Development and evaluation of an index to predict early postmenopausal bone loss

Abstract

An index to predict individual postmenopausal bone loss is presented. The index is developed by means of data from a 10-year prospective Norwegian study in which bone mass of the distal forearm was measured annually in 73 women. All the women were 47 years old and premenopausal at inclusion. Independent risk factors for postmenopausal bone loss were identified by applying multivariate regression analysis on anthropometric, biochemical, nutritional, and life-style variables measured at menopause. The analysis identified low body weight, reduced renal phosphate reabsorption, and smoking as significant independent risk factors, and by means of these three factors a predictive index for postmenopausal bone loss was developed. This index was validated by using data from a 10-year longitudinal Dutch study, in which bone mass of the proximal radius was measured annually in 86 women, aged between 49 and 57 years and perimenopausal at inclusion. We defined women with the highest index score as "high-risk persons." According to this definition approximately 25% of the perimenopausal women were classified as high-risk persons, and the estimated sensitivity/specificity/positive predictive power were 36%, 89%, and 74%, respectively, when used to select women with a postmenopausal bone loss above average. We conclude that the index may be helpful in identifying healthy perimenopausal women in whom bone mass measurements should be considered.