Major Adverse Kidney Events in the American Indian or Alaska Native Population with Diabetes
Anna Zemke, Kiara N. Mayhand, Radica Z. Alicic, Lindsey M. Kornowske, Cami R. Jones, Kenn B. Daratha, Christina L. Reynolds, Susanne B. Nicholas, Panayiotis Petousis, Leonid Shpaner, Joshua J. Neumiller, Keith C. Norris, Nisha Bansal, Katherine R. TuttleBackground:
Chronic kidney disease (CKD) and diabetes disproportionately affect the American Indian or Alaska Native population but understanding of major adverse kidney events (MAKE) in this population is limited.
Methods:
Electronic health records from the Providence health system identified the American Indian or Alaska Native adult population with diabetes during 2013-2022. A one-year window surrounding diabetes cohort entry was used to collect baseline data. Kaplan-Meier analyses assessed MAKE (≥40% estimated glomerular filtration rate [eGFR] decline, eGFR <15 mL/min/1.73 m 2 , dialysis or kidney transplant, and all-cause death) with propensity score matching (1:3) of American Indian or Alaska Native people to reference individuals (non-Hispanic White) by demographics and clinical characteristics. Cox proportional hazards modeling estimated associations between demographic, clinical, social, and healthcare utilization variables and MAKE.
Results:
The American Indian or Alaska Native population (N=6,103) was younger (mean±standard deviation age 54±15 years) with higher HbA1c (mean 7.4±2.2 mg/dL) compared to the reference population (N=354,283; age 62±14 years; HbA1c 6.9%±1.8%). During a median follow-up of 4.1 (interquartile range 2.0-6.4) years, the American Indian or Alaska Native population experienced a higher frequency of MAKE (26%, n=1,614) than the reference population (24%, n=85,920). With propensity score matching, MAKE survival estimates were significantly lower in the American Indian or Alaska Native population (p <0.0001). In the adjusted Cox model, increased MAKE risk was observed for the American Indian or Alaska Native population (versus reference population, hazard ratio [HR] 1.20, 95% confidence interval [CI] 1.15-1.27), higher social vulnerability index (HR 1.05, 95% CI 1.05-1.05), and hospitalization (HR 1.98, 95% CI 1.95-2.01), while more primary care visits were associated with lower MAKE risk (HR 0.78, 95% CI 0.77-0.79).
Conclusion:
The risk of MAKE was significantly increased in the American Indian or Alaska Native population with diabetes. Social factors and healthcare utilization importantly contributed to risk prediction for MAKE.