Renal fat fraction quantified by MRI-Dixon predicts the renal function progression in patients with type 2 diabetes mellitus: a prospective cohort study.

BACKGROUND: Lipid accumulation occurs in kidneys and leads to cellular damage. We aimed to investigate whether renal fat fraction (FF) is an independent predictor for the renal function progression in patients with type 2 diabetes mellitus (T2DM). METHODS: Renal fat fraction (FF) was measured by MRI-Dixon imaging technique. 174 patients with T2DM were included in cross-sectional analysis, and 111 patients were enrolled in longitudinal analysis. After followed up for a median of 18.0 months, 21 of the 111 participants developed renal composite endpoint events. The outcomes were 30% decrease in eGFR from baseline, progression to chronic kidney disease (CKD) (eGFR < 60 ml/min/1.73m2) or end-stage renal disease (ESRD) (eGFR < 15 ml/min/1.73m2). RESULTS: In cross-sectional analysis, multiple linear regression analysis showed that the renal FF was negatively correlated with baseline eGFR, with β coefficients (95% CI) of -16.03 (95% CI: -19.58 to -12.48) after confounders adjustment. In the prospective analysis, cox regression model revealed that each SD increment in baseline renal FF was associated with a higher incidence of renal composite endpoint events (HR = 3.78, 95% CI: 1.22–11.71). When renal FF values were categorized into tertiles, individuals in the highest tertile had an increased risk of renal composite endpoint events compared to those in the lowest tertile (HR = 18.27, 95% CI: 1.03-322.83). Furthermore, the combination of renal FF and uACR improved the predictive effect of renal composite endpoint events, with AUC of 0.78 (95% CI: 0.66–0.89). CONCLUSION: Renal FF was closely associated with the renal function and may serve as a predictor of the renal function progression in patients with T2DM.