I read with interest and mild disappointment the recent exchange of letters between Dr. McKenna [1] and van Daele and colleagues [2], who discussed bone mineral density in patients with non-insulin-dependent diabetes mellitus. Dr. McKenna asks whether "stress fractures in the foot occur more frequently than in controls," and van Daele and colleagues agree that this is an interesting question. In 1980, my colleagues and I at the Mayo Clinic conducted a population-based fracture incidence study in the diabetic population (986 cases) of Rochester, Minnesota, because concern had been raised about osteopenia in diabetic persons [3]. Our study was intended to determine the risk for both osteoporotic (nontraumatic spine, hip, humerus, and forearm fractures) and nonosteoporotic (for example, fractures occurring after substantial trauma) fractures, but fractures at all sites were included. Skeletal fracture rates in this patient population (most of whom had non-insulin-dependent diabetes mellitus) were lower than those of nondiabetic controls for many types of fractures. Of particular interest, we noted no increase in the number of osteoporotic fractures among patients with non-insulin-dependent diabetes mellitus. The only site at which fracture risk was increased was the medial malleolus (included in the "leg/ankle" category in the paper), a finding that might well be explained by obesity or neuropathy rather than by osteopenia. As far as we could determine from available records, these were not stress fractures but were instead the typical fractures of "turned ankles." Daly and colleagues [4] confirmed these observations by studying the epidemiology of ankle fractures in the larger Rochester, Minnesota, population. Metabolic studies have also failed to show meaningful calcium metabolic disturbances in non-insulin-dependent diabetes mellitus [5].