A urinary microRNA aging clock accurately predicts biological age.

Abstract

We aimed to develop and validate a urinary miRNA aging clock, positioning urine as a scalable, non-invasive aging-biomarker source. Using machine learning on 6331 adults, our clock achieved MAE ≈ 4.4 years (R² ≈ 0.79) in independent validation. The clock's key biomarkers included well-established geromiRs miR-34a-5p, miR-31-5p, miR-146a-5p, and miR-155-5p. While slightly less accurate than DNA-methylation clocks, our model outperformed blood-based miRNA and mRNA clocks, underscoring urinary miRNAs as promising, truly non-invasive biomarkers of biological age and disease risk.