Population-based reference values for kidney function and kidney function decline in 25- to 95-year-old Germans without and with diabetes

• Published in: Kidney international Vol. 106; no. 4; pp. 699 - 711
• Main Authors: Herold, Janina M., Wiegrebe, Simon, Nano, Jana, Jung, Bettina, Gorski, Mathias, Thorand, Barbara, Koenig, Wolfgang, Zeller, Tanja, Zimmermann, Martina E., Burkhardt, Ralph, Banas, Bernhard, Küchenhoff, Helmut, Stark, Klaus J., Peters, Annette, Böger, Carsten A., Heid, Iris M.
• Format: Journal Article
• Language: English
• Published: United States Elsevier Inc 01.10.2024
• Subjects: chronic kidney disease; diabetes; general population; kidney function; kidney function decline; reference values; chronic kidney disease; reference values; kidney function; general population; diabetes; kidney function decline
• ISSN: 0085-2538; 1523-1755; 1523-1755
• DOI: 10.1016/j.kint.2024.06.024

Understanding normal aging of kidney function is pivotal to help distinguish individuals at particular risk for chronic kidney disease. Glomerular filtration rate (GFR) is typically estimated via serum creatinine (eGFRcrea) or cystatin C (eGFRcys). Since population-based age-group-specific reference values for eGFR and eGFR-decline are scarce, we aimed to provide such reference values from population-based data of a wide age range. In four German population-based cohorts (KORA-3, KORA-4, AugUR, DIACORE), participants underwent medical exams, interview, and blood draw up to five times within up to 25 years. We analyzed eGFRcrea and eGFRcys cross-sectionally and longitudinally (12,000 individuals, age 25-95 years). Cross-sectionally, we found age-group-specific eGFRcrea to decrease approximately linearly across the full age range, for eGFRcys up to the age of 60 years. Within age-groups, there was little difference by sex or diabetes status. Longitudinally, linear mixed models estimated an annual eGFRcrea decline of -0.80 [95% confidence interval -0.82, -0.77], -0.79 [-0.83, -0.76], and -1.20 mL/min/1.73m2 [-1.33, -1.08] for the general population, “healthy” individuals, or individuals with diabetes, respectively. Reference values for eGFR using cross-sectional data were shown as percentile curves for “healthy” individuals and for individuals with diabetes. Reference values for eGFR-decline using longitudinal data were presented as 95% prediction intervals for “healthy” individuals and for individuals with diabetes, obesity, and/or albuminuria. Thus, our results can help clinicians to judge eGFR values in individuals seen in clinical practice according to their age and to understand the expected range of annual eGFR-decline based on their risk profile. [Display omitted]