The Aging Kidney Gets a Data-Driven Atlas: Machine Learning and Ferroptosis Lead a New Research Era

The kidney is one of the first organs to show its age. As glomeruli scar, tubules wither, and the delicate vascular network thins, renal reserve quietly erodes, leaving older adults dangerously vulnerable to acute kidney injury and chronic kidney disease. Now, a team of Chinese researchers has produced the most comprehensive quantitative map yet of how science has tackled this problem, analyzing 2,191 papers and 16,545 researchers published between 2000 and 2025 to chart the intellectual evolution of kidney aging research. Their findings, published in Frontiers in Nephrology, reveal a field in the midst of a dramatic paradigm shift—one where machine learning, ferroptosis, and Mendelian randomization are rapidly overtaking traditional clinical descriptions as the hottest topics on the block.

The study, led by Jieyan Wang, Mengning Yang, and Yifei Li from institutions including Southern Medical University and Xiamen University, combined data from the Web of Science Core Collection and Scopus databases. After merging and deduplicating records, the team deployed a suite of computational tools—Excel 365, CiteSpace, VOSviewer, and the Bibliometrix R package—to construct what they call a bibliometric atlas. The result is not just a literature review but a living map of who is publishing what, where the money and talent are flowing, and which molecular threads are pulling the field forward.

The numbers tell a striking story of acceleration. From 2000 to 2018, annual output remained below 100 papers, a period the authors describe as basic knowledge accumulation. Since 2018, the field has surged past that threshold, reaching a peak of 320 publications in 2025, with total annual citations hitting 8,577. This exponential growth, the researchers argue, signals that kidney aging has evolved from a niche descriptive pursuit into an accelerated phase of breakthrough discovery, driven by an influx of new researchers and shifting molecular paradigms.

Geographically, the landscape remains deeply uneven. The United States dominates with 587 papers and 25,229 citations, while China ranks second in volume with 399 papers but only fourth in citations—a gap the authors interpret as reflecting differences in research quality rather than sheer productivity. The collaboration network analysis identifies the United States as the core node, with the strongest bilateral pathway being US-China collaboration, followed by US-South Korea ties. Germany, though not among the top publishers, ranks second globally in total link strength, highlighting its role as a critical hub connecting the global research network. Meanwhile, the Mayo Clinic leads institutional metrics with 44 papers and 3,651 citations, ahead of the Chinese PLA General Hospital and the University of Pittsburgh.

The journal landscape reveals where the field’s knowledge foundation is anchored. The American Journal of Physiology-Renal Physiology tops the productivity rankings with 50 papers, while the Journal of the American Society of Nephrology and Kidney International each exceed 4,000 co-citations, establishing themselves as the core knowledge base. VOSviewer clustering sorted journals into five academic camps—geriatrics and physiology, clinical correlations, renal immunology, cellular and molecular biology, and drug development—while a dual-map overlay showed that citing journals cluster in molecular biology, immunology, and clinical medicine, with cited journals pointing back toward foundational disciplines like genetics and health sciences.

Among the 16,545 scholars identified, Xiang-Mei Chen from the PLA General Hospital leads in publication volume with 40 papers, while Anette Melk of Hannover Medical School tops citation impact with 379 citations. Yet the field’s social structure reveals a telling weakness: applying Lotka’s Law, the researchers found that over 75 percent of authors have published only a single paper, and the proportion of scholars with sustained output falls well below theoretical predictions. This suggests kidney aging research is still in a phase of rapid expansion and recruitment, lacking the core high-output leaders who typically anchor mature disciplines.

The thematic evolution is perhaps the most striking finding. Keyword analysis established aging (825 occurrences) and cellular senescence (330 occurrences) as the field’s central concepts, but the trajectory of research focus has shifted dramatically over time. Before 2015, keywords like diabetes mellitus, renal disease, and hypertension dominated, reflecting a macroscopic clinical perspective. Between 2016 and 2019, the field pivoted toward mechanistic exploration, with aging, oxidative stress, autophagy, and inflammation rising to prominence. Since 2020, the focus has sharpened to cellular precision—renal tubular epithelial cells, mitophagy, and organelle-level dysfunction—before the most recent emergence of machine learning, ferroptosis, and Mendelian randomization as the defining trends of 2025.

Cellular senescence itself sits at the mechanistic heart of this story. The authors explain that senescent cells—irreversibly arrested but metabolically active—acquire a senescence-associated secretory phenotype (SASP), releasing pro-inflammatory cytokines, chemokines, and profibrotic factors that trigger paracrine senescence in neighboring healthy cells. This self-amplifying cascade fuels extracellular matrix accumulation, nephron loss, and architectural remodeling, establishing a vicious cycle between cellular senescence and whole-organ decline. The translational promise is already visible: the landmark clinical trial by Hickson and colleagues, using the senolytic combination of dasatinib and quercetin to clear senescent cells in patients with diabetic kidney disease, ranks as the most co-cited reference in the entire dataset with 758 citations.

Among the emerging frontiers, ferroptosis—an iron-dependent form of cell death driven by lethal lipid peroxidation—has captured particular attention. In the aging kidney, iron accumulation, exhaustion of the system Xc/GSH/GPX4 axis, and ACSL4-mediated fatty acid enrichment create a heightened vulnerability to this non-apoptotic death pathway. Ferroptosis acts as a preemptive trigger in acute kidney injury and forms a pathological bridge to chronic disease progression by releasing damage-associated molecular patterns and activating innate immune cascades. Meanwhile, the gut-kidney-immune axis has emerged as a transformative paradigm, with uremic toxins like indoxyl sulfate and trimethylamine N-oxide driving premature senescence in proximal tubular cells, while beneficial short-chain fatty acids from commensal bacteria act as endogenous brakes against renal aging.

The convergence of computational intelligence and metabolic regulation, the authors conclude, marks a deep paradigm shift for the field. Machine learning algorithms combined with single-cell RNA sequencing and spatial transcriptomics can now deconvolve senescence heterogeneity across distinct nephron segments, revealing that proximal tubular cells, podocytes, and endothelial cells each follow distinct senescence trajectories. Mendelian randomization, using genetic variants as instrumental variables, is providing causal evidence linking gut microbiota, circulating metabolites, and ferroptosis-related genes to kidney function decline. Together, these approaches promise to transform clinical nephrology from late-stage symptom management into early, data-driven precision geroprotection—turning the aging kidney from an inevitable decline into a targetable biological process.

Subject of Research: Bibliometric mapping of cellular senescence and kidney aging research trends from 2000 to 2025

Article Title: Mapping the aging kidney: a bibliometric atlas of cellular senescence and future frontiers since the 20th century

Article References: Wang, J., Yang, M., Li, Y., Wan, Y., Zhan, X., & Liang, H. (2026). Mapping the aging kidney: a bibliometric atlas of cellular senescence and future frontiers since the 20th century. Frontiers in Nephrology, 6, Article 1905354. https://doi.org/10.3389/fneph.2026.1905354

Image Credits: AI Generated

DOI: 10.3389/fneph.2026.1905354

Keywords: kidney aging, cellular senescence, bibliometric analysis, chronic kidney disease, ferroptosis, machine learning, Mendelian randomization, SASP, mitochondrial dysfunction, gut-kidney axis, senolytics, renal fibrosis