Jiali Yao, E Shyong Tai, Xueling Sim, Falk Müller-Riemenschneider, Rob M van Dam
Higher total and central fat mass, especially visceral fat, was associated with elevated glucose levels, whereas a greater proportion of leg fat was associated with lower glucose levels in the Asian population. In addition, glycemic variability was larger at both high and low extremes of body fat levels. Our findings highlight that relationships between body fat and blood glucose control depend on aspects of glycemic dynamics (mean levels versus variability), body fat level, and body fat distribution, supporting the importance of accounting for these factors when developing metabolic health strategies.
BACKGROUND: Large-scale studies examining dose-response relationships between body fat and continuous glucose monitoring (CGM)-captured glycemic profiles in adults without diabetes are lacking.
METHODS: In this cross-sectional study, we recruited adults without diabetes from a multi-ethnic Asian cohort. Participants underwent dual-energy X-ray absorptiometry (DXA) scans to quantify body composition, followed by nine days of free-living CGM. CGM-derived mean blood glucose (MBG) and coefficient of variation (CV) were used as the primary metrics to capture glycemic level and glycemic variability, respectively.
RESULTS: We included 882 participants (mean age: 40.2 [SD: 13.8] years; 63% female) who contributed 7858 days of CGM for analysis. Associations of whole-body and regional body fat measures with MBG were predominantly linear and positive, with visceral fat percentage showing the strongest positive linear association with CGM-derived MBG (β per standard deviation of fat: 3.76 mg/dl; 95% CI: 2.87, 4.65). Leg fat percentage was an exception, exhibiting a negative linear association with MBG. U-shaped nonlinear associations were observed between most body fat measures and CGM-derived CV. For example, visceral fat mass (Bonferroni adjusted p-nonlinearity <0.001) was inversely associated with CV in the lower fat range below the inflection point (β: -1.73; CI: -2.62, -0.85) but positively in the higher fat range (β: 0.91; CI: 0.30, 1.51).
CONCLUSIONS: Higher total and central fat mass, especially visceral fat, was associated with elevated glucose levels, whereas a greater proportion of leg fat was associated with lower glucose levels in the Asian population. In addition, glycemic variability was larger at both high and low extremes of body fat levels. Our findings highlight that relationships between body fat and blood glucose control depend on aspects of glycemic dynamics (mean levels versus variability), body fat level, and body fat distribution, supporting the importance of accounting for these factors when developing metabolic health strategies.