Osteoporosis, an Underdiagnosed Disease

THE RESULTS FROM THE NATIONAL OSTEOPOROSIS RISK Assessment (NORA), the largest study of osteoporosis conducted to date in the United States, have been eagerly awaited by osteoporosis researchers and clinicians. NORA, which commenced in 1997, is a longitudinal observational study involving more than 200000 postmenopausal women. The study was initially designed to assess the association between osteoporosis risk factors and low bone mineral density (BMD) and to examine the relationship between BMD and other risk factors and shortterm (1-year) fracture incidence. In this issue of THE JOURNAL, Siris et al report the study’s initial findings of an unexpectedly high prevalence of unsuspected osteopenia and osteoporosis (as determined by BMD) among generally healthy postmenopausal women followed up in primary care practices. The study also confirms the utility of measurement of BMD at peripheral bone sites using several devices to predict short-term fracture risk. The importance of such findings, indicating a need for greater attention to the identification and management of osteoporotic risk, should not be underestimated. In NORA, 200160 postmenopausal women underwent peripheral bone densitometry or ultrasonography of heel, finger, or forearm in their physicians’ offices. They also completed questionnaires assessing risk factors and, approximately 1 year later, new skeletal fractures. Participants did not have aprevioushistoryofosteoporosis andwerenotcurrently treated with US Food and Drug Administration (FDA)–approved osteoporosis therapies except for estrogen replacement. Overall, approximately 40% of tested women had a peripheral BMD measurement denoting osteopenia (T score of −1 to −2.49), and approximately 7% had BMD in the osteoporotic range (T score −2.5) according to the World Health Organization classification for significant bone loss. At baseline, 11% of women reported at least 1 fracture after age 45 years. Among the 163979 participants with follow-up information, a BMD classification of osteoporosis was associated with a fracture rate approximately 4 times that of a normal BMD. Moreover, osteoporosis (defined by baseline BMD) increased the risk of incident fracture within 1 year by 2.7 times, and osteopenia increased the risk of incident fracture by 1.7 times. This study also confirms previously noted risk factors for osteoporosis, including age, history of previous fracture, smoking, and glucocorticoid use. Estrogen replacement, diuretic use, exercise, a high body mass index, and, interestingly, alcohol consumption decreased the likelihood of osteoporosis. In an assessment of race/ethnicity as a risk factor, Asian or Hispanic heritage was associated with an increased risk of osteoporosis, and African American heritage decreased the risk. However, as determined by BMD, 32% of African American women were osteopenic and 4% were osteoporotic, indicating a substantial absolute risk of osteoporotic fracture and suggesting that African American heritage may not confer full protection against osteoporosis. Assets of NORA include enrollment of a large number of geographically and ethnically diverse women; also, the women were primarily enrolled from clinical practice, providing a realistic cross-sectional and longitudinal assessment of osteoporosis risk in primary care practices in the United States. In addition, this study provides a stringent and scientifically rigorous evaluation of the utility of the peripheral densitometry/ ultrasonography technologies for screening large groups of women for bone loss, and generally supports the validity of these modes in screening of large populations for osteoporotic risk. Given the current attention to other predictors and mediators of fracture risk (eg, bone quality and turnover), reinforcement of the parameter of bone quantity (BMD) as a clinical mode for defining fracture risk is appropriate. A number of limitations of the study are addressed by the authors, including a potential underestimation of the true prevalence of osteoporosis in the study cohort by (1) exclusion of women with previous diagnosis or treatment of osteoporosis (except for hormone replacement therapy); (2) recording of risk factors and, particularly, fracture prevalence and incidence by self-report and recall without medical record or x-ray corroboration; and (3) the inability of the study to assess nonclinical (ie, asymptomatic) spine fracture. The latter criticisms are of concern because while self-reporting of fractures has been generally accurate, many osteoporotic spine fractures are unrecognized by both patient and practitioner. However, the lack of full documentation of skeletal fracture in the study is balanced and necessitated by the robustness of the study size and the desired use of a clinical prac-