Predicting the Occurrence of New Vertebral Fractures Using the Vertebral Bone Quality Score

Background: Previous studies have recognized the potential of the Vertebral Bone Quality (VBQ) score for predicting fractures. However, these studies often have lacked longitudinal perspectives and have not focused on community populations. Our study aimed to enhance the predictive capacity of the VBQ score by investigating its correlation with new vertebral fractures (NVFs) that were detected 11 years later in a community-based cohort and by developing a comprehensive prediction model. Methods: This study was a population-based study conducted in the Minami-Aizu area in Fukushima Prefecture, Japan. One hundred and thirty participants voluntarily underwent T1-weighted magnetic resonance imaging (MRI) of the lumbar spine in 2004 and 2015. VBQ scores were ascertained from the 2004 scans. NVFs that occurred between 2004 and 2015 were detected based on a >= 20% reduction in vertebral height on the midsagittal sections of the MRI. Other predictors that were considered included age, sex, body mass index, smoking history, heart disease, cerebrovascular disease, respiratory disease, and existing vertebral fractures (EVFs). A logistic regression analysis was conducted. Results: The logistic regression analysis indicated that the VBQ score, age, sex, and EVFs were significant predictors of NVFs. The prediction model showed an area under the curve of 0.84, suggesting excellent discriminatory power. The calibration capacity was confirmed using the Hosmer-Lemeshow test. Conclusions: The VBQ score was significantly correlated with the long-term incidence of NVFs in a community population. The prediction model exhibited satisfactory discrimination and calibration capacities, highlighting the use of the VBQ score as a potential tool for long-term prediction of NVFs.