Objectives The study investigated the association between occupational biomechanical risk factors and the occurrence of thumb carpometacarpal joint osteoarthritis (CMC1 OA) in construction workers.

Methods Male construction workers (n=237 525), participating in a Swedish occupational surveillance programme between 1971 and 1993, were followed between 1997 and 2019. CMC1 OA diagnoses were identified through linkage with national medical registries. Job title, smoking status, height, weight and age were collected from the surveillance programme. A job exposure matrix (JEM) was developed with exposure estimates on biomechanical risk factors for each occupational group. Relative risk (RR) of CMC1 OA diagnosis was calculated using a Poisson regression model.

Results There was an increased risk of CMC1 OA for all biomechanical risk factors (RR range 1.3–1.5). Exposure-response patterns were seen for repetitive wrist flexion and extension (low: RR 1.30 (95% CI 1.07 to 1.59), moderate: 1.32 (95% CI 1.07 to 1.62), high: 1.45 (95% CI 1.19 to 1.75)), wrist extension (low: 1.31 (95% CI 1.09 to 1.59), moderate: 1.41 (95% CI 1.17 to 1.70) and heavy lifting (low: 1.13 (95% CI 0.92 to 1.38), moderate: 1.45 (95% CI 1.18 to 1.77), high: 1.50 (95% CI 1.24 to 1.82). Electricians (1.29 (95% CI 1.03 to 1.89)), concrete workers (1.31 (95% CI 1.02 to 1.67)), plumbers (1.37 (95% CI 1.07 to 1.76)), sheet-metal workers (1.58 (95% CI 1.18 to 2.10)), wood workers (1.66 (95% CI 1.36 to 2.03)), repairers (1.75 (95% CI 1.06 to 2.90)) and glass workers (2.21 (95% CI 1.42 to 3.44) had an increased risk of CMC1 OA compared with the reference group.

Conclusion Wrist movements and hand loading were associated with CMC1 OA.