Abstract Purpose To assess associations between coronal knee alignment and regional epiphyseal bone mineral density (BMD) using computed tomography (CT)‐derived Hounsfield units (HU) and compare medial‐to‐lateral ratios (MLRs) across alignment groups, and as a secondary aim, to evaluate the association of sex and demographic factors with regional BMD patterns. Methods A retrospective search (2008–2025) identified patients with non‐contrast knee CT and standing anteroposterior long‐leg radiographs. Exclusions included prior femur or tibia fracture or surgery and metabolic bone disease other than osteopenia or osteoporosis. Mean trabecular HU were measured for medial and lateral femoral condyles (MFC, LFC), distal femur epiphysis (DFE), medial and lateral tibial plateaus (MTP, LTP) and proximal tibia epiphysis (PTE). Outcomes included DFE‐MLR (MFC/LFC), PTE‐MLR (MTP/LTP) and combined‐MLR (MFC + MTP/LFC + LTP). Alignment was categorized by hip–knee–ankle angle (HKAA) as major varus, varus, neutral, valgus or major valgus. Results The cohort included 312 patients (mean age 66.5 ± 9.3 years; 50.6% female); 99.4% underwent CT for robotic‐assisted arthroplasty planning. LFC HU exceeded MFC HU in 88.5% of cases ( p < 0.001), and MTP HU exceeded LTP HU in 92.9% ( p < 0.001). Increasing valgus correlated with decreasing MLR (DFE‐MLR r = −0.459; PTE‐MLR r = −0.322; combined‐MLR r = −0.509; all p < 0.001). Females demonstrated higher PTE‐MLR and combined‐MLR than males ( p ≤ 0.049). Conclusions Coronal knee alignment was strongly associated with compartmental epiphyseal BMD MLR, with lower MLR in valgus alignment and greater alignment sensitivity in females. These findings provide quantitative imaging context relevant to arthroplasty planning. Level of Evidence Level III, retrospective comparative study.
Klinger et al. (Wed,) studied this question.