2026-07-08
New evidence Still strong support
A 2026 review pooled 41 randomized trials (11,161 men) and found testosterone therapy did not raise prostate-cancer events (odds ratio 0.88, not significant).
The new study: Garcia-Becerra CA et al. (2026) — Meta 41 RCTs (n=11,161): TTh not associated with increased prostate-cancer events (OR 0.88) or clinically-significant prostate cancer (OR 1.13, both ns) - reinforces no-increased-risk.
What we already knew On the cancer question, the modern evidence is reassuring: the old fear that testosterone feeds prostate cancer is not supported. Randomized trials, large registries, and even prostate-cancer-survivor studies show no increased risk, and some show lower risk. The main caveat is that the trials weren't long enough to be the final word, so men on testosterone should still get routine PSA/prostate monitoring.
Why it matters: The strongest randomized-trial safety data yet — it reinforces, rather than changes, the modern reversal of the old 'testosterone causes prostate cancer' fear. Men on TRT should still get routine PSA monitoring.
2026-07-08
New evidence Still contested
The same 41-trial pooled analysis (11,161 men) found no increase in major heart events — heart attack, stroke, or cardiac death — with testosterone therapy.
The new study: Garcia-Becerra CA et al. (2026) — Meta 41 RCTs (n=11,161): TTh not associated with increased MACE (OR 0.83, 95% CI 0.52-1.32, ns) - supports short-to-mid-term cardiovascular safety (authors note long-term data still needed).
What we already knew Mostly reassuring, not fully settled. The big 2023 TRAVERSE trial found no increase in major cardiac events (heart attack, stroke, cardiac death) — putting an old scare to rest — but it did find more atrial fibrillation and blood clots in the lungs. So: probably safe for major events in the right patients, with a real, unresolved signal for irregular heartbeat and clots.
Why it matters: It nudges the cardiovascular-safety picture toward reassuring, but the verdict stays contested: an earlier large trial's signal of more atrial fibrillation and blood clots still stands, so this is 'reassuring on the big events, watch the rhythm/clot risk.'
2026-07-08
New evidence Still strong support
A 2026 meta-analysis of 9 trials (827 people) found that adding metformin to an exercise programme blunted the gains in cardiorespiratory fitness (VO2peak) and blood pressure that exercise produced on its own.
The new study: Etayo-Urtasun P et al. (2026) — Meta 9 studies (n=827): adding metformin to exercise attenuated VO2peak gain (MD -1.19) and blunted systolic/diastolic BP reductions vs exercise alone (GRADE moderate).
What we already knew Yes, blunting fitness and muscle gains in older adults — a real trade-off worth weighing.
Why it matters: Firms an already-strong finding: if you train specifically to improve fitness and blood pressure, metformin may partly cancel that benefit. Worth a conversation with your doctor about timing and whether you need it — not a reason to stop a prescribed drug on your own.
2026-07-08
New evidence Still strong support
A 2026 meta-analysis of 19 studies (91,000+ people) found higher Mediterranean-diet adherence cut major heart events, and a separate 7-year randomized trial agreed — though in that trial a low-fat diet worked about as well.
The new study: Volpe R et al. (2025) — Meta 19 studies (>91,000): higher Mediterranean-diet adherence lowered major CV events (RCT RR 0.44; cohort RR 0.95) and mortality; effects on BP and lipids were not significant.
What we already knew Yes — among the strongest diet evidence, mostly from fewer strokes.
Why it matters: Reinforces the Mediterranean diet's heart benefit, with an honest footnote: diet quality and how well you stick to it may matter more than the specific 'Mediterranean' label.
2026-07-08
New evidence Still strong support
A 2026 meta-analysis of 38 studies (about 1,100 people) found that ketone supplements gave a small but consistent boost to cognitive performance, with bigger doses helping more.
The new study: Bonnechere B et al. (2026) — Meta 38 studies / 29 protocols (n=1,117): exogenous ketones modestly improved cognition vs placebo (SMD 0.29, 95% CI 0.16-0.41), dose-dependent.
What we already knew Yes for fuel adequacy — ketones are a genuine, efficiently-used brain fuel, especially where glucose uptake is impaired, but "better than glucose in healthy brains" is not established.
Why it matters: Consistent with the idea that ketones are a usable backup fuel for the brain. The effect is modest, not dramatic — a gentle nudge, not a nootropic miracle.
2026-06-30
Verdict changed Leans support ↑ upgraded Strong support
What tipped the scales: Payab (2020) — MA: P. vulgaris significantly improved weight (SMD -0.88) in overweight/obese adults
What we already knew Yes — modest weight loss, but the active blocker is destroyed by baking and barely touches blood sugar.
Why it matters: Modest but consistent weight loss across trials. The catch for the bakery: the active amylase-blocker is destroyed by baking and barely moves blood sugar, so this is a supplement story, not a bread ingredient one.
2026-06-30
Verdict changed Insufficient ↑ upgraded Leans support
What tipped the scales: Nyambe-Silavwe (2016) — Crossover RCT: green-tea+polyphenol/fibre food cut postprandial glucose iAUC 27-49% and insulin 47% on a bread/starch load via amylase inhibition
What we already knew Probably a small effect — most human and lab tests show tea blunts the starch glucose rise, though the cleanest trial was null.
Why it matters: Tea polyphenols blunt the glucose spike from a starch meal (they inhibit amylase) in most trials — a usable, low-risk lever to pair with a bread/starch load. The cleanest trial was null, so effect size varies person to person.
2026-06-30
Verdict changed Insufficient ↓ downgraded Contested
What tipped the scales: Dalenberg JR, et al. (2020) — Sucralose+carb x10d lowered insulin sensitivity; sucralose alone & carb alone = no effect
What we already knew Unclear — evidence is mixed; any effect seems to show up mainly when paired with carbs or in certain gut-microbe types.
Why it matters: Any harm shows up mainly when sucralose is taken *with carbohydrate*, or in certain gut-microbiome types — sucralose alone is often metabolically neutral. So it's context-dependent, not a blanket 'sucralose wrecks insulin' effect.
2026-06-30
Verdict changed Insufficient ↑ upgraded Strong support
What tipped the scales: She J, et al. (2024) — RCT n=40: atorvastatin reduced GLP-1 & worsened glucose via gut bile-acid axis; UDCA rescued it
What we already knew Yes — statins appear to lower this hormone via a gut-bile pathway, though the human evidence rests on one small study.
Why it matters: Statins appear to lower GLP-1 through a gut–bile-acid pathway, coherent with their small diabetes-risk signal. Confident on direction, thin on magnitude — the direct human evidence is a single 40-person trial, so worth watching, not alarming.
2026-06-30
Verdict changed Leans against ↑ upgraded Leans support
What tipped the scales: Hooper (2016) — Reduced/modified dietary fat review: modest reduction in cardiovascular events with sustained fat modification.
What we already knew The evidence is genuinely split: RCT-only pooling shows no mortality or heart-attack benefit, but the fuller Cochrane analysis finds a ~17% drop in combined cardiovascular events, so the honest answer is mixed and depends on what is counted.
Why it matters: Genuinely split, and now leaning supportive: RCT-only pooling shows no mortality/heart-attack benefit, but the fuller Cochrane analysis finds ~17% fewer cardiovascular *events* when saturated fat is replaced with polyunsaturated. The answer depends on what you count.
2026-06-30
Verdict changed Contested ↓ downgraded Leans against
What tipped the scales: Guo 2021 (2021) — RS had NO direct effect on body weight/composition in humans
What we already knew Probably not — most trials show little to no weight loss, with one small promising study not yet replicated.
Why it matters: Resistant starch helps glucose and triglycerides but doesn't meaningfully reduce body weight. Set weight-loss expectations low for resistant-starch breads — the metabolic wins are real, the scale won't move much.
2026-06-30
Verdict changed Strong support ↓ downgraded Leans support
What tipped the scales: Puente-Fernandez (2023) — SR/MA: multi-ingredient (protein-based) supplementation added to exercise gave NO additional body-composition/performance benefit vs isocaloric control in mid/older adults.
What we already knew Probably yes — enough protein with resistance training preserves muscle in a calorie deficit, but protein alone won't substitute for the training.
Why it matters: It's the resistance *training* that preserves muscle in a calorie deficit — adding protein supplements on top gives little further benefit. So adequate protein helps, but protein alone is no substitute for lifting.
2026-06-30
Verdict changed Strong support ↓ downgraded Leans support
What tipped the scales: Appleby (2016) — EPIC-Oxford: vegetarians vs comparable non-vegetarians had no significant all-cause mortality difference
What we already knew Probably yes — fairly consistent, but it's an association and depends on food quality.
Why it matters: Still probably protective, but downgraded: a large cohort found no mortality difference for vegetarians versus comparable non-vegetarians. The *quality* of the plant foods matters more than the vegetarian label itself.
2026-06-30
Verdict changed Contested ↑ upgraded Leans support
What tipped the scales: Chen (2020) — Rotterdam + MA of 11 cohorts: higher animal protein assoc. with higher all-cause/CVD mortality; plant protein inverse.
What we already knew Possibly — midlife intake leans toward higher risk in several cohorts, but it's inconsistent and appears to reverse after 65.
Why it matters: In midlife, higher animal-protein intake leans toward higher mortality across several cohorts, and swapping in plant protein tracks lower risk. But it's observational and appears to reverse after ~65, so age changes the advice.
2026-06-30
Verdict changed Leans support ↓ downgraded Contested
What tipped the scales: Stote (2007) — Reduced meal frequency (1 vs 3 meals/day, isocaloric) RAISED LDL & total cholesterol; opposite direction in healthy adults.
What we already knew Unclear — one tiny seven-person study drove the early signal and later trials are mixed-to-null.
Why it matters: The early 'more, smaller meals → better lipids' signal came from one tiny 7-person study; later trials are mixed-to-opposite (one-meal-a-day actually *raised* LDL). Meal frequency is not a reliable lipid lever.
2026-06-30
Verdict changed Strong support ↓ downgraded Leans support
What tipped the scales: Johnstone 2008 (2008) — high-protein KETO lowered ad-lib intake vs moderate-carb (confounded by protein)
What we already knew Yes: in a tightly controlled inpatient trial people spontaneously ate ~550-700 fewer calories per day on a low-fat plant-based diet than on a low-carb ketogenic one, contradicting the carbohydrate-insulin model.
Why it matters: Downgraded: the famous inpatient trial still shows people spontaneously ate fewer calories on low-fat, but other controlled trials point the opposite way. So it's no longer a clean win for either low-fat or low-carb — context dependent.
2026-06-30
Verdict changed Refuted ↑ upgraded Leans support
What tipped the scales: Wang (2020) — 327,037 participants: consistent ~21% RR reduction per mmol/L LDL-C lowering across baseline LDL-C and risk strata.
What we already knew At the TRIAL level the size of LDL-lowering is a weak surrogate for benefit (R-squared near 0), yet per-mmol meta-analyses and Mendelian randomization still show LDL is causal — so this is a surrogacy gap, not proof LDL is harmless.
Why it matters: Important nuance: per-mmol meta-analyses and Mendelian randomization still show lowering LDL causally cuts cardiovascular risk. The earlier 'refuted' reflected a trial-level *surrogacy* gap, not LDL being harmless — LDL still matters.
2026-06-30
Verdict changed Insufficient ↓ downgraded Contested
What tipped the scales: NabizadehAsl (2022) — DB crossover: L. paracasei subsp. paracasei 431 +/- inulin did not robustly raise satiety hormones; effects on GLP-1 not significant.
What we already knew Unclear — evidence is mixed; the main trial found no effect on the hormone, appetite, or blood sugar.
Why it matters: The specific L. paracasei → GLP-1 claim doesn't hold up — the direct trials are null. GLP-1 boosts seen with *other* strains (e.g. L. reuteri) don't transfer, so probiotic choice matters and this one isn't the lever.
2026-06-30
Verdict changed Strong support ↓ downgraded Leans support
What tipped the scales: Struik (2020) — In energy-restricted T2D, very-low-carb and high-carb diets gave comparable fasting appetite; high-carb gave GREATER daily fullness, against keto.
What we already knew Likely — keto modestly curbs hunger by blunting the appetite spike of dieting, though some controlled trials find no real advantage.
Why it matters: Downgraded: keto curbs hunger only modestly, and controlled trials find the appetite edge often disappears once calories and protein are matched. The dramatic 'keto kills hunger' framing isn't carried by the cleaner trials.
2026-06-30
Verdict changed Insufficient ↑ upgraded Leans support
What tipped the scales: Cani (2006) — Landmark pilot: oligofructose promoted satiety and reduced energy intake in healthy humans, the human signal later attributed to GLP-1/PYY induction.
What we already knew Probably a small boost — animal and human trials lean toward inulin raising this appetite hormone, though one careful trial was null.
Why it matters: Fermentable inulin nudges satiety hormones up in both animals and humans — supportive overall. Expect a small effect, not a drug-like one, and one careful trial was null, so it's a gentle lever.
2026-06-30
Verdict changed Insufficient ↓ downgraded Contested
What tipped the scales: Huang (2024) — Network meta-analysis: IF/ADF and continuous CR produce comparable weight loss; no clear superiority of fasting beyond energy restriction.
What we already knew No clear advantage — when calories are matched, fasting adds no extra weight loss beyond simply eating less; trials split.
Why it matters: When calories are matched, intermittent fasting adds no extra weight loss beyond simply eating less — network meta-analysis finds it comparable to continuous restriction. Its real value is adherence, not metabolic magic.
2026-06-30
Verdict changed Insufficient ↑ upgraded Leans support
What tipped the scales: Stachowska (2020) — SR of prebiotic fibre (incl GOS) in NAFLD: favorable reductions in HOMA-IR, fasting insulin and BMI vs control
What we already knew Probably modestly yes overall, though the best trial found it shifts gut bacteria without actually improving blood-sugar handling.
Why it matters: GOS prebiotic fibre modestly improves insulin markers across trials — a real but small effect. Note the cleanest trial shifted gut bacteria without actually improving blood-sugar handling, so temper expectations.
2026-06-30
Verdict changed Refuted ↑ upgraded Leans against
What tipped the scales: Poffe (2019) — Ketone-ester post-exercise during 3-wk overload blunted overreaching and enhanced endurance performance vs control.
What we already knew Probably not — pooled trials show no net benefit, and the famous positive study was industry-run and failed to replicate.
Why it matters: Net, ketone esters don't improve performance and the famous positive study was industry-run and unreplicated — but a couple of recovery/overreaching contexts show real signals, so a flat 'useless' overstates it. Softened from refuted to leans-against.
2026-06-30
Verdict changed Contested ↑ upgraded Leans support
What tipped the scales: Deng et al. (2026) — [FT-verified] LPS->DNMT3->GLP-1R promoter hypermethylation->reduced receptor (resistance mechanism)
What we already knew Probably modestly yes — inflammation seems to dampen the hormone's signal, but this is shown mainly in animals and the lab.
Why it matters: Inflammation plausibly blunts GLP-1 signaling — newer work shows endotoxin can silence the GLP-1 receptor via methylation. But the evidence is mostly animal/mechanism, so it's a credible pathway, not yet proven in humans.
2026-06-30
Verdict changed Leans support ↓ downgraded Leans against
What tipped the scales: Eisa & Barood (2026) — MA: incretin drugs drive weight/lean-mass change; benefit comes from the pharmacologic agonist, not from stacking dietary incretin stimulation.
What we already knew Probably not — the drug already maxes out the receptor, so food-driven hormone seems redundant rather than additive.
Why it matters: Stacking food-driven GLP-1 on top of a GLP-1 drug looks redundant — the drug already saturates the receptor. Eat fibre and protein for their own sake, not expecting them to amplify a Wegovy/Ozempic-class medication.
2026-06-30
Verdict changed Leans against ↑ upgraded Leans support
What tipped the scales: Sebastian (2024) — Meta-analysis of RCTs: Mediterranean diet (a Blue-Zone-type pattern) reduced cardiovascular disease, the strongest interventional support for the pattern.
What we already knew Probably modestly yes that the eating pattern tracks longer life, but that the diet causes it is shaky and partly a bad-records artifact.
Why it matters: The eating pattern does track lower cardiovascular risk and mortality, so it's modestly protective. But 'Blue Zones cause longevity' is partly a bad-records artifact — the diet helps; the mystique oversells the causation.
2026-06-30
Verdict changed Strong support ↓ downgraded Leans support
What tipped the scales: Han Y, et al. (2018) — 121 Korean overweight adults tolerated 4g and 7g/dose d-allulose twice daily for 12wk without significant GI/diarrhea events at these moderate doses
What we already knew Probably yes — but only at large doses; the small amounts used to blunt blood sugar are well below that threshold.
Why it matters: The diarrhea is real only at large doses — controlled trials tolerated 4–7 g with no GI events. The small amounts used to blunt a glucose spike sit well under that threshold, which is reassuring for bread/CGM testing.
2026-06-30
Verdict changed Contested ↓ downgraded Refuted
What tipped the scales: Sun (2024) — Umbrella review of RCT meta-analyses: IF reduces total and LDL cholesterol vs controls, not raising LDL.
What we already knew No — most trials show no rise versus ordinary calorie cutting, with only one outlier suggesting harm.
Why it matters: Alternate-day fasting does not raise LDL; pooled trials show it slightly lowers cholesterol versus ordinary calorie-cutting. The earlier 'fasting harms lipids' worry rested on a single outlier and no longer holds.
2026-06-30
Verdict changed Leans against ↑ upgraded Strong support
What tipped the scales: Brooke-Taylor (2017) — Systematic review of 39 studies: A1 delays intestinal transit and triggers inflammation (rodents) and looser stool/discomfort (humans) vs A2.
What we already knew Yes — but much of the supporting evidence comes from A2-milk marketers, so weigh that conflict of interest.
Why it matters: The gut effect is now well-supported — A1 casein slows transit and irritates the gut in sensitive people. Caveat worth keeping in view: much of the evidence comes from A2-milk marketers, so weigh the conflict of interest.
2026-06-30
Verdict changed Insufficient ↓ downgraded Contested
What tipped the scales: Kullenberg de Gaudry (2019) — Systematic review of human A1 outcomes: evidence insufficient/low-quality for noncommunicable-disease harm; no cognition support.
What we already knew A1 milk reliably worsens GI symptoms and transit versus A2 in sensitive people, but the cognitive-speed benefit is weak, inconsistent, and largely from one research lineage.
Why it matters: A1 milk's *cognitive* harm now reads as unproven: the positive trials trace back to A2-milk-linked groups, while independent reviews rate the human cognition evidence low-quality. The firm, separate claim is that A1 worsens GI symptoms.
2026-06-29
Verdict changed Strong support ↓ downgraded Leans support
What tipped the scales: Schutz (2021) — Critical review: low-carb/keto weight loss not durably superior; weight regain common, frames them as recurrent 'fad diets'
What we already knew Probably yes — but most data is uncontrolled, from one company, and likely flatters those who stuck with it.
Why it matters: The program does drive weight loss, but most data are uncontrolled, from one company, and likely flatter the people who stuck with it. Downgraded — real effect, oversold evidence.
2026-06-29
Verdict changed Contested ↑ upgraded Leans support
What tipped the scales: Maissan P (2021) — Systematic review: sirtuin activity extends lifespan across organisms via NAD+-dependent regulation at the metabolism/cell-cycle interface (supportive synthesis).
What we already knew Unclear — evidence is mixed, only in yeast and worms, and a careful study saw it largely vanish.
Why it matters: Sirtuins plausibly mediate part of caloric restriction's lifespan effect — but only shown in yeast and worms, a careful study saw the effect largely vanish, and mammalian proof is lacking. Mechanism, not a human fact.
2026-06-29
Verdict changed Insufficient ↓ downgraded Contested
What tipped the scales: Miller (2011) — [FT-verified] Miller2011 NIA-ITP: resveratrol 300/1200ppm NO survival effect; rapamycin did. Robust null in healthy mice
What we already knew Correct — no human longevity outcome exists and human metabolic RCTs are inconsistent/mostly null, so his dismissal of a longevity benefit holds (though "no effect whatsoever" would overreach).
Why it matters: Lifespan extension showed up only in obese mice and failed the rigorous NIA replication; there's no human longevity data. 'Resveratrol extends life' is not supported — at best diet-context-specific in rodents.
2026-06-29
Verdict changed Refuted ↑ upgraded Leans against
What tipped the scales: Howitz (2003) — Origin claim: resveratrol lowers SIRT1 Km, stimulates p53 deacetylation, extends yeast lifespan 70%
What we already knew Probably not — four independent labs showed the original finding was a lab-test quirk, only ever seen in the lab.
Why it matters: The headline mechanism is essentially dead: the original 'resveratrol switches on SIRT1' result was a fluorophore-assay artifact, and four independent labs refuted it. Any SIRT1 link is indirect — the direct-activation story doesn't survive.
2026-06-29
Verdict changed Leans support ↑ upgraded Strong support
What tipped the scales: Migliavacca (2019) — Muscle NAD+ biosynthesis (NAMPT) and mitochondrial capacity reduced in human sarcopenia across three independent cohorts (n=119)
What we already knew Yes, agreed across many labs, though it's an observed link and the human drop varies by tissue.
Why it matters: That NAD+ falls with age is now well-replicated across labs and tissues. Crucial caveat: this is the *decline itself* — it is NOT proof that supplementing to raise NAD+ helps, which is the separate (and largely null) NMN/NR claim.
2026-06-29
Verdict changed Leans against ↑ upgraded Contested
What tipped the scales: Corley (2024) — Retrospective trial analysis: metformin reversed monocyte epigenetic age in virally-suppressed older PLWH; cell-type-specific, small
What we already knew No — the definitive trial (TAME) has not reported outcomes and existing human evidence is indirect/confounded, so "fuzzy/inconclusive" is accurate; corroborates the vault's existing contested state.
Why it matters: Still no human longevity trial (TAME is pending), but the evidence is now genuinely split rather than clearly negative. A promising hypothesis in non-diabetics — not yet something the data support acting on.
2026-06-29
Verdict changed Insufficient ↑ upgraded Strong support
What tipped the scales: Ference (2017) — EAS consensus: genetic, epidemiologic & RCT evidence that LDL causally drives ASCVD in a log-linear, cumulative-exposure manner (supports premise LMHR LDL is atherogenic).
What we already knew Yes — high "bad" cholesterol raises plaque risk however it got high; the popular claim that this pattern is uniquely safe isn't supported.
Why it matters: High LDL drives plaque however it got high — the consensus that LDL is causal is among the strongest in cardiology. The popular claim that the lean-mass-hyper-responder pattern is uniquely safe isn't supported; this is the honest counterweight to that narrative.
2026-06-29
Verdict changed Insufficient ↑ upgraded Strong support
What tipped the scales: Feldman D, Huggins S, Norwitz NG (2022) — Acute hyper-caloric high-fat overfeeding LOWERED LDL in LMHR - a LEM prediction confirmed
What we already knew Yes — the pattern is real, but its proposed explanation rests mostly on one research group's work.
Why it matters: The lean-mass-hyper-responder phenotype is real and the lipid-energy model fits it (overfeeding lowers their LDL, as predicted). But the support leans on one advocacy network — treat the mechanism as plausible, not settled. It does NOT mean the resulting high LDL is safe.
2026-06-29
Verdict changed Insufficient ↑ upgraded Leans support
What tipped the scales: Li (2021) — ESTHER cohort: GrimAge/PhenoAge acceleration independently predict all-cause mortality, validating clocks as biological-age biomarkers
What we already knew Probably modestly yes for predicting age, but it's only a correlation, not proof they track true biological ageing.
Why it matters: The clocks predict mortality better than chance, so they're useful biomarkers — but the link is correlational, not a validated readout of true biological age. Don't over-trust a single 'biological age' test result.
2026-06-29
Verdict changed Leans against ↓ downgraded Refuted
What tipped the scales: Chen (2024) — Meta-analysis of 8 NMN RCTs: no significant benefit on fasting glucose, insulin, HbA1c, HOMA-IR or lipids - human metabolic effect of raising NAD+ unsupported.
What we already knew No — it held in mice, but human trials that raised the molecule showed no benefit.
Why it matters: The mouse mechanism didn't translate: pooled human trials that raised NAD+ (via NMN/NR) showed no metabolic or mitochondrial benefit. A falling-NAD story that looked causal in mice now looks inert in people.
2026-06-29
Verdict changed Strong support ↓ downgraded Leans support
What tipped the scales: Waskiewicz (2026) — Critical review: carnivore/animal-based diets lack controlled evidence, raise micronutrient and gut-microbiota concerns; not recommended outside experimental use.
What we already knew Probably a modest hint, but the evidence is very thin — only a handful of self-selected people, no real trial.
Why it matters: Downgraded: the IBD-improvement signal is thin — a handful of self-selected people, no controlled trial — and animal studies show ketogenic diets can actually *worsen* colitis. Promising anecdote, not a recommendation.
2026-06-29
Verdict changed Insufficient ↑ upgraded Leans support
What tipped the scales: Lu (2020) — OSK reprogramming restores youthful DNA methylation and reverses vision loss in mice - core evidence for reversible epigenetic-info-loss thesis
What we already knew Probably yes, but only shown in animals so far — in people the drift looks correlational, not proven cause.
Why it matters: Partial reprogramming reverses age markers and restores function in MICE — striking mechanism. In humans the epigenetic drift still looks correlational, so this is a strong animal/mechanism story, not proven cause of human aging.