Should we consider Dupuytren's contracture as work-related? A review and meta-analysis of an old debate

BACKGROUND: In view of the conflicting opinions published, a meta-analysis was undertaken on epidemiological studies in order to assess any association between Dupuytren's contracture and work exposure. METHODS: Using the key words: "occupational disease", "work" and "Dupuytren contracture" without limitation on language or year of publication, epidemiological studies were selected from four databases (Pub-Med, Embase, Web of science, BDSP) after two rounds (valid control group, valid work exposure). A quality assessment list was constructed and used to isolate papers with high quality methodological criteria (scores of 13 or above, HQMC). Relevant associations between manual work, vibration exposure (at work) and Dupuytren's contracture were extracted from the articles and a metarisk calculated using the generic variance approach (meta-odds ratios, meta-OR). RESULTS: From 1951 to 2007, 14 epidemiological studies (including 2 cohort studies, 3 case-control studies, and 9 cross-sectional studies/population surveys) were included. Two different results could be extracted from five studies (based on different types of exposure), leading to 19 results, 12 for manual work (9 studies), and 7 for vibration exposure (5 studies). Six studies met the HQMC, yielding 9 results, 5 for manual work and 4 for vibration exposure. Five studies found a dose-response relationship. The meta-OR for manual work was 2.02[1.57;2.60] (HQMC studies only: 2.01[1.51;2.66]), and the meta-OR for vibration exposure was 2.88 [1.36;6.07] (HQMC studies only: 2.14[1.59;2.88]). CONCLUSION: These results support the hypothesis of an association between high levels of work exposure (manual work and vibration exposure) and Dupuytren's contracture in certain cases.