Correction

In the original paper [1], there is a mistake in the results of the occupational group analysis. This mistake occurred when the core data set was merged with the occupational group data. According to the modified occupational group analysis (see modified Table 1), OR for chemical processers and manufacturers of plastics products are no longer significantly elevated. Having worked more than 10 years as metal worker is associated with knee osteoarthritis (OR = 2.2; 95% CI 1.1-4.4). The knee osteoarthritis risk of plasterers, insulators, glaziers, terrazzo workers, construction carpenters, roofers, and upholsters approaches statistical significance in the long-duration category (OR = 3.7; 95 CI 0.9-15.2). For woodworkers, the knee osteoarthritis risk is no longer significantly elevated. Having worked more than 10 years as painter or varnisher is associated with knee osteoarthritis (OR = 9.6; 95 % CI 1.2-77.9). Finally, we find a significantly elevated OR of 3.2 (95% CI 1.1-9.1) among subjects having worked as physically exposed service workers (storemen, nurses, refuse collectors) for more than 10 years. When subjects with non-service work as main occupation (“blue-collar workers”) are compared with “white-collar workers”, the odds ratio for knee osteoarthritis is still significantly elevated (OR = 2.0; 95% CI 1.3-2.9).

Table 1 Occupational groups (reference group: service occupation as main occupation) and symptomatic knee osteoarthritis