O21.1 UPPER LIMB, BACK, AND KNEE PAIN AND THEIR RELATIONSHIP TO SOMATISATION: A COMMUNITY BASED SURVEY

K. T. Palmer¹, M. Calnan², J. Poole¹, D. Coggon¹.¹MRC Environmental Epidemiology Unit, University of Southampton, Southampton SO16 6YD, UK; ²MRC HSRC, Department of Social Medicine, Canynge Hall, Bristol BS8 2PR, UK

Introduction: The occupational psychosocial determinants of musculoskeletal pain have been much studied. By contrast, personal response to bodily symptoms (somatisation) has rarely been examined as a potential risk factor.

Methods: To investigate, we mailed a questionnaire to 5000 subjects aged 25–64 years, selected at random from the registers of five general practices in north Somerset, UK. Inquiries were made about pain in the upper limb, low back, and knee lasting ⩾1 day in the past 12 months; its duration and impact; depressive symptoms (SF-36 scale); and tendency to somatise (using the Brief Symptom Inventory; BSI). Associations of pain with psychological wellbeing were examined by Cox regression and expressed as prevalence ratios (PRs).

Results: Among 2550 respondents, 1257 reported upper limb pain in the past 12 months, including 343 whose pain was ‘longlived’ (lasted ⩾6 months) and 240 who had ‘disabling’ pain (found it difficult to sleep and get dressed and do daily chores). Longlived and disabling arm pain were more common in women, with increasing age, and among the unemployed. After adjustment, being in the highest versus the lowest third of the distribution for BSI score was associated with a PR of 5.3 (95% confidence interval I 3.6 to 7.9) for longlived pain, and 15.3 (7.8 to 30.0) for disabling pain. Back and knee pain were similarly common in the study group and showed similarly strong associations with BSI score. Substantially weaker associations were found with the SF-36 mental health score.

Discussion: Somatisation has been defined as the predisposition to amplify physiological sensations, or to misclassify symptoms of emotional arousal. Chronic pain may sensitise to physiological events and heighten bodily awareness; or somatisers may have a lower threshold for regarding somatic symptoms as disabling. To investigate the direction of causality, we intend following up respondents at an interval of 18 months to reassess their pain, disability, and propensity to somatise.

O21.2 GEOGRAPHIC DIFFERENCES OF RHEUMATOID ARTHRITIS AND WORK

R-SKoskela¹, R. Martikainen¹, M. Klockars², T. Klaukka³, P. Mutanen¹.¹Department of Epidemiology and Biostatistics, Institute of Occupational Health, Topeliuksenk. 41 a A, FIN-00250 Helsinki, Finland; ²Department of Public Health, University of Helsinki, Finland; ³Research Department, Social Insurance Institution of Finland

Introduction: Approximately 75 000 people in the Finnish population (1.5% of total, 8.5% of working age) suffer from rheumatoid arthritis (RA). Considerable geographical variations in the age adjusted prevalence of RA were found.

Objective: To compare the prevalence of RA and occupation, and some social and environmental factors, between the Finnish provinces.

Methods: Prevalence of specially compensated medication for RA in 2001 was derived from the national medication register, and data on occupations, environmental, and social factors from various regional statistics. All the variables were obtained for 448 municipalities and comparisons were made between the 12 provinces. Pearson correlation and linear regression analysis was used.

Results: The municipal age adjusted prevalence of RA varied from 0.6 to 2.9%; there were significant differences between the provinces (p<0.0001). The highest rates were in the eastern part of the country. After adjusting for provincial differences in RA, some positive associations between RA and work or other factors were found: work in agriculture and forestry (p<0.0001), cattle farming (p<0.01), unemployment (p<0.0001), poor housing (p<0.0001), and use of healthcare services (p<0.01). Alcohol consumption is an example of a negative association (p<0.05).

Conclusions: Differences in occupations may contribute to geographical variations of RA. Incidence analyses based on personal data of RA and of exposure are needed.

O21.3 HAND USE AND PATTERNS OF JOINT INVOLVEMENT IN OSTEOARTHRITIS

S. Solovieva¹, T. Vehmas², H. Riihimäki¹, K. Luoma³, P. Leino-Arjas¹.¹Finnish Institute of Occupational Health, Department of Epidemiology and Biostatistics, Helsinki, Finland; ²Finnish Institute of Occupational Health, Department of Occupational Medicine, Helsinki, Finland; ³Department of Radiology, Peijas Hospital, Helsinki University Central Hospital, Vantaa, Finland

Objective: To characterise the pattern of multiple joint involvement in osteoarthritis (OA) of the hand among middle aged women from two occupations with distinctively different hand loading.

Methods: Radiographs of both hands of 295 dentists and 248 teachers were examined. Each distal interphalangeal (DIP), proximal interphalangeal (PIP), first interphalangeal (IP), and metacarpophalangeal (MCP) joints were graded (OA severity scale 0–4) individually, by using reference images. The co-involvement of different hand joints was analysed by logistic regression.

Results: The DIP joints of all fingers were the most frequently involved joint groups, followed by the PIP joints. The most frequent DIP joint to be involved was in the fifth finger. No differences in the prevalence of any finger OA between dominant and non-dominant hand was observed in either occupation. The prevalence of OA of grade 2 or more in any finger joint as well as OA in any DIP joint was higher among teachers than dentists (58.7% v 48.5%, p = 0.02, and 58.1% v 47.5%, p = 0.01, respectively). The clustering of hand joint involvement was statistically highly significant for both occupations. If one DIP joint was involved (OA grade 2 or more), the risk of any other DIP joint in that hand was 3.30 (95% confidence interval (CI) 2.07 to 5.30) for dentists and 4.22 (2.42 to 7.42) for teachers. The clustering of hand joint involvement was also highly symmetrical for both occupations. However, the risk of more severe OA (grade 3 or more) in the working fingers (first to third) of the right hand was significantly elevated among the dentists compared with the teachers (odds ratio 2.61; 95% CI 1.03 to 6.59).

Conclusions: Our findings indicate that the prevalence of radiographically defined hand OA in women aged 45–63 years is high and often polyarticular. Hand use may have a protective effect on finger joint OA, whereas a continuing joint overload may lead to multiple joint impairment. Mechanical factors may be secondary to genetic or metabolic factors in the aetiology of OA.

O21.4 FACTORS THAT INFLUENCE JOB SURVIVAL IN WORKING AGE PATIENTS AWAITING SURGERY ON THE HIP OR KNEE

K. T. Palmer, P. Milne, J. Poole, D. Coggon.MRC Environmental Epidemiology Unit, University of Southampton, Southampton SO16 6YD, UK

Introduction: Major surgery on the hip or knee is relatively common in working aged adults, but the impact of such joint disorders on employment and the factors that influence job retention have been little studied.

Methods: To investigate, a questionnaire was mailed to 498 patients aged 16–64 years who were on a waiting list for hip or knee arthroplasty, hip periacetabular osteotomy, or knee arthroscopy at a hospital in Portsmouth, UK. Inquiry was made about work circumstances at the time the disorder began and subsequently, and on disability, job demands, work adjustments, and access to occupational health advice. Job loss was related to these factors using Cox regression, with the results expressed as hazard ratios (HRs) and 95% confidence intervals (CI).

Results: Of 370 (74%) respondents, 278 were in work when their joint problem began. Of these, 82 (30%) had left their original job because of the joint problem. This outcome was more common among younger workers, when more time had elapsed since disease onset, and with greater current disability (HR 1.7). It was also more frequent among those in non-sedentary work (HR 2.7, 95% CI 1.2 to 6.1) and those in ‘micro-firms’ (HR 1.9, 95% CI 1.2 to 3.0 for <10 v ⩾10 employees), but there was no elevated risk of job loss for reasons unrelated to hip or knee disease. After adjustment for potential confounders, employment outcome differed little according to access to an occupational health service or job modification.

Conclusions: Job loss is common in those with serious lower limb joint pathology. Absence of certain physical work demands is more compatible with job retention, as is the added flexibility that arises from working in a medium to large sized company. There was little evidence of benefit from occupational health advice or workplace adaptations. The observation that job retention is better in larger companies needs to be considered alongside the increasing tendency of employers to downsize and subcontract their activities to smaller business units. This trend may limit the opportunity for job preservation among workers with significant hip and knee disease (an increasing proportion of the total future workforce).