Aging Cell

The short-lived annual fish Nothobranchius furzeri (Nfu) is a powerful vertebrate model for aging research due to its rapid lifespan and accelerated development of age-associated phenotypes, including gliosis and lipofuscin accumulation. Here, we investigated the effects of dietary 1,3-1,6 {beta}-glucans (BGs), natural polysaccharides derived from Saccharomyces cerevisiae, on aging-related processes across multiple tissues, with particular focus on the brain. Chronic treatment with BG-fortified food reduced several hallmarks of aging in multiple organs. Mechanistically, BG treatment modulated pathways associated with autophagy, lysosomal function, protein oxidation, and inflammation. Both acute and chronic BG administration increased autophagic activity in the aging brain, although lipofuscin accumulation was not affected. To assess whether BGs act directly on neural tissue, we established an ex-vivo Nfu brain culture system that recapitulates the age-dependent decline in autophagy observed in vivo. In this model, acute BG treatment restored impaired autophagy and promoted mitochondrial and lysosomal biogenesis in aged brains. Proteomic analyses revealed increased mitochondrial respiration and modulation of V-ATPase components involved in autophagosome acidification. Depletion of microglia reduced but not eliminated this effect, suggesting direct action of BGs on neurons. To verify the validity of these findings in humans, we performed BG treatment in human iPSC-derived neurons under conditions of impaired autophagy and found an increase in survival. Together, these findings identify {beta}-glucans as modulators of autophagy, mitochondrial function, and inflammation, highlighting their potential to promote healthy aging.

Aging is associated with chronic, low-grade inflammation and progressive immune dysfunction. However, the current understanding of age-associated changes in dendritic cells across tissues is scarce. Studies exploring ageing-associated changes in dendritic cells (DCs) have reported either a general decline in the overall DC compartment or subset-specific alterations affecting cDC1, cDC2, and pDC populations across spleen, lung and liver, underscoring the considerable inconsistencies across tissues and studies. To underpin whether age-associated changes are extrinsic or intrinsic we investigated DCs across bone marrow, six peripheral tissues and in in vitro bone marrow derived DC cultures to examine the effects of aging on DC-poiesis, tissue distribution, and cellular states related to DC functionality and activation. We discovered that aging selectively alters DC development in the bone marrow by reducing cDC progenitor populations while preserving pDC-poiesis. In peripheral tissues, however, age-associated changes in DC homeostasis were strongly tissue-dependent. The most significant shifts in cDC1 and cDC2 frequencies occurred in barrier tissues, such as the lung and small intestine. In contrast, the spleen and liver exhibited more limited or variable changes. These quantitative alterations were accompanied by tissue-specific changes in phenotypic and activation-associated markers, including CD24, CD103, CD11b, MHCII, and CD86. Single-cell transcriptomic analyses of senescent p21-expressing DC across tissues and subsets indicated localized inflammatory states that aligned with local macrophage populations, pointing toward cell-extrinsic niches that contribute to local age-associated dysfunction. Notably, aged bone marrow retained the capacity to efficiently generate DCs in in vitro Flt3L cultures, and antigen-presenting function of BMDC to CD4 and CD8 T cells was maintained, pointing towards preserved cell-intrinsic functions, albeit subset-specific differences in activation and inhibitory receptor expression in response to different pattern-recognition receptor agonists. Collectively, our findings indicate that aging does not superimpose a uniform alteration module to the DC compartment across tissues, but instead promotes selective alterations in DC ontogeny and tissue-specific remodeling of DC phenotypes and cellular states.

The adult skeletal muscle regenerates robustly upon injury, but this regenerative capacity rapidly declines with age. In this study, we identify the lanthionine synthetase C-Like (LanCL) proteins, mammalian homologs of the bacterial peptide cyclase LanC, as positive regulators of muscle regeneration in middle-aged mice. In a barium chloride-induced injury model, we found the protein levels of LanCL1 and LanCL2 to increase during an early phase of regeneration in middle-aged (12-month-old) but not young adult (4-month-old) mice. Utilizing a mouse line lacking all three LanCL proteins (LanCL triple KO or LTKO), we examined a potential role of LanCL in injury-induced muscle regeneration. Consistent with an age-dependent function of LanCL, we observed a delayed regeneration of the tibialis anterior (TA) muscle after injury, as reflected by reduced sizes of regenerating myofibers in middle-aged (but not young) LTKO compared to age-matched WT mice. Although the pool size of quiescent satellite cells (Pax7+) was comparable between 12-month-old LTKO and WT muscles without injury, the number of Pax7+ cells was significantly higher in regenerating LTKO muscles at day 5 after injury, accompanied by drastically decreased numbers of MyoD+ and MyoG+ cells, as well as increased numbers of proliferating cells. In addition, we detected elevated expression of pro-inflammatory cytokines in regenerating LTKO muscles, while the number of macrophages was similar comparing LTKO and WT muscles. Taken together, our observations suggest that in aging muscles LanCLs are important for proper timing of inflammation resolution and regeneration upon injury. New & NoteworthyPhysiological roles of the mammalian homologs of bacterial LanC, LanCLs, are poorly understood. Our work uncovers a function of LanCLs in post-injury regeneration of aging skeletal muscles. Middle-aged LanCL triple KO mice displayed a delay in satellite cell differentiation and regenerative myofiber formation, as well as persistent inflammatory cytokine expression, suggesting that LanCLs may have an age-dependent role in modulating inflammation in the injured muscles to facilitate regeneration.

Aging is an inevitable risk factor for cardiovascular disease. Profound understanding of mechanisms underlying the early stages of cardiovascular aging is essential for the development of novel therapeutics. Therefore, animal models which closely reflect the human condition are highly sought after. Here, we investigated natural cardiovascular aging in a non-human primate, comparing healthy young-adult and aged common marmosets (Callithrix jacchus). Despite preservation of most cardiac functional parameters in aged animals, significant histological alterations were found including fibrosis and microvascular rarefaction. Molecular phenotyping by single-nuclei RNA-sequencing revealed activation of cardiac stress, pro-inflammatory and fibrotic gene programs in aged hearts. Importantly, proteomic analysis of cardiac extracellular vesicles revealed a cardioprotective cargo in young animals while functionally demonstrating pro-angiogenic properties on human cardiac microvascular endothelial cells. Finally, large vessel atherosclerosis was strikingly evident in aged animals and elucidated by bulk RNA-sequencing. Overall, the aging marmoset offers a great potential for translational cardiovascular research.

Chronic inflammation and cellular stress are hallmarks of aging, obesity, and type 2 diabetes (T2D), but whether these programs can be modulated by lifestyle intervention in late life, particularly in the presence of established metabolic disease, remain unknown. We profiled circulating immune cells from older adults with obesity and T2D (ages 66-83 years; n = 9) before and after a 6-month lifestyle intervention combining caloric restriction with exercise training. Participants showed substantial weight loss ([~]7%) alongside improvements in glycemic control, insulin sensitivity, and physical performance. Longitudinal single-cell transcriptomic and epigenomic profiling identified two major changes. First, intervention was associated with downregulation of inflammatory and endoplasmic reticulum (ER) stress transcriptional programs, with the most pronounced effects observed in CD14+ monocytes. DDIT3 (CHOP) was transcriptionally and epigenetically downregulated and its inferred regulatory network encompassed multiple inflammatory mediators. Second, naive CD4+ T, naive Treg, and naive B cells exhibited an upregulation of naive cell identity genes, with naiveness scores increasing after intervention, which declines with age in an independent healthy adult cohort. Together, these findings suggest that lifestyle intervention is associated with coordinated remodeling of both innate and adaptive immune compartments in older adults, revealing substantial plasticity of the aging immune system especially targeting ER stress, inflammation, and naive lymphocyte identity programs.

The aging population continues to grow, and along with it the need to better understand the unique characteristics of the aging brain. Despite the cortical thinning and myelin degeneration that occurs in later life, many individuals are resilient to these changes and are relatively spared from cognitive decline (so called "super-agers"). Longitudinal magnetic resonance imaging data from the Human Connectome Project allowed us to probe the underlying white matter connections that are strengthened or pruned to support fluid intelligence in younger and older adults, while also investigating physical activity as a potential behavioural intervention to promote these network reorganizations. Multivariate partial least squares analysis and the recent method of correlational tractography identified parts of the white matter structural connectivity network, measured with diffusion magnetic resonance imaging, where strengthening or pruning were associated with positive fluid intelligence trajectories and promoted by physical activity. This research improves our understanding of how the brain network adapts in later life, identifies physical activity as an accessible intervention to promote this adaptation, and enriches our fundamental models of what the brains neural network is capable of.

Autophagy is widely proposed to decline with age; however, direct evidence for this across cell and tissue types in humans remains limited. Furthermore, it remains unknown whether interventions that improve physiological health during aging can modify autophagic activity in humans. Here, we performed transcriptomic and functional autophagy analyses across subject-matched human cell types from a healthy aging cohort spanning the adult lifespan. RNA-seq of primary dermal fibroblasts and induced neurons (iNs) revealed increased transcription of many autophagy-related genes with age, most markedly in fibroblasts. The impact of age on autophagic activity, measured using autophagy flux assays, was cell type- and sex-dependent, and uncoupled from autophagy-gene transcription. Autophagy flux decreased with age in male fibroblasts, was unchanged in female fibroblasts, and increased in female iNs. In freshly isolated peripheral blood mononuclear cells (PBMCs), autophagy flux became more heterogeneous with age and trended higher in older individuals, independent of sex. Although autophagy flux levels did not match across different cell types, higher autophagy flux in all cell types was associated with reduced physical function in older adults ([≥]70 years). Importantly, autophagy flux decreased following 12 weeks of mild exercise in parallel with improved physical function. These findings indicate that autophagy is regulated in a cell type-, sex-and physiological function-dependent manner during human aging, and highlight PBMC autophagy flux as a potentially modifiable, blood-accessible readout of physiological state in older adults.

Background: While genetic factors strongly influence brain aging trajectories, variants conferring cognitive resilience remain poorly characterized. The neurokinin-3 receptor (NK3-R), encoded by Tachykinin Receptor 3 (TACR3), modulates cholinergic signaling in memory circuits vulnerable to aging. Previous studies linked the non-WT expression of the TACR3 variant rs2765 with cognitive decline and reduced volume of the hippocampus and basal forebrain, but systematic replication and mechanistic validation were lacking. Methods: We investigated rs2765 in the preregistered AgeGain cohort of cognitively healthy older adults (n=188) with independent validation in the ADNI cohort (n=809) which includes persons with and without Alzheimers Disease (AD) that show healthy cognition, mild cognitive impairment or dementia. Analyses integrated structural neuroimaging, longitudinal cognitive assessments, epigenetic aging (PhenoAge), genome-wide methylation profiling, and mechanistic validation through luciferase assays and cross-species protein expression studies. Results: The infrequent protective rs2765 WT variant, found in 12.8% of Europeans, conferred 49% slower cognitive decline (p = 0.002) for amyloid-positive individuals of the ADNI cohort and 3.7 years younger epigenetic age (p = 0.013, 95% CI: 0.79-6.67 years) in the cognitively healthy AgeGain cohort. WT carriers showed larger hippocampal and basal forebrain volumes across cohorts, with Allen Brain Atlas integration revealing these outcomes to occur exclusively in regions where TACR3 expression positively correlated with gray matter volume. Mechanistically, the non-WT variant ameliorated RBMX-mediated post-transcriptional regulation, reducing NK3-R protein expression by 25-40% in vitro and ex vivo murine brain slice models. Senescence-accelerated mice exhibited reduced endogenous NK3-R expression, phenocopying the predicted functional consequences of the variant. In AgeGain participants, genome-wide methylation profiling identified 2,313 differentially methylated CpGs affecting 228 pathways spanning glutamatergic signaling, acetylcholine receptor pathways, chromatin remodeling, and angiogenesis, suggesting coordinated molecular reprogramming from synaptic function to systemic aging. Conclusions: rs2765 WT confers resilience to age- and AD-related cognitive decline through RBMX-dependent regulation of NK3-R expression, with effects of remarkable size cascading from memory to systemic aging. rs2765 genotyping could stratify individuals for NK3-R modulator therapy (e.g., fezolinetant or senktides) and identify those maintaining function despite pathological burden, complementing APOE-based risk assessment in precision geromedicine.

Skeletal muscle function is central to the preservation of functional mobility. Given global shifts to an increasingly aged population, it is paramount that researchers and clinicians better understand the effectors of age-related functional decline. Muscle fatiguability acutely modifies skeletal muscle mechanics in ways that may affect joint stability. We have previously reported sex-specific reductions in cellular passive stress and modulus with fatigue in young males, but not females. Here, we assess whether older adults, who are more susceptible to fatigue during dynamic contractions, exhibit changes to cellular passive mechanics following fatiguing exercise. Muscle tissue biopsies were collected from 11 young and 11 older adults to measure passive stress and Youngs Modulus at the single fiber and bundle level. Biopsy samples were acquired from rested muscle and immediately following intermittent maximal contractions to task failure. Fatigue was associated with persistent reduction in elastic modulus that was specific to male participants, regardless of age. In muscle fiber bundles, containing both myofibrillar proteins and the extracellular matrix, fatigue-induced changes in modulus were largely negated, with the only significant change observed in young females, who demonstrated enhanced modulus with fatigue. Taken together our findings suggest a preservation of sex-based differences in the acute response to fatigue across the adult lifespan when measured at the myofilament level. However, further research is needed to understand how and whether these findings translate to the whole tissue level. New and noteworthyAcute modifications to muscle tissue mechanics are poorly understood but may have important impacts on functional outcomes in at-risk populations. Our findings suggest myocellular mechanics respond to acute fatigue stress in a sex specific manner that persists across the lifespan.

Early-life conditions can shape molecular ageing processes, yet to what extent developmental variation in telomere length (TL) influences ageing trajectories remains unclear, particularly in long-lived mammals. We investigated how early-life environmental conditions and maternal age relate to juvenile TL and short-term survival in two long-lived bat species, Myotis myotis and Rhinolophus ferrumequinum. Using novel long-term datasets spanning ten years in M. myotis and five years in R. ferrumequinum, we measured relative telomere length (rTL) in juvenile wing tissue and applied sliding window analysis to identify sensitive climatic periods during development. In both species, early-life rTL varied significantly among years and was associated with short-term climatic conditions, with rainfall predicting rTL in both species and temperature acting in opposing directions: longer rTL with warmer conditions in M. myotis, and longer rTL at intermediate temperatures in R. ferrumequinum. Maternal age at conception showed little association with offspring TL in either species, although a weak positive sex-specific longitudinal effect was detected in R. ferrumequinum. Despite clear environmental influences on early-life rTL, we found no evidence that early-life rTL or early-life telomere change predicted short-term survival. Together, these results indicate that early-life telomere variation in bats reflects climatic conditions during development, providing novel insights into how early-life exposures could contribute to inter-individual differences in ageing trajectories in long-lived mammals.

Familial longevity, quantified using the Longevity Relatives Count (LRC) score indicating the proportion of ancestral long-lived family members, associates with a pronounced 13 years delayed onset of cardiometabolic disease (CMD). Understanding the molecular basis of familial longevity therefore provides critical insights into mechanisms of cardiometabolic resilience. However, the combined metabolomics and proteomics profile associated with the delayed CMD onset observed in such long-lived family members is not understood yet. Hence, we integrated plasma metabolomics and proteomics in 495 participants from the Leiden Longevity Study to identify molecular signatures associated with (a contrast in) the LRC score. Metabolomics profiling captured 429 features, including amino acid derivatives, nucleosides, and lipid mediators, while proteomics quantified 374 proteins related to cardiovascular, metabolic, and inflammatory pathways. Three within-family analysis approaches were examined and overlapping findings were interpreted. We identified ten metabolites and nine proteins that are associated with increased familial longevity, exemplified by a high LRC score. High LRC scoring individuals exhibited lower levels of amino acid derivatives (prolylhydroxyproline, 5-hydroxy-tryptophan, asymmetric dimethylarginine), nucleosides (2-methylguanosine, 7-methylguanosine, pseudouridine), N-acetylneuraminic acid and quinolinic acid, indicating optimized extracellular matrix integrity, vascular function, and reduced neuroinflammatory activity. Lipid mediators, including elevated 6-keto-PGF1a and reduced 9-HOTrE/alpha-linolenic acid ratio, reflected preserved endothelial homeostasis and attenuated inflammatory signaling. At the proteome level, strong ancestral familial longevity is associated with immune regulators (RETN, NPPB, IGSF8), extracellular matrix components (EFEMP1, EPHB4), and adhesion/signaling molecules (LRP11, ICAM3, KIT, ADGRG2), highlighting coordinated regulation of inflammation, tissue remodeling, and regenerative capacity. Multi-omics pathway analyses indicated convergence on amino acid and nucleotide metabolism, lipid signaling, extracellular matrix remodeling, and receptor-mediated communication. Collectively, these multi-omics systemic signatures define a molecular framework of ancestral familial longevity characterized by reduced inflammation, preserved tissue integrity, and enhanced metabolic and regenerative processes. Our findings provide mechanistic insight into the biology of familial longevity and potentially cardiometabolic resilience.

Aging is the primary driver of chronic disease and mortality, requiring comprehensive frameworks for quantification of aging and nomination of longevity interventions. We developed mAge (multimodal age), a biological aging framework that integrates plasma proteomics, wearables, and mortality hazard to predict biological age, intrinsic capacity, and mortality risk. By combining proteomic and wearable data in UK Biobank samples, mAge exceeds unimodal baseline age prediction to 0.87 test R{superscript 2} and 2.3 years mean error, and reduces unimodal baseline mortality prediction error by 21%. We further constructed organ-and cell type-specific biological clocks that quantify aging across 49 distinct subsystems, revealing that cardiac, immune, and intracellular protein signatures benefit most from wearable integration. By mapping data to FDA-approved drug targets, we identified interventions, such as GLP-1 receptor agonists, gabapentin, and ACE inhibitors, that are associated with lower overall and subsystem-specific proteomic age and mortality risk or are associated with longer time-to-death and later age-at-death in longitudinal and deceased cohorts. mAge establishes a scalable framework for nominating and validating personalized longevity interventions, bridging continuous digital monitoring with molecular aging diagnostics.

Although remyelination, a central nervous system (CNS) regenerative process mediated by oligodendrocyte progenitor cells (OPCs), takes place in an inflammatory environment the long-term impact of inflammation on OPC remyelination capacity remains unclear. Here, we studied the short- and long-term impact of systemic inflammation on adult OPCs to assess whether transient inflammation triggers enduring chromatin remodelling indicative of inflammatory memory in OPCs. We observed long-lasting epigenetic modifications in response to both lipopolyssaccharide (LPS) and polyinosinic:polycytidylic acid (Poly(I:C)), but only LPS induced a tolerance-like memory. LPS-mediated tolerance-like memory enhanced OPC differentiation after demyelination in aged mice, reducing axonal damage. Our findings reveal OPC epigenetic memory of inflammation as a mechanism by which adult OPCs adapt to inflammatory challenges, which could be harnessed to reduce neuroinflammation and enhance remyelination efficiency in ageing and neurodegenerative diseases.

Aging is accompanied by a progressive decline in physiological function that contributes to chronic disease development. Biological clocks estimated from high-dimensional clinical and biological measurements may provide more granular tracking of the aging processes. Current biological clocks, however, have limited cross-ancestry generalizability and clinical applicability. Here, we developed a multi-ancestry biological clock (ClinBAG) using 22 routine blood and anthropometric biomarkers in 14,328 age- and sex-balanced individuals from the All of Us Research Program. We tested the association of ClinBAG with 434 traits and evaluated its ability to predict incident disease in 152,733 non-overlapping individuals. We also conducted genome-wide association studies in European (N=74,675), African (N=22,315), and Admixed American ancestry individuals (N=19,940). Among 190 neurological phenotypes, elevated ClinBAG was associated with cognitive decline, increased incidence of dementia (HR=1.020, p=1.6x10-5) and Parkinson's disease (HR=1.014, p=0.023), and decreased risk of migraine (HR=0.991, p=8.7x10-4). We also identified common (NPRL3) and ancestry-specific genetic loci (HBB in African-ancestry and FADS1/FADS2 in European-ancestry) for ClinBAG. Single-cell enrichment revealed that ClinBAG-associated genes are overexpressed in double-negative (DN) T cells in an age-dependent manner. This study presents a clinically applicable multi-ancestry biological age clock predicting neurological disease risk. Our findings also uncover population-specific genetic drivers, particularly involving erythropoiesis and DN T-cell-mediated neuroinflammation.

Psychedelic compounds such as lysergic acid diethylamide (LSD) are increasingly studied for their neuroplastic effects and potential relevance to brain aging and neurodegeneration. However, the molecular mechanisms linking psychedelic-induced plasticity to age-associated cognitive decline remain unclear. Brain aging and dementia are characterized by coordinated transcriptional programs that underlie synaptic dysfunction and altered neuron-glia interactions. If psychedelic-induced plasticity engages opposing molecular programs, it could counteract these conserved trajectories. In computational drug discovery, this concept has been formalized as the principle of transcriptional signature reversal, whereby compounds inducing gene expression states opposite to disease-associated programs may exert a therapeutic effect by counteracting disease-associated phenotypes. Here, we combine cross-species transcriptomic analyses with experimental validation to test whether LSD opposes conserved signatures of brain aging and dementia. By comparing transcriptional profiles induced by chronic LSD treatment in rodents with age- and dementia-associated gene expression changes in the human prefrontal cortex, we show that LSD induces gene expression patterns strongly anti-correlated with aging and neurodegeneration programs. This reversal is specific compared to other pharmacological perturbations and is reproducible across datasets and species. Moreover, LSD counteracts amyloid-{beta}-induced structural and molecular alterations in primary cortical neurons, linking transcriptomic opposition to functional rescue under neurodegenerative stress. Together, our findings suggest that LSD modulates molecular and cellular pathways associated with brain aging and neurodegeneration, linking systems-level gene expression changes to structural and functional resilience in neurodegeneration-relevant contexts.

Neurodegenerative diseases are often accompanied by systemic comorbidities, including changes in bone health, but the molecular relationship between neurodegeneration, skeletal decline, and cellular senescence remains poorly understood. In this study, we investigated sex-specific changes in circulating bone- and senescence-related proteins across the spectrum of Alzheimers disease(AD), Lewbody dementia(LB) and Parkinsons disease(PD). Plasma proteomic profiling was performed on samples from 408 participants deeply phenotyped for neurodegenerative diseases, followed by differential protein and pathway analyses. This study reveals sex-dependent alterations in bone and senescence-related circulating proteins in AD-related, PD and LB-related neurodegenerative diseases, providing insights into the complex relationship between neurodegeneration and bone health. Several candidate proteins were also associated with established plasma neurodegeneration biomarkers, particularly pTau181. Pathway analyses revealed shared mitochondrial and metabolic dysfunction across neurodegenerative diseases, with disease-specific features including vesicle trafficking disruption in AD and inflammatory-senescence pathways in LB, plus sex-divergent patterns in inflammatory signaling and bone-related pathways.

Cellular senescence is a tumor-suppressive program characterized by irreversible growth arrest; however, senescent cells can also promote inflammation and alter the tumor microenvironment through the senescence-associated secretory phenotype (SASP). Although SASP induction is regulated by pathways such as p38/NF-{kappa}B/I{kappa}B{zeta}, the mechanisms that restrain excessive or persistent SASP remain largely unknown. Here, we investigated the role of the Ets family transcription factor EHF in SASP regulation during cellular senescence. In IMR-90 human fibroblasts undergoing oncogene-induced senescence, EHF expression was upregulated after the onset of canonical senescence phenotypes. EHF knockdown did not substantially affect senescence establishment but increased SASP-related gene expression. Conversely, overexpression of full-length EHF suppressed SASP-related gene induction during senescence, whereas an ETS-domain-deficient EHF mutant failed to do so, suggesting that this EHF-mediated SASP suppression requires its DNA-binding domain. Furthermore, knockdown of NFKBIZ, which encodes I{kappa}B{zeta} and is induced downstream of NF-{kappa}B signaling, reduced EHF expression during senescence; however, NFKBIZ overexpression increased EHF and SASP-related gene expression. These results link EHF induction to the p38/NF-{kappa}B/I{kappa}B{zeta} inflammatory axis and support a model in which the inflammatory pathway that induces SASP also engages EHF as a negative regulator of SASP. Finally, conditioned medium from senescent cells promoted HCT116 cancer cell migration, and this activity showed a further increase after EHF knockdown. These findings suggest that EHF suppresses senescence-associated inflammatory responses and may function as a senomorphic effector that attenuates SASP-related inflammation without substantially affecting senescence establishment.

Machine learning (ML)- and artificial intelligence (AI)-based aging clocks are increasingly used to quantify physiological and molecular aging from omics and medical imaging data as distinct from chronological age. Here, we characterize a fundamental but underappreciated computational limitation of commonly used ML/AI regression models: systematic prediction bias and its propagation to downstream association estimates. We demonstrate that systematic prediction bias can distort, and in some cases reverse, biomedical conclusions drawn from aging-clock analyses. For example, it can produce spurious associations suggesting that older predicted brain age is linked to better cognitive performance, or that older epigenetic age is associated with better kidney function. To address this problem, we introduce a principled and broadly applicable ML/AI regression framework based on constrained optimization, ensuring trustworthy aging-clock estimation and biomedical inference.

Biological aging clocks are typically evaluated through competitive benchmarking, implicitly assuming that a single metric can sufficiently capture the complexities of aging1-6. Here, we tested an alternative hypothesis: that distinct clock types capture orthogonal dimensions of aging and therefore yield greater value when integrated. Using the Framingham Heart Study, we compared the immune-aging metric, IMM-AGE, with established DNA methylation clocks and found that integrated models consistently outperformed single-clock approaches. To investigate the basis of this complementarity, we derived IMMAGE-Epi, a 22-CpG methylation surrogate of IMM-AGE which exhibited minimal overlap with canonical epigenetic clock CpGs, suggesting that immune aging is associated with a distinct methylomic feature and pathway space rather than representing a reformulation of existing clock architectures. Together, our findings support an emerging multidimensional model of biological aging in which integrating orthogonal biological clocks may offer greater translational utility than competitive single-clock optimization.

tRNA-derived fragments (tRFs) are a class of small noncoding RNAs with emerging roles in stress responses and gene regulation, yet their dynamics across tissues and brain regions during aging remain poorly understood. Here, we systematically profiled age-associated changes in tRFs using three independent mouse small RNA-seq datasets spanning multiple organs and brain regions. Nuclear-encoded tRFs were the only small RNA class showing a strong, progressive accumulation with age, a pattern that was specific to the brain and broadly distributed across brain regions. Fragment length distributions, boundary profiles, and coverage maps were consistent with amplified cleavage at conserved sites, implicating angiogenin as a primary driver. Age-associated increases in specific tRF species, including 5CysGCA, 5GluCTC, and 5GlyGCC, were validated by northern blotting and RT-qPCR. Analogous upregulation was observed in human frontal lobe tissue from frontotemporal dementia patients and in cerebrospinal fluid from traumatic brain injury patients, suggesting that tRF accumulation is further amplified under neurological stress. Together, these findings establish nuclear-encoded tRFs as a small RNA class that accumulates selectively in the aging brain, with potential roles in neurodegeneration, and as biomarkers and targets for therapeutic intervention.