Osteoarthritis (OA) is a disease with a truly formidable impact. As the most common form of arthritis, it accounts for more dependency in walking, stair climbing, and other lower extremity tasks than any other disease, especially in the elderly (1). Its economic costs are impressive. Yelin (2) summarized data from several studies and estimated the cost of OA in the US at $15.5 billion (in 1994 dollars), roughly 3 times the cost of rheumatoid arthritis. More than half of the OA costs are due to work loss.

As a chronic disease with a multifactorial etiology that includes modifiable risk factors, OA could, in theory, be prevented. To weigh the prospects for preventing OA, investigators have turned to epidemiology, ie, the study of disease in human populations and its association with individual and environmental characteristics. The foundation of our understanding of OA epidemiology emanatcd from cross-sectional population-based studies conducted by Kellgren and Lawrence in the 1950s in Britain. After a long hiatus, a large number of cohort-based epidemiologic studies, many also community-based, were initiated in the mid-1980s and later. These have begun to provide a wealth of new insights into the risk factors for disease. Some of these studies have now begun to produce longitudinal information that could facilitate valid inferences about causation. When risk factors are consistently identified among epidemiologic studies as preceding disease oc-currence, these factors can reasonably be identified as causes of disease. The present review arises in the context of this burgeoning sct of high-quality studies,