Fracture rates calculated from fracture histories in normal postmenopausal women.

STUDY OBJECTIVE The aim was to estimate fracture rates and fracture prevalence from fracture histories in normal postmenopausal women. DESIGN Apparently healthy postmenopausal women were recruited by advertising in the media. Fracture histories were obtained by personal interview in 1983 and again by interview or questionnaire in 1988. Fracture rates were calculated prospectively and retrospectively. PARTICIPANTS 492 women (mean age 58.6 years) were selected from over 1000 applicants on the basis that they were suffering from no disease nor taking any therapy which might affect their bones. MEASUREMENTS AND MAIN RESULTS Retrospective premenopausal and postmenopausal fracture rates were calculated in 1983, prospective rates calculated from 1983 to 1988, and retrospective rates checked again on the second occasion. The retrospective and prospective fracture rates were very similar. The five year fracture rates were low and steady until the menopause, when they rose by a factor of 10 and reached a new plateau after about 15 years. The results were comparable to those obtained from hospital statistics. CONCLUSIONS The menopausal rise in fracture rates not only involves wrist fractures but most peripheral fractures, and probably reflects postmenopausal loss of trabecular bone. Meaningful fracture rates can be calculated from individual fracture histories in a well defined population. This technique may be particularly useful in developing countries where public health data may be incomplete.