Natural Course of Rotator Cuff Syndrome in a French Working Population
Julie Bodin,MSc,1Catherine Ha,MD, PhD,2Audrey Petit,MD,1,3Alexis Descatha,MD, PhD,4 Thierry Thomas,MD, PhD,5Marcel Goldberg,MD, PhD,4Annette Leclerc,PhD,4
and Yves Roquelaure,MD, PhD1,3
Background The aim was to assess the persistence of rotator cuff syndrome (RCS) in workers and to study associations with personal and work-related factors, job change, exposure change and treatment.
Methods At baseline, 274 workers suffered from RCS, of whom 150 were followed up.
Three groups were constituted: (1) a RCS recovery group, (2) a shoulder pain without RCS group, and (3) a RCS without recovery group. We studied the differences between groups (1) and (3) because of the low number of workers in group (2).
Results Recovery occurred in 61.5% of men and 51.4% of women. Factors related to work were associated with persistence of RCS in men whereas upper limb pain and absence of job rotation were associated with persistence of RCS in women.
Conclusion This descriptive study showed that a high percentage of workers recovered and several personal and work-related factors were associated with persistent RCS. Larger prospective studies are needed to confirm these results. Am. J. Ind. Med. 57:683–694, 2014. ß2014 Wiley Periodicals, Inc.
KEY WORDS: musculoskeletal disease; persistence; rotator cuff; personal factors;
work-related factors
INTRODUCTION
Rotator cuff syndrome (RCS) is a common upper extremity musculoskeletal disorder (UEMSD) in the working
population causing long periods of absence from work [Bonde et al., 2003; Silverstein et al., 2006; Miranda et al., 2008; Silverstein et al., 2008; Smith et al., 2009;
Roquelaure et al., 2011]. In 2011, work-related MSDs represented 79% of occupational diseases in France and shoulder disorders accounted for 37% of the MSDs [Caisse Nationale de l’Assurance Maladie des Travailleurs Salariés, 2012]. Previous studies have examined the factors associated with recovery or recurrence of shoulder pain in workers [Miranda et al., 2001; Cassou et al., 2002; Luime et al., 2004; Grooten et al., 2007; Herin et al., 2012;
Sadeghian et al., 2013]. However, factors associated with the recurrence of clinically diagnosed shoulder disorder have been little studied in workers [Bonde et al., 2003; Silverstein et al., 2006]. Using the data of the surveillance program for MSD implemented by the French Institute for Public Health Surveillance, we have previously reported that RCS was the most frequent disorder in the working population of a French region (274 workers [7.4%] suffered from RCS) [Roquelaure et al., 2011]. This survey was conducted in a large sample of
1LUNAMUniversite´,Universite´d’Angers,Laboratoire d’ergonomie etd’e´pide´miologie en sante´au travail (LEEST), Angers, France
2De´partement sante´travail, Institut de veille sanitaire, Saint-Maurice, France
3CHU Angers, Angers, France
4Universite´de Versailles St-Quentin, UMRS 1018, Population-Based Epidemiological Cohorts’ Research Platform,Villejuif, France
5CHU Saint-E´tienne, Service de Rhumatologie, Saint-E´tienne, France
Contract grant sponsor: French Institute for Public Health Surveillance, Saint-Maurice, France; Contract grant number: 9/25/2002-5; Contract grant sponsor: French National Research Agency; Contract grant number: SEST-06-36.
Disclosure Statement: The authors report no conflicts of interests.
CorrespondencetoJulie Bodin,MSc,Faculte´deMe´decine,Laboratoire d’Ergonomie et d’e´pide´miologie en sante´au travail, Rue Haute de Recule´e, 49045 ANGERS Cedex 01, France. E-mail: julie.bodin@univ-angers.fr
Accepted 7 February 2014
DOI10.1002/ajim.22315. Published online1March 2014 in Wiley Online Library (wileyonlinelibrary.com).
ß2014 Wiley Periodicals, Inc.
workers representative of the regional working population and used the recommendations of a European consensus to standardize the diagnosis of specific UEMSDs and the definition of their risk factors [Sluiter et al., 2001].
Our aim in the present study was to assess the persistence rate of RCS in these workers several years later and to study associations with (i) personal and work-related factors collected at baseline, (ii) change of job or company since baseline, (iii) change in physical, postural, psychosocial load, and time constraints, and (iv) shoulder treatment since baseline. The hypothesis of this study was that workers who had not recovered were more exposed to individual and work- related risk factors of RCS identified in the literature [Bernard, 1997; Malchaire et al., 2001; Van Rijn et al., 2010].
MATERIALS AND METHODS Study Population
This prospective study was based on two successive surveys of a large sample of workers in the French Pays de la Loire region. The region contains 5.6% of the French working population, and its diversified socioeconomic structure is close to that of France [Ha et al., 2009]. French salaried workers, including temporary and part-time workers, undergo a regularly scheduled health examination by an occupational physician in charge of the medical surveillance of a group of companies. Between 2002 and 2005, 83 occupational physicians (18% of occupational physicians of the region) volunteered to take part in the study. They were representative of the region’s occupational physicians in terms of working time and geographic and economic sectors covered. Subjects were selected at random (see [Roquelaure et al., 2006] for details) and a total of 3,710 workers (2,161 men and 1,549 women) were included. Women were slightly underrepresented in the sample (42% vs. 47% in the region).
Overall, the distribution of economic sectors and occupations in the study sample was close to that of the regional workforce [Roquelaure et al., 2006; Ha et al., 2009]. The analyses presented in this report are based only on the 274 workers who suffered from RCS at baseline (142 men and 132 women).
Follow-up of the cases initially included and still at work was conducted between 2007 and 2010. Retired people, people on parental leave or long-term sick leave and unemployed people were excluded. Several reminders were sent out to all occupational medicine units, and then to each occupational physician now responsible for the medical surveillance of at least one worker of the cohort. For workers who had changed occupational physician, the research team systematically contacted the last occupational physician responsible for their medical surveillance. After approval, each occupational physician received guidelines describing
the standardized clinical procedure that applied the method- ology and clinical tests of the European consensus criteria document for RCS, lateral epicondylitis, ulnar tunnel syndrome, carpal tunnel syndrome, De Quervain’s disease, and flexor-extensor peritendinitis or tenosynovitis of the forearm-wrist region [Sluiter et al., 2001]. The research team conducted a training program for the occupational physicians to standardize physical examinations at baseline and follow-up.
Rotator Cuff Syndrome
In cases of shoulder pain occurring during the preceding 12 months, the occupational physician conducted a physical examination to diagnose RCS using the standardized clinical procedure of the European consensus criteria document for diagnosing work-related UEMSD [Sluiter et al., 2001]. RCS was diagnosed if (i) there was at least intermittent pain in the shoulder region (without paresthesia), worsened by active elevation movement of the upper arm, as in scratching the upper back, currently or for at least 4 days during the preceding 7 days; and (ii) if at least one of the following shoulder tests was positive (Table I):
Painful arc on upper arm elevation;
Resisted shoulder abduction, external rotation, or internal rotation;
Resisted elbowflexion.
Physical examination at follow-up allowed classification of workers into three groups: (1) RCS recovery: workers without shoulder pain and without positive test results for RCS; (2) shoulder pain without RCS: workers with current shoulder pain or shoulder pain occurring at least for 4 days during the preceding 7 days or for at least 4 days in 1 week during the preceding 12 months without positive test results for RCS; and (3) RCS without recovery: workers with clinically diagnosed RCS.
The definition of RCS was the same at baseline and at follow-up.
Personal and Work-Related Variables The personal, organizational, biomechanical, and psy- chosocial factors studied were assessed at baseline by a self- administered questionnaire. Psychosocial work factors were assessed according to the validated French version of the Karasek Job Content Questionnaire [Niedhammer et al., 2006].
During the follow-up examination, the occupational physicians asked the worker if s/he had changed job in the same company and if s/he had changed company since baseline. In addition, the occupational physicians reported
TABLEI.ObjectiveTestsIncludedasCriteriaforRotatorCuffSyndromeAccordingtotheEuropeanConsensusCriteriaDocumentforDiagnosingWork-RelatedUEMSD NameoftestKindoftest Startingpositionç patient Starting positionçdoctorDescriptionassignmentPositivewhen Painfularctest(abduction^ elevation)R&L
Activeshouldergirdle,for rotatorCuffsyndrome Standingwitharms hangingdownwardsand thumbsdirectedtoventral side Standinginfrontof thepatient
‘‘Liftyourarmstothesideuptoshoulder height,turnthepalmsofyourhandsup- wardsandliftthearmsuntilyourhands touchaboveyourhead’’
Painisfeltduringapartof themovement(somewhere between608and1208 abduction) Activeelevationtest1:Apley’s scratchtest(abduction/external rotation)compareRwithL
Activeshouldergirdle,for rotatorCuffsyndrome StandingStandingbehind patient
‘‘Takeyourhandbehindyourheadand touchtheupperpartofyourother shoulderbladewiththetopofyour fingers’’
Localpaininshoulder duringorattheendofthe movement Activeelevationtest2:Apley’s scratchtest(abduction/internal rotation)compareRwithL
Activeshouldergirdle,for rotatorCuffsyndrome StandingStandingbehind patient
‘‘Takeyourhandbackwardsandtouch thelowerpartofyourothershoulder bladeonyourbackwiththetipofyour middlefinger’’
Localpaininshoulder duringorattheendofthe movement Activeelevationtest3:Apley’s scratchtest(adduction)compare RwithL
Activeshouldergirdle,for rotatorCuffsyndrome StandingStandinginfrontof thepatient
‘‘Grabthetopofyourothershoulder’’Localpaininshoulder duringorattheendofthe movement Resistedabductionglenohumeral jointR^L
Isometricresistance,for rotatorCuffsyndrome Sitting,Lupperarmin 10^208abduction StandingonLsideof patient ForL:Rhandstabilizestopofshoulder,L handonlateralupperarmandforceis builtuptowardsadduction‘‘Keepyour arminthispositionandresistmyforce’’
Localpaininshoulder (supraspinatemuscle) Resistedexternalrotation glenohumeraljointR-L
Isometricresistance,for rotatorCuffsyndrome Sitting,Lupperarmagainst body;elbowin908flexion, forearminneutralposition StandingonLside patient
ForL:Rhandcontrolspositionofelbow;L handagainstdorsalsideforearmand forceisbuiltuptowardsinternalrotation glenohumeraljoint‘‘Keepyourelbow againstyourbodyandresistmyforce’’
Localpaininshoulder (infraspinatemuscle) Resistedinternalrotation glenohumeraljointR^L
Isometricresistance,for rotatorCuffsyndrome Sitting,Lupperarmagainst body;elbowin908flexion, forearminneutralposition Standinginfrontof thepatient
ForL:Rhandcontrolspositionofelbow;L handagainstventralsideofforearmand forceisbuiltuptowardsexternalrotation oftheglenohumeraljoint‘‘Keepyour elbowagainstyourbodyandresistmy force’’
Localpaininshoulder (subscapularmuscle) Resistedelbowflexion test(¼speed’stest)R^L
Isometricresistancebiceps brachiimuscle,forrotator Cuffsyndrome Sitting,Rupperarmin908 anteflexion,forearm supinated,elbowinlight flexedposition StandingonLsideof thepatient
RhandstabilizesRglenohumeraljoint,L handplacedonventralsideofRforearm andforceisbuiltuptowardselbow extension.‘‘Keepyourarminthisposition andresistmyforce’’
Localpainoverinsertion bicepstendon R,right;L,left.
whether physical load, postural load, psychological load, and time constraints had decreased, remained the same or increased during the preceding 12 months and reported if the worker had had a sick leave due to upper-limb pain during the preceding 12 months. Information was recorded regard- ing history of shoulder surgery, infiltration of corticosteroids, shoulder physiotherapy, and any other treatment for RCS.
Only treatments during the follow-up period and involving the same shoulder side as the initially diagnosed RCS were analyzed.
Statistical Analysis
Descriptive analyses were performed separately for men and women to take into account gender differences in exposure to work constraints [Silverstein et al., 2009]. We first used the Chi-square test (or Fisher exact test when one or more expected values were less thanfive) to identify baseline variables associated with loss to follow-up. Associations between the health status group at follow-up and the personal, medical, and work-related variables assessed at baseline and/
or follow-up were studied. Because of the small number of workers involved in the“shoulder pain without RCS”group, we did not study this group. The Mann–Whitney test was used for continuous variables because of the small size of the groups and the non-normality of the distributions. The Chi-square test (or Fisher exact test) was used for categorical variables. Associations were considered significant for P<0.05. SAS (version 9.3) was used for all statistical analyses. Each subject provided informed written consent to participation in this study at baseline and the study received approval from France’s National Committee for Data Protection,first in 2001, and then in 2006.
RESULTS
Of the 274 workers with RCS at baseline, 40 were excluded because of retirement, unemployment, death, parental leave, or long-term sick leave, four were followed up by a non-participating occupational physician and 41 were lost to follow-up (Fig. 1). Moreover,five workers refused to participate and 34 workers did not undergo the second physical examination in the absence of a mandatory examination scheduled before the end of the follow-up period. Finally, 150 workers (78 men and 72 women) were followed up. The follow-up rate did not differ between genders (54.9% for men vs. 54.5% for women). Comparisons between workers followed up or not showed a significant difference in age (P¼0.033 in men and P¼0.045 in women): 43.7% of men without follow-up were over 50 years versus 21.8%, the corresponding values for women were 46.7% and 27.8%, respectively (Table II). Moreover, more
men lost to follow-up had had daily shoulder pain during the preceding 12 months at baseline (42.2% vs. 21.1%, P¼0.007) and more women lost to follow-up had had low back pain during the preceding 12 months (68.3% vs. 50.0%, P¼0.033). The workers lost to follow-up of both genders did not differ in terms of occupational category or length of service in the job. Men lost to follow-up had had high visual demand (28.1% vs. 11.5%,P¼0.012) and more women lost to follow-up had had low decision authority (63.3% vs.
44.3%, P¼0.030). Mean follow-up after baseline was 4.2 years (standard deviation: 1.3, range 2.1–8.1 years).
The men included in the study were mainly skilled blue collar workers whereas the women were mainly low-grade white collar workers (Table II). Men were more exposed to work pace dependent on colleagues’ work, overtime hours and use of vibrating hand tools. Women were more exposed to high repetitiveness of tasks and seated work.
Forty-eight of the 78 men with RCS at baseline (61.5%) had totally recovered (without shoulder pain), 7 (9.0%) had shoulder pain without RCS, and 23 (29.5%, 6 bilateral cases) had persistent RCS. The corresponding values for women were 37 (51.4%), 9 (12.5%), and 26 (36.1%, 4 bilateral cases), respectively. There was no significant difference between genders (P¼0.445). Moreover, there was no significant difference according to the duration of follow- up (<4 years and4 years). At least two clinical tests were positive in 90% of cases at follow-up.
Variables Assessed at Baseline
Because of the small number of workers involved in the
“shoulder pain without RCS”group (7 men and 9 women), only the differences between the “RCS recovery” and the
“RCS without recovery”groups were studied.
FIGURE1. Study population flow-chart.
TABLE II. Comparison of Workers’ Baseline Characteristics According to Gender
Men (n¼142) Women (n¼132)
With follow-up
(n¼78)
Without follow-up
(n¼64)
P-value
With follow-up
(n¼72)
Without follow-up
(n¼60)
P-value
n % n % n % n %
Personal factors and medical history
Age 0.033a 0.045a
<40 23 29.5 17 26.6 17 23.6 14 23.3
40^44 16 20.5 9 14.1 12 16.7 10 16.7
45^49 22 28.2 10 15.6 23 31.9 8 13.3
50 17 21.8 28 43.7 20 27.8 28 46.7
Overweight (25^30 kg /m2) or obesity (30 kg /m2) 37 50.0 36 58.1 0.348a 26 38.2 24 40.0 0.838a Shoulder pain
Lasting more than 1month during the preceding 12 months 35 46.1 36 56.3 0.229a 34 48.6 35 59.3 0.223a
Daily during the preceding 12 months 16 21.1 27 42.2 0.007a 20 28.6 21 35.6 0.394a
Elbow pain during the preceding 12 months 30 38.5 21 32.8 0.485a 30 41.7 19 31.7 0.236a
Hand/wrist pain during the preceding 12 months 34 43.6 18 28.1 0.057a 30 41.7 30 50.0 0.338a
Neck pain during the preceding 12 months 45 58.4 31 48.4 0.235a 43 59.7 43 71.7 0.152a
Low back pain during the preceding 12 months 59 75.6 53 82.8 0.298a 36 50.0 41 68.3 0.033a
Upper extremity musculoskeletal disordersc 19 24.4 11 17.2 0.298a 18 25.0 11 18.3 0.357a
Length of service in the job (years) 0.739b 0.253b
<1 3 3.8 3 4.8 3 4.2 8 13.3
1^2 8 10.3 10 15.9 6 8.3 4 6.7
3^10 21 26.9 14 22.2 24 33.3 15 25.0
>10 46 59.0 36 57.1 39 54.2 33 55.0
Occupational category 0.256b 0.975a
Manager, professional, technician, craftsman 24 30.8 13 20.3 13 18.1 11 18.3
Low-grade white collar 5 6.4 3 4.7 32 44.4 28 46.7
Skilled blue collar 32 41,0 37 57.8 9 12.5 6 10.0
Unskilled blue collar 17 21.8 11 17.2 18 25.0 15 25.0
Economic sector 0.193b 0.436b
Agriculture 0 0.0 1 1.6 0 0.0 1 1.7
Industries 34 43.6 26 40.6 29 40.3 20 33.3
Construction 6 7.7 11 17.2 1 1.4 0 0.0
Trade and services 38 48.7 26 40.6 42 58.3 39 65.0
Factors related to work organization
Paced work 11 14.7 10 16.1 0.813a 10 14.3 7 12.1 0.713a
Work pace dependent on automatic rate 15 19.7 7 11.5 0.191a 11 15.7 7 11.9 0.530a
Work pace dependent on other technical organization 21 27.6 16 25.8 0.810a 16 22.5 7 12.3 0.133a
Work pace dependent on colleagues’ work 32 43.2 19 31.1 0.149a 18 25.7 15 25.9 0.985a
Work pace dependent on quantified targets 43 57.3 31 49.2 0.340a 41 59.4 25 43.1 0.067a
Work pace dependent on demand of customers 36 47.4 19 31.7 0.064a 29 40.3 25 43.1 0.745a
Work pace dependent on permanent controls or surveillance 19 25.0 18 29.5 0.555a 23 32.9 15 25.9 0.389a Job/task rotation (1job rotation per week) 33 44.0 24 40.0 0.640a 27 40.3 27 49.1 0.331a
Overtime hours 46 59.7 37 57.8 0.817a 28 38.9 32 55.2 0.064a
Work with temporary workers 23 29.5 23 35.9 0.414a 17 23.6 18 30.0 0.408a
High visual demand 9 11.5 18 28.1 0.012a 21 29.2 13 21.7 0.327a
Variable weekly working time 38 49.4 34 53.1 0.655a 35 50.0 31 51.7 0.850a
(Continued)
Among the men, there were more unskilled blue collar workers in the “RCS without recovery” group and more managers, professionals, technicians, and craftsman in the
“RCS recovery”group (Table III). However, the difference was not significant. Compared to the“RCS recovery”group, men who had not recovered more often reported biomechan- ical constraints, such as high perceived physical exertion, and high time constraints, such as high repetitiveness of tasks (4 hr/day), paced work, work pace dependent on automatic rate, permanent controls, or surveillance. No difference was observed for self-reported shoulder movements (arms above the shoulder, arms abducted (60–90°) and holding hand behind the trunk). Moreover, 47.8% of men who had not recovered were exposed to 2–5 of the above biomechanical constraints compared to 21.3% of men who had recovered (P¼0.053). They more often worked with temporary workers and they were more often exposed to low decision authority and job strain. Compared to the“RCS recovery” group, women without recovery more often reported shoulder pain lasting more than 1 month during the preceding 12 months, elbow pain during the preceding 12 months and the preceding 7 days and hand/wrist pain during the
preceding 12 months. Women who had recovered were more often exposed to job rotation (Tables III and IV).
Stratification on duration of follow-up (<4 years and4 years) decreased the statistical power but did not change the meaning of the results.
Variables Assessed at Follow-Up
When comparing those who reported a job change in the same company, we found no statistically significant differ- ence in RCS status at follow-up, whatever the gender (Table V). However, six female workers in the “RCS recovery”group had changed company (16.2% vs. 0.0% in the“RCS without recovery”group,P¼0.038). Four of these had changed economic sector. Men in the “RCS without recovery”group more frequently reported an increase in their physical (8/22 vs. 3/47,P¼0.002) and postural load (7/22 vs.
2/44,P¼0.014) during the preceding 12 months compared to the“RCS recovery”group. Eighty-three percent of men who had recovered had had similar physical loads during the preceding 12 months and 75.0% had identical postural loads TABLE II. (Continued.)
Men (n¼142) Women (n¼132)
With follow-up
(n¼78)
Without follow-up
(n¼64)
P-value
With follow-up
(n¼72)
Without follow-up
(n¼60)
P-value
n % n % n % n %
Working postures and biomechanical constraints
High repetitiveness of tasks (4 hr/day) 25 32.1 26 40.6 0.289a 36 50.7 26 43.3 0.400a
High perceived physical exertiond 27 35.1 24 37.5 0.764a 17 23.6 23 38.3 0.067a
Posture with the arms above the shoulder (2 hr/day) 19 24.4 24 37.5 0.090a 15 21.1 10 16.7 0.517a Posture with the arms abducted (60^908;2 hr/day) 16 20.5 16 25.0 0.524a 17 23.9 16 26.7 0.721a
Holding hand behind the trunk (2 hr/day) 5 6.4 4 6.3 1.000b 8 11.3 2 3.3 0.108b
Use of hand tools (2 hr/day) 44 56.4 42 65.6 0.264a 36 50.7 28 46.7 0.645a
Use of vibrating hand tools (2 hr/day) 20 25.6 15 23.4 0.762a 6 8.3 4 6.7 0.755b
Pushing or pulling load (2 hr/day) 18 23.4 18 28.1 0.520a 11 15.3 11 18.3 0.639a
Working seated (4 hr/day) 13 16.7 11 17.2 0.934a 31 43.1 20 33.3 0.253a
Psychosocial factors at work
High psychological demand 44 56.4 39 60.9 0.586a 37 52.1 28 46.7 0.535a
Low skill discretion 50 64.1 37 57.8 0.444a 51 70.8 41 69.5 0.867a
Low decision authority 27 34.6 14 21.9 0.096a 31 44.3 38 63.3 0.030a
Low supervisor support 33 44.6 31 48.4 0.652a 28 41.2 35 58.3 0.053a
Low coworker support 15 19.2 13 20.3 0.872a 14 20.3 19 32.8 0.110a
Job strain 22 28.2 20 31.3 0.692a 22 31.4 22 37.3 0.484a
aChi-squared test, in bold,P-value<0.05.
bFisher’s exact test, in bold,P-value<0.05.
cLateral epicondylitis,ulnar tunnel syndrome,carpal tunnel syndrome,De Quervain’s disease, and flexor^extensorperitendinitis or tenosynovitis of the forearm-wrist region.
dRating Perceived Exertion Borg scale15 for men and14 for women.
TABLE III. Work-Related Characteristics at Baseline According to Health Group and Gender
Men Women
Recovery RCS (n¼48)
RCS without recovery (n¼23)
P-value
Recovery RCS (n¼37)
RCS without recovery (n¼26)
P-value
n % n % n % n %
Occupational category 0.152a 0.817a
Manager, professional, technician, craftsman 18 37.5 4 17.4 8 21.6 4 15.4
Low-grade white collar 3 6.2 1 4.3 15 40.6 12 46.1
Skilled blue collar 20 41.7 10 43.5 5 13.5 2 7.7
Unskilled blue collar 7 14.6 8 34.8 9 24.3 8 30.8
Factors related to work organization
Paced work 3 6.4 7 31.8 0.009a 7 20.0 1 3.8 0.122a
Work pace dependent on automatic rate 5 10.6 7 30.4 0.050a 6 17.1 3 11.5 0.720a
Work pace dependent on other technical organization 12 25.0 7 31.8 0.552b 7 19.4 5 19.2 0.983b
Work pace dependent on colleagues’ work 19 41.3 10 45.5 0.746b 9 25.7 7 26.9 0.915b
Work pace dependent on quantified targets 25 53.2 15 68.2 0.240b 23 65.7 12 48.0 0.170b
Work pace dependent on demand of customers 24 50.0 9 40.9 0.479b 15 40.5 9 34.6 0.634b
Work pace dependent on permanent controls or surveillance 5 10.6 11 50.0 <0.001b 12 34.3 9 34.6 0.979b Job/task rotation (1job rotation per week) 21 46.7 11 47.8 0.928b 20 55.6 3 13.0 0.001b
Overtime hours 30 63.8 12 52.2 0.350b 16 43.2 9 34.6 0.491b
Work with temporary workers 9 18.8 12 52.2 0.004b 8 21.6 7 26.9 0.627b
Variable weekly working time 26 55.3 9 39.1 0.203b 16 45.7 13 50.0 0.740b
Working postures and biomechanical constraints
High repetitiveness of tasks (4 hr/day) 11 22.9 11 47.8 0.034b 17 47.2 14 53.9 0.607b
High perceived physical exertionc 13 27.7 13 56.5 0.019b 9 24.3 5 19.2 0.632b
Posture with the arms above the shoulder level (2 hr/day) 11 22.9 6 26.1 0.770b 6 16.2 6 24.0 0.521a Posture with the arms abducted (60^908) (2 hr/day) 8 16.7 6 26.1 0.359a 8 21.6 7 28.0 0.565b
Holding hand behind the trunk (2 hr/day) 3 6.3 2 8.7 0.656a 4 10.8 3 12.0 1.000a
Biomechanical exposured 0.053b 0.565b
0 21 44.7 5 21.7 12 33.3 8 32.0
1 16 34.0 7 30.4 11 30.6 5 20.0
2^5 10 21.3 11 47.8 13 36.1 12 48.0
Use of hand tools (2 hr/day) 28 58.3 14 60.9 0.839b 17 47.2 14 53.8 0.607b
Use of vibrating hand tools (2 hr/day) 11 22.9 9 39.1 0.155b 4 10.8 0 0.0 0.136a
Pushing or pulling load (2 hr/day) 11 22.9 7 31.8 0.429b 3 8.1 5 19.2 0.257a
Working seated (4 hr/day) 9 18.8 3 13.0 0.739a 21 56.8 9 34.6 0.083b
Psychosocial factors at work
High psychological demand 28 58.3 13 56.5 0.885b 15 40.5 16 61.5 0.101b
Low skill discretion 28 58.3 18 78.3 0.100b 25 67.6 21 80.8 0.245b
Low decision authority 9 18.8 14 60.9 <0.001b 16 44.4 13 50.0 0.665b
Low supervisor support 20 43.5 9 42.9 0.962b 16 44.4 10 41.7 0.832b
Low coworker support 7 14.6 5 21.7 0.507a 7 20.6 6 23.1 0.817b
Job strain 11 22.9 11 47.8 0.034b 9 25.0 11 42.3 0.150b
RCS, rotator Cuff syndrome.
aFisher exact test, in bold,P-value<0.05.
bChi-squared test, in bold,P-value<0.05.
cRating Perceived Exertion Borg scale15 for men and14 for women.
dIncluded variables: high repetitiveness of tasks (4 hr/day), high perceived physical exertion (15 for men and14 for women), posture with the arms above the shoulder (2 hr/day), posture with the arms abducted (60^908,2 hr/day) and holding hand behind the trunk (2 hr/day).