Introduction: The measurement of lumbar bone mineral density (BMD) in DISH patients using Dual-energy X-ray absorptiometry (DXA) scans may be prone to overestimation due to the presence of osteophytes in the anterior lumbar column. Spine surgeons are increasingly interested in utilizing routine CT or MRI scans to assess vertebral BMD, rather than relying primarily on DXA. This study aimed to investigate the diagnostic effectiveness of DXA, CT, and MRI in assessing lumbar BMD for individuals with diffuse idiopathic skeletal hyperostosis (DISH).
Methods: Retrospective analysis of 105 DISH patients and 116 age-matched controls with lumbar spinal stenosis was conducted. Lumbar BMD was evaluated using DXA-based T-scores, MRI-based vertebral bone quality (VBQ) scores, and CT-based Hounsfield unit (HU) values. Patients were categorized into three BMD subgroups. Lumbar osteophyte categories were determined by Mata score. Demographics, clinical data, T-scores, VBQ scores, and HU values were collected. Receiver operating characteristic (ROC) analysis identified VBQ and HU thresholds for diagnosing osteopenia/osteoporosis using DXA in controls. Correlations between VBQ, HU, and T-scores were analyzed.
Results: Age, gender, and BMI showed no significant differences between DISH and control groups. DISH patients had higher T-scores and Mata scores. VBQ and HU did not significantly differ between groups. In controls, VBQ and HU effectively diagnosed osteopenia/osteoporosis (AUC=0.857 and 0.910, respectively) with cutoffs of 3.0 for VBQ and 104.3 for HU. DISH had lower osteopenia/osteoporosis prevalence using T-scores (30.5% vs. 41.4%, P < 0.05), but no significant differences using VBQ (42.9% vs. 43.8%, P＞0.05) and HU (41.9% vs. 42.2%, P＞0.05). Correlations revealed moderate correlations between HU and T-scores in DISH (r=0.642, P＜0.001), and strong in controls (r=0.846, P＜0.001). Moderate negative correlations were observed be-tween VBQ and T-scores in DISH (r=-0.450, P＜0.001), and strong in controls (r=-0.813, P＜0.001).
Discussion: In this study, DEXA, CT, and MRI were employed to analyze the disparity in lumbar BMD between DISH group and the control group. Our results showed that: 1) DEXA has the potential to overestimate lumbar BMD due to the presence of osteophytes anterior to the lumbar spine, and the degree of overestimation increases with the size of the osteophyte; 2) MRI-based VBQ scores and CT-based HU values could potentially function as additional diagnostic tools for assessing cancellous bone mineral density in the lumbar region when T-score derived from DEXA is overestimated in patients with DISH. This study represents the first report on the diagnostic efficacy of VBQ scores and HU values for assessing bone mass quality in DISH patients regardless of osteophytes.