviations, and internal con-
sistencies (Cronbach α) for the variables within the
E.M. Langballe et al. / The relationship between burnout and musculoskeletal pain 183
three burnout dimensions (exhaustion, disengagement
and professional efficacy) and musculoskeletal pain
measures by gender and occupational group are pre-
sented in Table 2.
For all four dimensions, the Cronbach αwas accept-
able in all groups, indicating good internal reliability
and validity. Skewness and kurtosis ranged from −1.2
to 1.3 and from 1.2 to 2.6 respectively for the 14 MBI-
GS items in each of the 14 groups (female and female
groups across seven occupations). For the five muscu-
loskeletal pain items skewness ranged from 0.2 to 2.7
and kurtosis from −1.3 to 3.6, except in four groups, in
which the item measuring pain in the upper back had
a kurtosis between 5.2 and 8.1. With the exception of
the high kurtosis in one of the items in four groups,
the assumptions of normality were not violated, and
the conditions for estimating model parameters with
the maximum likelihood method were considered to be
met.
3.2. The model fit to the observed data
The path analysis of the four-factor model using the
Cholesky factorization procedure showed a reasonably
good model fit to the observed data both when there
were no model constraints across groups and when the
factor loadings were set to be invariant. The result of
the invariancetest is presented in Table 3.
The configural invariance criteria were considered
metby the good modelfitshownin the multigroupanal-
ysis with no common parameters across groups. When
the two models were compared, the chi square differ-
ence test was significant, indicating measurement non-
invariance across groups. However, alternative meth-
ods to evaluate measurement invariance such as delta
CFI and model fit suggested adequate measurement in-
variance. The delta CFI test was well below the −0.01
limit suggested by Cheung and Renvold [11,53], and
the multigroup model exhibited an excellent model fit
to the data [7]. According to Byrne [8], instruments are
often group specific in the way they operate and, thus,
to expect the final model to be completely identical for
each group would be unrealistic. Since the homogene-
ity of variance assumption was not rejected by the delta
CFI test and model fit indices, group comparison of the
relationship between burnout and musculoskeletalpain
was based on the so called metric invariance, or weak
invariance [40],where the factor loadings are set to be
invariant across groups. Invariant factor loadings on
theintendedlatent construct for exhaustion, disengage-
ment, professional efficacy and musculoskeletal pain,
loaded between 0.65–0.84, 0.62–0.77, 0.48–0.69 and
0.41–0.83, respectively.
3.3. The burnout and musculoskeletal pain
relationship
There was a significant, positive relationship be-
tween musculoskeletal pain and exhaustion in all the
occupational groups and for both females and males.
When compared on a common metric, the strength of
the relationship in one group can be evaluated against
the strengths in the other groups. The regression coef-
ficients presented in Table 4 revealed that the strength
of the relationship ranged from 0.26 in female physi-
cians to 0.64 in male bus drivers. Interestingly, in
male and female bus drivers, male IT workers, church
ministers and nurses increased disengagement scores
were associated with decreased musculoskeletal pain.
Relatively weak relationships with both positive and
negative signs were found between professional effica-
cy and musculoskeletal pain among male nurses, male
teachers, and IT workers of both genders.
In general, the results of the multigroup regression
analysis showed larger differences in the strengths of
therelationships between differentoccupational groups
than between females and males within the same pro-
fession. The strength of the significant associations in
this study maximally differs by 0.16 between females
and males within the same profession (bus drivers)
whereas the discrepancy in the association is much
larger between some of the occupational groups (e.g.
lawyers and physicians compared to bus drivers). The
overall range in the strength of association between the
three burnout dimension and musculoskeletal pain is
illustrated in Fig. 2.
4. Discussion
The primary research questions were: Is there a re-
lationship between symptoms of burnout and muscu-
loskeletal pain? If so, how strong is that relationship
forfemales and males in different occupationalgroups?
The relationship between burnout and musculoskeletal
pain was examined among lawyers, physicians, nurses,
teachers,churchministers, bus drivers,andinformation
technology (IT) employees (N =4507). The cross-
sectional design of the present study may be consid-
ered a limitation because it precludes causal interpre-
tations. However, the data collection procedures and
the acceptable response rate in most groups allow for
generalization of the findings to these specific occu-
pations in Norway. In five of the seven occupational
groups the association between exhaustion and muscu-
184 E.M. Langballe et al. / The relationship between burnout and musculoskeletal pain
Table 2
Mean scores on observed variables with standard deviations for Exhaustion, Disengagement, Professional Efficacy and Musculoskeletal
Pain, and Cronbach α(Exhaustion/Disengagemeny/Professional efficacy/Musculoskeletal pain) for each occupational group and gender (F
=Females, M =Males)
Exhaustion Disengagement Professional efficacy Musculoskeletal Pain Cronbach α
Lawyers F 2.41 0.79 1.99 0.80 2.17 0.54 0.88 0.63 0.84/0.79/0.75/0.80
M 2.26 0.74 1.90 0.71 2.08 0.51 0.56 0.41 0.82/0.73/0.77/0.64
Bus drivers F 2.57 1.00 2.31 0.99 2.26 0.58 1.41 0.78 0.87/0.81/0.64/0.82
M 2.40 0.98 2.30 0.91 2.15 0.58 1.16 0.70 0.87/0.78/0.68/0.79
IT workers F 2.50 0.89 2.30 0.86 2.24 0.59 1.01 0.66 0.87/0.81/0.79/0.78
M 2.36 0.89 2.26 0.89 2.12 0.56 0.73 0.57 0.88/0.81/0.76/0.76
Physicians F 2.55 0.83 1.94 0.75 2.08 0.54 0.71 0.47 0.85/0.80/0.77/0.66
M 2.42 0.79 2.04 0.70 2.06 0.53 0.56 0.39 0.85/0.75/0.78/0.60
Teachers F 2.80 0.86 2.06 0.81 2.19 0.50 0.98 0.65 0.84/0.81/0.77/0.63
M 2.62 0.90 2.21 0.83 2.08 0.49 0.73 0.56 0.87/0.78/0.74/0.75
Church
Ministers F 2.59 0.80 1.98 0.80 2.08 0.53 0.95 0.59 0.82/0.80/0.76/0.75
M 2.36 0.80 1.83 0.68 2.10 0.50 0.60 0.43 0.84/0.78/0.76/0.64
Nurses F 2.48 0.86 1.99 0.76 2.11 0.50 0.97 0.65 0.86/0.79/0.72/0.79
M 2.45 0.82 2.23 0.82 2.08 0.55 0.72 0.52 0.85/0.80/0.75/0.72
Note: Range for the bumout dimensions: 1–5: ‘do not agree’–‘totally agree.’ Range for musculoskeletal pain: 0–3: ‘not at all’–‘severe’.
Table 3
Multigroup regression analysis of the four factor model – burnout (three) dimensions and musculoskeletal pain (one dimension).
Males and females in each of the seven occupational (14 groups) are included in the analyses
Multigroup regression analysis. 14 groups χ
2df RMSEA (90%CI) NNFI CFI ∆χ2(df) ∆CFI
Null model 58694.3 2394
No common parameters 4247.7 2044 0.064 (0.061–0.067) 0.955 0.961
Invariant factor loadings 4754.9 2239 0.066(0.063–0.068) 0.953 0.957 507.2 (195) −0.004
Table 4
Multigroup analysis of the association between the three burnout dimensions and muscu-
loskeletal pain, regression coefficients. Common metric completely standardized solution
Exhaustion Disengagement Professional Efficacy
Lawyers Females 0.40∗−0.16 0.02
Males 0.35∗−0.04 0.03
Bus drivers Females 0.61∗−0.42∗−0.05
Males 0.64∗−0.26∗0.11
IT workers Females 0.49∗−0.18 −0.25∗
Males 0.43∗−0.30∗−0.17∗
Physicians Females 0.26∗0.06 −0.01
Males 034∗−0.13 0.08
Teachers Females 0.50∗0.04 −0.12
Males 0.37∗−0.08 0.16∗
Church ministers Females 0.52∗−0.21 0.08
Males 0.40∗−0.22∗0.01
Nurses Females 0.54∗−0.18 0.12
Males 0.45∗−0.15∗−0.21∗
∗p<0.05.
Note: Factor loading are invariant across groups in the multigroup analysis.
loskeletal pain was higher among female employees.
In line with previous research [26] the present findings
showed that burnout and musculoskeletal pain often
co-occur and to a slightly higher degree for females
than males within the same occupation. Even so, the
results indicated that the association between the three
burnout dimensions and musculoskeletal pain varied
more between occupations than between males and fe-
males within the same profession. This suggests that
factorsrelated to occupation may be strongerpredictors
of a co-occurrence of these two health complaints than
gender.
The results showed that the exhaustion dimension
of burnout was significantly and positively associat-
ed with musculoskeletal pain irrespectiveof profession
and gender. The relationship was moderate to high in
E.M. Langballe et al. / The relationship between burnout and musculoskeletal pain 185
Exhaustion
Disengagement
Professional
efficacy
Musculoskeletal
pain
*0.26-0.64
***-0.17-0.16
**-0.15--0.42
Fig. 2. The range of the significant relationship between the three burnout dimensions and musculoskeletal pain tested in the present study.
Standardized solution. Note: ∗significant in all groups, ∗∗significant in five groups, ∗∗∗significant in four groups.
all groups, and particularly high in bus drivers. Inves-
tigations of burnout in bus drivers are scarce (see [13,
35]), and the impact of stress and work-related psy-
chosocial factors on musculoskeletal pain is poorly un-
derstood [35]. Results of the present study suggest
that musculoskeletal pain and burnout symptoms are
likely to co-occur in bus drivers. Hence, a compre-
hensive approach, considering both the physical and
the psychological work environment,may be the most
efficient way to overcome these sorts of health prob-
lems in this occupational group. Moreover, the rela-
tionship between exhaustion and musculoskeletal pain
was relatively high in information technology work-
ers and nurses of both genders, and in female teachers
and church ministers. Static working positions among
IT-workers and physically demanding work tasks such
as heavy lifting among nurses may increase the risk
for musculoskeletal pain. Interactions between biome-
chanical and psychosocial factors may therefore only
partly explain the findings in both these occupational
groups. Physical explanatory factors in the work envi-
ronment are less obviousin female teachers and church
ministers. Wehypothesizethat emotionallydemanding
work tasks are the most important factors in the two
latter groups, but further research is needed.
Interestingly the disengagement dimension of burn-
outwas associated with decreasedmusculoskeletalpain
whereas in a previous study high cynicism levels were
associated with increased musculoskeletal pain [26].
However, because this burnout dimension were mea-
sured with different items and analyzed with different
statistical procedures it is difficult to compare the re-
sults, but the contradictory findingsindicate that the re-
lationshipbetween disengagement andmusculoskeletal
pain should be further investigated. Nevertheless a sig-
nificant negative relationship between disengagement
and musculoskeletal pain was evident in the samples
of bus drivers,male IT workers, church ministers, and
nurses. The strength of this relationship was moderate
to low, with the strongest association found among fe-
male bus drivers. It is noteworthy that an increase in
disengagement scores in these occupations was associ-
ated with a decreasein musculoskeletal pain. The find-
ings may imply that whereas exhaustion may lead to
physical tension and musculoskeletal pain, disengage-
ment from work may have the opposite effect. Disen-
gagement or cynicism is considered a way to cope with
feelings of exhaustion caused by strain at work [10,23,
24,48] and is a significant component of burnout since
it may have negative effects on the employee’s produc-
tivity and health [52]. On the other hand, it has been
argued that disengagementor cynicism is first and fore-
most a coping strategy, and should not be considered
part of the burnout syndrome [31]. The negative rela-
tionship between disengagement and musculoskeletal
pain found in this study supports the assumption that
disengagement may be regarded as a coping mecha-
nism. At the same time, the high correlations between
exhaustion and disengagement may indicate that the
two dimensions are indeed related concepts that can
be viewed as parts of the same syndrome. Thus, there
is a need to learn more about potential positive out-
186 E.M. Langballe et al. / The relationship between burnout and musculoskeletal pain
comes of different coping mechanisms fordealing with
stress [21].
Professional efficacy was weakly and inconsistent-
ly (i.e., both positively and negatively) were associat-
ed with musculoskeletal pain; a result that is in line
with Schaufeli and Enzmann’s [47] conclusion that this
third burnout dimension is the least strongly related to
potential correlates.
4.1. Final conclusion
The findings from the present study indicate a co-
occurrence of musculoskeletal pain, exhaustion, and
disengagement. The association between exhaustion
and musculoskeletal pain was particularly strong and
found in all occupational groups. The strength of the
relationship varies by occupation and partly by gender.
Hence, the present study contributes to increase our
knowledge of how burnout may be differently relat-
ed to physical outcomes such as musculoskeletal pain.
The interaction between burnout and musculoskeletal
pain may intensify the total experience of illness, im-
plying that more interdisciplinary research on the in-
teraction between the psychosocial and physiological
mechanisms involved in the burnout process is needed.
More needs to be learned about how burnout and mus-
culoskeletal pain are expressed and it would be valu-
able to obtain qualitative data from the participants in
future quantitative studies [21].
Thedemonstratedassociation of burnoutandmuscu-
loskeletal pain may inform occupational specific pre-
ventive,diagnostic,and considerations for intervention.
Rehabilitationprofessionals (i.e. physiotherapists) may
have a particularly important role to play in this re-
gard. Complex relationships between psycho-social
and physical working conditions that are either not
sufficiently investigated [18] or not discovered by the
physician in the relatively brief initial meetings with
the patient may become evident during the more long
lasting treatment provided by rehabilitation personnel.
Awareness of these relationships may promote more
precise diagnosis and treatment strategies.
Acknowledgments
The writing of this paper was supported by the
aid of EXTRA funds from the Norwegian Foundation
for Health and Rehabilitation through the Norwegian
Council for Mental Health, and by Diakonhjemmet
Hospital. The data collection was financed by the Re-
search Institute of the Norwegian Medical Association.
References
[1] K.A. Alexanderson, K.E. Borg and G.K Helsing, Sickness ab-
sence with low-back, shoulder, or neck diagnoses: An 11-year
follow-up regarding gender differences in sickness absence
and disability pension, WORK 25 (2005), 115–124.
[2] G.A. Ariens, W. van Mechelen, P.M. Bongers, L.M. Bouter
and G. van der Wal, Psychosocial risk factors for neck pain:
a systematic review, American Journal of Industrial Medicine
39 (2001), 180–193.
[3] B.P. Bernard, ed., Musculoskeletal Disorders and Workplace
Factors: A Critical Review of Epidemiologic Evidence for
Work-Related Musculoskeletal Disorders of the Neck, Upper
Extremity, and Low Back, (NIOSH Publication No. 97-141),
US Department of Health and Human Services, Cincinnati,
1997.
[4] K. Bingefors and D. Isacson, Epidemiology, co-morbidity,
and impact on health-related quality of life of self-reported
headache and musculoskeletal pain-A gender perspective, Eu-
ropean Journal of Pain 8(2004), 435–450.
[5] K.A. Bollen, Structural Equations with Latent Variables, John
Wiley and Sons, Oxford, England, 1989.
[6] M.W. Browne and R. Cudeck, Alternative ways of assessing
model fit, in: Testing Structural Equation Models, L.J. Bollen
KA, ed., Sage Publications, Newbury Park, California, 1993,
pp. 136–162.
[7] B.M. Byrne and S.M. Stewart, The MACS approach to testing
for multigroup invariance of a second-order structure: A walk
through the process, Structural Equation Modeling 13 (2006),
287–320.
[8] B.M. Byrne, The Maslach Burnout Inventory: Validating
Factorial Structure and Invariance Across Intermediate, Sec-
ondary, and University Educators, Multivariate Behavioral
Research 26 (1991), 583–605.
[9] D. Carnes, S. Parson, D. Ashby, A. Breen, N.E. Foster, T.
Pincus, S. Vogel and M.Underwood, Chronic musculoskeletal
pain rarely presents a single body site: results from a UK
population study, Rheumatology 46 (2007), 1168–1170.
[10] C. Cherniss, Professional Burnout in the Human Service Or-
ganizations, Praeger, New York, 1980.
[11] G.W. Cheung and R.Rensvold, Evaluating goodness-of-fit in-
dexes for testing measurement invariance, Structural Equation
Modeling 9(2002), 233–255.
[12] J. Cohen, P. Cohen, S.G. West and L.S. Aiken, Applied Mul-
tiple Regression/Correlation Analysis for the Behavioral Sci-
ences (3rd ed.), Lawrence Erlbaum Associates, Mahwah, NJ,
US, 2003.
[13] C.B. Cunradi, B.A. Greiner, D.R. Ragland and J.M. Fisher,
Burnout and alcohol problems among urban transit operators
in San Francisco, Addictive Behaviors 28 (2003), 91–109.
[14] P.F. de Jong, Hierarchical regression analysis in structural
equation modeling, Structural Equation Modeling 6(1999),
198–211.
[15] E. Demerouti, A.B. Bakker, I. Vardakou and A. Kantas, The
convergent validity of two burnout instruments: A multitrait-
multimethod analysis, European Journal of Psychological As-
sessment 19 (2003), 12–23.
[16] A. Diamantopoulos and J.A. Siguaw, Introducing LISREL: A
Guide for the Uninitiated, Sage publications, London, 2000.
[17] D.M. Dimitrov, Comparing groups on latent variables: A
structural equation modeling approach, WORK 26 (2006),
429–436.
[18] R.M. Epstein, C.G. Shields, S.C. Meldrum, K. Fiscella, J. Car-
roll, P.A. Carney and P.R. Duberstein, Physicians’ responses
E.M. Langballe et al. / The relationship between burnout and musculoskeletal pain 187
to patients’ medically unexplained symptoms, Psychosomatic
Medicine 68 (2006), 269–276.
[19] H.R. Eriksen and H. Ursin, Sensitization and subjective health
complaints, Scandinavian Journal of Psychology 43 (2002),
189–196.
[20] H.R. Eriksen, C. Ihlebaek and H. Ursin, A scoring system for
subjective health complaints (SHC), Scandinavian Journal of
Public Health 1(1999), 63–72.
[21] S. Folkman, Positive affect and the other side of coping, Amer-
ican Psychologist 55 (2000), 647–654.
[22] B.F. French and W.H. Finch, Confirmatory Factor Analysis
Procedures for the Determination of Measurement Invariance,
Structural Equation Modeling 13 (2006), 378–402.
[23] I.A. Friedman, Multiple pathways to burnout: Cognitive and
emotional scenarios in teacher burnout, Anxiety, Stress and
Coping: An International Journal 9(1996), 245–259.
[24] E.R. Greenglass, R.J. Burke and R. Konarski, The impact
of social support on the development of burnout in teachers:
Examination of a model, Work and Stress 11 (1997), 267–278.
[25] G. Grossi, J.J. Soares, J. Aengesleva and A. Perski, Psychoso-
cial correlates of long-term sick-leave among patients with
musculoskeletal pain, Pain 80 (1999), 607–619.
[26] T. Honkonen, K. Ahola, M. Pertovaara, E. Isometsa, R.
Kalimo, E. Nykyri, A. Aromaa and J. Lonnqvist, The asso-
ciation between burnout and physical illness in the general
population-results from the Finnish Health 2000 Study, Jour-
nal of Psychosomatic Research 61 (2006), 59–66.
[27] W.E. Hooftman, Gender differences in the relations between
work-related physical and psychosocial risk factors and mus-
culoskeletal complaints, Scandinavian Journal of Work, Envi-
ronment, and Health 30 (2004), 261–278.
[28] R.H. Hoyle, Structural equation modeling: Concepts, issues,
andapplications, Sage Publications, Inc, Thousand Oaks, CA,
US, 1995.
[29] L. Hu and P.M. Bentler, Evaluating model fit, in: Structural
Equation Modeling: Concepts, Issues, and Applications, R.H.
Hoyle, ed., Sage Publications, Inc, Thousand Oaks,CA,US,
1995, pp. 76–99.
[30] L. Hu and P.M. Bentler, Cutoff criteria for fit indexes in co-
variance structure analysis: Conventional criteria versus new
alternatives, Structural Equation Modeling 6(1999), 1–55.
[31] T.S. Kristensen, M. Borritz, E. Villadsen and K. Christensen,
The Copenhagen Burnout Inventory: A new tool for the as-
sessment of burnout, Work and Stress 19 (2005), 192–207.
[32] E.M. Langballe, E. Falkum, S.T. Innstrand and O.G. Aasland,
The factor Validity of the Burnout Inventory-general Survey
in Representative Samples of Eight Different Occupational
Groups, Journal of Career Assessment 14 (2006), 370–384.
[33] H. Leventhal and P.A. Mora, Is there aScience of the Processes
Underlying Health and Illness Behaviors? A Comment on
Maes and Karolyn, Applied Psychology: An International
Review 54 (2005), 255–266.
[34] S. Linton, Occupational psychological factors increase the risk
for back pain: A systematic review, Journal of Occupational
Rehabilitation 11 (2001), 53–66.
[35] J.L. M. Tse, R. Flin and K. Mearns, Bus driver well-being
review: 50 years of research, Transportation Research Part F:
Traffic Psychology and Behaviour 9(2006), 89–114.
[36] W.S. Marras, K.G. Davis, C.A. Heaney, A.B. Maronitis and
W.G. Allread, The influence of psychosocial stress, gender,
and personality on mechanical loading of the lumbar spine,
Spine 25 (2000), 3045–3054.
[37] W.S. Marras,State-of-the-art research perspectives on muscu-
loskeletal disorder causation and control: the need for an in-
tergraded understanding of risk, Journal of Electromyography
and Kinesiology 14 (2004), 1–5.
[38] S.E. Mathiassen, Expo[not-so]sure, Reflections on a Tricky
Issue in musculoskeletal research, Scandinavian Journal of
Work,Environment, and Health 30 (2004), 173–177.
[39] S. Melamed, A. Shirom, S. Toker, S. Berliner and I.Shapira,
Burnout and Risk of Cardiovascular Disease: Evidence, Pos-
sible Causal Paths, and Promising Research Directions, Psy-
chological Bulletin 132 (2006), 327–353.
[40] W. Meredith, Measurement invariance, factor analysis and
factorial invariance, Psychometrika 58 (1993), 525–543.
[41] H. Miranda, E. Viikari-Juntura, S. Heistaro, M. Heliovaara
and H. Riihimaki, A population study on differences in the
determinants of a specific shoulder disorder versus nonspecific
shoulder pain without clinical findings, American Journal of
Epidemiology 161 (2005), 847–855.
[42] H. Miranda, J.E. Gold, R. Gore and L. Punnett, Recall of
prior musculoskeletal pain, Scandinavian Journal of Work,
Environment, and Health 32 (2006), 294–300.
[43] E. Røysamb, K. Tambs, T. Reichborn-Kjennerud, M. C. Neale
and J. R. Harris, Happiness and Health: Environmental and
Genetic Contributions to the Relationship Between Subjective
Well-Being, Perceived Health, and Somatic Illness, Journal of
Personality and Social Psychology 85 (2003), 1136–1146.
[44] T. Reichborn-Kjennerud, C. Stoltenberg, K. Tambs, E.
Roysamb, E. Kringlen, S. Torgersen, and J.R. Harris, Back-
neck pain and symptoms of anxiety and depression: a
population-based twin study, Psychological Medicine 32
(2002), 1009–1020.
[45] A. Rozanski and L.D. Kubzansky, Psychologic functioning
and physical health: a paradigm of flexibility, Psychosomatic
Medicine 67(Suppl 1) (2005), 47–53.
[46] M. Salanova, S. Llorens, M. Garc´
ıa-Renedo, R. Burriel, E.
Bres´
o and W.B. Schaufeli, Towards a Four-Dimensional Mod-
el of Burnout: A Multigroup Factor-Analytic Study Including
Depersonalization and Cynicism, Educational and Psycholog-
ical Measurement 65 (2005), 807–819.
[47] W.B. Schaufeli and D. Enzmann, The Burnout Companion to
Study and Practice: A Critical Analysis, Taylor & Francis,
London, 1998.
[48] W.B. Schaufeli, D. Enzmann and N. Girault, Measurement
of burnout: A review, in: Professional Burnout: Recent De-
velopments in Theory and Research, W.B. Schaufeli and C.
Maslach, eds, Series in applied psychology: Social issues
and questions, Taylor &Francis, Philadelphia, PA,US, 1993,
pp. 199–215.
[49] N. Schutte, S. Toppinen, R. Kalimo and W. Schaufeli, The
factorial validity of the Maslach Burnout Inventory-General
Survey (MBI-GS) across occupational groups and nations,
Journal of Occupational and Organizational Psychology 73
(2000), 53–66.
[50] A. Shirom and S. Melamed, Does burnout affect physical
health? A review of the evidence, in: Research Companion
to Organizational Health Psychology, A-S. G.Antoniou and
C.L.Cooper, eds, 2005, pp. 599–622.
[51] J.F. Soares and B. Jablonska, Psychosocial experiences among
primary care patients with and without musculoskeletal pain,
European Journal of Pain 8(2004), 79–89.
[52] S. Sonnentag, Burnout research: Adding an off-work and day-
level perspective, Work and Stress 19 (2005), 271–275.
[53] R.J. Vandenberg and C.E. Lance, A review and synthesis of
the measurement invariance literature: Suggestions, practices,
and recommendations for organizational research, Organiza-
tional Research Methods 3(2000), 4–69.
188 E.M. Langballe et al. / The relationship between burnout and musculoskeletal pain
[54] N. Warren, Workstress and musculoskeletal disorders etiolo-
gy: The relative role of psychosocial and physical risk factors,
WORK 17 (2001), 221–234.
[55] H.A.H. Wijnhoven, H.C.W. de Vet and H.S.J. Picavet, Preva-
lence of Musculoskeletal Disorders is Systematically Higher
in Women Than in Men, Clinical Journal of Pain 22(8) (2006),
717–724.
[56] V. Woods, Work-related musculoskeletal health and social
support. Occupational Medicine 55(3) (2005), 177–189.
