Case definitions of knee osteoarthritis in 4,151 unselected subjects: relevance for epidemiological studies

The aims of the present study were threefold: to examine the distribution of knee joint osteoarthritis in a large, standardized radiological study; to examine the relationships between self reported knee pain and radiological OA; and to examine the natural history of radio-morphological change over age in individuals without radiological features of OA. The Copenhagen Osteoarthritis Study - COS is a substudy of the Copenhagen City Heart Study, a longitudinal regional health survey. From the third inclusion of the CCHS (1992-1994) 4,151 subjects were selected for subsequent standardized radiography of the pelvis, the knees, the hands, the wrists, and the lumbar spine. Images were analyzed and knee joint osteoarthritis (OA) was classified according to the radiographic atlas of Kellgren and Lawrence. Joint space width (JSW) was measured at three sites within both the medial and the lateral compartment. For the entire cohort the prevalence of radiological knee joint OA of all grades was 38.7% for men and 44.2% for women. Age stratification documented increasing knee joint OA both in regard to prevalence and morphological severity. Knee pain was universally correlated to the Kellgren and Lawrence severity of OA. In a subgroup with no features of radiological OA, a significant and linear decline in JSW with increasing age was found. We found a clear relationship between self-reported knee pain and radiological osteoarthritis. Pain was proportionally related to the severity of change. We also demonstrated a significant diminishing of joint space width with increasing age in individuals without apparent radiological degeneration.