Prediction of Fractures: is Using the Lowest T-score of a Single Lumbar Vertebra Just as Good as the Composite Lumbar Spine T-score?

Program: Abstracts - Orals, Featured Poster Presentations, and Posters
Session: SAT 248-267-Osteoporosis II
Clinical
Saturday, June 15, 2013: 1:45 PM-3:45 PM
Expo Halls ABC (Moscone Center)

Poster Board SAT-260
Yewxin Teh*1, Diantha Howard2, Rose C Christian1 and Edward S Leib1
1University of Vermont College of Medicine, Burlington, VT, 2Vermont Center for Clinical and Translational Science, Burlington, VT
Prediction of Fractures: is Using the Lowest T-score of a Single Lumbar Vertebra Just as Good as the Composite Lumbar Spine T-score?

Yew-Xin Teh MD.1, Diantha Howard2, Rose Christian MD.1, Edward Leib MD.1

1University of Vermont College of Medicine, Burlington, VT

2Vermont Center for Clinical and Translational Science, Burlington, VT

Background: In postmenopausal women and men age 50 years and older, the World Health Organization classification using T-score of -2.5 or lower in the lumbar spine, hip or forearm is diagnostic of osteoporosis(1). Many artifacts can alter the lumbar spine bone density measurement. Current guidelines recommend the use of T-scores of at least 2 lumbar vertebrae for the diagnosis of osteoporosis(2). The role of a single lowest lumbar vertebra T-score in predicting fracture risk remains to be determined. We performed a retrospective study comparing the utility of a single lowest lumbar vertebra T-score to the composite lumbar spine T-score in predicting fracture risk. 

Methods: A retrospective, case-control study (ratio = 1:2) of 6750 postmenopausal women and men age 50 years and older who underwent dual-energy x-ray absorptiometry evaluation between 2000 and 2010 was performed. Individuals on osteoporosis treatment were excluded. History of fracture was self-reported by subjects. Odds ratio (OR) for fracture was calculated for single lowest lumbar vertebra T-score (group A), composite L1-L4 lumbar spine T-score (group B), and composite T-score of at least 2 evaluable lumbar vertebrae, defined as no more than 1 standard deviation T-score difference between 2 contingent vertebrae (group C). Logistic regression analysis of different independent variables for fracture risk was performed.

Results: There were 2250 subjects with a history of fracture. Odds ratio for fracture for those with an osteoporosis diagnosis in group A was 1.90 (95% confidence interval 1.67-2.18), group B 1.76 (1.49-2.07), and group C 1.81 (1.56-2.10), and 2.1 (1.78-2.49) with femoral neck T-score. Using logistic regression analysis of BMI and number of clinical risk factors as independent variables of fracture risk, OR for fracture in group A was 2.08 (1.81-2.39), group B 1.92 (1.62-2.27), and group C 1.96 (1.68-2.28), and 2.3 (1.94-2.73) with femoral neck T-score. Odds ratio remained similar for all groups when stepwise logistic regression within each sex group was performed.

Discussion: Our study demonstrates that the OR for fracture using the lowest T-score of a single lumbar vertebra is comparable to composite lumbar spine T-score. This shows that the single lowest lumbar vertebra T-score correlates well with prevalent fractures. Since our study evaluates prevalent fractures only, the predictive ability of the single lumbar vertebra for future fractures cannot be concluded from this study.

(1) Kanis JA, Melton LJ III, Christiansen C, Johnston CC, Khaltaev N. The diagnosis of osteoporosis. J Bone Miner Res. 1994;9(8):1137-1141. (2) Lewiecki EM, Gordon CM, Baim S, Leonard MB, Bishop NJ, Bianchi ML et al. International Society for Clinical Densitometry 2007 Adult and Pediatric Official Positions. Bone. 2008 Dec;43(6):1115-21.

Nothing to Disclose: YT, DH, RCC, ESL

*Please take note of The Endocrine Society's News Embargo Policy at http://www.endo-society.org/endo2013/media.cfm