Relationship between historical height loss and vertebral fractures in postmenopausal women

Abstract

The aim of this study was to evaluate the relationship between historical height loss (HHL) and prevalent vertebral fractures (VF) in postmenopausal Moroccan women and to estimate its accuracy as a clinical test for detecting VF. Two hundred eighty-eight postmenopausal women were studied. All subjects had bone density measurements and spinal radiographs. Vertebral bodies (T4-L4) were graded using the semi-quantitative method of Genant. HHL was calculated as the difference between a patient's tallest recalled height and the current measured height. The mean age was 58.4 +/- 7.8 years. Thirty-one percent of patients were osteoporotic, and 46.5% had VF. Patients with VF had lost more height than those without VF (median, 2.0 cm (0.26-3.3) vs 0.96 cm (0.33-2.4), p < 0.05). In univariate analysis, HHL was positively correlated to both number and grade of prevalent VF (p < 0.05). The area under the receiver operating characteristics curve for the ability of HHL to detect VF was 0.60 (95% confidence interval (CI), 0.52, 0.69). Our HHL threshold for detecting VF was >1.5 cm, its sensitivity was 58%, and its specificity was 61%. The positive predictive value was 53%, and the negative predictive value was 65%. With HHL >1.5 cm, positive likelihood ratio was 1.49 with 95% CI, 1.07, 2.06. Our results demonstrate significant positive associations between HHL, VF, number of VF, and grade of VF. However, this relationship is not clinically pertinent. Consequently, HHL cannot be used as a reliable clinical test for detecting VF in postmenopausal Moroccan women.