| Summary: | Background. The aim of a longitudinal, retrospective study was to establish variables predicting fracture incidence over a decade. Methods. The study sample comprises a group of 457 postmenopausal women aged over 55 years, recruited from the database of an outpatient osteoporotic clinic. Several variables with potential influence on bone status, including the measurement of body height and hip bone densitometry, were collected. BMD at the femoral neck (FN BMD) was established using a Prodigy device (Lunar, GE, USA). Current body height was compared with the maximal historical body height in early adulthood, as reported by the patient. Results. Three hundred and ninety-four women did not have fractures during the follow up, and 63 subjects presented fractures. Subjects with fracture had lower FN BMD with a T-score of −1.86 ± 1.04 compared to −1.44 ± 0.89 in those without fractures (<i>p</i> < 0.001). Mean height loss (HL) was 3.47 ± 2.11 cm in fractured subjects and 2.50 ± 2.47 cm in unfractured ones, and differed significantly, <i>p</i> < 0.01. Fracture incidence was significantly related to age, rheumatoid arthritis, falls, and previous fractures. In the multivariable analysis using logistic regression, FN BMD, baseline fracture, and HL were identified as the significant predictors of fractures of follow up. Conclusions. Osteoporotic fractures in postmenopausal women were predicted by FN BMD, prior fracture(s), and HL easily established during physical examination.
|
|---|