Physical, psychosocial, and individual risk factors for neck/shoulder pain with pressure tenderness in the muscles among workers performing monotonous, repetitive work.
ABSTRACT Cross-sectional study.
To evaluate the effect of individual characteristics and physical and psychosocial workplace factors on neck/shoulder pain with pressure tenderness in the muscles.
Controversy prevails about the importance of workplace factors versus individual factors in the etiology of pain in the neck and/or shoulders.
Study participants were 3123 workers from 19 plants. Physical risk factors were evaluated via video observations, and psychosocial risk factors were assessed with the job content questionnaire. Other procedures included symptom survey, clinical examination, and assessment of health-related quality of life (SF-36). The main outcome variable, neck/shoulder pain with pressure tenderness, was defined on the basis of subjective pain score and pressure tenderness in muscles of the neck/shoulder region.
The prevalence of neck/shoulder pain with pressure tenderness was 7.0% among participants performing repetitive work and 3.8% among the referents. We found an association with high repetitiveness (prevalence ratio 1.8, 95% confidence interval 1.1-2.9), high force (2.0, 1.2-3.3), and high repetitiveness and high force (2.3, 1.4-4.0). The strongest work-related psychosocial risk was high job demands (1.8, 1.2-2.7). Increased risk was also associated with neck/shoulder injury (2.6, 1.6-4.1), female gender (1.8, 1.2-2.8), and low pressure pain threshold (1.6, 1.1-2.3). Neck/shoulder pain was strongly associated with reduced health-related quality of life.
Work-related physical and psychosocial factors, as well as several individual risk factors, are important in the understanding of neck/shoulder pain. The findings suggest that neck/shoulder pain has a multifactorial nature. Reduced health-related quality of life is associated with subjective pain and clinical signs from the neck and shoulders. The physical workplace factors were highly intercorrelated, and so the effect of individual physical exposures could only be disentangled to a minor degree.
- SourceAvailable from: Chang-Zern Hong[show abstract] [hide abstract]
ABSTRACT: To review recent clinical and basic science studies on myofascial trigger points (MTrPs) to facilitate a better understanding of the mechanism of an MTrP. English literature in the last 15 years regarding scientific investigations on MTrPs in either humans or animals. Research works, especially electrophysiologic studies, related to the pathophysiology of MTrP. (1) Studies on an animal model have found that a myofascial trigger spot (MTrS) in a taut band of rabbit skeletal muscle fibers is similar to a human MTrP in many aspects. (2) An MTrP or an MTrS contains multiple minute loci that are closely related to nerve fibers and motor endplates. (3) Both referred pain and local twitch response (characteristics of MTrPs) are related to the spinal cord mechanism. (4) The taut band of skeletal muscle fibers (which contains an MTrP or an MTrS in the endplate zone) is probably related to excessive release of acetylcholine in abnormal endplates. The pathogenesis of an MTrP appears to be related to integrative mechanisms in the spinal cord in response to sensitized nerve fibers associated with abnormal endplates.Archives of Physical Medicine and Rehabilitation 08/1998; 79(7):863-72. · 2.36 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: Our aim was to determine the predictive value of some psychometric instruments for the development and persistence of musculoskeletal pain. In 452 subjects, pain in the shoulder, neck, and low back during the preceding year was assessed at baseline, one year, and two years' follow-up. Psychological distress, depression, self-efficacy beliefs, subjective work prognosis, disability, and work characteristics were assessed at baseline. The best predictor of future pain was disability. The psychometric measures did not predict changes in pain. The explanatory power of the variables in the multivariate analyses was low. Perceived disability in persons with musculoskeletal symptoms should be acknowledged early. The traditional research paradigm focusing on a few hypothetically relevant variables does not take into account the reflective, dynamic, reciprocal nature of human behavior during the process of recovery from or development of a pain problem.Journal of Occupational and Environmental Medicine 06/1998; 40(5):445-53. · 1.85 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: The purpose of this study was to determine aspects of validity of a short self-report inventory, based on physical and psychological symptoms that, according to other studies, are most prevalent in women. It was hypothesized that such symptoms form a single entity, which would be consistent with the view that they represent a proxy for general distress in women. Method: One hundred and one women were approached, all of whom filled in the "common symptoms in the general population of women" (CSGP) instrument and the multidimensional self-report inventory Symptom Check List 90 (SCL90) for external comparison. The results indicate that the CSGP scale has high internal consistency (alpha = 0.80). The strongest correlations were found between the CSGP scale and the SCL90 subscales assessing "somatization", "depression" and "anxiety". In the factor analysis, the items on the CSGP scale were widely dispersed among the factors represented by the main SCL90 subscales, and therefore could not be considered to represent any particular subscale. Our conclusion is that the CSGP scale seems to assess a single entity. Other researchers suggest that such an entity represent a proxy for general distress in women. This will have implications for the interpretation of the common symptoms represented by the scale, which women report in clinical settings.Scandinavian Journal of Public Health 01/2000; 27(4):311-7. · 1.97 Impact Factor
SPINE Yolume 27, Numbcr 6, pr 660-667
©2002, Lirrineorr Williams & Wilkins, Ine.
Physical, Psychosocial, and Individual Risk Factors for
NeckJShoulder Pain With Pressure Tenderness in the
Muscles Among Workers Performing Monotonous,
Johan Hviid Andersen, PhD,* Anette Kaergaard, PhD,* Poul Frost. PhD,*
Jane Frølund Thomsen, PhD,t Jens Peter Bonde, DMedSc,§ Nils Fallentin, PhD,t
Vilhelm Borg, MSc,t and Sigurd Mikkelsen, DMedSct
Study Design. Cross-sectional study.
Objectives. To evaluate the effect of individual charac-
teristics and physical and psychosocial workplace factors
on neckJshoulder pain with pressure tenderness in the
Summary of Background Data. Controversy prevails
about the importance of workplace factors versus individ-
ual factors in the etiology of pain in the neck and/or
Methods. Study participants were 3123 workers from
19 plants. Physical risk factors were evaluated via video
observations, and psychosocial risk factors were as-
sessed with the job content questionnaire. Other proce-
dures included symptom survey, ciinical examination,
and assessment of health-related quality of life (SF-36).
The main outcome variable, neckJshoulder pain with
pressure tenderness, was defined on the basis of subjec-
tive pain score and pressure tenderness in muscles of the
Results. The prevalence of neckJshoulder pain with
pressure tenderness was 7.0% among participants per-
forming repetitive work and 3.8% among the referents.
We found an association with high repetitiveness (preva-
lence ratio 1.8, 95% confidence interval. 1.1-2.9), high
force (2.0, 1.2-3.3), and high repetitiveness and high force
(2.3, 1.4-4.0). The strongest work-related psychosocial
risk was high job demands (1.8, 1.2-2.7). Increased risk
was also associated with neckJshoulder injury (2.6, 1.6-
4.1), female gender (1.8, 1.2-2.8), and low pressure pain
threshold (1.6,1.1-2.3). NeckJshoulder pain was strongly
associated with reduced healthcrelated quality of Iife.
Conclusions. Work-related physical and psychosocial
factors, as well as several individual risk factors, are im-
portant in the understanding of neckJshoulder pain. The
findings suggest that neckJshoulder pain has a multifac-
torial nature. Reduced health-related quality of Iife is as-
sociated with subjective pain and ciinical signs from the
neck and shoulders. The physical workplace factors were
highly intercorrelated, and so the effect of individual
physical exposures could only be disentangled to a minor
From the "Deparrment of Occupational lv1edicine, Herning Hospital,
Herning, the tDeparrmenr of Occupational lv1edicine, Copenhagen
Counry Hospital, Glostrup, the tNational Institute of Occupational
Health, Copenhagen, and the §Deparrment of Occupational lv1edicine,
Aarhus University Hospital, Aarhus, Denmark.
Sllpporred by grams from the Danish Working Environmenr Fund and
the Danish lv1edical Research COllncil.
Acknowledgmenr date: Januaty 4, 2001.
Revision date: May 7, 2001.
Acceptance date: July 2, 200l.
Device starus category: l.
Conflict of inrerest category: 14.
degree. [Key words: objective assessment of physical job
exposure, PRIM study, pressure tenderness in the mus-
cles,ciinicalexamination of neckJshoulder muscles]
Muscle pain in the neck and shoulder girdle is a common
complaint.9,33,35,48 There are no generally accepted cri-
teria for classification. In occupational medicine um-
brella terms, such as repetitive strain injury (RSI), cumu-
lative trauma disorder (CTD), occupational
cervicobrachial disorder (OCD), or work-related muscu-
loskeletal disorder (WMSD), have been used.6,19 From a
population-based perspective, others have argued that
these are misleading terms, implying a single uniform
etiology, and that they should be avoided.34No specific
pathoanatomic or pathophysiologic mechanisms have as
yet been uncovered, and opinions on muscle pain range
from pure skepticism8,17,47 to discussions about a neu-
ropathic origin16 or a nociceptive dysfunction.20Aside
from the difficulties in defining health effects on the basis
of muscle pathology, defining occupational exposure
also renders problems. Most occupational epidemiologic
studies are based on comparisons between workers in
heavy industrial occupations and workers in light indus-
try or office environments,1,2,4,5,12,18,25,26,37,45 and most
studies are heterogeneous with regard to both the expo-
sure measures and the outcome measures.46Three main
dimensions on which physical exposure may be quanti-
tatively measured have been suggested: level or magni-
tude of force, repetitiveness, and duration of exposure.44
However, most studies provide only scarce information
on these dimensions. The role of psychosocial factors as
contributors to musculoskeletal disorders is uncertain,
but several studies suggest a positive association between
neck pain and high job demands, low social support, low
job control, and low job satisfaction.3,32 Prospective
studies based on questionnaires have found that disabil-
ity, bur not workload,14 have little influence on one's
own work situation, but not repetitiveness,13 and psy-
chological distress and psychosomatic problems, but not
occupa1:on,3° were predictors of neck pain. Population-
based studies have shown that tender points in the mus-
NecklShoulder Pain Risk Factors' Andersen et al
Table 1. Industries, Recruited Participants, Number of Women, Mean Age, Number of Referents, and Mean Value of
N/No,No, (%) of
No, (%) of
% Time Without
% Time With
3 Plants (women's undervvear,
children's wear, and
Plastic and paper production
Production of life-saving
2 Cardboard production
2 Postal-sorting offices
34.4 (11 ,2)
265 (98,9)38.4 (10.5)26 (9.7)16,2 (8,0-33.011.0 (0.6-2.0)
97 (60-100)90 (57-100)
510/342 (67,1)11 (3.2) 42 (9.5)116 (33.9)10.5 (1-34.0)1.5 (0.2-3.0)60 (7-100)44 (3-87)
4174/3123 (74.8) 1823 (58.4) 38.6 (10.7)788 (25.2)14.4 (0.4-36,0) 1.2 (0.1-4,0)
• Repetitivity is the number af shoulder movements per minute; force is force requirements fo,- the shoulder joint (0-4 an a 5-point ordinal seale), Lack af reeovery
time is the percentage af time within eyeles without micropauses for >2 seeonds, Neek flexion is the percentage af time with neek flexed >20°,
t Referents were blue-eollar workers at the plant with nonrepetitive work tasks,
cles are associated with psychological distress as well as
characteristics of somatizationy,36 No evidence exists
on ergonomic interventions for preventing neck pain.31
This article reports the tindings of a nationwide cross-
sectional study in which uniform methods were used to
assess exposure and health outcomes applied across a
variety of working populations. The objectives were 1)
to examine the risk for neck/shoulder pain with pressure
tenderness in relation to physical workplace factors, psy-
chosocial factors, and individual characteristics and 2) to
assess the level of health-related quality of life in relation
to neckJshoulder pain.
Recruitment af Subjects. A total of 4162 workers at 19
different workplaces were invited to participate in the Project
on Research and Intervention in Monotonous Work (PRIM)
health study in 1994 and 1995. All workers in production
facilities at these plants were invited to participate. Initial plant
selecrion aimed at obtaining a variety of repetitive work tasks
with large contrasts in exposure profiles. 15 The workplaces
included four food processing companies, three textile plants,
seven other manufacturing, and five service companies.
Type of industry and baseline characteristics are shown in
Table 1. Approval was obtained from the local research ethics
committee and participants gave written informed consent.
Physicaf Exposure Quantification. Task-related exposure
groups were established afrer repeated visits to the 19 factories,
extensive walk-throughs, and reviews of existing job data. All
repetitive work tasks were evaluated, and on the basis of sim-
ilarity in physical exposure level five or six task-related expo-
sure groups were established at each of the 19 factories, giving
a total of 103 exposure groups. The number of work tasks
inc!uded in an exposure group varied from one to 18. A total of
300 repetitive work tasks were evaluated, and the median num-
ber of tasks per exposure group was two. Workers with non-
repetitive tasks (e.g., varied office work, internal transporta-
tion, blue-collar supervisory work, control-room work, and
maintenance work) were included as an unexposed reference
group. All workers included were unskilled blue-collar work-
ers, and the reference group did not differ from the group with
repetitive work on educationallevel or salary.
Data concerning the magnirude or intensiry and time varia-
tion (repetitiveness) of exposure were obtained using a video-
based observation method. Random samples of workers were
selecred from each ofthe 103 exposure groups and a median of
three observations for work task conducted.15Duration ofex-
posure and time spent performing tasks were provided by
workers' self-reports, and estimates of cumulative physical ex-
posure over the workday were then obtained by allocating the
exposure assessment for the different tasks (based on group
means) to the proportional distribution of work tasks over the
whole day. A detailed description of the exposure assessment
strategy and some aspecrs ofthe reliability have been published
The observed physical exposures inc!uded number of shoul-
der movements/minute (summed median values from the indi-
vidual participants' reports of time spent in different exposure
groups), percentage of time with neck flexion >20°, and per-
centage of cycle time spent with no upper arm support or rest
662Spine· Volume 27· Number 6·2002
for>2 seconds (Iack of shoulder recovery time). Force require-
ments were subjectively assessed and computed by the observer
using S-point ordinal scales (0-4), relative to maximum
voluntary contraction (MVC, <10%, 10-29%, 30-49%,
50-79%, and 2:80% MVC). In a substudy of the PRIM co-
hort, observer-estimared force seerned to correspond well with
the mean peak EMG activity.22
1Vlean values for the four exposure values and ranges are
shown for the 14 different types of industry or service in
Table 1. The four ergonomic exposure values turned out to
be highly correlated, and correlation coefficients were as
follows: repetitiveness/force = 0.47, repetitiveness/neck
flexion = 0.59, repetitiveness/shoulder recovery time =
0.72, force/necl< flexion = 0.47, force/recovery time = 0.59,
and neck flexion/recovery time = 0.84. The values for each
exposure were divided into three levels on the basis of score
distributions: the reference group was assigned the value of O,
and the repetitive group was assigned 1 (Iow) or 2 (high): re-
petitiveness (l-IS movements per minutel16-40 movements
per minute), force «
10% of MVC/2:10% of MVC), neck
flexion >20° «66% of time/2:66% of time), and lack of re-
covery time «80% of time/2:80% of time).
Psychosocial Work Characteristics. Psychosocial risk fac-
tors were assessed using the Whitehall II version of Karasek's
job content questionnaire.27The answer to each item was di-
chotomized and given a raw score of 1 or O, and three scales
were constructed as raw score summations: job demands (3
items, Cronbach's alpha = 0.53), job control (14 items, Cron-
bach's alpha = 0.72), and social support from colleagues and
supervisors (6 items, Combat's alpha = 0.53). Each scale was
dichotomized: high demand = high score on rwo items, low
control = high score on five items, and low social support =
high score on three items).
Individual Characteristics. Data on several individual char-
acteristics, including age and gender (Table 1), were collected.
Leisure time activity was categorized into none or light physical
activity (according to none or light physical activiry for <2 .
hours/week or light activity for 2-4 hours/week) and physically
active (light physical activity for >4 hours/week or 2-4 hours
with hard physical activiry or hard physical activiry for >4
Self-reported infiammatory rheumatic disease (yes/no) was
recorded. Slenderness was calculated as the ratio of shoulder
width and height. Those scoring below the 25% percentile
were c1assified as slender. Body mass index was calculated from
measured weight and height. Pressure pain threshold was mea-
sured by means of an Algometer (Somedic, Stockholm, Swe-
den). Pressure was applied with an increased rate of50 kPa S-l
through a circular rubber-coated pressure head (area 1 cm2).
The mean values of measures on the tibia and vastus medialis
muscle were thought to give an expression ofthe person's over-
all pain threshold, independent of neck/shoulder pain with
pressure tenderness. These values were dichotomized for the
purpose of analysis using the 25'h percentile as the cutoff point
«411 kPa for women and <665 kPa for men).
Neck/shoulder injury (no/yes) was derived from the ques-
tion: Have you ever had long-Iasting disorders in the neck
and/or shoulders after an accident?
Personality traits were measured by the dimension"intrinsic
effort" from Siegrist's effort-reward model.38 The 29 items
from the four subscales, "need for approval," "competitive-
ness," "disproportionate irritability," and "inability to with-
draw from work," were summed up to an index and dichoto-
mized into high versus low intrinsic effort on the basis of the
25thpercentile of the mean values.
Outcome Measure. Neck/shoulder pain was assessed by a
numerical box complaint scale from O(no complaints at all) to
9 (pain as bad as could bel, indicating, respectively, severity of
pain at worst, levelof average pain within rhe last 3 months,
severity of impairment in daily activiries within the last 3
months due to neck and/or shoulder pain, and levelof average
pain within the last 7 days.23 The scores were summed up,
giving a possibie range of 0-36 (Cronbach's alpha = 0.93).
On-site physical examination of the neck and upper extrem-
ities was performed by three teams of rwo or three physicians
according to a protocol.24 The examiners \vere blinded with
respect to the study participants' exposure and health status.
Palpation tenderness was scored on a 0-3-point scale and di-
chotomized into indisputable/severe palparion renderness
(score 2 or 3), with withdrawal or jump sign at the palpation,
and no or minor palpation tenderness (score Oor 1). The inter-
rarer reliability berween three examiners was tested in a sub-
group of 60 participants from one of the three centers. Multi-
rater kappa values ranging from 0.45 (SE 0.08) in right
infraspinatus to 0.57 (SE 0.06) in right trapezius were ob-
tained. Consistency bet\veen the three centers was assessed in-
directly by testing if the c1inical findings among exposed groups
and referents not reporting neck/shoulder pain were at the
same level in the three centers. There were no major or system-
atic differences bet\veen centers.
Our definition of neck!shoulder pain with pressure tender-
ness was as follows: pain score 2:12 in the neck or shoulder
combined with indisputable/severe palpation tenderness in the
neck muscles or upper trapezius border, and in the side-specific
supraspinatus or infraspinatus muscle.24The intention was to
identify persons who had substantial pain complaints and who
were tender at severallocations in the same region.
Health-related quality of life was assessed by the SF-36
health survey, covering eight health dimensions: physical func-
tioning, role physical functioning, bodily pain, general health,
vitality, social functioning, emotional role, and mental
Statisticai Analysis. The prevalence ratio for dominant neck!
shoulcler pain \vith pressure tenclerness was analyzed in rela-
tion to the generic physical risk factors: frequency of shoulder
movements, force requirements, percentage of movements with
neck fiexed > 20°, and percentage of time with continuous load
(Iack of recovery time). The combination of repetitiveness with
each of the three other generic risk factors was then analyzed.
Two by two analyses of force, neck flexion, and lack of recov-
ery time contained no more information than the generic fac-
tors alone, and the results are not shown here. All the analyses
of the physical risk factors were repeated without the reference
group to test for an internal exposure-response association in
the group with repetitive work. The crude associations were
adjusted by means of Cox proportional hazards model with a
constant risk period.39In the adjusted model all three dimen-
sions (physical exposure, psychosocial exposure [demand, con-
trol, and social support], and individual determinants) were
included in the model. We tested for interaction bet\veen de-
mand, control, and social support, but the interaction terms did
Neck/Shoulder Pain Risk Factors· Andersen et al663
Table 2. Physical Risk Factors for Neck/Shoulder Pain With Pressure Tenderness Among Industrial Workers
Physical Risk Factorn
Repetitiveness (shoulder movements/min)
Low (1-15 movements/minl
High (16-40 movements/min)
Low <10% of MVC
High 2:10% of MVC
Neck flexion (proportion of task cycle time
with neck flexed >20°)
Low «66% of time)
High (2:66% of time)
Lack of recovery time (proportion of task
cycle time without micropauses)
Low «80% of time)
High (2:80% of time)
Repetitiveness and force
Low repetitiveness and low force
High repetitiveness and low force
Low repetitiveness and high force
High repetitiveness and high force
Repetitiveness and neck flexion 2:20°
Low repetitiveness and low time
High repetitiveness and low time
Low repetitiveness and high time
High repetitiveness and high time
Repetitiveness and recovery time
Low repetitiveness and high recovery
Low repetitiveness and high recovery
Low repetitiveness and low recovery
Low repetitiveness and low recovery
PPR = prevalence proportion ratio.
• Adjustment for individual factors: age. gender. neck/shoulder injury, body mass index, intrinsic ef!ort, physical leisure time activity, slenderness, rheumatic
diseases, and psychosocial factors.
t Exciusion of nonexposed reference group. Same adjustments as with PPRadj.
*Reference category inciudes workers with nonrepetitive work tasks.
not contribute to the model, and only the main effects were kept
in the analysis. All other terms were retained in the models
independent of the magnitude of effect or levelof significance.
In the analysis of neck/shoulder pain and social functioning,
the population was divided into three grollps: those without
neck/shoulder pain, those with neck/sholllder pain without
pressure tenderness, and the those with neck/shoulder pain and
pressure tenderness according to the aforementioned criteria.
The SF-36 profile was scored using the original O-IOD-point
The overall particip'ltion rate was 74.8%. A total of 185
participants (6.2%) met our criteria for neck/shoulder
pain with pressure tenderness in the dominant neck/
shoulder region. In the group with monotonous repeti-
tive work the prevalence was 7.0%, and in the reference
group the prevalence was 3.8% (prevalence proportion
ratio [PPR] = 1.9; 95% confidence interval 1.3-2.9).
Increased risks of neck/shoulder pain with pressure ten-
derness were found for all the generic physical risk fac-
tors and the combinations with repetitiveness (Table 2).
The adjusted risks ranged from 1.7 for neck f1exion to
2.0 for force, but there was no significant contribution of
force requirements when the nonexposed reference
group was omitted (PPR = 0.9; 0.7-1.3). The combina-
tion ofrepetitiveness and force was the onlycombination
that revealed higher risk estimates than the generic fac-
tors alone with an eIevated risk of 2.3 (1.4-4.0). High
job demands and low job control were associated with
neck/shoulder pain, but social support was not (Table 3).
Several of the individual risk factors contributed signifi-
cantly to the regression model (Table 4). Women had an
elevated risk of 1.8. Low pain pressure threshold and
high intrinsic effort were both associated with neckl
shoulder pain with pressure tenderness. The strongest
risk factor in this anaIysis was neck/shoulder injury with
a prevalence ratio of 2.6 (1.6-4.1) in the fully adjusted
model. Neck/shoulder pain was not associated with age,
664Spine' Yolume 27 • Number 6 • 2002
Table 3. Psychosocial Risk Factors for Neck/Shoulder
Pain With Pressure Tenderness Among
PPR = prevalence proportion ratio; Cl
•.4djustment for all three psychosocial factors, physical factors and individual
factors: age, gender, neck/shoulder injury, body mass index, intrinsic effort,
physical leisure time activity, slenderness, and rheumatic diseases.
slenderness, body mass index, or rheumatic diseases. In
these data physical leisure activity at the levelof light
Dhysical activity for >4 hours/week or hard physical ac-
.ivity for 2-4 hours/week did not seem to be strongly
associated with a reduction in neck/shoulder pain.
The SF-36 O-lOa-point scoring produced the profile
shown in Table 5. The mean scores on all eight dimen-
sions of the SF-36 decreased significantly from the group
Table 4. Individual Risk Factors for Neck/Shoulder Pain
With Pressure Tenderness Among Industrial Workers
Individual Risk Factor
Pain pressure threshold
Body mass index
Physical leisure activity
None ar light physical
PPR = prevalence proportion ratio; Cl = confidence interval.
• Adjustment for physical workplace factors, psychosocial factors, and indi-
vidual factors: age, gender, neck/shoulder injury, body mass index, intrinsic
effort, physical leisure time activity, slenderness, and rheumatic diseases.
withoLlt neck/shoulder pain to the group with pain but
no pressure tenderness, and the mean scores were further
redllCed in the third group, who had neck/shoulder pain
and tenderness. The largest difference in mean score was
found for bodily pain (mean score difference 30.0), but
also one of the mental health components, vitality,
showed a difference of 21.8. Only a minor difference of
9.1 was seen for physical functioning in this working
Neck/shoulder pain with pressure tenderness was inde-
pendently related to work-related physical and psychos-
ocial factors, earlier experience of neck/shoulder injury,
female gender, low pain pressure threshold, and high
intrinsic effort. Increasing self-reports of pain and pres-
sure tenderness in the muscles were related to decreased
health-related qua!ity of !ife on all SF-36 scales.
A number of limitations to the study must be men-
tioned. A cross-sectional design inherently involves selec-
tion bias. Workers employed in the most physically de-
manding jobs may represent a survivor population, and
workers who previously had problems may have been
shifted to less demanding jobs. Both instances could re-
sult in an underestimation of the risle The participation
rate was 74.8%. The only available data on nonpartici-
pants were age and gender, which did not differ from the
participants. If workers with neck/shoulder pain were
more inclined to participate in the study, a bias toward
higher risk estimates could be the result as long as this
selection was stronger in the high level exposure groups.
There appeared, however, to be no tendency for a higher
participation among workers in the most strenuous jobs
This study benefits from uniform and objective mea-
surements of physical risk factors in a cohort that
stretches across job functions in different industrial and
service settings. Despite these variations, it was only pos-
sible to disenrangle the importance of the different phys-
icai exposure variables to a minor degree. All the generic
physical risk factors were associated with neck/shoulder
pain, and the only combination that carried an extra risk
was high force with high repetition. All other combina-
tions showed risk estimates at the same level as the sep-
arate generic risk factors alone. In real work situations,
repetitiveness, force requirements, posture, and lack of
recovery time were found to be strongly correlated with
each other. Attempts to examine physical workplace fac-
tors by dividing exposure into repetitiveness, force, static
work, and extreme postures may often lead to associa-
tions mainly based on the type of measurements, their
specificity, and the focus of the study.6
Assessment of exposures in this study was made on
the basis of 300 observations of 103 grouped repetitive
tasks. There are several assumptions behind this method.
First, the estimated values at task group level are repre-
sentative for all workers performing these tasks, and sec-
NeckJShoulder Pain Risk Factors' Andersen et al665-
Table 5. Health-Related Quality of ute Assessed by the SF-36 Among Participants Without Neck/Shoulder Pain, With
Neck/Shoulder Pain Without Pressure Tenderness, and Those With Neck/Shoulder Pain and Pressure Tenderness
Nec!<lShoulder Pain Without
(n = 1940-1992)
Neck/Shoulder Pain With
(n = 179-185)
Without Nec!<lShouJder Pain
(n = 730-751)
Cl = confidence interval.
ond, the observation period (about 10-15 minutes) is
representative of all time spent on that task. It is a limi-
tation that no actual measurements were made in the
task reference group. This group consisted of workers
who were mainly performing tasks that can be charac-
terized as unskilled, but more varied blue-collar work,
and the video-based observation method is not easily
applied to workers doing different kinds of jobs, moving
around, and changing their postures and functions. In
this study it was not possibie to evaluate the importance
of level and duration of shoulder elevation. Although it
was measured, variation in shoulder posture variables
between tasks did not exceed variation within tasks.15
The associations with high job demands and low job
control were in accordance with other studies.3We
found no evidence for the 'high strain' hypothesis in our
data. One of the difficulties in determining the relative
importance of psychosocial and physical risk factors is
that psychosocial factors are usually measured at the in-
dividuallevel, and physical risk factors are measured at a
task or group level. Studies that have not found an effect
of physical factors or occupation probably reilect a lack
af specificity in the measurement of physical risk
Women had a higher risk than men. Other studies
have also found a higher occurrence of neck/shoulder
pain among women.6,19 A Swedish study suggests that
common physical and psychological symptoms among
women are a proxy for general distress in women, which
could be the explanation for the remaining greater risk
among women after adjusting for workplace factors.28A
prospective follow-up study of forearm pain also found
that psychological distress and psychosocial and physical
workplace factors were related to the onset of nonspe-
cific forearm pain.34Our definition of neck/shoulder
pain with pressure tenderness in the muscles of the neck/
shoulder region was by no means thought of as a defini-
tion of a clinical entity. The prevalence rate of 6.2% in
our cohort is smaller than in most other studies of neck/
shoulder pain, in which the outcomes have most often
been tension neck syndrome,4,37 defined more broadly
from questionnaires alone, or in combination with tight-
ness of the muscles and/or tender spots. Prevalence rates
have been reported as high as 40%,37 56% for trapezius
myalgia,42 and 24% with tension neck syndrome in a
study ofelectronic workers.21Studies in which outcomes
were based purely on self-reports show higher preva-
lences. Many studies have relied on the assessment of
pain by the Nordic Questionnaire,29 in which most oc-
cupational groups have had prevalences of pain between
50% and 80% within the last 12 months. Using outcome
measures that are toa broad can influence the under-
standing of pain as well as coping in society as a whole.
Our use of pressure tenderness as part of the outcome
measure was intended as a way of verifying regional pain
status and severity. Tender points have been found to be
a general measure of distress, and pressure tenderness in
the muscles may well be a measure that samples elements
ofstress and distress,33,36 including that which emanates
from outside the workplace. Our strict controlling in the
analysis for pain pressure threshold in the lower extrem-
ities was introduced to adjust for individual distress, and
the results favored an independent effect ofpain pressure
The analysis of health dimensions from the SF-36 re-
vealed that more frequent complaints and clinical signs
were associated with lower scores on all aspects of
health-related quality of life, especially those subscales
indicating somatic iIIness. Participants with neck/
shoulder pain and pressure tenderness had scores com-
parable with those found in chronic somatic diseases.7
In conclusion, our study emphasizes that neck/shoulder
pain has a multidimensional nature.lO,33 This is in accor-
dance with viewpoints on fatigue and myalgia syndromes,
whichfavor a dimensional distribution ofmuscle pain.43In
thisstudypain was associated with health-related qualityof
life, and, as has been argued for simple backache,40 pain in
';6Spine· Volume 27· Number 6·2002
che neck and shoulder region should also be understood
within the framework ofa biopsychosocialmodel. From an
occupational health viewpoint, this calls for a much
broader perspective on prevention.
The authors thank Erik Overgaard, Kurt Rasmussen, Ja-
nus Kvamm, Birgit Juul Christensen, Jesper Bælum, and
Charlotte Brauer for helping to collect the data.
III Key Points
• This study examined the effects of joint exposure
to individual, physical, and psychosocial factors on
necklshoulder pain with pressure tenderness in a
• Physical job exposures were measured by a vid-
eo-based observational method.
• Neck/shoulder pain was associated with repeti-
tiveness, force, and neck posture, but even more so
with individual characteristics. Health-related
quality oflife was affected among participants with
severe pain and substantial palpation tenderness.
1. Andersen JH, Gaardboe O. Prevalence of persistenr neck and upper limb
pain in ahistorical cohort of sewing machine operators. Am J Ind Med
2. AndersenJH, Gaardboe O. Musculoskeletal disorders of rbe neck and upper
limb among sewing macbine operators: a elinicai investigation. Am J Ind
3. Ariens GAM, van Mechelen W, Bongers PM, er al. Psychosocial risk factors
for neek pain. Am J Ind Med 2001;39:180-93.
4. Bergqvist U, Wolgast E, Nilsson B, et al. Musculoskeletal disorders among
visual display terminal workers: individual, ergonomic, and work organiza-
tional facrors. Ergonomics 1995;38:763-76.
5. Bergqvist U, Wolgast E, Nilsson B, et al. The influence of VDT work on
musculoskeletal disorders. Ergonomics 1995;38:754-62.
6. Bernard B, ed. Musculoskeletal disorders and workplace factors: a critieal
review of epidemiological evidenee for work-related museuloskeletal disor-
ders of the neck, upper extremity, and low back (Publication no. 97-141).
Cincinnati, OH: Nationallnstitute ofOecupationalSafety and Health, 1998.
Bjorner JB, Tbunedborg K, Kristensen TS, er al. The Danish SF-36 Healrh
Survel': translation and preliminary validiry studies. J Clin Epidemiol1998;
8. Bohr T. Problems with myofascial pain syndrorne and fibromyalgia syn-
jrome. Neurology 1996;46:593-7.
9. Croft P. Soft tissue rheumatism. In: Silman AJ, I-Iochberg MC. eds. Epide-
miology of the Rheumatic Diseases. Oxford: Oxford University Press, 1999:
10. Croft P. Is life becoming more of a pain? People may be getting more willing
to report pain. BMJ 2000;320:1552-3.
11. Croft P, Schollum J, Silman A. Population srudy of tender poinr COUllts and
pain as evidence of fibromyalgia. BMJ 1994;309:696-9.
12. Ekberg K, Bjorkqvist B, Malm P, et al. Case-control study of risk facrors for
disease in the neck and sboulder area. Occup Environ Med 1994;51:262-6.
13. Eriksen W, Natvig B, Knardahl S, et al. Job eharacterisrics as predictors of
neck pain: a 4-year prospecrive srudy. J Oceup Environ Med 1999;41:893-
14. Estlander AM, Takaia EP, Viikari-Junrura E. Do psychological factors pre-
dict changes in musculoskeleral pain? A prospective, t\vo-year follow-up
srudy of a working population. J Occup Environ Med 1998;40:445-53.
IS. Fallentin N, Juul-Kristensen B, Mikkelsen S, er al. Physieal exposure assess-
ment in monotonous repetirive work: the PRIM srudy. Scand J Work Envi-
ron Health 2001;27:2]-9.
16. Gunn Cc. Radiculopathic pain: diagnosis and treatmenr of segmental irri-
tation or sensitization. J Musculoskeleral Pain 1997;5:119-34.
]7. Hadler NM. Repetitive upper-extremity morions in the workplace are not
bazardous. J Hand Surg 1997;22:19-29.
18. Hagberg M, Wegman DH. Prevalence rates and odds ratios ofshoulder-neek
diseases in differenr occupational groups. Br J Ind Med 1987;44:602-10.
19. Hagberg M, Silverstein B, Wells R, et al. In: Kuorinka I, Forcier L, eds. Work
Related Museuloskelctal Disorders (WMSD's): A Reference Book for Pre-
venrion. London: Taylor & Francis, 1995:17-137.
20. Hong CZ, Simons DG. Pathophysiologic and e1ectrophysiologic meebanisms
of myofascial trigger poinrs. Areh Phys Med Rehabil 1998;79:863-72.
21. Jonsson BG, Persson J, Kilbom ..\. Disorders of the cervicobraehial region
among temale workers in rhe e1ecrfonics inciusrry: a two-year fol1ov</ up. Int
J Ind !vled 1988;3:1-12.
22. Juul-Kristensen B, Fallentin N, Hansson G..\, et al. Force demands during
manual and mcchanical deboning af poulery estimared by e!ecr[omyogra-
phy, force tranducers, observer-based force esrimarions and e1ectrogoniom-
cref regisrrations. IntJInd Erg 2002, in press.
23. Kaergaard A, Andersen JH, Rasmussen K, et al. Idenrification of neck-
silOulder disorders in a I year follow-llp srudy: validarion of a qucsrionnaire-
based method. Pain 2000;86:305-10.
24. Kaergaard A, Andersen JH. Musculoskeletal disorders ofthe neek and shoul-
defs among sewing machine operators: prevaience, incicience and prognosis.
Occup Environ Med 2000;57:528-34.
25. Kamwendo K, Linton SJ, Moritz U. Neck and sholllder disorders in medicai
secretaries: I. Pain prevalence and risk factors. Scand J Rehabil Med 1991;
26. Kamwendo K, Linron SJ, Moritz U. Neck and sboulder disorders in medicai
secretaries: II. Ergonomical work environment and symptom profile. ScandJ
Rehabil Med 1991;23:135-42.
27. Karasek R, TheorelI T. Healthy Work. New York: Basic Books, 1990.
28. Krantz G, Ostergren PO. Women's health: do common symptoms in women
mirror general distress or specific disease entities? Seand J Public Health
29. Kuorinka I, Jonsson B, Kilbom A, et al. Standardised Nordic questionnaires
for the analysis of museuloskeletal symptoms. Appl Ergonomics 1987;18:
30. LecIerc A, Niedhammer I, Landre MF, et al. One-year predictive factors for
various aspects of neck disorders. Spine 1999;24:1455-62.
31. Linton SJ, van Tulder MW. Preventive interventions for back and neck pain
problems: what is the evidence? Spine 2001;26:778-87.
32. Linton Sj. A review of psyehological risk factors in back and neck pain. Spine
33. MaefarIane GJ. Generalized pain, fibromyalgia and regional pin: an epide-
miologieal view. Baillieres Best Pract Res Clin Rheumatol 1999;13:403-14.
34. MacfarIane GJ, Hunt 1M, Silman AJ. Role of mechanieal and psychosocial
factors in the onset of forearm pain: prospective population based srudy.
35. Makela M, Heliovaara M, Sievers K, et al. Prevalence, determinants, and
eonsequenees of chronic neck pain in Finland. Am J Epidemiol 1991;134:
36. McBeth J, MaefarIane GJ, Benjamin S, et al. The association bet\veen tender
points, psychologieal distress, and adverse ehildhood experiences. Arthritis
37. Ohlsson K, Attewell RG, Palsson B, et al. Repetitive industrial work and
neck and upper limb disorders in femaies. AmJ Ind Med 1995;27:731-47.
38. Siegrist J, Peter R. Measuring effort-reward imbalance at work: guidelines.
Diisseldorf, Germany: University of Diisseldorf, 1996.
39. Thompson ML, Myers JE, Kriebel D. Prevalenee odds ratio or prevalence
ratio in the analysis of eross seetional data: what is to be done? Occup
Environ Med 1998;55:272-7.
40. Waddell G. The Back Pain Revolution. Edinburgh: Churchill Livingstone,
41. Ware JE Jr, Snow KK, Kosinski M, et al. SF-36 Health Survey: Manual and
Interpretation Guide. Boston: Healtb Institute, New England Medicai Cen-
42. Veiersted KB, Westgaard RH, Andersen P. Electromyographic evaluation of
muscular work pattern as a predictor of trapezius myalgia. Scand J Work
Environ Health 1993;19:284-90.
43. Wessely S, HotopfM. Is fibromyalgia adistinet elinicai entity? Historicai anr
epidemiological evidence. Baillieres Best Praet Res Clin RbeumatoI1999;13:
44. Westgaard RH, Winkel J. Ergonomie intervention research for improved
musculoskeletal health: a critieal review. Int J Ind Erg 1997;20:463-500.
45. Viikari-Juntura E, Riihimaki H, Tola S, et al. Neck trouble in machine operat-
ing, dynamic physical work and sedemary work: a prospective srudy on occu-
pational and individ"al risk lactors.J Clin EpidemioI1994;47:14[1-22.
46. Vingård E, Nachemson A. Work-related influences on neck and low back
pain. In: Nachemson A, Jonsson E, eds. Neck and Back Pain: The Sciemific
Evidence olCauses, Diagnoses and Treatmem. PhiJadelphia: Lippincott Wil-
liams & Williams, 2000:97-126.
47. Wolens D. The myolJsciaJ pain syndrome: a critical appraisaL State ol the
Art Reviews: Phys lvled Rehabil 1998;[2:299-316.
48. Wolle F, Ros, K, Anderson J, et aL The prevaJence and characteristics ol
fibromyalgia in the general population. Arthritis Rheum 1995;38:[9-28.
Point of View
Hilkka Riihimaki, OMedSc, MSc
Finnish Institute af Occupatianal Health
Oepartment af Epidemialagy and 8iastatistics
According to recentsurveys, there seems to be an increas-
ing trend in the prevalence ofneck/shoulder pain, at least
in European countries. Suggested explanations include
the ongoing change toward a more sedemary lifestyle
and the cominuously growing use of information tech-
nology in work. This development raises concern about
prevemion of neck/shoulder problems, but several recent
reviews have shown that the evidence of risk factors of
neck/shoulder pain is still sparse and mainly based on
AIso, this article reports the results based on cross-
sectional data with inherent biases in such studies. The
data are from the baseline of the PRIM study, a Danish
large-scale epiGemiologic study focusing on the associa-
tion of monotonous repetitive work and neck and upper
limb problems. The unequivocal strength of the study is
the rigorous quamitative assessment of physical expo-
sure at work. In this respect the study is among the high-
est-ranking in its field. The exposure assessment resulted
in 43 exposure variables, some of which had to be disre-
garded because individual variation within exposure
groups was toa large compared with the between-group
variation.! Variables describing the posture/motion of
the shoulder joint were lost. Valid assessment of expo-
sure is the Achilles heel in epidemiologic studies on mus-
Device status category: 1.
ConfJict of inrerest categoty: 12, 14.
NeckJShoulder Pain Risk Factors' Andersen et al66:
Address correspondence to
Johan Hviid Andersen, PhD
Department of OCCltpational lvfedicine
Gl. Landevej 61
DK-7400 Herning, Denmark
culoskeletal disorders as the PRlj\;l research group has
stated.1A major problem is uncertainty in hazard iden-
titication, which may lead into wasted effort in exposure
assessmem. This research group has good data to assist
in resolving this problem. However, help is also needed
from experimental research on physiologic responses of
the musculoskeletal system to physicalloading.
Another good feature of this study is defining the
health outcome on the basis of symptoms and c!inical
tindings. Comparability with other studies is lost; how-
ever, t o ~ little has been done to improve the health out-
come measures beyond the simple question of subjective
pain experience. One price that must be paid is to in-
crease the size of epidemiologic studies; with stricter def-
initions of neck disorders the occurrence rates faII, as
with the PRIM study. Yet one important goal for epide-
miologists is to identify clinically relevant disorder enti-
ties that are etiologically and prognostically different.
The PRIM study provides a good opportunity to also
study the validity of various outcome definitions.
With great interestlam looking forward to the results
of the longitudinal phase af the PRIM study.
1. Fallemin N, Juul-Kristensen B, Mikkelsen S, er al. Physical exposure assess-
ment in monotonous physical work: rhe PRIM srudy. Scand J Work Enviton