Proximal Femur Bone Density Predicts Fractures in Women Better Than That of Lumbar Spine

August 23, 2007 — Proximal femur bone density measurements consistently outperformed lumbar spine measurements for global fracture prediction in women, according to the results of a historical cohort study reported in the August 13/27 issue of the Archives of Internal Medicine.
"Bone density measurement with dual-energy x-ray absorptiometry [DXA] is widely used for fracture risk assessment," write William D. Leslie, MD, MSc, from the University of Manitoba in Winnipeg, Canada, and colleagues from the Manitoba Bone Density Program. "Discordance between measurement sites is common, but it is unclear how this affects fracture prediction."
Using a database containing all clinical DXA test results for the province of Manitoba, Canada, the investigators identified a study cohort of 16,505 women aged 50 years or older at the time of baseline DXA of the spine and hip. The mean period of observation was 3.2 ± 1.5 years.
After bone density testing, each woman's longitudinal health service record was evaluated for the presence of fracture codes. The likelihood ratio test was used to determine the improvement in fracture prediction from Cox proportional hazards models with bone density covariates from a single site or from combined sites.
For osteoporotic fracture, age-adjusted hazard ratios (HRs) per SD ranged from 1.61 (95% confidence interval [CI], 1.39 - 1.87) for the lumbar spine to 1.85 (95% CI, 1.70 - 2.01) for the total hip. HRs were intermediate for the femoral neck (HR, 1.76; 95% CI, 1.62 - 1.92) and trochanter (HR, 1.77; 95% CI, 1.63 - 1.92).
Using the minimum bone density measurement was no better than using a hip measurement alone in predicting fractures. As long as the total hip measurement was included in a fracture prediction model for the overall population, none of the other measurements added significant predictive value. In the prediction of spinal fractures alone, the spine was the most useful site.
Limitations of the study include ascertainment of fractures from administrative health data and inclusion of few women of nonwhite race or ethnicity, limiting generalizability.
"Proximal femur bone density measurements consistently outperformed lumbar spine measurements for global fracture prediction," the study authors write. "In this cohort, the total hip was the best site for overall fracture assessment."
The CHAR/GE Healthcare Development Awards Programme funded this study in part. Dr. Leslie has disclosed various financial relationships with Merck Frosst Canada and Proctor & Gamble Pharmaceuticals.
Arch Intern Med. 2007;167:1641-1647.