Abstract
PURPOSE: Relative fat mass (RFM) is a new metric used for obesity assessment. We aim to investigate the association between RFM and vitamin D deficiency in patients with diabetes. METHODS: A total of 5,128 participants with diabetes mellitus from the NHANES 2007-2018 and an external Chinese validation cohort of 238 subjects from the Affiliated Wujin Hospital of Jiangsu University were analyzed. Logistic and linear regression, subgroup and curve fitting analyses were performed to assess the relationships between RFM and vitamin D deficiency risk as well as serum 25(OH)D levels. Receiver operating characteristic (ROC) and decision curve analysis (DCA) were applied to compare diagnostic efficacy among RFM, body mass index (BMI), waist circumference (WC), and height. RESULTS: Vitamin D deficiency prevalence increased with rising RFM levels (P<0.001). Higher RFM was significantly associated with increased risk of vitamin D deficiency (OR = 1.056, 95%CI= (1.039, 1.073), P<0.001) and lower 25(OH)D levels (β=-0.662, 95%CI= (-0.852, -0.471), P<0.001) in patients with diabetes. ROC and DCA indicated that RFM yielded the highest discrimination for vitamin D deficiency (AUC = 0.626), outperforming BMI (0.592), WC (0.567), and height (0.492). The associations remained robust in various subgroups and were confirmed in the external Chinese population. CONCLUSIONS: RFM is superior to conventional obesity measures in identifying individuals with diabetes at high risk for vitamin D deficiency. RFM may help to improve clinical risk stratification and management.