RESUMEN
PURPOSE: Knee pain and osteoarthritis (OA) are common and often lead to disability among older adults. Existing published evidence, however, utilizes differing criteria to define studies' knee OA populations. We, therefore, aimed to determine whether differences exist in the characteristics of individuals with the presence of knee pain and different diagnostic criteria for knee OA. METHODS: The Promoting Independence in Seniors with Arthritis (PISA) study is a longitudinal observational study of individuals with and without knee pain and knee OA recruited from the orthopaedics clinic of the Universiti Malaya Medical Centre and the local hospital catchment. Patients were diagnosed with OA based on the American College of Rheumatology (ACR) criteria, the presence of knee pain, and a history of physician-diagnosed knee OA. Psychosocial parameters were measured using validated measures for social participation, independence, and ability to perform activities of daily living, and life satisfaction. RESULTS: Of the 230 included participants, mean age was 66.9 years (standard deviation: 7.2) and 166 (72.2%) were women. Kappa agreement between ACR criteria and knee pain was 0.525 and for ACR and physician-diagnosed OA it was 0.325. Binomial logistic regression analysis showed that weight, anxiety, and handgrip strength (HGS) were predictive of ACR OA. Knee pain was only predicted by HGS but not weight and anxiety. Physician-diagnosed OA was predicted by weight and HGS but not anxiety. HGS was predictive of ACR OA, knee pain, and physician-diagnosed OA. CONCLUSION: Our study showed that the characteristics of patients with OA are different, physically and psychosocially, depending on the criteria used. Poor agreement was observed between radiological diagnosis and the other diagnostic criteria. Our findings have important implications for the interpretation and comparison of published studies using different OA criteria.