Long-Term Study Confirms Fractures in Postmenopaus

Newswise — A long-term study of over 170,000 women in the U.S. confirms that previously identified risk factors in postmenopausal women, as well as a few surprising factors, predict future fracture risk, according to research presented this week at the American College of Rheumatology Annual Scientific Meeting in Washington, DC.

Postmenopausal women have markedly reduced levels of estrogen, leading to loss of bone and damage to bone quality. As a result, there are over 34 million Americans with low bone mass and 10 million with osteoporosis who are susceptible to fractures. These fractures, many of which are associated with a high rate of morbidity, increased mortality, and impaired quality of life, amount to 1.5 million each year in the U.S.

To evaluate predictors of new fractures in women from throughout the U.S. who were enrolled from their primary care physicians' practices, researchers tapped the large and diverse population of women participating in the National Osteoporosis Risk Assessment (NORA) study. At baseline, NORA enrolled over 200,000 postmenopausal women over age 50 with no prior diagnosis of osteoporosis and no history of osteoporosis-specific medications. These women received an initial bone mineral density test and completed a baseline survey.

Out of the 170,314 women completing follow-up surveys at year one, two and five, 7,989 reported new fractures. The strongest predictors of fracture proved to be a history of fracture after age 45, indications of osteoporosis and osteopenia as determined by bone density testing, increasing age, and poor/fair self-rated health. Surprisingly, self-reported loss of height and depression also predicted incident fractures. Compared to Caucasians, Black and Asian women had lower odds of fracture. The use of estrogen at the outset of the study also was associated with lower risk.

"This very large study of women ranging in age from 50-99 confirms that risk factors traditionally associated with future fracture risk should be taken very seriously by physicians and patients alike," said Ethel Siris, MD, Madeline C. Stabile Professor of Clinical Medicine, College of Physicians and Surgeons at Columbia University, New York, NY, and the principal investigator in the study. "It underscores the importance of assessing these risk factors periodically over time and of performing bone mineral density tests in menopausal women, according to current guidelines, something not being done for the majority of women in the U.S. The NORA data also reinforces the need for physicians to screen for depression during routine primary care, as this is a common and serious problem that also turns out to predict risk for fractures."

The American College of Rheumatology is the professional organization for rheumatologists and health professionals who share a dedication to healing, preventing disability and curing arthritis and related rheumatic and musculoskeletal diseases. For more information on the ACR's annual meeting, see http://www.rheumatology.org/annual.

Presentation Number: 1417

Predictors of Five-Year Incident Fractures: Results From the National Osteoporosis Risk Assessment Program (NORA)

Ethel S. Siris1, Elizabeth Barrett-Connor2, Paul D. Miller3, Thomas W. Weiss4, Colleen A. McHorney5. 1Columbia University, College of Physicians and Surgeons, New York, NY; 2UC at San Diego, La Jolla, CA; 3Colorado Center for Bone Research, Lakewood, CO; 4Merck & Co., Inc, West Point, PA; 5Merck & Co., Inc., West Point, PA

PURPOSE: Few contemporary longitudinal cohort studies exist that track fractures over several years among postmenopausal women. The NORA study afforded an opportunity to model predictors of fracture over five years among postmenopausal women initially untreated for osteoporosis.

METHODS: From 1997-1999, NORA enrolled over 200,000 women from 49 US states and DC in a longitudinal registry of PMW. Eligibility criteria were: 50 or older, at least six months post last menses, no prior diagnosis of osteoporosis, no BMD testing in the preceding year, and no specific osteoporosis medication use. Baseline BMD testing at a peripheral site was performed. At the Year 1, 2, and 5 surveys, women reported incident fractures. Relative risk (RR) for incident fractures across the five surveys was calculated using Cox regression models. We used baseline demographic and osteoporosis risk variables as independent variables.

RESULTS: A total of 170,134 participants completed at least one of the follow-up surveys, and a total of 7,989 incident fractures were reported. The strongest predictors of fracture across the five years were: (1) history of a fracture prior to baseline (RR=2.35, CI=2.23-2.48); (2) T-score <=-2.50 ("osteoporosis" ) (RR= 2.39, CI=2.22-2.59); (3) T-score <=-1.0 to -2.49 ("osteopenia" ) (RR=1.66, CI=1.57-1.74); (4) poor/fair (vs. excellent) self-rated health (RR=1.71, CI=1.57-1.87); (5) older age compared to age 50-65 (range of RR of 2.02, CI=1.75-2.33, for age 85+ to RR of 1.12, CI=1.04-1.19, for age 65-69; (6) self-reported loss of height compared to 0% loss of height (RR=1.64, CI=1.13-2.38 for 10% + loss of height; RR=1.24, CI=1.05-1.48 for 5%-10% loss of height; RR=1.11, CI=1.06-1.17 for 1%-5% loss of height); (7) depression score in the lower fifth of the scale-score distribution (RR=1.26, CI=1.17-1.37). Compared to Caucasians, Black (RR=0.56, CI=0.48-0.64) and Asian (RR=0.58, CI=0.42-0.80 respectively) had lower odds of fracture. Use of estrogen at baseline (vs. no history of estrogen therapy) was also associated with a lower risk of fracture (RR=0.87, CI=0.83-0.92).

CONCLUSIONS: Some traditional fracture risk factors were corroborated with the NORA five-year fracture data (e.g., age, T-score, black and Asian ethnicity). Interestingly, self-reported loss of height and depression remained significant predictors despite statistical control for other variables. These findings should prove useful in developing fracture risk-prediction models for use in research and practice

Disclosure Block: E.S. Siris, Merck & Co., Inc, 5; Pfeizer Pharmaceuticals, 5; Eli Lilly, 5; Amgen, 5; Novaritis, 5; Proctor & Ganble, 5; E. Barrett-Connor, Merck & Co., Inc, 5; Eli Lilly, 5; Pfizer Pharmaceuticals, 5; P.D. Miller, Merck & Co., Inc, 5; Novartis Pharmaceuticals, 5; Pfizer Pharmaceuticals, 5; Amgen 2, 5; Proctor & Gamble Pharmaceuticals, 5; Aventis Pharmaceuticals, 5; Roche Pharmaceuticals, 5; Eli Lilly, 5; T.W. Weiss, Merck & Co., Inc, 3; C.A. McHorney, Merck & Co., Inc, 3.