Abstract
The multifactorial control of serum uric acid levels is discussed from an epidemiological point of view and the principles at issue related to the epidemiology of osteoarthrosis. It is shown that in osteoarthrosis the clinical characteristics vary from one joint group to another. Association of x-ray changes with pain varies between joints and between sexes for the various joints. The disease tends to be commoner in females but not in all age groups. There is an association between obesity and osteoarthrosis in the hands and knees but not the hips or ankles. These differences raise questions about the homogeneity of generalised osteoarthrosis. Such questions might be answered by epidemiological studies which consider individual joint groups and are carefully directed towards specific hypotheses, the development of which could be facilitated by the use of iconic models.