Background/Purpose: Predicting which individuals are at risk to experience a fracture and modify that risk is important in preventative health.  The WHO’s Fracture Risk Assesment (FRAX) defines osteopenia and osteoporosis in terms of the femoral neck T-score and allows use of total proximal femur,  but the lumbar spine site is frequently used in clinical practice and its use in diagnosis is supported by the International Society of Clinical Densitometry and the National Osteoporosis Foundation.  Our study aim is to quantify the impact of spine bone mineral density (BMD) on fracture risk prediction and determine the positive predictive value of fracture prediction by using the lowest BMD value at the femoral neck, total hip, or lumbar spine and compare this to the femoral neck alone.   

Methods: We performed a retrospective cross-sectional analysis of our database of 15,033 post-menopausal women combining clinical risk factors (CRF) and bone density (BMD) results collected over 9.2 years utilizing a GElunar densitometer.  We validated our database by showing that age, low BMD and CRFs in our population correlate with presence of fracture.  We performed a logistic regression to assess the contribution of age, BMI, number of CRFs, T-score, and WHO osteoporosis classification category to the presence of fracture.  T-scores were differentiated as femoral neck, total hip, and lumbar spine.

Results:   The percent of subjects with fracture was lower than that reported in other studies.  In individuals with normal BMD at the femoral neck, there were few who were osteoporotic at the lumbar spine (0.7%) and more who were osteopenic at the femoral neck and osteoporotic at the lumbar spine (9.7%).  In patients whose T-scores are 1 or 2 osteoporosis categories lower at the lumbar spine than femoral neck, there is an approximately 30% increased risk of fracture when compared with the femoral neck alone.  For patients less than 60 years old, the odds ratio of having a fracture based on presence of lumbar spine osteoporosis was greater than the odds ratio based on femoral neck osteoporosis.  This reversed for those ≥ 65 compared with those <65 years old.  For each age category, the presence of osteoporosis measured at the total hip correlated best with presence of hip fracture and was better than taking the lowest T-score at any of the three sites (femoral neck, total hip, or lumbar spine). 

Conclusion: It is most important to measure BMD at the lumbar spine in younger, post-menopausal women for fracture prediction. In our population total hip BMD is the best predictor of fracture. The spine BMD appears to be a better predictor of fracture than femoral neck in women 60 years and younger. When using the WHO Fracture Prediction Tool (FRAX), we recommend that the 10 year fracture prediction be adjusted when the lumbar spine T-score is 1-2 osteoporosis categories lower than the femoral neck T-score.

---

« Back to 2012 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/utility-of-spine-bone-mineral-density-in-fracture-prediction-within-the-fracture-risk-assessment-tool-frax/