Nature Aging

Clonal hematopoiesis (CH) increases with age, but whether different somatic clones represent an ageing phenotype or exert distinct systemic effects is unclear. In 450,587 UK Biobank participants, including 46,324 with plasma proteomics, we compared clonal hematopoiesis of indeterminate potential (CHIP) and mosaic loss of chromosome Y (mLOY) or X (mLOX) across biological ageing, incident disease, and circulating proteins. Despite shared age dependence, these alterations showed distinct disease spectra: non-DNMT3A CHIP was associated with broad multisystem disease burden, mLOY with a more focused respiratory, musculoskeletal and cardiovascular profile, whereas mLOX lacked broad age-related disease associations. Clone burden mapped to distinct proteomic programs: mLOY to neutrophil degranulation and extracellular-matrix remodeling, non-DNMT3A CHIP to myeloid immune regulation, and mLOX unexpectedly to cytotoxic lymphocyte/NK-cell responses. Mendelian randomization supported selected protein-disease relationships. Thus, age-related hematopoietic clones are not interchangeable markers of ageing but define alteration-specific systemic programs associated with distinct disease vulnerabilities.

Background Phenotypic age acceleration (PhenoAgeAccel), derived from PhenoAge, and MetaboHealth are composite exposures of biological ageing and metabolic health associated with dementia-related outcomes. Whether these associations are causal and reflect the exposures, constituent biomarkers, or both remains unclear. Methods This study included UK Biobank participants of White British genetic ancestry. MetaboHealth was derived from nuclear magnetic resonance (NMR) metabolomics and PhenoAgeAccel from clinical biomarkers and chronological age. Genome-wide association studies (GWAS) were conducted for MetaboHealth (n=272,568) and PhenoAgeAccel (n=274,077). Independent genome-wide significant variants were used as genetic instruments in two-sample Mendelian randomisation (MR) with FinnGen all-cause dementia summary statistics. Inverse-variance weighting was the primary MR method. Causal network analysis estimated relationships among constituent biomarkers and dementia. Findings GWAS identified 126 and 141 independent genome-wide significant variants for MetaboHealth and PhenoAgeAccel, of which 109 and 141 were retained as genetic instruments. MR found no evidence of a causal effect of genetically predicted MetaboHealth (per unit: OR 0.83, 95% CI 0.49-1.42; p=0.51) or PhenoAgeAccel (per year: OR 0.99, 95% CI 0.95-1.02; p=0.44) on all-cause dementia, with consistent findings across sensitivity analyses and robust MR methods. Lower lymphocyte percentage and higher NMR-derived glucose had direct relationships with dementia in the joint constituent-biomarker network. Interpretation MR provided no evidence that either composite exposure causally influenced dementia. The network prioritised lymphocyte percentage and NMR-derived glucose, supporting examination of composite exposures alongside their constituent biomarkers. Funding NIHR, UKRI, MRC, UK Dementia Research Institute, Innovate UK, and European Union. Full funding details are provided in the acknowledgements.

Prolonged fasting induces marked metabolic adaptations, but whether such time-limited yet intensive interventions engage molecular processes related to biological aging in humans remains unclear. To address this, we investigated physiological and epigenetic responses to a 12-day medically supervised fasting intervention with longitudinal follow-up in 32 participants. Fasting elicited coordinated systemic physiological changes across metabolic and hematological parameters, with partial persistence at one month. Genome-wide DNA methylation analyses revealed modest but detectable CpG-level changes, primarily emerging at follow-up and distributed across genomic contexts. In parallel, epigenetic aging clocks showed clock-specific and time-dependent responses, with substantial inter-individual variability. Lower baseline epigenetic age acceleration was consistently associated with greater fasting-induced weight loss, suggesting a link between epigenetic state and metabolic responsiveness. Together, these findings indicate that prolonged fasting induces coordinated physiological adaptations alongside structured changes in epigenetic aging measures that are not captured by conventional clinical biomarkers. This study highlights epigenetic clocks as integrative molecular readouts for probing aging-related responses to short-term metabolic interventions, while also delineating their current interpretive limits.

Age is a major risk factor for many diseases, but the biological processes driving aging are heterogeneous across individuals. Efforts to untangle differences between chronological and biological age have focused on identifying age-associated markers, such as omics clocks. Many omics features, including proteins, are strongly associated with age, and genetics contribute to variance in these measures. However, few studies have identified genetic drivers of interindividual variability in omics changes over time. Using longitudinal proteomics data (Olink 3k) from the Multi-Ethnic Study of Atherosclerosis (MESA), we calculated a protein slope for each individual (n=2,007) and protein (n=2,737) across 3 visits spanning 14-18 years, then conducted a genome-wide analysis for each slope, both with and without adjusting for baseline protein level. Subsets in UK Biobank (UKB; n=948) and CARDIA (n=1,328) with longitudinal proteomics data were used for replication. We considered additional methods for modeling of protein change and variability, including linear mixed models, SNP-by-age interactions, and variance quantitative trait loci. Without baseline adjustment, only 19 proteins (20 credible sets) had a slope pQTL in MESA, with poor replication in UKB and CARDIA. With baseline adjustment, 607 proteins (698 credivle sets) had a slope pQTL and over 70% replicated in CARDIA and/or UKB; such baseline adjusted models may, however, be subject to collider bias. Longitudinal and cross-sectional interaction models identified fewer than 14 pQTLs, suggesting they were generally underpowered; but 73% of proteins with a variance pQTL also had a slope pQTL. By examining effect direction concordance, replication rate, directed acyclic graphs, and signal overlap with other models we demonstrate that many baseline-adjusted slope pQTLs may be arising due to model misspecification or regression to the mean. Overall, our results highlight considerations for modeling strategies of change phenotypes and build on understanding of potential genetic mechanisms influencing interindividual proteome changes over time.

Endothelial cells (ECs) express organ-specific gene programs supporting tissue homeostasis and resilience. However, the mechanisms by which aging reshapes these organ-specific endothelial programs and how the resulting changes affect tissue homeostasis, resilience, and disease susceptibility remain largely unknown. Herein, we performed single-cell RNA sequencing of ECs harvested from five organs across the lifespan and found that aging progressively erodes organ-specific endothelial programs while inducing shared interferon-responsive and antigen-presentation programs across organs. Although vascular subtype identity and conserved capillary subset identity were mostly preserved, these organ-specific transcriptional programs were broadly attenuated with aging, indicating erosion of organ-specific endothelial identity to be a fundamental feature of endothelial aging. Importantly, these alterations were associated with declines in specialized EC functions, including alveolar barrier maintenance in the lung, scavenging activity in the liver, angiogenic capacity in the heart, and homeostatic programs in the kidneys and the brain, suggesting that age-related EC alterations compromise tissue homeostasis and resilience in multiple organs. Furthermore, we established a single-cell aging index for alveolar capillary ECs in mice and humans, revealing stress-associated endothelial activation to potentially be an intermediate state linking functional deterioration to cellular senescence, and also demonstrating marked heterogeneity in aging states among ECs of the same chronological age. Notably, alveolar capillary ECs exhibited progressive functional decline before reaching a senescent-like state, suggesting endothelial dysfunction to precede overt cellular senescence as the organism ages. Collectively, our findings establish progressive erosion of organ-specific endothelial programs as a central feature of vascular aging and provide a conceptual framework for elucidating how endothelial aging contributes to tissue dysfunction and reduced resilience across organs.

Aging is caused, fully in large part, by the progressive accumulation of damage, yet quantifying age-related damage across tissues and conditions remains a challenge. Here, we present a computational framework to quantify damage from standard RNA-sequencing data. It captures four classes of aberrant transcript structures, including premature termination upon intron retention, domain-disrupting splice variants, repeat elements, and gene fusion events, each reflecting distinct forms of RNA integrity loss. Using this method, we revealed a robust age-associated increase in transcriptomic damage across tissues. To integrate these measurements into a unified biomarker, we constructed a transcriptomic damage-based aging (tDamAge) clock using machine learning models trained across mouse tissues or human peripheral blood. It could predict age and detect transcriptomic shifts under both pro-aging and anti-aging conditions. Progeroid models exhibited accelerated tDamAge, whereas interventions such as caloric restriction, rapamycin, and methionine restriction lowered tDamAge. Cross-dataset analysis showed that diverse anti-aging interventions converge on shared transcriptomic signatures, particularly RNA processing and chromatin organization pathways, and these age-associated patterns could be reversed by interventions. We further identified elevated damage age acceleration in Alzheimers disease and observed rejuvenation-like reductions during embryonic development. Together, our findings establish transcriptomic damage as a causal, quantifiable and biologically interpretable feature of aging and demonstrate that tDamAge could detect age progression, acceleration, deceleration, and reversal.

Circadian timing influences human physiology and disease risk, yet scalable measures of molecular circadian phase are lacking. Here we infer circadian phase from circulating blood biomarkers in UK Biobank. Among 3,228 plasma biomarkers, 58% exhibit significant diurnal variation, with harmonic modeling identifying acrophase clustering consistent with canonical circadian patterns and independent constant-routine datasets. Machine-learning models trained on plasma proteomics predict sampling time (R2=0.68) and retain substantial accuracy with ~60 proteins. We define a novel construct, circadian acceleration (CA), as deviation from the population-average phase; CA is temporally stable, associates with chronotype and shift work, and responds to environmental perturbation. CA is heritable (h2SNP=0.10) and genetically correlated with chronotype and accelerometry-derived sleep traits. These results establish plasma proteomics as a scalable approach for population-level molecular circadian phenotyping.

The female reproductive system is one of the first major organ systems to show signs of age-related decline, and menopause is associated with increased risk of several diseases, including osteoporosis and cardiovascular disease. Menstrual fluid contains a mixture of blood and endometrial tissue and is a noninvasive biological sample type that has immense potential for diagnostics related to female reproductive aging. However, existing epigenetic aging clocks show limited performance in hormone-dependent tissues such as the endometrium. At Xella Health, we collected menstrual fluid (MF) samples, from a diverse patient cohort (n=66) and quantified genome-wide 5mC methylation levels. We then developed a novel, deep learning-based epigenetic aging clock that is optimized for performance in menstrual fluid and endometrial tissue. Our model, the Xella Clock, outperforms other widely used epigenetic aging clocks at predicting chronological age from MF data and on endometrial tissue. The model is a useful tool for advancing the study of female reproductive aging and can be used to examine associations between endometrial age acceleration and clinical factors.

Rapamycin extends lifespan across species, yet its cell-type-specific benefits and vulnerabilities remain unclear at whole-organism scale. Here, we present the Rapamycin Fly Cell Atlas (Rapa-FCA), a whole-organism single-nucleus transcriptomic atlas of Drosophila spanning both sexes, multiple ages, 18 cell classes, and 181 cell types. Rapamycin elicited a highly heterogeneous response, with prominent effects in reproductive, digestive, and neuromuscular systems and modest responses in most neuronal populations. Across diverse tissues, we identified a rapamycin-sensitive Convergent Aging Trajectory (CAT), marked by Fkbp12 enrichment and mTORC1-linked metabolic programs, including glycolysis and lipid synthesis. CAT-high nuclei accumulated with age and were preferentially reduced by rapamycin, especially in females, consistent with stronger female lifespan extension. By integrating CAT abundance, aging-clock predictions, and nucleus-ratio changes, we mapped sex- and cell-type-specific geroprotection effects of rapamycin. Together, the Rapa-FCA provides an organism-wide framework for resolving how rapamycin reshapes cellular aging across sex, tissue, and cellular state.

Aging is a progressive decline in biological function that is proposed to be driven by the accumulation of epigenetic noise and the loss of epigenetic information. Among epigenetic readouts, DNA methylation has been extensively used to develop aging clocks, machine learning models that predict age from molecular data. However, DNA methylation clocks are relatively difficult to interpret and remain distant from gene regulatory networks, a gap that can be complemented by clocks built from another epigenetic layer: chromatin accessibility profiled by ATAC-seq. Existing chromatin accessibility clocks predict age from bulk ATAC-seq data, thereby averaging over the epigenetic heterogeneity across cells that drives aging. We hypothesized that a chromatin accessibility clock trained at the level of individual cell types, using pseudobulk profiles derived from single-nucleus ATAC-seq (snATAC-seq) data, would be particularly useful for characterizing cell type-specific aging. We focused on the brain, a highly heterogeneous tissue whose diverse cell types age asynchronously, and assessed how well cell type-specific accessibility clocks can predict chronological age, capture rejuvenation from genetic perturbation, and detect age acceleration in age-associated neurodegenerative disease. To this end, we introduce a set of cell type-specific and all-cell aging clocks built from snATAC-seq profiles of the prefrontal cortex (PFC) of 357 human donors (15 to 100 years), which generalize to accurately predict age across brain regions and species. Beyond healthy aging, these PFC clocks captured the rejuvenating effects of SIRT6 overexpression in mouse liver and cell type-specific age acceleration in Alzheimer's disease (AD) and Parkinson's disease, with microglial age acceleration correlating most strongly with pathology among major cell types, and with female oligodendrocytes and OPCs showing the largest sex differences in age acceleration. Interpreting the clocks further revealed the regulatory elements, genes, pathways, and motifs underlying these signals across species, disease, and perturbation, including repression of the NF-kB pathway in SIRT6 transgenic mice, upregulation of immune and inflammatory pathways in severe AD, and conserved age-predictive peaks related to histone regulation, metabolism, and neuronal survival across brain regions and species. Together, these results establish PFC snATAC-seq aging clocks as a generalizable tool that accurately predicts age and captures cell type-specific perturbation effects of rejuvenation and disease on the epigenetic landscape, providing both a means to evaluate perturbations and insight into the epigenetic mechanisms of aging and disease.

Aging is associated with progressive tissue dysfunction and impaired repair after injury. In the retinal pigment epithelium (RPE), these changes contribute to age-related macular degeneration (AMD), yet the mechanisms limiting repair remain incompletely understood. Here, we establish a longitudinal human embryonic stem cell (hESC)-derived RPE aging model that recapitulates key features of aged human donor RPE and combine it with mosaic cell ablation to assess injury-responsive repair. Although aged RPE cells initiate DNA synthesis after injury, they exhibit impaired mitotic progression, uncoupling S-phase entry from epithelial repopulation. Transcriptomic profiling links this defect to a senescence-associated program marked by inflammatory signaling and suppressed mitotic networks. Pharmacologic reduction of senescent cells with Navitoclax shifts aged RPE toward a younger transcriptional profile but does not induce repopulation by itself. Instead, senolytic treatment primes aged RPE for repair, improving epithelial density and homeostatic function only in response to injury, a strategy we term "senolytic priming." These findings establish a human stem-cell-derived platform for investigating age-associated epithelial repair failure and show that aged human RPE retains latent repair capacity that can be re-enabled by targeting cellular senescence.

Aging is a major risk factor for neurological disease, yet the molecular architecture of human brain aging remains poorly defined. Here, we analyzed more than 10,000 cerebrospinal fluid (CSF) proteomes across multiple cohorts and proteomic platforms to develop a 249-protein CSF aging clock that accurately predicted chronological age and generalized across independent datasets. CSF brain-age acceleration was increased across diverse neurological diseases, associated with blood-brain barrier (BBB) dysfunction, and predictive of longitudinal cognitive decline, neuroimaging progression and dementia conversion. A simplified 30-protein panel retained similar prognostic performance. Biologically, the clock resolved two opposing programs: pro-aging activation of immune, vascular/BBB, extracellular matrix and coagulation pathways, marked by CHI3L1, CD14, VWF, LRG1 and LTBP2, and collapse of anti-aging neuronal-maintenance programs, marked by NPTX2, COL1A2, NID1, CDH8 and PENK. Brain-wide single-cell and regional mapping linked these programs to disease-vulnerable compartments. These findings establish a CSF-based molecular framework for quantifying biological brain aging and predicting neurological disease progression.

Environmental exposures are major modifiable determinants of human aging, yet the evidence remains fragmented across organ-agnostic summaries and rarely confronts population inequity. Here we present an exposomic atlas of pan-organ aging in ~300,000 UK Biobank adults, mapping 164 environmental and behavioural exposures onto biological aging of the whole body and nine organ subsystems. Comprising 1,476 systematically tested exposure-subsystem associations, the atlas reveals that environmental effects on human aging are pervasively organ-specific, with 65.9% of exposures acting divergently across organ subsystems. This landscape resolves into nine navigable modules that preserve organ selectivity, predict 23 major age-related diseases, and expose distinct dimensions of health inequity. In-silico analyses further show that priorities for ameliorating aging are target-dependent rather than universal, diverge markedly from the whole-body ranking (Kendall's {tau} = 0.52 to 0.39), reorder substantially across population strata, with findings externally validated in an ethnically distinct cohort. The atlas establishes an organ-resolved and target-aware foundation for precision environmental health.

Biological age estimates are increasingly used to study aging, disease risk, and mortality, yet their predictive uncertainty is rarely quantified. Consequently, conventional age-gap measures can treat deviations as equally informative even when the underlying biological age predictions differ substantially in reliability. We developed a framework for uncertainty-aware biological aging that generates calibrated prediction intervals and individualized probabilities of accelerated or decelerated aging alongside point estimates. We applied this framework to the UK Biobank Pharma Proteomics Project, evaluating three composite and eleven organ-specific biological age clocks. Predictive uncertainty varied substantially both within and across clocks, revealing that apparently extreme age gaps can differ markedly in the strength of evidence supporting accelerated or decelerated aging. In particular, low-accuracy clocks, including many organ-specific clocks, provided little evidence for confidently accelerated or decelerated aging. Beyond biological age gaps, prediction-interval width was independently associated with disease risk and mortality, particularly for composite, brain, and immune clocks, suggesting that predictive uncertainty captures an additional dimension of biological aging that may reflect increased molecular heterogeneity and dysregulation associated with aging and disease. We replicated these findings in Biobank Japan and an independent clinical cohort from Stanford. By incorporating individual-specific predictive uncertainty, our framework provides a more informative characterization of biological aging and enables improved individual-level risk stratification for disease prevention and longitudinal monitoring.

Cognitively normal older adults are often regarded as a homogeneous population in preventive and disease-modifying clinical trials for dementia. However, longer-term cognitive aging outcomes vary substantially in this population, and this variability remains poorly understood. Here, we leveraged rich multimodal, multi-domain biomarker profiles from a large prospective cohort (N=1,136), and Deep Embedded Clustering, to cluster cognitively normal older adults into biologically distinct subgroups. Input data included cortical thickness derived from MRI, plasma Alzheimer's disease (AD) biomarkers, plasma inflammatory biomarkers, and vascular measures. The deep clustering algorithm identified three biologically distinct subgroups within the sample, stratified along a gradient of neurobiological burden (low, intermediate, and high). Cortical thinning and inflammatory burden were the primary drivers of clustering assignments. The High-Burden subgroup showed significantly worse memory and executive function, elevated cardiometabolic comorbidity, and markedly higher rates of conversion to mild cognitive impairment or dementia within two years. The results were validated in an independent external sample. The study reveals marked variability among individuals who are otherwise all defined as cognitively normal, and provides a data-driven stratification framework for enriching disease-modifying and preventive trials by identifying cognitively normal individuals at high risk for future cognitive decline.

Aging is characterized by highly reproducible alterations across multiple layers of the epigenetic landscape, including DNA methylation, chromatin accessibility, and histone modifications. However, it remains unclear to what extent these age-associated changes are interconnected and coordinated coherently. To investigate the genome-wide distribution and interplay of age-associated epigenetic alterations, we generated whole-genome bisulfite sequencing (WGBS) data from blood samples of 120 healthy donors. Integration with ATAC-seq data revealed no clear relationship between age-related changes in DNA methylation and chromatin accessibility. We further examined the association of these alterations with age-dependent changes in CTCF occupancy and histone modifications, including H3K27ac, H3K27me3, H3K4me1, H3K4me3, and H3K9me3, but observed very little corresponding changes in chromatin states. Collectively, our integrative genome-wide analysis revealed only limited association between age-associated epigenetic alterations in DNA methylation, chromatin accessibility, and histone modifications, arguing against a broadly coordinated remodeling of the aging epigenome.

Systemic aging and metabolic overload remodel the vasculature; however, how endothelial cells integrate these stresses across organs remains unclear. Using multi-organ single-cell and spatial transcriptomics with functional validation, we mapped endothelial and hematopoietic responses in adipose tissue, skeletal muscle, liver, and heart. Organ-specific endothelial transcriptional features were relatively preserved, whereas chronic stress selectively reconfigured regulatory programs: aging induced a conserved Irf/Stat-centered endothelial program, while high-fat diet engaged organ-biased lipid and remodeling programs. Spatial analysis revealed perivascular niches centered on aging-associated interferon-stimulated endothelial activation, with neighboring immune and stromal cells expressing C3 and LRP1-associated signals. Rather than simply amplifying inflammation, these niches contained mechanisms that restrained IFN activation, as C3 depletion upregulated vascular IRF7 expression. In parallel, the IFN downstream effector BST2 promoted anti-inflammatory macrophage differentiation and suppressed atherosclerosis. These findings define vascular inflammaging as an organ-resolved niche process in which endothelial IFN activation is coupled to local inflammatory restraint. HighlightsO_LIAging induces a shared endothelial type I IFN program across organs. C_LIO_LIA high-fat diet triggers organ-biased endothelial remodeling programs. C_LIO_LIPerivascular interferon niches couple inflammation with local restraint. C_LIO_LIIFN-induced endothelial BST2 promotes CD200R-associated macrophage regulatory features. C_LI

Type 2 diabetes is linked to systemic inflammation driven by metabolic stress and aging. Although pancreatic inflammation associated with these factors is well documented, the dynamics of immune cell populations and their molecular changes remain poorly understood. We characterized immune cell alterations in the pancreas and pancreatic islets during Western diet (WD) feeding and aging using imaging mass cytometry (IMC) and single-cell RNA sequencing (scRNA-seq). Spatial and transcriptional analyses were performed to define immune cell subtype composition, activation states, and inferred cell-cell communication programs under metabolic and age-related stress conditions. Our analyses identified expansion of an F4/80low macrophage subtype and activated effector-like CD8+ T cells throughout the pancreas during WD feeding and aging. Within pancreatic islets, single-cell RNA sequencing identified a type 1 interferon-responsive macrophage population with low F4/80 expression that expanded during overnutrition. Notably, the type 1 interferon responses elicited by these stressors diverged: aging was associated with a more canonical type 1 interferon response, whereas overnutrition induced a broader response that included STAT3-associated transcriptional programs. We further provide evidence for enhanced cytokine-mediated communication between macrophages and a CD8+ cytotoxic T-cell population under overnutrition and aging. These findings show that metabolic stress and aging remodel pancreatic inflammation through overlapping but distinct immune mechanisms, involving expansion of F4/80low macrophages, activation of divergent type 1 interferon programs, and enhanced macrophage-CD8+ T-cell communication. Together, these findings suggest that distinct therapeutic approaches may be required to preserve islet function in type 2 diabetes driven by metabolic stress versus aging.

DNA methylation-based epigenetic clocks estimate biological age from methylation profiles, and the difference between predicted biological age and chronological age is commonly described as age acceleration (AA). We compared AA across eight cancer types, lung, colorectal, breast, thyroid, bone marrow and blood, kidney, uterus, and head and neck, using seven epigenetic clocks and 5,528 publicly available samples. Across the 56 cancer type clock combinations, tumor tissues showed higher average AA than normal tissues in 44 comparisons. The uterus cohort showed the clearest deviation from this overall trend, with normal samples exhibiting higher AA for six of seven clocks. Analyses of paired normal and tumor samples generally showed higher predicted ages and greater variability in tumor samples. We additionally examined age-associated methylation changes and the ability of clock CpGs to distinguish tumor from normal tissue. Several discriminatory CpGs were shared across cancer types and frequently showed tumor-associated hypermethylation at cancer-related loci. Small subsets of top-ranked CpGs captured substantial discriminatory information. Age-stratified subsampling preserved the main AA patterns, suggesting that chronological-age differences did not explain the observed tumor-normal differences. Overall, these findings highlight broad cancer-associated alterations in epigenetic aging together with substantial cancer type- and clock-specific heterogeneity.

Spatial transcriptomics enables analysis of molecular organization with anatomical context. Existing spatial differential expression methods are restricted to within-sample inference, forcing between-sample comparisons to rely on approaches adapted from single-cell RNA-seq. Here, we establish a scale-aware inference framework for spatial differential expression by modeling compositional constraints and variation in total RNA abundance rather than removing them through normalization, enabling calibrated between-sample inference at cell-level resolution. Our method produces more reliable results in simulated data and different spatial platforms. When applied to aged mouse brains, the analysis reveals converging aging-associated programs involving cellular signaling, membrane homeostasis, and neurovasculature across independent datasets.