Abstract
Background/Objectives: This retrospective observational cohort examined whether opportunistic CT-derived Hounsfield units (HU) of the lumbar spine and proximal femur together with serum 25-OH-vitamin D can predict sacral insufficiency fractures (SIF) and osteoporosis. No interventional procedures were performed. Methods: Consecutive suspected SIF cases over 3 years (n = 253) were assigned to SIF (n = 98) or controls without SIF or spine/hip fractures (n = 155). HU were measured using ellipsoidal ROIs at L1-L3 and an irregular area ROI across the entire proximal femoral cancellous bone; vitamin D was quantified; ROC analyses assessed discrimination. HU cut-points were referenced via HU-to-QCT/CTXA conversions. Results: SIF patients had markedly lower HU than controls (lumbar 44.84 vs. 105.66 HU; femoral 47.0 vs. 148.0 HU). Diagnostic performance was excellent (AUC 0.98 for SIF discrimination using lumbar HU; AUC 0.98 for osteoporosis prediction using femoral HU). Vitamin D deficiency (<20 ng/mL) was highly prevalent (92.9%) with lower means in SIF (3.72 vs. 8.24 ng/mL). Within SIF, patients with hip fracture had femoral HU ≈ 14.2 vs. 70.6 without hip fracture; effect sizes were very large. Conclusions: Opportunistic HU assessment from routine CT provides a rapid, reproducible surrogate of bone density that distinguishes SIF with near-perfect accuracy and identifies osteoporosis. HU thresholds around ~96-98 are consistent with osteoporotic ranges and can be implemented to trigger metabolic evaluation and early osteoanabolic therapy where appropriate.