Kidney Week

Abstract: SA-PO1100

Discrepancy between Lifetime Risk of ESKD vs. 3-Year Risk of CKD Progression in US Adults with Diabetes

Session Information

• CKD: Epidemiology, Risk Factors, and Prevention - 3
  October 26, 2024 | Location: Exhibit Hall, Convention Center
  Abstract Time: 10:00 AM - 12:00 PM

Category: CKD (Non-Dialysis)

• 2301 CKD (Non-Dialysis): Epidemiology, Risk Factors, and Prevention

Authors

• Obi, Yoshitsugu, University of Mississippi School of Medicine, Jackson, Mississippi, United States

Yoshitsugu Obi, MD, PhD, FASN

• Zhu, Xiaoqian, University of Mississippi School of Medicine, Jackson, Mississippi, United States

Xiaoqian Zhu, PhD

• Tio, Maria Clarissa, University of Mississippi School of Medicine, Jackson, Mississippi, United States

Maria Clarissa Tio, MD, MPH, FASN

• Yen, Timothy E., University of Mississippi School of Medicine, Jackson, Mississippi, United States

Timothy E. Yen, MD

• Hall, Michael E., University of Mississippi School of Medicine, Jackson, Mississippi, United States

Michael E. Hall, MD, MS

• Dossabhoy, Neville R., University of Mississippi School of Medicine, Jackson, Mississippi, United States

Neville R. Dossabhoy, MD, FASN

• Shafi, Tariq, Houston Methodist, Houston, Texas, United States

Tariq Shafi, MBBS, MHS, FASN

Background

The CKD Prognosis Consortium developed a model predicting CKD progression (i.e., ≥40% eGFR decline or end-stage kidney disease [ESKD]) in three years (Grams, Diabetes Care 2022). However, the risk of CKD progression is cumulative over time and the lifetime risk of ESKD is also important to assess. We sought to compare the 3-year CKD progression risk with the lifetime ESKD risk using the Swedish prediction model (Østergaard, CJASN 2022).

Methods

Among 5,284 US adults with diabetes who participated in the 1999-2020 NHANES, we estimated the lifetime ESKD risk by calibrating baseline hazards for mortality and ESKD in the Swedish prediction model. Since ESKD incidence was not available, we used the Kidney Failure Risk Equation and obtained the calibration factor for ESKD. We then evaluated their correlation by comparing risk percentiles using spearman’s rho and risk quartiles using concordance kappa.

Results

Mean age was 63 years, 47% were women, median HbA1c was 7.0%, and median diabetes duration was 12 years. Mean eGFR was 82 (SD, 21) mL/min/1.73 m², with 65% and 32% having moderate and severe albuminuria, respectively. Median [interquartile range] 3-year CKD progression risk was 2.6% [1.5% to 5.2%], and median lifetime ESKD risk was 1.6% [0.9% to 3.0%]. There was only a fair correlation between the two risk estimates (rho=0.49, Figure) and only a slight agreement between the risk quartiles (kappa=0.17). The correlation and concordance were poorer among people with eGFR ≥60 mL/min/1.73 m² (rho=0.34; kappa=0.11). Additionally, the lifetime ESKD risk percentiles were higher than the 3-year CKD progression risk percentiles among people with longer life expectancies, and vice versa.

Conclusion

Shorter term risk of CKD progression may not be sufficient to identify individuals with high lifetime ESKD risk. Clinicians should consider lifetime ESKD risk in addition to short-term CKD progression risk when counseling people with diabetes.

Funding

• Other NIH Support