Aims: The aim of this study is to examine the dose-response relationships between age, "lifestyle factors" (body mass index, tobacco smoking, sports), and symptomatic knee osteoarthritis in a population-based case-control study. Additionally, the study aims to investigate the mode of interaction between body mass index (BMI) and physical workload (occupational kneeling/squatting and lifting/carrying of loads) with respect to the risk of symptomatic knee osteoarthritis.
Methods: In five orthopedic clinics and five practices, 295 male patients aged 25-70 with radiographically confirmed knee osteoarthritis associated with chronic complaints were recruited. The control group comprised 327 male control subjects. In a structured personal interview, body weight at different ages, body height, cumulative amount of smoking, and cumulative duration of different sports activities until the date of first diagnosis of knee osteoarthritis were elicited. Adjusted odds ratios (OR) and 95% confidence intervals (CI) were calculated using unconditional logistic regression analysis. An interaction analysis for the parameters BMI and kneeling/squatting respective lifting/carrying of loads was performed. Population attributable risks (PAR) for knee osteoarthritis were determined for BMI solely and for the combination of BMI with occupational kneeling/squatting and lifting/carrying of loads, respectively.
Results: Age and overweight were strongly associated with the diagnosis of knee osteoarthritis. Compared with persons less than 35 years old, persons who were at least 65 years old had an odds ratio (OR) of 19.0 (95% CI 6.1-58.7) for knee osteoarthritis. Persons with a BMI > or = 28.41 kg/m2 had a strongly elevated risk of knee osteoarthritis (OR 10.8; 95% CI 4.8-24.3) compared to persons with a BMI < 22.86 kg/m2. Heavy tobacco smoking (> or = 55.5 pack years) was associated with a decreased knee osteoarthritis risk in comparison with never-smoking (OR 0.2; 95% CI 0.1-0.5). Ball games (handball, volleyball, basketball) and cycling were associated with symptomatic knee osteoarthritis (OR 4.0; 95% CI 1.8-8.9 and OR 3.7; 95% CI 1.7-7.8 in the highest category of cumulative duration, respectively); to a weaker degree jogging, swimming, and soccer also were positively related to symptomatic knee osteoarthritis. Combining the two parameters, BMI and kneeling/squatting into one variable led to a multiplicative interaction mode for symptomatic knee osteoarthritis. For persons with elevated BMI in combination with moderate to high exposure to occupational kneeling/squatting, the population attributable risk (PAR) was 4%. The PAR for elevated BMI in combination with moderate to high exposure to occupational lifting/carrying of loads was 7%.
Conclusions: In accordance with the literature, we find a strong association between BMI and knee osteoarthritis risk. Considering the relatively high prevalence of occupational manual materials handling, prevention of knee osteoarthritis should not only focus on body weight reduction, but should also take into account work organizational measures particularly aiming to reduce occupational lifting and carrying of loads.