Assessing fat fraction and R2* value of lumbar spine based on MRI as a marker of bone mineral density.

Abstract

The study aimed to investigate the correlation between fat fraction (FF), R2* value of vertebrae based on IDEAL-IQ sequence and bone mineral density (BMD) based on QCT, and their diagnostic value for low BMD and osteoporosis. Subgroups were divided according to different gender, age, BMI and bone mass to compare the differences in parametric variables. One-way ANOVA, independent samples t-test, correlation coefficient analysis, linear regression analysis and ROC curve analysis were performed. Significant differences were found in FF among different bone mass groups, and between different gender and age groups. While R2* only had significant difference between different gender groups and males with different age. BMD was significantly negatively correlated with FF, especially in women, and FF significantly negatively affected BMD after controlling for gender, age and BMI. There was mildly positive correlation between BMD and R2* in men, and R2* significantly positively influenced BMD controlling for the confounders. In addition, FF was positively correlated with age, whereas R2* was negatively correlated with age in men. FF had high diagnostic efficacy for low bone mass and osteoporosis, while R2* alone was weakly diagnostic. Vertebral FF can be served as a potential important imaging biomarker for assessing low BMD and osteoporosis, and R2* of males can be utilized as an complementary parameter for evaluating osteoporosis. The IDEAL-IQ sequence has the potential to be used as an accessory examination in the diagnosis of osteoporosis, assessment of treatment efficacy and prediction of fracture risk.