Genome Med. 2025 Jun 25. 17(1): 70
Lieke M Kuiper,
Michelle M J Mens,
Julia W Wu,
Jaap Goudsmit,
Yuan Ma,
Liming Liang,
Albert Hofman,
Trudy Voortman,
M Arfan Ikram,
Jeroen G J van Rooij,
Joyce B J van Meurs,
Mohsen Ghanbari.
BACKGROUND: MicroRNAs are small non-coding RNAs that regulate gene expression post-transcriptionally and show differential expression in various tissues with aging phenotypes. Detectable in circulation, extracellular microRNAs reflect (patho)physiological processes and hold promise as biomarkers for healthy aging and age-related diseases. This study aimed to explore plasma extracellular microRNAs as a biological aging indicator and their associations with health outcomes using population-level data.
METHODS: We quantified plasma expression levels of 2083 extracellular microRNAs using targeted RNA-sequencing in 2684 participants from the population-based Rotterdam Study cohort. The training and test sets included 1930 participants from the advanced-aged initial and second subcohort (RS-I/RS-II; median age: 70.6), while the validation set comprised 754 participants from the middle-aged fourth subcohort (RS-IV; median age: 53.5). Based on 591 microRNAs well-expressed in plasma, we examined differential expression of microRNAs with chronological age, PhenoAge-a composite score of age and nine multi-system blood biomarkers-the frailty index, and mortality. Next, elastic net models were employed to construct composite microRNA-based aging biomarkers predicting chronological age (mirAge), PhenoAge (mirPA), frailty index (mirFI), and mortality (mirMort). The association of these aging biomarkers with different age-related health outcomes was assessed using Cox Proportional Hazard, linear regression, and logistic regression models in the test and validation sets.
RESULTS: We identified 188 microRNAs differentially expressed with chronological age within the RS-I/RS-II advanced-aged population (ntraining = 1158, ntest = 772), of which 177 microRNAs (94.1%) were replicated in the middle-aged RS-IV subcohort (nvalidation = 754). Moreover, 227 miRNAs showed robust associations with PhenoAge, 61 with FI, and 16 with 10-year mortality independent of chronological age. Subsequently, we constructed four plasma microRNA-based aging biomarkers: mirAge with 108, mirPA with 153, mirFI with 81, and mirMort with 50 miRNAs. Elevated scores on these microRNA-based aging biomarkers were associated with unfavorable health outcomes, including lower subjective physical functioning and self-reported health and increased mortality and frailty risk, but not with first- or multi-morbidity. Overall, larger effect estimates were observed for mirPA, mirFI, and mirMort compared to mirAge.
CONCLUSIONS: This study describes distinct plasma microRNA-aging signatures and introduces four microRNA-based aging biomarkers with the potential to identify accelerated aging and age-related decline, providing insights into the intricate process of human aging.
Keywords: Aging; Biological age; Biomarker; Frailty; MicroRNA; Mortality