Cervical Spine Computed Tomography Hounsfield Units Accurately Predict Low Bone Mineral Density of the Femoral Neck

Study Design: A retrospective cohort. Objective: The objective of this study to determine the correlation between Hounsfield unit (HU) measurements from the C4 vertebral body and dual-energy x-ray absorptiometry (DXA) T-score. Summary of Background Data: Recent attention has turned to the utilization of HU measurements from computed tomography (CT) as a potential screening method for low bone mineral density (BMD). We hypothesized that cervical spine CT HU measurements will correlate with BMD measurements conducted with DXA scans of the femoral neck. Material and Methods: Patients with cervical CT and femoral neck DXA scans at 1 institution were included in the study. HUs were manually measured from the cancellous bone in the C4 vertebrae by 1 author blinded to DXA scans. HU measurements were compared with femoral neck DXA T-scores for the entire population. Results: A total of 149 patients with 149 cervical CT and femoral neck DXA scans were included in the study. The low BMD group (osteoporotic and osteopenic combined) showed a significant difference in HU compared with the normal groups within the study (P<0.0001). A low BMD screening value of 447 HU captured over 95% of patients with low BMD within our study and provided a sensitivity of 92% and negative predictive value of 82.1%. The male and female intrasex analysis demonstrated a significant difference between the low BMD and normal BMD groups with P=0.001 and P=0.0001, respectively. Conclusions: HU measurements taken from the C4 vertebral body on CT scan correlate with low BMD of the femoral neck as determined by DXA scan T-scores. Screening values of 447 HU captured 95% of patients with low BMD, with a high degree of sensitivity, and negative predictive value of 80%. Utilization of cervical spine HU as a screening method provides a simple, quick, and easily assessable screening tool for assessing low BMD. Level of Evidence: Level III-diagnostic.