#908 The effect of obesity and fat distribution on estimated glomerular filtration rate (eGFR): A Mendelian randomization and image data analysis study

Abstract Background and Aims Obesity is intimately linked to the initiation and progression of chronic kidney disease, but questions on causality still remain. The diagnosis of chronic kidney disease is often based solely on creatinine- and/or cystatin C-based estimated glomerular filtration rate (eGFR), but both these endogenous biomarkers are also known to be related to body composition. This study aims to further investigate how obesity causally affects eGFR, derived from both biomarkers. Method Methods used were: Traditional cross-sectional analyses and magnetic resonance imaging data analyses performed in the POEM study (502 subjects, all aged 50 years), and two-sample Mendelian randomization using already published summary data. Outcome variables for all analyses were eGFR indexed to body surface area, based on creatinine and cystatin C, respectively. In both the traditional cross-sectional and the Mendelian randomization analyses body mass index and waist circumference were used as exposure variables. In the image data analyses, eGFR was regressed nonparametically on fat content and tissue volume for each 3D voxel. Results were visualized as a correlation “imiomics” map. Results Results from the POEM study showed in an adjusted model a negative correlation between cystatin C calculated eGFR and both body mass index [beta = −0.190 (95% CI: −0.280 to −0.100)] and waist circumference [beta = −0.160 (95% CI: −0.250 to −0.060)]. In contrast, a positive association was found for eGFR when using creatinine levels [beta = 0.120 (95% CI: 0.030 to 0.210) for body mass index; beta = 0.160 (95% CI: 0.070 to 0.260) for waist circumference]. This pattern was repeated in the image data analysis, which was visualized in the imiomics map: both visceral and subcutaneous adipose tissue were negatively correlated to eGFR based on cystatin C but positively correlated to eGFR based on creatinine (Fig. 1). Mendelian randomization implied a negative causal effect of obesity-related measures on cystatin C-based eGFR, with statistically significant estimates for both body mass index [beta = −0.031 (95% CI: −0.037 to −0.026)] and waist circumference [beta = −0.038 (95% CI: −0.045 to −0.031)]. No statistically significant effects was found on creatinine-based eGFR (Fig. 2). Conclusion This study indicates a causal negative effect of obesity on cystatin C-based eGFR, but not on eGFR calculated from creatinine. This discrepancy is further highlighted by analyses of adipose tissue from magnetic resonance imaging. These findings thus warrant further research on the impact of using the endogenous biomarkers creatinine and cystatin C to estimate renal function when assessing the link between obesity and chronic kidney disease.