A cGAS/STING tradeoff undercuts a popular target — and decades-old exposures write the biological-age bill
This week's throughline was tradeoffs and decoupling rather than fresh mechanism. The cleanest organismal test yet of cGAS/STING inhibition — a full knockout in killifish — reduced senescence markers without extending lifespan, undercutting a target this digest has covered enthusiastically in prior cycles. A parallel mouse study found that naked mole-rat hyaluronan synthase, which reliably extends healthspan, does nothing for age-related hearing loss — the same lesson in a different system: a validated longevity intervention does not automatically generalize across organs. On the human side, two large observational studies quantified how early — in infancy and in midlife — cardiometabolic exposures compound: sugar rationing in the first 1,000 days of life and five low-risk cardiovascular factors at age 42 each predicted double-digit differences in disease, mortality or frailty five to seven decades downstream. And Austad, Kaeberlein and Miller mined the NIA Interventions Testing Program's own dataset to propose biochemical shortcuts that could let mouse geroscience move faster than a full lifespan study allows.
Survival to age 90 in men with all five low-risk cardiovascular factors at age 42 versus none, in a 50-year Finnish cohort follow-up (n=2,690); frailty prevalence at a mean age of 79 was 2% versus 25% in the same two groups. Observational, men only, single cohort — not a trial, and risk factors were measured once in midlife rather than tracked over time.
Where the human evidence is
Count by statusEvidence-stage map
Evidence stage × investigator promise (1–10)Comparative biology: healthspan gains that don't travel to every organ, and a genomic clue to Peto's paradox
Naked mole-rat hyaluronan synthase improves healthspan and lifespan in transgenic mice but does not protect against age-related hearing loss — aging phenotypes decouple even within one validated intervention
Transgenic mice expressing naked mole-rat Has2 produce very-high-molecular-mass hyaluronan and, per this group's prior work, gain improved healthspan, reduced inflammaging and extended lifespan. Here, despite cochlear nmrHas2 mRNA elevated roughly 780-fold over controls, auditory brainstem response testing showed statistically indistinguishable, significant age-related hearing loss in transgenic and wild-type mice from 3 to 12 months (P≥0.26 between genotypes at 12 months). The authors propose tissue-specific hyaluronidase degradation of the protective polymer, or absence of the CD44 receptor needed for its anti-inflammatory signaling in cochlear hair cells, as likely explanations. The result is useful precisely because it is negative: a single validated pro-longevity intervention does not generalize to every age-related phenotype, and healthspan/lifespan gains can be mechanistically decoupled from organ-specific functional decline. Female mice only, and a supplemental arm of a reproductive-aging study rather than a purpose-built hearing trial.
Read the paperCancer-associated genomic loci are more evolutionarily conserved and carry distinct m6A methylation patterns in longer-lived mammals — a genomic contribution to Peto's paradox
Peto's paradox — cancer risk not scaling with body size or lifespan despite larger, longer-lived animals accumulating far more cell divisions — has several proposed genomic explanations; this adds evolutionary conservation of cancer-associated sequence as one. Mapping human cancer-associated loci onto whole-genome alignments across 58 mammals (23 classified longer-lived, 35 shorter-lived), the authors found no overall difference in evolutionary rate between the groups genome-wide, but cancer-associated sites specifically were more conserved in longer-lived species. Longer-lived mammals also showed disproportionate m6A RNA methylation enrichment at cancer-associated transcripts relative to their own lower global m6A baseline, concentrated on tumor-suppressor transcripts. The authors are explicit that this is a correlative genomic pattern — causal molecular mechanisms remain unproven, and the m6A data derive from male samples only — so it nominates a mechanism for future testing rather than demonstrating one.
Read the paperHuman evidence: metabolic exposure in infancy and in midlife each write a 50-year bill
Being in the first 1,000 days of life during Britain's post-war sugar rationing predicts lower disease incidence, lower mortality, and younger biological age five decades later
British sugar rationing (roughly 40g/day, close to current dietary guidance) ended in September 1953, and infant sugar intake roughly doubled afterward — creating quasi-random exposure by birth month that the authors use to isolate early-life sugar restriction from confounded diet-quality and caloric-restriction effects. Among 64,809 UK Biobank participants born October 1951–March 1956 and followed to 2023, exposure during the first 1,000 days was associated with 9% lower incidence of 83 aging-related hallmark diseases (HR 0.91, 95% CI 0.88–0.94), 19% lower all-cause mortality (HR 0.81, 95% CI 0.69–0.93, roughly 60% mediated through disease prevention), 1.0–1.2 years younger biological age across multiple clocks, and a 47-protein signature enriched for AMPK/longevity signaling with suppressed mTOR activity. This is a natural experiment, not a randomized trial: birth-month assignment approximates but does not guarantee randomization, adult diet was measured only once, the cohort is predominantly White European and UK-Biobank-selected, and a 1–2 year clock difference, while statistically robust at this sample size, is of uncertain meaning for any individual patient.
Read the paperFive low-risk cardiovascular factors at age 42 predict roughly six-fold higher survival to 90 and a ten-fold lower rate of frailty half a century later
In a Finnish cohort of 2,690 healthy men enrolled at mean age 42 in the 1960s–70s and followed for over 50 years, men with all five low-risk factors at baseline (non-smoking, BMI<25, systolic BP<140 mmHg, total cholesterol<6.0 mmol/L, 1-hour post-load glucose<9.0 mmol/L) reached age 90 at roughly six times the rate of men with none of the five (42% vs 7%). Among the 907 survivors reassessed at mean age 79, frailty prevalence was 2% versus 25%, and self-reported happiness scores were modestly higher (8.0 vs 7.5), in the all-low-risk versus zero-low-risk groups. The size and durability of this gradient — a single midlife measurement, read out five decades later across survival, frailty and subjective wellbeing simultaneously — is unusual for an observational cohort and reinforces that cardiometabolic risk compounds across the full lifespan rather than acting mainly in old age. Men only, a single Finnish cohort from a specific era, and risk factors were measured at one time point rather than tracked longitudinally, so how later-life changes in risk factors would modify the trajectory is not addressed.
Read the paper