Full text of DOI:10.1207/S15327558IJBM0902_04

INTERNATIONAL JOURNAL OF BEHAVIORAL MEDICINE, 9(2), 122–138
Copyright © 2002, Lawrence Erlbaum Associates, Inc.
Perceived Work Stress,
Overcommitment, and Self-Reported
Musculoskeletal Pain:
ACross-Sectional Investigation
Ljiljana Joksimovic, Dagmar Starke,
Olaf v.d. Knesebeck, and Johannes Siegrist
Theobjectiveof thisstudywas to analyze associations ofthree indicators ofperceived
work stress (physical job demand, low control at work, and an imbalance between ef-
fort and reward), and of overcommitment, a personal pattern of coping with work de-
mands, with musculoskeletal pain. A standardized questionnaire measuring these
conditions in addition to self-reported musculoskeletal pain at different locations was
administered to a group of 316 male and female employees of a public transport enter-
prise. After we adjusted for confounding effects of age, sex, socioeconomic status,
shift work, and negative affectivity, we observed elevated prevalence odds ratios in
employees who scored high on overcommitment, who were exposed to physical job
demand, and, to a lesser extent, who reported psychosocial work stress. Results have
implications for a more comprehensive approach to primary and secondary preven-
tion of musculoskeletal pain.
Key words: musculoskeletal pain, physical job demand, control at work, effort–reward
imbalance, overcommitment
This research was supported by a grant from The Federal Ministry of Education and Sciences in Ger-
many (Public Health Research Program, Project C1–NRW). We are grateful to the employees of the
company, in particular to the chief occupational physician, Dr. Manuela Berghof, for their support. We
also thank three anonymous reviewers for their helpful comments of an earlier version of this article.
Ljiljana Joksimovic, Dagmar Starke, Olaf v.d. Knesebeck, and Johannes Siegrist, Department of
Medical Sociology, Heinrich-Heine-University, Duesseldorf, Germany.
CorrespondenceconcerningthisarticleshouldbeaddressedtoLjiljanaJoksimovic, Departmentof
Medical Sociology, Heinrich-Heine-University, Universitätsstrabe 1, P.O. Box 101007, D-40001
Duesseldorf, Germany. E-mail: joksimov@uni-duesseldorf.de

WORK-STRESS AND MUSCULOSKELETAL PAINWORK
123
Musculoskeletal pain is a common medical complaint and a frequent cause of
sickness absence in industrialized countries (Brogmus, Sorock, & Webster,
1996). According to recent studies, the most common musculoskeletal symp-
toms are neck and back pain. In general populations, the 1-year prevalence was
found to vary between 30% and 40% (Bovim, Schrader, & Sand, 1994) and be-
tween 32% and 62% (Berger-Schmitt, Kohlmann, & Raspe, 1996) for neck and
back pain, respectively. Prevalence data in working populations are even higher
(Cole & Hudak, 1996; Eriksen, Natvig, Knardahl, & Bruusgaard, 1999;
Magnavita, Bevilacqua, Mirk, Fileni, & Castelino, 1999; Magnusson, Pope,
Wilder, & Areskoug, 1996; Riihimäki, 1999). Additional sites of frequent
musculoskeletal complaints are the shoulder, hip, and upper and lower extremi-
ties. In all these instances, interventions based on principles of behavioral medi-
cine are instrumental in reducing pain and in improving coping with ill health
(Eriksen & Ursin, 1999). Yet, more information is needed on specific environ-
mental and personal factors that contribute to the development of
musculoskeletal pain and, thus, may define targets of intervention. Among envi-
ronmental factors, working condition, in particular physical workload, plays a
prominent role (Ariëns, van Mechelen, Bongers, Bouter, & van der Wal, 2000;
Bernard, 1997; Hagberg et al., 1995; Magnavita et al., 1999; Walsh, Varnes, Os-
mond, Styles, & Coggon, 1989). More recently, a stressful psychosocial work
environment was shown to be associated with elevated risk of musculoskeletal
pain (Bongers, de Winter, Kompier, & Hildebrandt, 1993; Foppa & Noack,
1996). Among these conditions, a low degree of task control at work seems to be
particularly relevant (Bongers et al., 1993; Hollmann, Heuer, & Schmidt, 2001;
Houtman, Bongers, Smulders, & Kompier, 1994). In addition to situational fac-
tors, personal ways of coping with demands have to be taken into account
(Bongers et al., 1993; Eriksen & Ursin, 1999).
This study was undertaken to explore associations between a new measure of
psychosocial stress at work, effort–reward imbalance, and musculoskeletal pain
in a working population with high prevalence of these conditions (Siegrist,
1996). We chose this new measure of work stress because it allows to the re-
searcher to distinguish between perceived situational work stressors and per-
sonal coping characteristics at the conceptual and operational level. This model
maintains that the availability of an occupational status is associated with recur-
rent options of contributing and performing, of being rewarded or esteemed, and
of belonging to some significant group (e.g., colleagues). Yet, these potentially
beneficial effects are contingent on a basic prerequisite of exchange in social
life, that is, reciprocity. Effort spent at work is part of a socially organized ex-
change process to which society at large contributes in terms of rewards. Re-
wards are distributed by three transmitter systems: money, esteem, and career
opportunities. The model of effort–reward imbalance claims that lack of reci-
procity between the costs and gains (i.e., high-cost/low-gain conditions) defines

124
JOKSIMOVIC ET AL.
a state of emotional distress with special propensity to autonomic arousal and
neuroendocrine stress responses. For instance, having a demanding but unstable
job or achieving at high level without being offered any promotion prospects are
examples of a particularly stressful working context. High-cost/low-gain condi-
tions are maintained under the following three circumstances. First, when an al-
ternative choice in the labor market is not available, anticipated costs of
disengagement outweigh the costs of accepting inadequate benefits. Therefore,
unrewarding jobs are held. Second, people may accept unfair job arrangements
for a certain time for strategic reasons, because they may improve their chances
for career promotion and related rewards at a later stage.
A third reason for a continued mismatch between efforts and rewards points
to a particular pattern of coping with work-related demands and rewards. People
characterized by an excessive job involvement, as previously described in the
Type A behavior pattern, may expose themselves to recurrent sustained activa-
tion. In a refined concept of excessive job involvement, termed overcommitment
(Siegrist, 1996), a cognitive–motivational pattern has been identified where a
high level of overcommitment prevents people from accurately assessing the
costs and gains of demanding challenges. Because the experience of mastering
demands is rewarding for them, they are likely to overestimate their coping re-
sources and underestimate the efforts required for mastering demands. In the
long run, overcommitted people suffer from their inability to withdraw from
work obligations. Despite a positive outcome expectancy, cumulative load of
their excessive efforts reduces their potential of recovery and increases their sus-
ceptibility to ill health (Siegrist, 2001; Ursin & Eriksen, 2001).
Therefore, we analyze the role of these two components of the model of ef-
fort–reward imbalance, perceived situational work stressors as measured by a ra-
tio between demands and rewards (see Methods section), and the personal
coping pattern of overcommitment, in explaining musculoskeletal symptoms at
different sites. Additional explanatory factors are physical workload and low
control at work (Karasek & Theorell, 1990). To adjust for possible confounding
effects, age, sex, socioeconomic status, and shift work are taken into account in
multivariate analyses. In addition, a measure of negative affectivity is included
to reduce bias due to the substantial methodological problem of common
method variance (Abraham, 1999).
METHODS
Participants
A group of female and male employees of a large public transport company in a
major city in Germany were invited to participate in this cross-sectional study.

WORK-STRESS AND MUSCULOSKELETAL PAINWORK
125
Participants provided written informed consent. Data were collected during two
working days in a separate facility under standardized and controlled conditions.
Selection criteria for participation were (a) age (35 years), (b) duration of em-
ployment in the company at least 2 years, and (c) fluency in German language.
Restriction to midlife (35–60 years) was decided because effects of work stress
are expected to be more prevalent in this group compared with a younger group
(Morano, 1993).
Although a majority of employees were bus and subway drivers, we did not
restrict participation to this employment group but invited repair personnel and
administrative personnel to participate as well. The study was undertaken with
support from the company’s occupational health department. A notice about vol-
untary participation was put up in two of four bus drivers’ working sites and in
the administration building of the company. This means that about 1,000 em-
ployees fulfilling sample criteria had a chance of being informed about the
study. However, because data collection was restricted to 2 days and the survey
was conducted before or after regular working time, the size of the eligible pop-
ulation was far smaller. A total of 316 employees participated (268 men and 48
women). Although we cannot exclude a selection bias due to a high level of per-
ceived work stress or health complaints, the participants’ feedback indicated that
their interest in a scientific survey and the financial incentive offered were
equally strong motivations for participation.
We did not intend to recruit a representative sample, but it turned out that the
proportions of the different occupational groups of the sample were well compa-
rable to the company’s composition of workforce. The fact that few women par-
ticipated reflects the low proportion of women employed in this enterprise,
although women in our sample were clearly underrepresented. Mean age of par-
ticipants was 44.8 years (7.6) among men and 43.7 (6.4) among women. Half of
the participants were bus or subway drivers, 23% were white-collar workers,
and 19% were blue-collar workers in repair services. Almost all participants
were full-time employed. Mean employment time in the current job was 16.1
(8.6) years, and shift work was frequent among participants (78%; the majority,
46%, involving night shift). In addition, 63.1% had a low level of education (el-
ementary school), and 21.8% had a household income lower than 3000 DM per
month after taxes.
Measures
Perceived Work Stress. Low control at work was measured by three items
representing the Job Decision Latitude scale derived from the Job Content Ques-
tionnaire, a widely used measure of psychosocial work stress in terms of high de-
mand and low control (Karasek et al., 1998). Physical job demand and effort–re-

126
JOKSIMOVIC ET AL.
ward imbalance were assessed by a Likert-scaled standardized questionnaire.
Items are answered in two steps. First, participants agree or disagree on whether the
item content describes a typical experience of their work situation. Subsequently,
participants who agree are asked to evaluate to what extent they usually feel dis-
tressed by this typical experience (4-point Likert scale). Physical job demand is
measured by one item, whereas five items define a scale of effort (quantitative and
qualitative load, increase of work demands in recent past). Eleven items define a
scale of occupational rewards (financial rewards, esteem rewards, and rewards re-
lated to promotion prospects and job security). Psychometric properties of the
questionnaire including factorial structure and internal consistency of scales were
reported by Niedhammer, Siegrist, Landre, Goldberg, and Leclerc (2000), Peter et
al. (1998), and Siegrist (1996). In this study, internal consistency (Cronbach’s α =
0.72 and 0.86, respectively) and factorial structure of the Effort and Reward scales
were consistent with previous analyses. The ratio of the sum score extrinsic effort
(e) and the reversed sum score of reward (r) was computed according to the follow-
ing formula: e / (r × c), where c defines a correction factor for different numbers of
items in the nominator and denominator. As a result, a value of 1.0 indicates ef-
fort–reward balance, whereas values >1.0 indicate the critical condition of high ef-
fort and low reward. Based on this threshold, a binary variable of the extrinsic
model component was defined: effort–reward ratio, high or low.
Overcommitment. The personal pattern of coping with work demands
termed overcommitment was measured by a recently developed short version (six
items, see the appendix) of the original 29-item version of a psychometric test
(Starke et al., 2002). In previous studies, overcommitment was associated with ele-
vated risks of coronary heart disease, indicators of the metabolic syndrome, coro-
nary reocclusion after percutaneous transluminal coronary angioplasty, and burn-
out (for review, see Schnall, Belkic, Landsbergis, & Baker, 2000; Siegrist, 2001).
This study offers an opportunity to explore associations of overcommitment with a
different health indicator, musculoskeletal pain. The short version of the
overcommitment scale was developed because previous confirmatory factor analy-
ses that used the original 29-item version showed that its explanatory power was
confined largely to the Inability to Withdraw From Work Obligations subscale, a
subscale that was particularly well replicated across different data sets
(Niedhammer et al., 2000). Therefore, confirmatory factor analysis based on data
from two large international samples and from two smaller German samples was
performed where the unidimensionality of the scale, its internal consistency, and
the variance explained *[Goodness-of Fit-Index (GFI), Adjusted-Goodness-of-Fit
Index (AGFI), Root-Mean-Square Error of Approximation (RMSEA).] were ana-
lyzed by using maximum likelihood method (Starke et al., 2002). In the current

WORK-STRESS AND MUSCULOSKELETAL PAINWORK
127
sample, respective values were α = 0.73, GFI = 0.98, AGFI = 0.95, and RMSEA =
2
0.06, and loadings (α ) ranged from 0.31 to 0.84. In line with previous analyses, a
binary variable was computed with scores in the upper tertile defining a critical
level of overcommitment (value = 1) versus the remaining group with scores indi-
cating low or moderate overcommitment (value = 0).
Negative Affectivity. Several authors now recommend adjusting the effects
of psychosocial measures on health indicators in cross-sectional studies for
mood-dispositional variables such as negative affectivity (Abraham, 1999; Burke,
Brief, & George, 1993). Negative affectivity describes a tendency to react with
negative mood to all types of self-reported items. Thus, this tendency may inflate
the previously mentioned associations. In this study, negative affectivity was as-
sessed by the Negative Affect scale (eight items) of the Profile of Quality of Life in
Chronically Ill People (Siegrist, Starke, Laubach, & Brähler, 2000). Construct va-
lidity was documented by a unidimensional factor structure with all items loading
on a single factor. The scale was shown to be highly reliable in several studies
(Cronbach’s α ranging from 0.82 to 0.93), and its psychometric quality was repli-
cated in a representative sample of healthy men and women in Germany (Siegrist et
al., 2000). Again, we constructed a binary measure with scores in the upper tertile
defining a high level of negative affectivity.
Musculoskeletal Symptoms. A standardized, widely tested questionnaire
measuring the frequency of musculoskeletal pain at different sites of the body was
administered (Berger-Schmitt et al., 1996). This questionnaire was applied in sev-
eral representative health surveys in Germany, thus providing reference informa-
tion (Berger-Schmitt et al., 1996). Participants were asked to assess in which part of
the body (neck, shoulder, elbow, hand, back, hip, knee, foot) they had experienced
recurrent pain during the previous 12 months and again during the previous 7 days.
In this study, we relied on data on the previous 12 months because this information
may describe chronic musculoskeletal pain more accurately. We dichotomized re-
spective variables with the category “yes” defining presence and “no” defining ab-
sence of musculoskeletal symptoms.
Data Analysis
All analyses were conducted on a personal computer with SPSS Version 9.0 (SPSS
Institute, Chicago, IL). First, univariate and bivariate analyses were performed.
Student’s t test in case of normal distribution or Mann–Whitney U test was applied

128
JOKSIMOVIC ET AL.
for continuous data, and a chi-square test was applied in case of categorical vari-
ables to characterize the two groups of interest (musculoskeletal pain present vs.
absent). To test the hypothesis of associations of three indicators of perceived work
stress (physical load, low control, high ratio of effort–reward imbalance) and of
overcommitment with musculoskeletal symptoms, we performed multivariate lo-
gistic regression analyses. In the first model, the four predicting variables were in-
cluded, they were adjusted for each other, and respective odds ratios of
musculoskeletal symptoms at different sites were calculated. In this model, odds
ratios were adjusted for age and sex. In the second model, education and household
income, as indicators of socioeconomic status, and shift work were introduced as
confounders. The final model included negative affectivity as a measure of report-
ing bias. With one exception to be discussed, no significant changes in the odds ra-
tios were observed during this final step. Results therefore are given for Models 1
and 2. Multivariate odds ratios and 95% confidence intervals are displayed. In view
of the relatively small sample size, a level of significance p < .05 was considered
appropriate.
RESULTS
Table1describesthesampleintermsofperceivedworkstressvariables.Inthissam-
ple of employees of an urban transport enterprise, 33.4% reported physical job de-
mand,27.4%fromlowcontrolatworkand15.1%fromanimbalancebetweeneffort
and reward. Thirty-six percent scored high on overcommitment and 30.6% scored
high on negative affectivity. Moderate correlations were found between negative
affectivity and overcommitment (r = .33) and between negative affectivity and ef-
fort–reward ratio (r = .35), thus indicating no substantial overlap.
The 1-year prevalence of musculoskeletal symptoms was relatively high: 70.4%
for back pain, 60.1% for neck pain, and 52.9% for shoulder pain. Symptoms related
to hip (24.3%) and to upper (36.2%) and lower extremities (46.2%) were reported
less frequently. It should be kept in mind that these figures include all participants
who experienced recurrent pain at least once during the past 12 months. If restricted
TABLE 1
Perceived Work Stress Variables in 316 Employees of a Traffic Company
Variable
Items
Rating
Range
M
SD
Effort
Reward
Overcommitment
Physical work demand
Control
5
11
6
1
3
1–5
1–5
1–4
1–5
1–5
5–25
11–54
6–24
1–5
12.48
24.25
12.27
1.67
3.93
8.94
3.75
1.07
2.38
3–14
4.72

129

130
JOKSIMOVIC ET AL.
to pain experienced during the past 7 days, prevalence rates are substantially lower.
Table 2 shows results of bivariate analyses between the six confounding variables
and different sites of musculoskeletal symptoms. Due to the relatively homogenous
sampleintermsofage,nosubstantialage-relatedeffectswereobserved.Concerning
sex, women more often reported neck pain and pain in upper extremities compared
with men. The two indicators of socioeconomic status were not consistently related
to musculoskeletal symptoms, although there was a tendency toward more symp-
toms in the less educated group. Shift workers (i.e., bus and subway drivers and re-
pairpersonnel)reportedhigherlevelsofpain,inparticularneck,shoulder,andupper
extremitypain. Asexpected, ahighlevelofnegativeaffectivitywasconsistentlyas-
sociated with prevalence of reporting musculoskeletal pain.
Table3indicatesfindingsforthefourpredictingvariablesonwhichthetestofthe
hypothesiswasbased.Ascanbeseen,eachindicatorofperceivedworkstressshows
some association with musculoskeletal pain, but differences in general are not sub-
stantial. Yet, the personal coping pattern of overcommitment clearly is related to
symptom reporting, because differences are marked in four of six conditions of
musculoskeletal pain.
To test the hypothesis, results of logistic regression analyses based on the two
models are given in Table 4. Odds ratios and confidence intervals refer to expo-
sure status, that is, high physical job demand, low control, high ratio of ef-
fort–reward imbalance, and high level of overcommitment. Reference categories
with an odds ratio of 1.0 were low physical job demand, high control, low ratio
of effort–reward imbalance, and low level of overcommitment. In Model 1, odds
ratios of the four predicting variables were adjusted for each other and, in addi-
tion, for age and sex. Model 2 introduced education, household income, and
shift work as additional confounders. As can be seen, overcommitment was
more strongly and more consistently associated with musculoskeletal pain, com-
pared with the three indicators of perceived work stress. Further adjustment for
negative affectivity, although not affecting all remaining results, slightly re-
duced the odds ratio of overcommitment for lower extremities from 1.93 to 1.88
(confidence interval = .98 – 3.62; p = .057).
DISCUSSION
This study provides preliminary evidence of an association of indicators of per-
ceived work stress and of a personal pattern of coping with work demands,
overcommitment, with musculoskeletal symptoms at different bodily sites in a
sample of middle-age public transport employees. Associations were relatively
strongest for overcommitment, where the risk of experiencing recurrent neck and
hip pain clearly was elevated. Physical job demand and, to a lesser extent, low con-
trol and effort–reward imbalance also were related to the experience of

131

132

WORK-STRESS AND MUSCULOSKELETAL PAINWORK
133
musculoskeletal pain. Results were adjusted for the effects of age, sex, socioeco-
nomic status, and occupational group (shift work). Moreover, a measure of nega-
tive affectivity was introduced given the challenge of common method variance in
a cross-sectional study design. Data were obtained from standardized,
psychometrically validated questionnaires and were collected under standardized
conditions. Information obtained through these measures allowed us to distinguish
between (a) physical and psychosocial work stressors, (b) perceived situational
stressors related to the work environment and a personal coping characteristic, and
(c) different bodily sites of experienced musculoskeletal pain.
To our knowledge, no other study has analyzed the two components of this
newly developed model of psychosocial stress at work, the ratio of effort–reward
imbalance and overcommitment, in relation to musculoskeletal pain. Our findings
are in line with previous evidence documenting an important role of personal cop-
ing characteristics in explaining musculoskeletal pain, in addition to work envi-
ronment-related stressors (Eriksen & Ursin, 1999). Although it is possible that
musculoskeletal complaints affect the level of overcommitment, this is rather un-
likely given a relatively high interpersonal stability over time of this pattern
(Siegrist, 1996).
In physiological terms, musculoskeletal pain produced by overcommitment
can be explained as a result of recurrent activation of low-threshold motor units in
specific muscles after sustained activity of the autonomic nervous system
(Lundberg, 1999; Westgaard, Holte, & Vasseljen, 2000). Previous studies showed
that the more central regions of the body, especially neck and shoulder, are more
vulnerable to enhanced tension and its adverse effects on musculoskeletal symp-
toms (Theorell, Harms-Ringdahl, Ahlberg-Hulten, & Westin, 1991; Toomingas,
Theorell, Michélsen, & Nordemar, 1997; Westgaard, Jensen, & Hansen, 1993). In
this context, it is interesting to note that the two components of the work stress
model of effort–reward imbalance also are associated with musculoskeletal pain in
more central bodily regions, such as neck and hip regions. Although it seems plau-
sible to attribute hip pain to osteoarthritis, there is independent evidence of an as-
sociation of psychosocial factors with hip pain (Wolfe, 1999).
Several limitations of this investigation must be mentioned. Because the study
design is not prospective, no causal inference can be drawn. Furthermore, no ob-
jective measure of muscle tension (e.g., derived from electromyographic record-
ings) was available in investigating a healthy population at the worksite.
Although we applied a well-tested standardized questionnaire that differenti-
ates between symptoms at different sites and in different time intervals, we were
not able to explore the dynamics of pain perception, symptom reporting, and cog-
nitive defense mechanisms that may be involved (Ursin & Eriksen, 2001). Thus,
our data on musculoskeletal symptoms must be considered rather crude approxi-
mate measures of more varied and more subtle processes resulting in subjective
health complaints. However, there is some indication of the validity of the mea-

134
JOKSIMOVIC ET AL.
sures of musculoskeletal pain used in this study. For instance, the prevalence rates
that we observed were in accordance with the ones reported in previous investiga-
tions (Eriksen et al., 1999; Foppa & Noack, 1996; Magnusson et al., 1996;
Riihimäki, 1999). In addition, our results are consistent with previously reported
findings concerning associations of musculoskeletal symptoms with physical
workload (Ariëns et al., 2000), low control at work (Theorell et al., 1991), and ad-
verse work conditions such as shift work (Kleiven, Boggild, & Jeppesen, 1998;
Koda et al., 1991; Ueda et al., 1989). Indirect evidence of the validity of the find-
ings is given by the fact that psychosocial work stress as measured in this approach
was associated with additional indicators of self-reported health complaints, such
as depressive mood, general subjective health status, or sleep disturbances (results
not shown in detail).
Because perceived work stress was measured only once, no inference on ex-
posure over time can be made. Moreover, our measure of physical work demand
was limited. This may have resulted in an underestimation of its contribution to-
ward explaining musculoskeletal pain. Moreover, we did not test the full job de-
mand–control model (Karasek et al., 1998) but restricted our measure to the
component of control or decision latitude. This decision was made in view of
some conceptual overlap between the two respective measures: the component
of effort of the effort–reward imbalance model and the component of demand in
the demand–control model.
A further limitation of this study is that the sample was limited in size (in par-
ticular, few employed women), and a selection bias cannot be ruled out. As
mentioned, main reasons for participation were interest in supporting a scientific
study, financial incentive, and personal concern about work and health. Even if
we assume that this latter reason was the main driving force to participate, two
opposite effects on results produced by this selection bias could be expected. On
one hand, this may have contributed to an overestimation of the hypothesized ef-
fect, but on the other hand, the really exhausted and stressed employees and
those with very busy schedules may have been underrepresented in the study
sample.
In view of these limitations, further studies are needed to validate these find-
ings. Larger samples would allow more complex statistical analyses such as tests
of interaction terms between predicting variables, in particular the effort–reward
imbalance ratio and overcommitment or interaction terms between psychosocial
and physical workload. A better control of possible Type 1 error is needed as well
as demonstration of effects in a prospective study design. Meanwhile, our group
has started a much larger prospectively designed investigation in this important
field of inquiry.
Despite the limitations discussed, this study documents associations of per-
ceived work stress and the coping pattern overcommitment with musculoskeletal
symptoms at different sites in an otherwise healthy, middle-age, employed popula-

WORK-STRESS AND MUSCULOSKELETAL PAINWORK
135
tion. In view of an important role of musculoskeletal pain in absenteeism, our re-
sults underline the importance of implementing stress-reducing measures at work,
at both the behavioral and structural level.
REFERENCES
Abraham, R. (1999). Negative affectivity: Moderator or confound in emotional dissonance-outcome re-
lationship? Journal of Psychology, 133, 61–72.
Ariëns, G. M., van Mechelen, W., Bongers, P. M., Bouter, L. M., & van der Wal, G. (2000). Physical risk
factors for neck pain. Scandinavian Journal of Work, Environment and Health, 26, 7–19.
Berger-Schmitt, R., Kohlmann, T., &Raspe, H. (1996). Rückenschmerzenin Ost-und Westdeutschland
[Neck pain in East and West Germany]. Gesundheitswesen, 58, 519–524.
Bernard, B. P. (1997). Musculoskeletal disorders and work place factors (DHHS [NIOSH] Publication
No. 97–141). Cincinnati, OH: U.S. Department of Health and Human Services.
Bongers, P. M., de Winter, C. R., Kompier, M. A., & Hildebrandt, V. H. (1993). Psychosocial factors at
work and musculoskeletal disease. Scandinavian Journal of Work, Environment and Health, 19,
297–312.
Bovim, G., Schrader, H., &Sand, T. (1994). Neckpaininthegeneralpopulation. Spine, 19, 1307–1309.
Brogmus, G. E., Sorock, G. S., & Webster, B. S. (1996). Recent trends in work-related cumulative
trauma disorders of the upper extremities in the United States: An evaluation of possible reasons.
Journal of Occupational and Environmental Medicine, 38, 401–411.
Burke, M. J., Brief, A. P., & George, J. M. (1993). The role of negative affectivity in understanding the
relations between self-reports of stressors and strains: A comment on the applied psychology litera-
ture. Journal of Applied Psychology, 78, 402–412.
Cole, D. C., & Hudak, P. L. (1996). Prognosis of nonspecific work-related musculoskeletal disorders of
the neck and upper extremity. American Journal of Industrial Medicine, 29, 657–668.
Eriksen, W., Natvig, B., Knardahl, S., & Bruusgaard, D. (1999). Job characteristics as predictors of neck
pain: A 4-year prospective study. Journal of Occupational and Environmental Medicine, 41,
893–902.
Eriksen, H., & Ursin, H. (1999). Subjective health complaints: Is coping more important than control?
Work & Stress, 13, 238–252.
Foppa, I., &Noack, R. H. (1996). Therelationofselfreportedbackpaintopsychosocial, behavioral, and
health-related factors in a working population in Switzerland. Social Science and Medicine, 43,
1119–1126.
Hagberg, M., Silverstein, B., Wells, R., Smith, M. J., Hendrick, H. W., Carajon, P., et al. (1995).
Work-related musculoskeletal disorders (WMSDs): A reference book for prevention. London:
Tayler & Francis.
Hollmann, S., Heuer, H., & Schmidt, K-H. (2001). Control at work: A generalized resource factor for the
prevention of musculoskeletal symptoms? Work & Stress, 15, 29–39.
Houtman, I. L. D., Bongers, P. M., Smulders, P. G. W., & Kompier, M. A. J. (1994). Psychosocial stress-
ors at work and musculoskeletal problems. Scandinavian Journal of Work, Environment and
Health, 20, 139–145.
Karasek, R., Brisson, C., Kawakami, N., Houtman, I., Bongers, P., &Amick, B. (1998). TheJobContent
Questionnaire:Aninstrumentforinternationally comparativeassessments ofpsychosocial jobchar-
acteristics. Journal of Occupational Health Psychology, 3, 322–355.
Karasek, R., & Theorell, T. (1990). Healthy work. New York: Basic Books.
Kleiven, M., Boggild, H., & Jeppesen, H. J. (1998). Shift work and sick leave. Scandinavian Journal of
Work, Environment and Health, 24(Suppl. 3), 128–133.

136
JOKSIMOVIC ET AL.
Koda, S., Hisashige, A., Ogawa, T., Kurumatani, N., Dejima, M., Miyakita, T., et al. (1991). An epide-
miological study on low back pain and occupational risk factors among clinical nurses. Sangyo
Igaku, 33, 410–422.
Lundberg, U. (1999). Stress responses in low-status jobs and their relationship to health risks:
Musculoskeletal disorders. In N. E. Adler, M. Marmot, B. S. McEwen, & J. Stewart (Eds.), Socio-
economic status and health in industrial nations: Social, psychological, and biological pathways
(pp. 162–172). New York: New York Academy of Sciences.
Magnavita, N., Bevilacqua, L., Mirk, P., Fileni, A., & Castelino, N. (1999). Work-related
musculoskeletal complaints in sonologist. Journal of Occupational and Environmental Medicine,
41, 981–988.
Magnusson, M. L., Pope, M. H, Wilder, D. G., & Areskoug B. (1996). Are occupational drivers at an in-
creased risk for developing musculoskeletal disorders? Spine, 21, 710–717.
Morano, J. (1993). The relationship of workplace social support to perceived work-related stress among
staff nurses. Journal of Post Anesthesia Nursing, 8, 395–402.
Niedhammer, I., Siegrist, J., Landre, M. F., Goldberg M., & Leclerc A. (2000). Étudea des qualités
psychométriques de la version francaise du modèle du Déséquilibre Efforts/Récompenses
[Psychometric properties of the French version of the effort–reward questionnaire]. Révue
d’Épidémiologie et de Santé Publique, 48, 419–437.
Peter, R., Alfredsson, L., Hammar, N., Siegrist, J., Theorell, T., & Westerholm, P. (1998). High effort,
low reward and cardiovascular risk factors in employed Swedish men and women–Baseline results
from the WOLF-study. Journal of Epidemiology and Community Health, 52, 540–547.
Riihimäki, H. (1999). Musculoskeletal diseases–A continuing challenge for epidemiologic research.
Scandinavian Journal of Work, Environment and Health, 25(Suppl. 4), 31–35.
Schnall, P., Belkic K., Landsbergis P., & Baker D. (2000). The workplace and cardiovascular disease.
Philadelphia: Hanley & Belfus.
Siegrist, J. (1996). Adverse health effects of high effort/low reward conditions. Journal of Occupational
Health Psychology, 1, 27–41.
Siegrist, J., Starke, D., Laubach, W., & Brähler, E. (2000). Soziale Lage und gesundheitsbezogene
Lebensqualität: Befragungsergebnisse einer repräsentativen Stichprobe der deutschen Bev”lkerung
[Social status and health related quality of life: Results of a survey in a representative German sam-
ple]. In M. Bullinger, J. Siegrist, and U. Ravens-Sieberer (Eds.), Lebensqualitätsforschung aus
medizinpsychologischer und soziologischer Perspektive [Research of quality of life from a medical
psychological and medical sociological view] (pp. 307–319). Göttingen: Hogrefe.
Siegrist, J. (2001). A theory of occupational stress. In J. Dunham (Ed.), Stress in the workplace: Past,
present and future (52–66). London: Whurr.
Starke, D., Joksimovic, L., Knesebeck, v.d. O., Peter, R., Niedhammer, E., Siegrist, J. (2002). Die
Entwicklung einer Kurzversion der intrinsischen Modellkomponente Overcommitment. [Develop-
ment of short version of the intrinsic component overcommitment]. Manuscript submitted for publi-
cation.
Theorell, T., Harms-Ringdahl, K., Ahlberg-Hulten, G., & Westin, B. (1991). Psychosocial job factors
and symptoms from the locomotor system—A multicausal analysis. Scandinavian Journal of Reha-
bilitation Medicine, 23, 165–173.
Toomingas, A., Theorell, T., Michélsen, H., & Nordemar, R. (1997). Stockholm MUSIC I Study Group.
Associations between self-rated psychosocial work conditions and musculoskeletal symptoms and
signs. Scandinavian Journal of Work, Environment and Health, 23, 130–139.
Ueda, T., Hashimoto, M., Yasui, I., Sunaga, M., Higashida, T., & Hara, I. (1989). A questionnaire study
onhealthoftaxidrivers—Relationstoworkconditionsanddailylife. Sungyo-Igaku, 31, 162–175.
Ursin, H., & Eriksen, H. (2001). The cognitive activation theory of stress. Manuscript submitted for pub-
lication.

WORK-STRESS AND MUSCULOSKELETAL PAINWORK
137
Walsh, K., Varnes, N., Osmond, C., Styles, R., & Coggon, D. (1989). Occupational causes of low-back
pain. Scandinavian Journal of Work, Environment and Health, 15, 54–59.
Westgaard, R. H., Holte K. A., & Vasseljen O. (2000). Shoulder pain and work/leisure time exposures of
female workers in the service industry [Abstract]. International Journal of Behavioral Medicine,
7(Suppl. 1), 169–170.
Westgaard, R. H., Jensen, C., & Hansen, K. (1993). Individual and work-related risk factors associated
with symptoms of musculoskeletal complaints. International Archives of Occupational and Envi-
ronmental Health, 64, 405–413.
Wolfe, F. (1999). Determinants of Womac function, pain and stiffness scores: Evidence for the role of
low back pain, symptom counts, fatigue and depression in osteoarthritis, rheumatoid arthritis and
fibromyalgia. Rheumatology (Oxford), 38, 355–361.

138
JOKSIMOVIC ET AL.
APPENDIX
Short Version of the
Overcommitment Questionnaire