NIH Funding Freeze Eases, Grant Pipeline May Reopen: Biomedical researchers got a measure of relief after the White House finally approved the NIH’s 2026 budget apportionments, allowing the agency to begin spending more freely on research grants. The move followed criticism from lawmakers, especially Representative Rosa DeLauro, who said grant awards had slowed to a near standstill. NIH Director Jay Bhattacharya told Congress that funding is now available and promised that all of it will be spent on strong science by year’s end. He said new grants should begin flowing soon and that NIH is also working to fill leadership vacancies created during recent turmoil and staffing cuts. Although Bhattacharya downplayed reports of a severe slowdown, agency staff and outside observers say shortages and budget delays have significantly disrupted grant processing and administration across NIH. (Science)

Dopamine’s Identity Crisis: At the upcoming Dopamine Society meeting in Seville, neuroscientists are set to debate whether the long-dominant reward prediction error model still explains dopamine’s role in the brain. For decades, researchers saw dopamine as a signal linking cues to rewards, helping animals and people learn what to seek. But newer tools for measuring dopamine in living brains have revealed a far more complicated picture. Studies now suggest dopamine also responds to movement, novelty, threats, attention, memory, and social behavior. Some scientists argue the classic framework can be expanded to fit these findings; others think it has become too stretched and needs replacement. The stakes are high: theories of dopamine shape how researchers and clinicians think about ADHD, addiction, and other disorders, making this more than an abstract scientific fight. (Nature)

Epigenetic Clocks Track Survival, Not Just Age: A new Nature Aging paper argues that how a person’s epigenetic clocks change over time may matter more than any one-time reading. Using longitudinal data from the InCHIANTI cohort, the researchers found that changes across several DNA methylation clocks were independently associated with long-term mortality over as much as 24 years of follow-up. That matters because the field has produced many “biological age” clocks, but there has been ongoing debate over whether repeated measures actually add meaningful clinical information. This study strengthens the case that trajectories, not just snapshots, could be useful in aging research and perhaps eventually in medicine. The finding does not mean epigenetic clocks are ready for routine clinical use, but it does suggest they are moving closer to relevance for real-world survival prediction. (Nature)

DeepStrataAge Finds Aging Happens in Sex-Specific Phases: Researchers reporting in npj Aging unveiled DeepStrataAge, a deep-learning DNA methylation clock trained on 29,167 samples and 12,234 CpG sites. The model reportedly reached an accuracy of 1.89 years and, more importantly, was designed to be interpretable rather than just predictive. Using SHAP analyses, the team says it uncovered wave-like aging dynamics, with an early-life module, a midlife transition, and later-life remodeling, with different timing in males and females. That is a useful shift in emphasis. Instead of treating aging as one smooth, linear process, the paper suggests biological aging may unfold in distinct phases, potentially opening the door to better-targeted interventions. It is still a clock, not a therapy, but it pushes the field toward a more mechanistic view of how aging progresses over the lifespan. (Nature)

Killifish Study Suggests Aging Has a Behavioral Architecture: A Stanford-led study highlighted in EurekAlert tracked short-lived African killifish continuously across their lives and found that aging appeared to proceed in discrete steps rather than as a smooth decline. The striking implication is that everyday behavior by midlife may already contain clues about lifespan. In the release, researchers describe how movement patterns helped reveal the “architecture” of aging, suggesting that behavioral readouts can expose hidden physiological transitions long before death. For longevity science, that is intriguing because it hints that aging may be measurable through continuous passive observation rather than only blood draws, imaging, or tissue assays. Fish are not humans, of course, and the work is still best understood as a model-system insight. But it adds to a growing sense that aging may have identifiable stages and signatures visible in motion itself. (EurekAlert!)

Six Tiny RNA Signals in Blood May Forecast Near-Term Longevity: Science News reports on a study in Aging Cell suggesting that levels of six piRNAs in blood may help identify which older adults are most likely to survive the next two years. In a cohort of more than 1,200 people aged 71 and older, the RNA panel reportedly predicted short-term survival with up to 86 percent accuracy, outperforming age, cholesterol, physical activity, and more than 180 standard health measures. That is notable because the hunt for practical blood-based aging biomarkers has often produced signals that are statistically interesting but clinically fuzzy. These piRNAs are not yet ready for the doctor’s office, and the results still need replication. But the work points toward a future in which aging prognosis might rely less on broad health proxies and more on compact molecular signatures that reflect resilience directly. (Science News)

Wearables May Catch Brain Aging Clues Earlier Than Expected: Longevity.Technology highlights new work suggesting that consumer wearables may be able to detect subtle changes in cognition and mood before they become clinically obvious. The most informative signals were not exotic digital biomarkers but familiar variables: sleep, heart rate, weather, and pollution exposure. That framing is important. It suggests that brain aging may be legible through ordinary physiological and environmental patterns captured passively over time. The article also notes limits: only a small subset of outcomes clearly beat a population-average prediction model, so this is better read as proof of feasibility than as a finished monitoring system. Even so, the direction is compelling for longevity medicine. If validated in larger cohorts, passive wearable data could become an early-warning layer for cognitive decline, shifting monitoring from occasional clinic visits to continuous observation embedded in daily life. (Longevity.Technology)

Gut Microbes Could Help Explain Why GLP-1 Drugs Work Better for Some People: A Longevity.Technology report points to growing evidence that the gut microbiome may influence how strongly people respond to GLP-1 drugs. The basic idea is biologically plausible: fiber-eating gut bacteria generate metabolites that help stimulate GLP-1 release, while dysbiosis and inflammation may blunt the system. The article frames this as one possible reason some patients see dramatic benefits while others get weaker results. It also notes that GLP-1 drugs themselves may reshape the microbiome, making the relationship potentially two-way. For the longevity crowd, this matters because GLP-1 drugs are increasingly discussed not only for diabetes and obesity but for broader metabolic and healthy-aging effects. The field is not at the point of microbiome-personalized GLP-1 prescribing yet, but this is the kind of mechanistic clue that could push it there. (Longevity.Technology)

Alzheimer’s Risk Mapping Gets a Huge Real-World Data Upgrade: A Columbia-led team has launched a new metaplatform, M3AD, that integrates nearly 10 million electronic health records across major health systems in New York City, Chicago, and Miami, including about 60,000 people with Alzheimer’s disease and related dementias. According to the EurekAlert release, the goal is to capture the messy, long-term interplay of chronic disease, behavior, social conditions, and environment that precedes dementia. That matters because Alzheimer’s risk is rarely driven by one variable, and many datasets are too narrow to model the full path to disease. A platform of this scale could improve risk prediction and help identify more realistic intervention windows. The caveat is that better data infrastructure is not the same thing as a treatment breakthrough. But for aging neuroscience, this is the kind of systems-level resource that can reshape the questions researchers ask. (EurekAlert!)

AI Model Pulls Telomere Clues From Ordinary Biopsy Slides: Phys.org reports on a new computational model called TLPath that predicts telomere length from routine histopathology images. The model was trained on 5,263 slides from 18 tissue types donated by 919 individuals and reportedly outperformed chronological age as a predictor in multiple tissues. The logic is elegant: if telomere shortening leaves detectable traces in the architecture of cells and tissues, image analysis may provide a scalable surrogate for a key aging-related metric without specialized telomere assays. That could be especially useful because hospitals and biobanks already contain vast numbers of digitizable tissue slides. As always, there is a gap between technical promise and clinical deployment, and the approach will need broader validation. Still, it is a strong example of AI turning existing medical archives into aging data rather than waiting for new infrastructure to be built. (Phys.org)

Blood Chemistry and Gut Microbes Power New Biological Age Models: A new paper in Aging describes two neural-network models for forecasting biological age: one based on standard blood biochemistry and another on gut microbiota composition. The blood model uses seven indicators with sex-specific inputs, while the microbiome model uses 45 bacterial species. Both achieved mean absolute errors of roughly six to seven years and R² values above 0.8, while also being made more interpretable through SHAP analysis. That combination of parsimony and explainability is notable. The field has plenty of black-box aging models, but fewer that try to remain clinically practical and transparent. The paper does not claim to have solved biological age measurement, and the cohort context matters. Still, the study shows how routine lab work and gut microbial profiles are increasingly being treated as parallel windows onto the aging process. (Aging)

Simple Strength Tests Still Look Like Hard-Nosed Longevity Predictors: One of the more practically useful stories of the week comes via SciTechDaily: a University at Buffalo-led study following more than 5,000 women ages 63 to 99 found that stronger grip strength and faster chair-rise performance were associated with lower mortality risk over eight years. The headline number is clean and memorable: every additional seven kilograms of grip strength was linked to a 12 percent lower mortality rate. The appeal here is not novelty so much as translation. Longevity research can drift toward complex biomarkers and costly profiling, but grip and sit-to-stand tests are cheap, quick, and already familiar to clinicians. They are not destiny, and association is not intervention. But they remain a powerful reminder that in aging science, straightforward functional capacity may tell us as much about survival as many more sophisticated measurements. (SciTechDaily)