The relationship between cortical bone involution and fracture occurrence in an affluent aging American white population.

A study of peripheral cortical bone mineral status, as measured by single-photon absorptiometry, was initiated in Sun City, Arizona, in 1982. Affluent, active white women over age 50 and men over age 60 participated in up to eight annual sessions involving the measurement of the left mid-distal radius and the collection of questionnaire data. The data were retrospectively analyzed with a focus on the relationship between the bone mineral variables (BMVs: mass, areal density, density, width, and second moment of area) and postmature (after age 50) fracture history. Fractures were classified as either retrospective (incurred prior to baseline session) or prospective (incurred during the interval between baseline and final sessions). A distinction was made between hip fractures, which are often debilitating, and non-hip fractures, which, in the case of this sample, were sufficiently "benign" to allow resumption of more or less normal physical activities. The following results were obtained: (1) Women with low baseline bone mass incurred about three times as many retrospective benign fractures as the women with high baseline bone mass. (2) Women with retrospective benign fractures were about three times as likely as nonfracture women to sustain prospective benign fractures; consideration of BMV status did not improve predictive power. (3) Rate of interval bone loss is no different between the prospective fracture women and their nonfracture counterparts; the fracture group, however, shows no evidence of new bone formation via periosteal apposition (as inferred from bone width changes), whereas the nonfracture group is characterized by modest width increases. (4) The men have more massive bones and fewer fractures than the women, with no significant correlations between BMVs and fracture status. These results reaffirm that BMVs, particularly mass, are highly correlated with postmature fracture status in women, but the BMVs are better in "predicting" past fracture than in predicting future first fracture. This would be expected if systemic postfracture remodeling processes either promote resorption or inhibit new bone formation, with the latter response strongly implicated in this study