Arto T. Toivanen

Objective. Several studies have shown that knee OA is associated with obesity, physical stress at work, traumatic knee injuries, heredity and female gender. However, the body of such evidence comes from cross-sectional or case–control studies, and from only a few follow-up studies, mostly of short duration. Based on the nationwide Mini-Finland Health Survey, we analysed the potential risk factors for prediction of incident knee OA in the long term. Methods. Focused on major health problems, the survey was carried out in 1978–80 in a sample of 8000 subjects, representative of the Finnish population aged ⩾30 years. Altogether 823 subjects free from knee OA at the baseline were re-examined in 2000–01, and after the intervening 22 years 94 new cases of knee OA were found. Knee OA was diagnosed on both occasions by physicians using information on disease histories, symptoms and standardized clinical examinations. Results. The risk of developing knee OA was strongly associated with BMI (kg/m²); adjusted for age and gender and other covariates, and compared with the reference category (BMI < 25.0); the relative odds ratios (ORs) with 95% CIs were 1.7 (95% CI 1.0, 2.8) and 7.0 (95% CI 3.5, 14.10) for subjects with BMIs 25.0–29.9 and ⩾30.0, respectively. Similarly, the adjusted OR for the heaviest category of physical stress at work was 18.3 (95% CI 4.2, 79.4) compared with the lightest category, and 5.1 (95% CI 1.4, 19.0) for permanent complaints due to past knee injury. Conclusions. This prospective study confirms the roles of obesity, heavy work load and knee injury in the aetiology of knee OA.

To determine the agreement between clinical diagnosis and different radiological grading scales of knee osteoarthritis (OA) in an epidemiological study. Health 2000 Survey is an extensive population study focusing on major health problems in a representative sample of 8028 Finns over 30 years of age. In the survey, physicians diagnosed knee OA on the basis of physical status, symptoms, and medical history. A total of 130 participants (mean age 60 years, 68% female) were re-examined 1 year later (Kuopio OA 2000 Study) to determine the agreement between clinical and radiological diagnosis as well as between three different radiological grading scales (Kellgren and Lawrence, Ahlbäck, and Piperno). Weight-bearing knee radiographs were taken and graded by a radiologist in Kuopio University Hospital. The history of knee symptoms was obtained using the Western Ontario MacMaster (WOMAC) and Lequesne questionnaires. Knee OA was diagnosed clinically in 17.7% and radiologically in 24.6-30% of participants. The strength of agreement was moderate (kappa values 0.34-0.54) between the clinical and the radiological diagnosis and substantial (0.62-0.78) between the different radiological scales. Those subjects identified by any diagnostic method as having OA in either knee reported significantly more symptoms and disability than the other subjects. The agreement between the clinical diagnosis performed in a large population study and radiological grading scales was only moderate. By contrast, the agreement between different radiological scales was substantial.