Skip Navigation


Occupational Medicine Advance Access originally published online on April 20, 2006
Occupational Medicine 2006 56(4):243-250; doi:10.1093/occmed/kql016
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
56/4/243    most recent
kql016v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (4)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Walker-Bone, K.
Right arrow Articles by Palmer, K. T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Walker-Bone, K.
Right arrow Articles by Palmer, K. T.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2006. Published by Oxford University Press on behalf of the Society of Occupational Medicine. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

Risk factors for specific upper limb disorders as compared with non-specific upper limb pain: assessing the utility of a structured examination schedule

Karen Walker-Bone, Isabel Reading, David Coggon, Cyrus Cooper and Keith T. Palmer

MRC Epidemiology Resource Centre, Southampton General Hospital, Southampton, UK

Correspondence to: Keith T. Palmer, MRC Epidemiology Resource Centre, Southampton General Hospital, Tremona Road, Southampton SO16 6YD, UK. Tel: +44 23 8077 7624; fax: +44 23 8070 4021; e-mail: ktp{at}mrc.soton.ac.uk


   Abstract
Top
 Abstract
Introduction
 Methods
 Results
 Discussion
 Conflicts of interest
 References
 
Background Few community-based epidemiological investigations of upper limb disorders (ULDs) have classified cases by validated procedures involving a structured clinical examination.

Aim To compare risk factor profiles for different diagnostic categories of ULD using one such examination scheme.

Methods A questionnaire about upper limb pain and demographic, occupational and psychosocial risk factors was mailed to 10 264 adults from two English general practices, followed by standardized physical examination in those with arm or neck pain. Logistic regression was used to compare those with specific ULDs and non-specific arm pain with those who had no neck or arm symptoms.

Results There was a 59% response rate. A total of 1197 subjects with arm or neck pain underwent standardized physical examination and were classified as having one or more of 11 specific ULDs or non-specific regional pain. Among these, 250 subjects with specific ULDs and 176 with only non-specific arm pain were compared with 2248 subjects who had no neck or arm symptoms. Certain physical risk factors were more strongly associated with specific disorders than with non-specific pain. In comparison with pain-free subjects, the odds ratios (ORs) in keyboard users (≥1 h versus <1 h/day) were 3.1 (95% CI 1.3, 7.8) for hand–wrist tendonitis but 1.3 (0.8, 2.1) for non-specific hand–wrist pain. Other differential associations were found with age, sex, manual versus non-manual employment and smoking. Unexpectedly, low vitality was similarly associated with both specific disorders and non-specific pain.

Conclusion These findings suggest that the schedule may usefully distinguish disorders that differ in their association with physical risk factors.

Keywords      Classification; occupational; soft tissue rheumatism


   Introduction
Top
 Abstract
 Introduction
Methods
 Results
 Discussion
 Conflicts of interest
 References
 
Musculoskeletal disorders of the upper limb are a major cause of sickness absence, disability and health care expenditure [1,2]. Much epidemiological research has been conducted on upper limb disorders (ULDs), as summarized in systematic reviews, textbooks and reports [2–6]. However, interpretation of the findings is limited by the multiplicity of approaches to case definition [7,8]. Symptoms may arise from various pathologies in the neck and arm or occur in the absence of well-defined pathology. At present, it cannot necessarily be assumed that specific disorders have the same causes or require the same management as non-specific pain and there is a need to differentiate them. However, in epidemiological investigations there has been little attempt to classify cases in this detail and according to validated assessment procedures. Only recently have consensus-driven diagnostic criteria been developed [9,10] and incorporated into structured assessment schedules. One such scheme, the Southampton examination schedule for ULDs, has been found to be repeatable between observers and to agree with a consultant rheumatologist's independent opinion in the hospital setting, and to be sufficiently reproducible in the general population to be usable in epidemiological research [11,12].

In the absence of a suitable gold standard, assessing a diagnostic scheme's validity is more challenging. However, one approach is to test its utility in distinguishing categories of illness that differ in their risk factors, or in their prognosis or response to treatment [13]. Applying the first of these criteria, we compared patterns of association with psychosocial and occupational risk factors when the Southampton examination schedule for ULDs was used to classify subjects from a community sample who reported upper limb pain. Our particular interest was whether associations with risk factors differed between subjects classed as having specific disorders as compared with non-specific arm pain.


   Methods
Top
 Abstract
 Introduction
 Methods
Results
 Discussion
 Conflicts of interest
 References
 
In 1998–2000, a postal questionnaire was mailed to all 10 264 men and women aged 25–64 years who were registered with two general practices in Southampton, with the exception of 156 subjects who were recently bereaved or suffering a terminal illness. The questionnaire asked about current neck or upper limb pain, sensory symptoms in the upper limb, feelings of well-being and (for those in employment) about physical and psychosocial aspects of their work.

For each anatomical site in the arm, questions were posed about occupational activities that were expected to stress the local tissues physically. For the shoulder, we asked about working with hands held above shoulder height (for >1 h per typical working day versus less) and carrying weights on one shoulder or lifting or carrying weights of ≥5 kg in one hand; for the elbow, about repeated bending and straightening of the joint (for >1 h/day versus less) and for the hand/wrist, about use of a keyboard or typewriter (for <1, 1–4, >4 h/day) and other tasks involving repeated movement of the wrists or fingers (for >4 h/day).

Questions on symptoms were based on the Standardized Nordic Questionnaire [14], and the vitality subsection of the Short-Form 36 (SF-36) [15], while those on psychosocial aspects of work (demands of work, control over work and support at work) were based on the Karasek model [16].

All respondents with pain in the past 7 days were invited to an assessment, conducted by a research nurse or physiotherapist, which was normally completed within 4 weeks of returning the postal questionnaire. At the assessment, they were asked if they were still in pain and at which site(s), and were examined according to the standardized Southampton examination schedule [11]. This examination included inspection and palpation of the upper limbs, measurement of the range of shoulder and neck movements and clinical provocation tests (including Finkelstein's test). All clinical assessments were conducted without knowledge of subjects' exposures to risk factors (which were ascertained by questionnaire, prior to interview).

Analysis was restricted to those subjects who held a job at the time of completing the postal questionnaire, and who had arm pain at examination that had been present at the same anatomical site when they answered the postal questionnaire or were completely free from neck and arm symptoms at the time of answering the postal questionnaire. A few subjects who had sustained an upper limb fracture in the past 12 months were excluded from the analysis (see Figure 1).


Figure 1
View larger version (17K):
[in this window]
[in a new window]
  		Figure 1. Response pattern and subjects included in the analysis.

 
Case definition, as in earlier reports [11,12], was based on the interview and examination. Subjects were classified by a pre-defined algorithm according to whether or not they had a specific disorder of the wrist–hand [tenosynovitis or De Quervain's disease; osteoarthritis of thumb base or distal interphalangeal (DIP) joints], a specific disorder of the elbow (medial or lateral epicondylitis), a specific disorder of the shoulder (rotator cuff tendonitis, adhesive capsulitis, bicipital tendonitis, subacromial bursitis or acromioclavicular joint dysfunction) or non-specific pain at one or more of these sites.

Bilateral involvement and multiple diagnoses were quite common. The individual, rather than the limb, was the unit of statistical analysis, so that those symmetrically affected by a disorder were counted singly. In addition, to simplify interpretation, subjects with both specific and non-specific disorders were excluded.

The figure illustrates the exclusions applied and the three groups finally analysed—namely (i) questionnaire respondents who were free of neck and upper limb symptoms at baseline; (ii) subjects with one or more specific disorders at the shoulder, elbow or wrist/hand, but without non-specific pain and (iii) subjects with non-specific pain in the absence of any specific disorder. The reference for each analysis was subjects from the first group—those who were asymptomatic when they answered the postal questionnaire. Specific disorders of the elbow were combined and considered together, as were specific disorders of the shoulder. But at the wrist/hand, a further distinction was drawn between those with tenosynovitis or De Quervain's disease (with or without osteoarthritis) and those with osteoarthritis only.

Associations of occupational and other risk factors (age, sex, smoking habits, SF-36 score, social class, relevant physical activities) with the various specific and non-specific outcomes were explored by logistic regression and expressed as adjusted ORs with associated 95% confidence intervals (CIs). The main contrasts drawn were between the associations with specific disorders, as compared with non-specific pain at the same anatomical sites.


   Results
Top
 Abstract
 Introduction
 Methods
 Results
Discussion
 Conflicts of interest
 References
 
Questionnaires were mailed to 10 264 subjects (see Figure 1), and usable returns were received from 6038 respondents. These included 1413 subjects who were not in work at the time, 61 who had fractured an upper limb bone in the past 12 months and 394 who complained only of numbness and tingling in the upper limb. Among the remaining 4170 subjects, 2248 had been free of upper limb or neck symptoms in the past week. A further 1197 reported pain and agreed to be examined. As described elsewhere, those who answered the questionnaire were similar in age, sex and employment rate to national and regional profiles, while those who attended the examination were similar to non-responders in age, sex and the anatomical distribution and severity of upper limb symptoms as judged by its impact on everyday activities [17].

In total, 496 of these subjects had continuing pain at the shoulder, elbow or wrist when subsequently interviewed and examined. These comprised 250 subjects who were classified as having only specific disorders; 176 with only non-specific pain and 70 (subsequently excluded) with a combination of specific and non-specific diagnoses. Altogether, 188 subjects had one specific diagnosis, 43 had two and 19 had three or more. The most common specific diagnoses were shoulder capsulitis (n = 143), rotator cuff tendonitis (103) and painful osteoarthritis of the thumb base or DIP joints (77).

Table 1 compares associations with risk factors between subjects classified as having a specific disorder of the shoulder and those classified as having only non-specific shoulder pain. In comparison with non-specific pain, a specific diagnosis was more often associated with being female (OR 2.4 versus 1.4), aged ≥55 years (OR 4.1 versus 2.3), a current smoker (OR 1.5 versus 0.4) and a blue-collar worker (OR 1.6 versus 0.4), although CIs were sometimes wide or overlapping. By contrast, there was a stronger association between non-specific shoulder pain and work with the arm elevated (OR 4.9 versus 1.6). Poor or absent support at work appeared a greater risk factor for specific shoulder disorders than for non-specific shoulder pain. Risks of both specific and non-specific outcomes were elevated in those in the worst versus the best third of SF-36 vitality (ORs 3.4 and 2.5, respectively).


View this table:
[in this window]
[in a new window]
  		Table 1. Association between risk factors and different patterns of disorder at the shoulder

 
At the elbow (Table 2), there were fewer distinctive associations by diagnostic category, although being in blue-collar work tended to be more strongly associated with epicondylitis than with non-specific elbow pain (OR 2.5 versus 1.5). Risks of both outcomes were similarly elevated in those with a poor SF-36 vitality score (lowest third of the distribution), and those whose occupation required repeated elbow bending or involved poor peer support or high work demands.


View this table:
[in this window]
[in a new window]
  		Table 2. Association between risk factors and different patterns of disorder at the elbow

 
At the hand–wrist, there were some notable differences of association between diagnostic groups (Table 3). Specific disorders of the wrist were far more strongly associated than non-specific pain with advancing age and with a poor SF-36 vitality score. Hand–wrist tendonitis (tenosynovitis or De Quervain's disease) was also associated with use of a keyboard or typewriter to a greater extent than other diagnostic groups (OR for ≥1 h versus <1 h 3.1, 95% CI 1.3–7.8). Finally, symptomatic osteoarthritis appeared to be associated more strongly than were the other diagnostic categories with repetitive movements of the wrists or fingers, and with work entailing high demand and low support.


View this table:
[in this window]
[in a new window]
  		Table 3. Association between risk factors and different patterns of disorder at the wrist

 

   Discussion
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
Conflicts of interest
 References
 
The diagnostic categories defined by this study's classification scheme showed some evidence of differential associations with pattern of work activity. In general, physical risk factors were more strongly associated with specific disorders than with non-specific pain. Thus, for example, specific disorders at the shoulder and elbow were more frequently associated with being in a blue-collar occupation than non-specific shoulder or elbow pain, and hand–wrist tendonitis was more commonly associated with repetitive keyboard work than non-specific hand–wrist pain. However, the pattern was not entirely consistent. Non-specific shoulder pain showed a stronger association with work with the arm elevated, and the risks associated with repeated elbow bending were little different for epicondylitis and non-specific elbow pain. Moreover, associations with indices of mental health tended to be similar for specific and non-specific disorders at the same site.

Comparison of our findings with earlier research is difficult, as the explicit contrast in risk factors between arm pain with and without specific physical signs has seldom been addressed previously using clear-cut case definitions. Several of the risk factors selected for study have well-recognized associations with upper limb pain overall (work with the arms elevated, for example, has often been linked with pain and soft tissue rheumatism in the shoulder) [2–4,18–20]; and our findings on specific disorders are consistent with several studies that have reported associations with age (shoulder disorders, epicondylitis, tenosynovitis and symptomatic wrist–hand osteoarthritis) [21–23], with blue-collar occupation (epicondylitis) [24,25] and with certain physical activities (shoulder disorders, tenosynovitis) [26–28]. But these studies did not report also on risk factors for non-specific pain at the same sites.

Nonetheless, our findings on physical risk factors are mostly consistent with prior knowledge. Of more surprise was the lack of differential association with mental health—a reasonable prior hypothesis being that physical risk factors are more influential in the presentation of illness associated with pathology in the arm and psychological risk factors in that of non-specific pain. In one previous study, hip osteoarthritis, which is characterized by demonstrable local tissue pathology, showed little association with poor mental health [29], whereas for non-specific regional pain disorders like low back pain, such associations have consistently been observed. Our finding that specific disorders showed clear associations with psychosocial risk factors may reflect greater awareness of associated symptoms in those with low vitality, reverse causation with a negative effect of the arm disorder on mental health, or, less likely perhaps, a greater vulnerability to specific disease in those with poor mental health. The case definition required symptoms reported at the postal questionnaire stage still to be present at examination, so a fourth possibility is that psychological factors favour symptom persistence, irrespective of the underlying pathology.

The response among those selected for study was incomplete, with attrition at both postal questionnaire and examination stages. However, as the main focus was an internal comparison between those with relevant signs and those without, we would not expect response bias to have had a major influence on the patterns observed. It also seems unlikely that our findings derive from errors or biases in the conduct of the physical examination. All examinations were conducted blind to information supplied on risk factors, and signs elicited within the schedule have previously been shown to be reasonably repeatable in both hospital and community settings [10,11].

We found, as others have done [30,31], that subjects often had symptoms and signs at multiple sites. This complexity and the relatively large number of diagnoses considered under the umbrella term of ‘ULD’ posed challenges, both in interpreting the findings in subjects who carried several diagnostic labels and in preserving sufficient numbers for comparisons of interest. In comparing associations with risk factors according to the way in which subjects were classified by the examination, we simplified by excluding from analysis those individuals who were diagnosed as having both a specific disorder and non-specific pain. To preserve power to detect associations we then combined specific disorders at each anatomical site. This lumping together of diagnostic categories by site could bias risk estimates towards the null if physical risk factors were associated with only some of the disorders that were grouped. If so, the net effect would be to make specific disorders appear more similar to non-specific ones in their patterns of association, and would not explain the differences that we found. On the other hand, we necessarily drew multiple comparisons, so some of the observed associations could have arisen by chance. Moreover, limited study power sometimes resulted in wide and overlapping CIs for risk estimates. For these various reasons, it would be prudent to test the findings in further studies.

The differences in association with risk factors between groups distinguished by the classification scheme are consistent with differences in pathogenesis, but alternatively, the presence of physical signs might simply be an index of severity. Whatever the explanation, our findings suggest that certain categories of ULD can usefully be distinguished by structured physical examination, as judged by their differential association with risk factors. To further evaluate our classification scheme it will be useful to test for differences in prognosis during follow-up [13], and we intend making this the next step of inquiry.


   Conflicts of interest
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 Conflicts of interest
References
 
None declared.


   References
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 Conflicts of interest
 References
 

  1. Jones JR, Hodgson JT, Clegg TA. Self-Reported Work-Related Illness in 1995. Norwich: Health and Safety Executive, HMSO, 1998.

  2. National Institute for Occupational Health and Safety. Musculoskeletal Disorders and Workplace Factors. A Critical Review of Epidemiologic Evidence for Work-Related Musculoskeletal Disorders of the Neck, Upper Extremity, and Low Back. Cincinnati, OH: US Department of Health and Human Sciences/NIOSH, 1997, Publication No. 97-141.

  3. Hagberg M, Silverstein B, Wells R et al. Work-Related Musculoskeletal Disorders (WMSDs): A Reference Book for Prevention. London: Taylor & Francis, 1995.

  4. National Research Council and the Institute of Medicine. Musculoskeletal Disorders and the Workplace: Low Back and Upper Extremities. Panel on Musculoskeletal Disorders and the Workplace and Commission on Behavioural and Social Sciences and Education. Washington, DC: National Academy Press, 2001.

  5. Hagberg M, Wegman DH. Prevalence rates and odds ratios of shoulder-neck diseases in different occupational groups. Br J Ind Med 1987;44:602–610.[Web of Science][Medline]

  6. Bongers P, Kremer AM, ter Laak J. Are psychosocial factors, risk factors for symptoms and signs of the shoulder, elbow or hand/wrist? A review of the epidemiological literature. Am J Ind Med 2002;41:315–342.[CrossRef][Web of Science][Medline]

  7. Buchbinder R, Goel V, Bombardier C et al. Classification systems of soft tissue disorders of the neck and upper limb: do they satisfy methodological guidelines? J Clin Epidemiol 1996;49:141–149.[CrossRef][Web of Science][Medline]

  8. Van Eerd D, Beaton D, Cole D et al. Classification systems for upper-limb musculoskeletal disorders in workers: a review of the literature. J Clin Epidemiol 2003;56:925–936.[CrossRef][Web of Science][Medline]

  9. Harrington JM, Carter JT, Birrell L et al. Surveillance case definitions for work-related upper limb pain syndromes. Occup Environ Med 1998;55:264–271.[Abstract/Free Full Text]

  10. Sluiter JK, Rest KM, Frings-Dresen MH. Criteria document for evaluating the work-relatedness of upper-extremity musculoskeletal disorders. Scand J Work Environ Health 2001;27(Suppl. 1):1–102.[Web of Science][Medline]

  11. Palmer K, Walker-Bone K, Linaker C et al. The Southampton examination schedule for the diagnosis of musculoskeletal disorders of the upper limb. Ann Rheum Dis 2000;59:5–11.[Abstract/Free Full Text]

  12. Walker-Bone K, Byng T, Linaker C et al. Reliability of the Southampton examination schedule for the diagnosis of upper limb disorders in the general population. Ann Rheum Dis 2002;61:1103–1106.[Abstract/Free Full Text]

  13. Coggon D, Martyn C, Palmer KT, Evanoff B. Assessing case definitions in the absence of a diagnostic gold standard. Int J Epidemiol 2005;34:949–952.[Abstract/Free Full Text]

  14. Kuorinka I, Jonsson BG, Kilbom A et al. Standardised Nordic questionnaire for the analysis of musculoskeletal symptoms. Appl Ergon 1987;18:233–237.[CrossRef][Web of Science][Medline]

  15. Ware JE. SF-36 Health Status Questionnaire. Boston, MA: Institute for the Improvement of Medical Care and Health, New England Medical Center Hospital, Quality Quest Inc., 1989.

  16. Karasek RA. Job demands, job decision latitude, and mental strain: implications for job redesign. Adm Sci Q 1979;24:285–307.[CrossRef][Web of Science]

  17. Walker-Bone K, Palmer KT, Reading I, Coggon D, Cooper C. Prevalence and impact of musculoskeletal disorders of the upper limb in the general population. Arthritis Rheum 2004;51:642–651.[CrossRef][Web of Science][Medline]

  18. Smedley J, Inskip H, Trevelyan F et al. Risk factors for incident neck and shoulder pain in hospital nurses. Occup Environ Med 2003;60:864–869.[Abstract/Free Full Text]

  19. Harkness EF, Macfarlane GJ, Nahit ES et al. Mechanical and psychosocial factors predict new onset shoulder pain: a prospective cohort study of newly employed workers. Occup Environ Med 2003;60:850–857.[Abstract/Free Full Text]

  20. Leclerc A, Chastang JF, Niedhammer I et al. Incidence of shoulder pain in repetitive work. Occup Environ Med 2004;61:39–44.[Abstract/Free Full Text]

  21. Allander E. Prevalence, incidence and remission rates of some rheumatic diseases and syndromes. Scand J Rheumatol 1974;3:145–153.[Web of Science][Medline]

  22. Hamilton PG. The prevalence of humeral epicondylitis: a survey in general practice. J R Coll Gen Pract 1986;36:464–465.[Web of Science][Medline]

  23. Lawrence LRC, Hochberg MC, Kelsey JL et al. Estimates of the prevalence of selected arthritis and musculoskeletal disorders in the United States. J Rheumatol 1989;16:427–441.[Web of Science][Medline]

  24. Kurppa K, Viikari-Juntura E, Kuosma E et al. Incidence of tenosynovitis or peritendinitis and epicondylitis in a meat-processing factory. Scand J Work Environ Health 1991;17:32–37.[Web of Science][Medline]

  25. Haahr JP, Andersen JH. Physical and psychosocial risk factors for lateral epicondylitis: a population based case-referent study. Occup Environ Med 2003;60:322–329.[Abstract/Free Full Text]

  26. Bernard BP, ed. Musculoskeletal Disorders (MSDs) and Workplace Factors. Cincinnati, OH: US Department of Health and Human Services, 1997.

  27. Armstrong TJ, Fine LJ, Goldstein SA et al. Ergonomic considerations in hand and wrist tendinitis. J Hand Surg 1987;12A:830–837.[Medline]

  28. Kryger AI, Andersen JH, Lassen CF et al. Does computer use pose an occupational hazard for forearm pain; from the NUDATA study. Occup Environ Med 2003;60:14e.[Abstract/Free Full Text]

  29. Croft P, Lewis M, Wynn-Jones C et al. Health status in patients awaiting hip replacement for osteoarthritis. Rheumatology 2002;41:1001–1007.[Abstract/Free Full Text]

  30. Takala J, Sievers K, Klaukka T. Rheumatic symptoms in the middle-aged population in south-western Finland. Scand J Rheumatol 1982;47(Suppl.):15–29.

  31. Urwin M, Symmons D, Allison T et al. Estimating the burden of musculoskeletal disorders in the community: the comparative prevalence of symptoms at different anatomical sites, and the relation to social deprivation. Ann Rheum Dis 1998;57:649–655.[Abstract/Free Full Text]


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
56/4/243    most recent
kql016v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (4)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Walker-Bone, K.
Right arrow Articles by Palmer, K. T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Walker-Bone, K.
Right arrow Articles by Palmer, K. T.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?