Heberden's nodes and osteoarthrosis of the hip.

no significant association between osteoarthrosis of the weight bearing joints (hips and knees) and Heberden's nodes in a controlled study.2 My main criticism of the McGoldrick and O'Brien paper is that they do not compare like with like. The prevalence and severity of the nodes in their axillary, superomedial, and protrusion groups combined (representing primary OA) is indeed higher than in the congenital dislocation of the hip, dysplasia, and superolateral groups combined (representing secondary OA). On the other hand, the combined female to male ratio is 32/18 among the probands in the first three groups and 15/22 among the probands in the last three groups. As is also pointed out by the authors this represents a clear difference in favour of women in the groups representing primary OA. It is universally acknowledged that Heberden's nodes by themselves are more common among women, regardless of the presence of OA in other joints. Thus it would have been better if the authors had analysed their data by considering the prevalence of the nodes among the two genders separately in order to eliminate the sex bias.