Erratum to: Case–control study of knee osteoarthritis and lifestyle factors considering their interaction with physical workload

Erratum to: Int Arch Occup Environ Health (2010) 83:291–300 DOI 10.1007/s00420-009-0486-6

In the original paper, there is a mistake in the calculation of population-attributable risks: Applying formula (3) as reported by Coughlin et al. (1994) [AR = P(E\D) * ((RR − 1)/RR), where P(E\D) is the proportion of exposed cases], for persons with elevated BMI in combination with moderate to high exposure to occupational kneeling/squatting, the population attributable risk (PAR) was not 4% (as stated in the abstract, the results section, and the discussion), but 19%. Furthermore, the PAR for elevated BMI in combination with moderate to high exposure to occupational lifting/carrying of loads was not 7%, but 24%. With correct PAR values, the last part of the “Results” section “Population attributable risks (PAR) for BMI and physical workload” should read as follows:

• The adjusted population attributable risk (PAR) for a BMI of 22.86 or more compared with a BMI of less than 22.86 was 59% (no table). The adjusted PAR for kneeling/squatting for 4,757 h or more was 17% (no table). The adjusted PAR for occupational lifting and carrying of weights ≥5,120 kg*hours was 23%.

• When population attributable risks were calculated for the combination of BMI elevations and occupational exposures, for persons with a BMI ≥24.92 kg/m² exposed to kneeling/squatting for 4,757 h or more, the PAR was 19%. The population attributable risk for the combined exposure to BMI ≥24.92 kg/m² and occupational lifting/carrying of weights ≥5,120 kg*hours was 24%. When persons with occupational lifting/carrying of weights ≥630 kg*hours were regarded as exposed, the population attributable risk for the combined exposure to BMI ≥24.92 kg/m² and occupational lifting/carrying increased to 37%.