What has been read, and what it moved

In an open-label phase 1-2 study with no comparison group, 20 people with heart failure (ejection fraction 35 per cent or less) were given patches of engineered heart muscle made from donor stem cells, with drugs to prevent rejection. Three died during the study (one each of vasoplegia, covid-19 and aortic dissection) and one had a heart transplant. Of the 16 given the highest dose, 12 had reached the 3-month interim follow-up: the treated wall was 4.5 mm thicker, the heart's ejection fraction was 3.9 percentage points higher (P = 0.04) and a quality-of-life score 6.7 points higher (P = 0.06). Every patient had at least one adverse event.

Wolfram-Hubertus Zimmermann, Stephan Ensminger, Ingo Kutschka, Christina Paitazoglou, Tim Seidler, Soren Brandenburg, Stefan D. Anker, and 28 others (35 in all), for the BioVAT-HF Investigators. New England Journal of Medicine 394(20), 1991-2001 (2026), published online 2026-05-27, issue dated 2026-05-28; trial registered as NCT04396899. Read it.

Bears on: Heart muscle grown from stem cells is grafted onto a person's failing heart.

What it moved: description on Heart muscle grown from stem cells is grafted onto a person's failing heart, later. A small trial in people, with no comparison group, was under way. It is the first grafted human heart muscle found alive in a person. to later. That small trial in people, with no comparison group, reported in the New England Journal of Medicine on 2026-05-27: of 20 people given the patches, three died during the study and one had a heart transplant; among 12 given the highest dose who had reached the 3-month interim check, the heart's pumping (ejection fraction) was 3.9 percentage points higher and the treated wall 4.5 mm thicker. The 2025 case remains the first grafted human heart muscle found alive in a person., on 17 October 2026: [reading 2026-10-17] The description said the trial was under way; it reported on 2026-05-27. The entry was already at first human on the 2025 case, so the stage does not move, and the new paper is an interim, uncontrolled report, not an approval.; estimate.why on Heart muscle grown from stem cells is grafted onto a person's failing heart, (nothing; a dated note added at the end of the reasoning; no year changed) to [Note 2026-10-17, from the weekly reading of that date: the reasoning above assumes the small trial reports around 2030. It reported an interim analysis on 2026-05-27 (zimmermann-2026): pumping and wall thickness were higher at 3 months in the highest-dose group, with three deaths among 20 people and every patient having at least one adverse event. The years are unchanged; John may want to weigh at the next review whether the reporting year in the reasoning (and so the earliest, 2031) should move earlier, and what the deaths and the need for rejection drugs mean for the latest year.], on 17 October 2026: [reading 2026-10-17] A dated note that the estimate's premise about when the first trial reports has been overtaken; no year changed.

Added under the weekly reading of 2026-10-17. The heart-patch entry was already at first human on the 2025 single-case report in Nature (its stage record unchanged); this paper is a larger human result of the same kind and is the next rung's sort of evidence only in the sense that a regulator would look at it. Because it is an interim, uncontrolled analysis and the next rung is an approval, no stage moves. The paper is not named as the entry's stage record, so it is not marked as having settled a stage.

Not yet verified against a source: Only the abstract was open to read (the full text is behind a payment page); the 6 to 52 months of follow-up that press releases mention is not in the abstract, which reports an interim analysis at 3 months.

The record is data/papers/zimmermann-2026.toml in the public repository.

The authors applied six published blood-protein ageing clocks to stored blood samples from 42 people with a lung-scarring disease (idiopathic pulmonary fibrosis) in an earlier 12-week, placebo-controlled phase 2a trial of the drug rentosertib, and found that all six predicted a lower biological age in the treated groups than in the placebo group. They call the findings exploratory and say the clocks alone cannot tell slower ageing from an improved diseased lung; the strongest effect was in the 30 mg twice-daily group.

Alex Zhavoronkov, Fedor Galkin, Shan Chen, Feng Ren, Alex Aliper, Mikhail Durymanov, Denis Sidorenko, Hui Cui, Jing-Dong J. Han, Hao Xu, Xiaodong Liu, Zuojun Xu, and 7 others, ending with Mahdi Moqri, Alexander Tyshkovskiy and Vadim N. Gladyshev (19 in all). Nature Biotechnology (2026), online 2026-09-07, pages 1 to 13. Read it.

Bears on: A treatment slows a biological age clock in a controlled trial.

What it moved: notes on A treatment slows a biological age clock in a controlled trial, (nothing; text added at the end of the notes) to Added 2026-10-17, from the weekly reading of that date: a 2026-09-07 paper in Nature Biotechnology (zhavoronkov-2026) applied six blood-protein clocks to stored samples from a placebo-controlled phase 2a trial of the lung-scarring drug rentosertib in 42 people, and all six predicted a lower biological age in the treated groups. The authors call the findings exploratory and say the clocks cannot separate slower ageing from an improved lung; the clocks were run on samples already collected, so, like the 2023 CALERIE paper, it was not a clock analysis planned before the data were unblinded, and it does not settle a rung. Statements by Eli Lilly and Novo Nordisk that clocks show people on their weight-loss drugs ageing more slowly (reported by MIT Technology Review on 2026-10-06) are company statements, were not opened, and are not the kind of record the test asks for. No stage or year was changed., on 17 October 2026: [reading 2026-10-17] Adds the rentosertib clock paper and notes why it does not settle a rung, in line with John's ruling of 2026-10-17 on the CALERIE paper.

Added under the weekly reading of 2026-10-17. It is not the record of the stage for the entry on a treatment slowing a biological age clock: the clock analysis was run on samples already collected in a trial that had been published, and the authors call it exploratory, so it was not planned before the data were unblinded, which the entry's test asks for (the same reason John gave on 2026-10-17 for the 2023 CALERIE paper). The first author is the founder and chief executive of the company that makes the drug.

Not yet verified against a source: The paper's full text was read for its abstract, its statements of exploratory use and its discussion; the number of comparisons reaching significance (reported elsewhere as 21 of 54) was not checked in the tables.

The record is data/papers/zhavoronkov-2026.toml in the public repository.

Using national prescription counts from September 2023 to September 2024, the authors found that Wegovy fills rose 136 per cent (from 367,371 in February 2024 to 868,301 in September 2024), and that the March 2024 approval to reduce heart events and Medicare coverage of it were associated with about 592,624 extra fills. Six months after Medicare plans began covering Wegovy, only 32,732 prescriptions had been filled, under 1 per cent of the roughly 3.6 million Medicare beneficiaries estimated to be eligible.

Christopher Scannell, John A. Romley, Rebecca Myerson, Dana Goldman, Dima Mazen Qato. The American Journal of Managed Care 32(7), 392-395 (2026), published 2026-07-09. Read it.

Bears on: A weight-loss drug is approved to prevent heart attacks and strokes in people without diabetes.

What it moved: notes on A weight-loss drug is approved to prevent heart attacks and strokes in people without diabetes, (nothing; text added at the end of the notes) to Added 2026-10-17, from the weekly reading of that date: the first published uptake figure for the heart-prevention purpose is a study in the American Journal of Managed Care (scannell-2026, 2026-07-09), using national prescription counts to September 2024. Wegovy fills rose 136 per cent after the approval of 2024-03-08 and Medicare coverage, yet six months after Medicare plans began covering it only 32,732 prescriptions had been filled, under 1 per cent of an estimated 3.6 million eligible beneficiaries. It is one country and one payer, and the data end in September 2024. It does not change the stage., on 17 October 2026: [reading 2026-10-17] Records the first published uptake figure, which the notes said was still to come.; estimate.why on A weight-loss drug is approved to prevent heart attacks and strokes in people without diabetes, (nothing; a dated note added at the end of the reasoning; no year changed) to [Note 2026-10-17, from the weekly reading of that date: the first published uptake figure for this purpose (scannell-2026) is under 1 per cent of eligible Medicare beneficiaries six months after coverage began, with data to September 2024. That points the other way from the earliest year of 2027 above, which assumed uptake might already count. The years are unchanged; John may want to weigh moving the earliest wide-use year later at the next review.], on 17 October 2026: [reading 2026-10-17] A dated note that the first uptake figure undercuts the premise behind the earliest wide-use year; no year changed.

Added under the weekly reading of 2026-10-17. It is the first published uptake figure for the weight-loss drug's heart-prevention purpose, and it covers one country and one payer, ending in September 2024. It points the other way from the estimate's earliest wide-use year of 2027, which John may want to weigh; no year was changed.

Not yet verified against a source: Only the abstract and body text of the journal's page were read, not the tables; the eligible-population figure is the authors' estimate from other sources.

The record is data/papers/scannell-2026.toml in the public repository.

The registry lists a phase 2 pilot in the United States of an investigational eye drop, 0.4 per cent ZOC2017217, against a drops-with-no-active-ingredient control, twice a day for 24 weeks in about 40 people aged 50 or over with age-related cataract, quadruple-masked and randomised. Its start date is 2026-02-10, the entry shows 8 people enrolled, and its primary completion is estimated for 2027-02; outcomes include best-corrected vision, contrast sensitivity and lens opacity at weeks 4, 12 and 24.

Ocusun Ophthalmic Pharmaceutical (Guangzhou) Co., Ltd. (sponsor). ClinicalTrials.gov registry entry NCT07395986, first posted 2026-02-09, last updated 2026-08-11. Read it.

Bears on: Eye drops dissolve the clouded protein of a cataract.

What it moved: description on Eye drops dissolve the clouded protein of a cataract, repeat the effect, and no human trial has been published. to repeat the effect, and no human trial has been published. A controlled pilot trial of a drop derived from lanosterol (ZOC2017217, from the company Ocusun) in about 40 people has been running in the United States since 2026-02-10, with its first results expected from 2027-02; its makers also cite an unpublished exploratory study in China., on 17 October 2026: [reading 2026-10-17] The description said no human trial had been published, which is still true, but left out the registered trial under way. The stage stays at shown in animals: the first-human step asks for a peer-reviewed controlled report of clearer lenses in people.; estimate.why on Eye drops dissolve the clouded protein of a cataract, (nothing; a dated note added at the end of the reasoning; no year changed) to [Note 2026-10-17, from the weekly reading of that date: the reasoning above rests on there being no registered human trial. That premise no longer holds: a randomised, double-masked, placebo-controlled phase 2 pilot of a lanosterol-derived drop has been registered and running since 2026-02-10 (ocusun-2026), with a first patient announced on 2026-07-14, 8 people enrolled by the registry's last update and a primary completion estimated for 2027-02. The earliest first-human year (2029), set on the premise that a trial would start within three years, was therefore met early. The years are unchanged; whether the first-human years should move, and whether this company's drop answers the replication question about the 2015 dog result, are John's to weigh.], on 17 October 2026: [reading 2026-10-17] A dated note that the estimate's premise of no registered human trial no longer holds; no year changed.

Added under the weekly reading of 2026-10-17. The registry's status was read from the registry's own data; the reading had described the trial as active but not recruiting. It is a registered controlled human trial of a lanosterol-derived drop, which the entry's estimate said did not exist; it has not reported, so the stage stays at shown in animals.

Not yet verified against a source: The registry entry does not name lanosterol; that the drop is derived from it, and the company's exploratory study in China in which 50.0 to 86.4 per cent reached 20/40 vision, are from the sponsor's and its research partner's release of 2026-07-14 (carried by BioSpace and read), which cites no paper.

The record is data/papers/ocusun-2026.toml in the public repository.

The company says its ZEUS trial of ziltivekimab, a monthly antibody against the inflammatory messenger interleukin-6, in over 6,300 people with hardened arteries, chronic kidney disease and raised inflammation, lowered the expected inflammation markers but did not lower heart attacks, strokes or cardiovascular deaths against placebo (hazard ratio 0.99, 95 per cent confidence interval 0.88 to 1.11). Overall rates of side effects were similar, serious infections were more frequent on the drug, and deaths from any cause did not differ. Full results are to be presented at a scientific meeting later in 2026; two further trials in heart failure and after a heart attack are expected to read out in the first half of 2027.

Novo Nordisk A/S. Company announcement, 2026-07-31 (reference PR260731-ZEUS). Read it.

Bears on: A drug that acts only on inflammation is approved to prevent heart attacks.

What it moved: notes on A drug that acts only on inflammation is approved to prevent heart attacks, (nothing; text added at the end of the notes) to Added 2026-10-17, from the weekly reading of that date: on 2026-07-31 the company Novo Nordisk announced that its ZEUS trial of ziltivekimab, an antibody against the inflammatory messenger interleukin-6, in over 6,300 people with hardened arteries, chronic kidney disease and raised inflammation, lowered inflammation markers but not heart attacks, strokes or cardiovascular deaths (hazard ratio 0.99; novo-nordisk-2026). It is the largest test since the 2017 canakinumab trial of a drug acting on inflammation alone, it rests on a company announcement and not yet a peer-reviewed report, and it is a second mark against, beside the 2025 colchicine trial, for the yearly review. No stage or year was changed., on 17 October 2026: [reading 2026-10-17] Adds the ZEUS result beside the 2025 colchicine trial, as the reading proposed.

Added under the weekly reading of 2026-10-17. The largest test since the 2017 canakinumab trial of a drug that acts on inflammation alone, and it found no benefit on heart events; the notes on the anti-inflammatory-drug entry now carry it beside the 2025 colchicine trial. The estimate years were not changed.

Not yet verified against a source: A company announcement of top-line results, not a peer-reviewed report; the full results were not available.

The record is data/papers/novo-nordisk-2026.toml in the public repository.

Two people with a rare inherited disease of very short chromosome ends (a telomere biology disorder) were given their own blood stem cells, treated outside the body with a virus carrying the gene ZSCAN4, which lengthens telomeres, and followed for 24 and 5 months. The telomeres of the treated stem cells were 1.06 to 1.34 times longer in the dish; in one person the average length in a class of white blood cells rose from 3.6 to 6.7 thousand base pairs, and both people's blood counts improved. The authors say no firm conclusions can be drawn from two patients. The abstract reports no rate of cancer.

Kasiani C. Myers, Stella M. Davies, Carolyn Lutzko, Robin Wahle, David D. Grier, Geraldine Aubert, Kevin Norris, and 6 others, the last Minoru S. H. Ko (13 in all). NEJM Evidence 4(3) (2025), published 2025-02-25; trial registered as NCT04211714. Read it.

Bears on: A gene therapy lengthens the ends of chromosomes in an old animal, with no rise in cancer.

What it moved: estimate.why on A gene therapy lengthens the ends of chromosomes in an old animal, with no rise in cancer, (nothing; a dated note added at the end of the reasoning; no year changed) to [Note 2026-10-17, from the weekly reading of that date: the first sentence above is wrong as written. A registered gene-therapy trial in people with a telomere disease did exist, for a different gene, ZSCAN4 (registry entry NCT04211714), and it published two patients in 2025 (myers-2025); the checks listed on this entry also name a 2019 trial of a telomerase gene therapy registered in Colombia. The ZSCAN4 report does not meet this entry's first-human test on a plain reading: the gene is ZSCAN4 and not telomerase, the cells were treated outside the body, there are two patients, and no cancer rate is given. The years are unchanged. Whether the ZSCAN4 report should count for the first-human rung is John's ruling; if it does, the first-human estimate falls away with the move.], on 17 October 2026: [reading 2026-10-17] A dated note that the estimate's first sentence ('no registered trial of a telomerase gene therapy in people existed') is wrong as written, and why the ZSCAN4 report does not move the stage; no year changed.; notes on A gene therapy lengthens the ends of chromosomes in an old animal, with no rise in cancer, (nothing; text added at the end of the notes) to Added 2026-10-17, from the weekly reading of that date: the one human result of lengthened telomeres from a gene therapy is a 2025 report of two people with a rare telomere disease, whose own blood stem cells were treated outside the body with the gene ZSCAN4 (myers-2025); in one the average length in a class of white blood cells rose from 3.6 to 6.7 thousand base pairs. It was not entered as the first-human rung, for the four reasons given in the estimate's dated note. FOR JOHN: whether that report counts is a reading of the test., on 17 October 2026: [reading 2026-10-17] Records the ZSCAN4 report and leaves the first-human question for John.

Added under the weekly reading of 2026-10-17. It does not meet the first-human test of the chromosome-end gene-therapy entry on a plain reading: the gene is ZSCAN4, not telomerase; the cells were treated outside the body; there are two patients; and the abstract gives no cancer rate. Whether it counts is for John; the entry was left at shown in animals.

Not yet verified against a source: Only the abstract and the notes were open to read (the full text is behind a payment page), so the follow-up beyond the abstract, the cancer monitoring and the full author list were not seen.

The record is data/papers/myers-2025.toml in the public repository.

The company reports interim results from the first three people with open-angle glaucoma given its eye-injected reprogramming therapy ER-100 (first dose level, with a daily antibiotic to switch the genes on for 56 days) in an open-label safety trial: no dose-limiting toxicity and no serious adverse events by day 56, and a safety board's recommendation to move to the second dose level. In two of the three people, at least 28 per cent of tested points in the visual field improved by more than 4 decibels at day 56; the company calls this a preliminary signal in a trial designed to measure safety.

Life Biosciences, Inc.. Company press release, Boston, 2026-10-08, with data presented at a meeting of the American Academy of Ophthalmology; trial registered as NCT07290244. Read it.

Bears on: A person receives cell-resetting genes in a registered trial; Partially resetting cells reverses ageing in a living animal.

What it moved: notes on A person receives cell-resetting genes in a registered trial, (nothing; text added at the end of the notes) to Added 2026-10-17, from the weekly reading of that date: on 2026-10-08 the company announced interim results from the first three people with glaucoma (life-biosciences-2026-interim): no serious adverse events by day 56, a safety board's recommendation to move to the second dose level, and a preliminary improvement in the visual field in two of the three. It is a company announcement of an open-label safety trial; this entry counts the dose, which was already on the record, so the stage does not change., on 17 October 2026: [reading 2026-10-17] Notes the interim data of 2026-10-08, which the record of 2026-10-16 did not mention.; description on Partially resetting cells reverses ageing in a living animal, reverse ageing rather than slow it. The first person was given it, in one eye, in June 2026, with no result yet reported; a separate leaf holds that dose. to reverse ageing rather than slow it. The first person was given it, in one eye, in June 2026. On 2026-10-08 the company announced interim results from the first three people with glaucoma: no serious side effects by day 56 and a preliminary improvement in the visual field in two of them. It is a company announcement of an open-label safety trial, not a peer-reviewed report, and it measures sight and not a marker of biological age; a separate entry holds the dose., on 17 October 2026: [reading 2026-10-17] The sentence 'with no result yet reported' was out of date from 2026-10-08. The announcement is a company release of preliminary, uncontrolled data on sight, not the peer-reviewed report of a measured reversal that the first-human step asks for, so the stage stays at shown in animals.

Added under the weekly reading of 2026-10-17. An open-label company announcement of a visual-function change in two of three people is not the peer-reviewed report the cell-resetting-in-animals entry asks for at its first-human rung, and it measures sight, not a marker of biological age; the entry stays at shown in animals. The first-dose record of 2026-06-09 is life-biosciences-2026.

Not yet verified against a source: The release is the sponsor's own and has not been checked against the conference presentation, which was not opened; the visual-field figures are the company's.

The record is data/papers/life-biosciences-2026-interim.toml in the public repository.

Thirty-one adults with scarred fatty liver disease (median age 56) were randomly given either the two-drug course dasatinib plus quercetin, taken three days a week for three weeks and repeated across three seven-week cycles, or a matching placebo. The pre-stated main measure, at least one stage less scarring on paired liver biopsies without the disease worsening, was met by 47 per cent of treated people against 7 per cent on placebo (P = 0.02); the disease resolved in 53 per cent against 7 per cent. Sequencing of single cells from the biopsies showed fewer signs of aged cells and of scar-forming cells in the treated group. Side effects were more frequent on the drugs (82 per cent against 43 per cent) and all passed on their own.

Mijra Koning, Artemiy Kovynev, Marcos F. Fondevila, Eliza J. M. Ruhe, Michiel C. Mommersteeg, Christian Ramakers, Barbara A. Hutten, Bart Verwer, Marije Vlug, Teaco Kuiper, Maarten van den Berg, Anne Vrieze, and 23 others, ending with Tamar Tchkonia, Bernd Schnabl, James L. Kirkland and Abraham S. Meijnikman (35 in all). Nature Metabolism (2026), online 2026-10-01, pages 1 to 12; trial registered as NCT05506488. Read it.

Bears on: A drug that clears aged cells beats a sham treatment in a controlled trial; A drug clears aged cells from a person.

What it moved: description on A drug clears aged cells from a person, human, on a measurement. No such drug is approved anywhere, and the one company drug to reach a controlled trial on a clinical outcome, for knee osteoarthritis, failed in 2020. to human, on a measurement. No such drug is approved anywhere. The first company drug to reach a controlled trial on a clinical outcome, for knee osteoarthritis, failed in 2020; a small academic controlled trial of the same two drugs in scarred fatty liver read out positive on 2026-10-01 (31 people, a biopsy measure), the first result on a clinical-type measure in this class that went the other way., on 17 October 2026: [reading 2026-10-17] The sentence 'the one company drug to reach a controlled trial on a clinical outcome ... failed in 2020' was out of date once the Amsterdam liver trial read out positive on 2026-10-01. The stage is unchanged: the 2019 kidney report stays the record of first human.; estimate.why on A drug clears aged cells from a person, (nothing; a dated note added at the end of the reasoning; no year changed) to [Note 2026-10-17, from the weekly reading of that date: the sentence above that the one such trial to read out so far failed is out of date. A small academic controlled trial of the same two drugs in scarred fatty liver read out positive on 2026-10-01 (koning-2026). The years are unchanged; John may want to weigh it at the next review of the ranges.], on 17 October 2026: [reading 2026-10-17] A dated note that the estimate's premise that the one controlled trial so far failed no longer holds; no year changed.; notes on A drug clears aged cells from a person, (nothing; text added at the end of the notes) to Added 2026-10-17, from the weekly reading of that date: in the Amsterdam liver trial (koning-2026), single-cell sequencing of liver biopsies showed fewer signs of aged cells in the treated group than the placebo group. That is a supporting measurement, not a replacement for the 2019 kidney report, and the stage stays where it was., on 17 October 2026: [reading 2026-10-17] Records the supporting senescence measurement in the new liver trial.; description on A drug that clears aged cells beats a sham treatment in a controlled trial, peer-reviewed journal in 2025. The follow-on trial against the standard drug missed its main goal in 2025 and the company let its staff go, so the result stands but nothing is approved. to peer-reviewed journal in 2025. The follow-on trial against the standard drug missed its main goal in 2025 and the company let its staff go, so the result stands but nothing is approved. A second controlled result came on 2026-10-01: in a phase 2 trial in Amsterdam, 31 people with scarred fatty liver were given a by-mouth course of two drugs that clear aged cells (dasatinib and quercetin) or a placebo, and 47 per cent of those treated, against 7 per cent on placebo, had at least one stage less scarring on biopsy without the disease worsening., on 17 October 2026: [reading 2026-10-17] Adds the Amsterdam liver trial (Nature Metabolism, 2026-10-01), a second controlled result for the class that the description of 2026-10-16 did not know. The stage was already first human, so it does not move.; notes on A drug that clears aged cells beats a sham treatment in a controlled trial, 2022 eye result is the first controlled human result in the drug class, and it is a small trial of one dose. to 2022 eye result was the first controlled human result in the drug class, a small trial of one dose; the 2026 liver result (koning-2026) is a second, with a by-mouth two-drug course in a different organ. It too is small (31 people) and in one condition, run by an academic group, with two of its authors holding patents on drugs of this class., on 17 October 2026: [reading 2026-10-17] Replaces 'the first controlled human result in the drug class' with the two controlled results now on record.; estimate.why on A drug that clears aged cells beats a sham treatment in a controlled trial, (nothing; a dated note added at the end of the reasoning; no year changed) to [Note 2026-10-17, from the weekly reading of that date: the reasoning above counts one positive controlled trial. Two now exist, the 2022 eye trial and the 2026-10-01 liver trial (koning-2026), each small and in a single organ. The years are unchanged; John may want to weigh the second trial at the next review of the ranges.], on 17 October 2026: [reading 2026-10-17] A dated note that the premise 'one positive controlled trial exists' is out of date; the years are John's and were not changed.

Added under the weekly reading of 2026-10-17 (the first reading's flagged items, applied by hand). The page read is the journal's own. The trial is a second controlled result for the class on a clinical-type measure, in the liver; the entry for a drug that clears aged cells beating a sham was already at first human on the 2022 eye trial, so it moves no stage. Two authors (Tchkonia, Kirkland) declare patents on senolytic drugs held by the Mayo Clinic, as the paper's competing-interests statement says. The estimate years were not changed.

Not yet verified against a source: The page was read in full for its abstract, methods and funding; whether the number of authors and the order of the last few match the printed PDF was not checked.

The record is data/papers/koning-2026.toml in the public repository.

The company says the United States regulator has lifted its stop on the late-stage MAGNITUDE trial of its in-body gene editor nexiguran ziclumeran in people with heart disease caused by the protein transthyretin (about 1,200 people, a single infusion against placebo). The stops on this trial and its sister trial in nerve disease were imposed on 2025-10-29 after one person given the drug had a very large rise in liver enzymes and bilirubin; the sister trial's stop was lifted on 2026-01-27. The changes agreed with the regulator are closer liver-test monitoring, guidance on short steroid treatment, exclusion of people with certain liver abnormalities, and, for this trial, exclusion of people with a recent unstable heart or an ejection fraction under 25 per cent.

Intellia Therapeutics, Inc.. Company press release, Cambridge, Massachusetts, 2026-03-02, filed with the United States securities regulator as exhibit 99.2 to a Form 8-K. Read it.

Bears on: A gene is edited inside a person's body to stop an age-related disease.

What it moved: description on A gene is edited inside a person's body to stop an age-related disease, patient's liver was harmed, and restarted in March 2026 with closer liver monitoring; nothing is approved yet. to patient's liver was harmed, and restarted in March 2026 with closer liver monitoring; nothing is approved yet. That patient, a man in his eighties, died on 2025-11-05. Early trade reports linked the death to the liver event; the company's chief executive said in January 2026 that it followed an infection after a bowel perforation and was not caused by liver failure. The cause is not yet confirmed in a written company statement., on 17 October 2026: [reading 2026-10-17] Adds the death in the heart trial and the two accounts of its cause, which the description of 2026-10-16 left out. The company's release of 2026-03-02 (intellia-2026) confirms the stop of 2025-10-29 and the restart; the death and its cause come from trade reports of the company's own remarks. The stage does not change.

Added under the weekly reading of 2026-10-17 to carry the correction to the gene-editing entry's description. The restart date in the entry (March 2026) matches the release. The pages opened for the death were the trade press, not the company.

Not yet verified against a source: The release does not mention the patient's death. That a patient in the trial died on 2025-11-05 is from the trade press (CGTLive, 2025-11-09, citing the company); that the death was put down to sepsis after a bowel perforation and not to liver failure is the chief executive's account at a January 2026 investor conference as reported by Clinical Trials Arena (2026-01-15). Neither the company's written statement of the cause nor the trial registry's note of it was found.

The record is data/papers/intellia-2026.toml in the public repository.

The company says it has renewed its agreement with the Ace Alzheimer Center in Barcelona to keep running its plasma-exchange-with-albumin programme (AMBAR) there, which has operated since 2021; that AMBAR clinic centres are opening in hospitals in Madrid, Mallorca and Alicante; that more than 5,000 plasma exchanges with albumin have been done since the programme began; and that the protocol is included in the medical guidelines of the American Society for Apheresis. A registry with real-world data is planned. The release describes the 2020 trial as having slowed progression by 61 per cent in some patients, a figure that is the company's own wording.

Grifols, S.A.. Company press release, Barcelona, 2026-03-12. Read it.

Bears on: Replacing the liquid part of a person's blood slows an age-related disease.

What it moved: notes on Replacing the liquid part of a person's blood slows an age-related disease, (nothing; text added at the end of the notes) to FOR JOHN, added 2026-10-17 from the weekly reading of that date: the case for the clinic rung. The test accepts a named clinical setting that gives the treatment for that purpose in routine care and publishes that practice. On 2026-03-12 the supplier Grifols announced (grifols-2026) that the Ace Alzheimer Center in Barcelona has run its plasma-exchange-with-albumin programme (AMBAR) for people with Alzheimer's disease since 2021, that new clinics are opening in hospitals in Madrid, Mallorca and Alicante, that more than 5,000 exchanges have been done, and that the protocol is in the guidelines of the American Society for Apheresis. On a plain reading that could be the clinic rung from 2021 or 2026-03-12. It was not moved because the source is the supplier's release and not the clinical setting's own publication of its practice, the apheresis society's category for Alzheimer's disease was not seen, and a stage never moves down, so a mistake here could not be undone. Whether a company-run clinic programme counts as routine care, and whether the clinic rung's test should be widened to say so, is John's call. If it counts, the entry moves to in clinical use and its estimate (earliest 2030) must be rewritten. This entry stays at first human., on 17 October 2026: [reading 2026-10-17] Writes the case for the clinic rung into the notes with its source and leaves it for John, as ruled; the stage did not move.

Added under the weekly reading of 2026-10-17. The entry's clinic-rung test accepts a named clinical setting that gives the treatment in routine care and publishes that practice; this release may be that for a named centre in Barcelona, but it is the supplier's account. The entry stays at first human; the case for the clinic rung is written into its notes for John.

Not yet verified against a source: A company's release, not the clinical setting's own publication of its practice; the apheresis society's guideline was not opened, so the category it gives for Alzheimer's disease is not known.

The record is data/papers/grifols-2026.toml in the public repository.

In a double-blind, placebo-controlled phase 2a trial, 38 older people having a hip or knee replacement were randomly given four infusions of a plasma protein fraction from young donors (GRF6021) or salt water, around the day of surgery; 36 finished. Blood tests showed that the immune and inflammatory response to the surgery differed between the groups. The authors call it a first proof of principle in humans that a young plasma fraction regulates immune and inflammatory responses in older people; it was not designed to show a clinical benefit.

Brice Gaudilliere, Lei Xue, Amy S. Tsai, Xiaoxiao Gao, Tiffany N. McAllister, Martha Tingle, Gladys Porras, Igor Feinstein, Dorien Feyaerts, Franck Verdonk, Maximilian Sabayev, Julien Hedou, and 13 others, ending with Stuart B. Goodman, Nima Aghaeepour and Martin S. Angst (25 in all). Journal of Translational Medicine 23, article 183 (2025), published 2025-02-14; trial registered as NCT03981419. Read it.

Bears on: Young blood rejuvenates an old animal.

What it moved: description on Young blood rejuvenates an old animal, caregivers rated daily function better, an unconfirmed result), and no controlled trial has found a benefit. to caregivers rated daily function better, an unconfirmed result), and no controlled trial has found a benefit on an age-related function. A 2025 Stanford trial in 38 older people having a joint replacement found that four infusions of a young-donor plasma fraction changed their immune and inflammatory response to the surgery, which is not a benefit to an age-related function., on 17 October 2026: [reading 2026-10-17] The description's 'no controlled trial has found a benefit' was true but narrower than it looked; the nearest controlled human result is now named. The stage stays at shown in animals.; notes on Young blood rejuvenates an old animal, (nothing; text added at the end of the notes) to Added 2026-10-17, from the weekly reading of that date: the nearest controlled human result is gaudilliere-2025, a double-blind placebo-controlled phase 2a trial (NCT03981419) of a young-donor plasma fraction in 38 older patients having a hip or knee replacement, which the authors call a first proof of principle that such a fraction regulates immune and inflammatory responses in older people. Its measure is the response to surgical injury and not an age-related function, so the first-human test is not met on a plain reading. Whether such a measure could count is John's ruling. No stage or year was changed., on 17 October 2026: [reading 2026-10-17] Adds the nearest controlled human trial to the notes and leaves the question of whether it counts for John.

Added under the weekly reading of 2026-10-17. It is the nearest controlled human result to the young-blood entry's first-human test, but its measure is the response to surgical injury, not an age-related function, so the entry stays at shown in animals; John decides whether such a measure could count.

Not yet verified against a source: The main clinical measures (fatigue, pain, joint function over six weeks) were listed but their results were not read in the tables.

The record is data/papers/gaudilliere-2025.toml in the public repository.

The two companies say the United States regulator has approved an under-the-skin weekly autoinjector form of lecanemab as the starting dose as well as the maintenance dose (500 mg a week as two injections to start; 360 mg a week after 18 months), so that a whole course can be taken at home; a United States launch as a starting dose was planned for late August 2026. The approval rests on exposure matching the drip form in a long-term extension of the main trial, with similar rates of brain swelling and bleeding and some injection-site reactions. The release gives no figure for how many people have been treated.

Eisai Co., Ltd., Biogen Inc.. Company press release, Tokyo and Cambridge, Massachusetts, 2026-07-13, as carried by BioSpace. Read it.

Bears on: An antibody clears amyloid from the brain and slows the decline of Alzheimer's disease.

What it moved: description on An antibody clears amyloid from the brain and slows the decline of Alzheimer's disease, The effect is modest, the drug needs infusions and brain scans, and some patients bleed or swell in the brain; so it is in clinical use, not wide use. to The effect is modest, the drug needs frequent brain scans, and some patients bleed or swell in the brain; so it is in clinical use, not wide use. It has needed infusions, but on 2026-07-13 the United States regulator approved an under-the-skin weekly autoinjector as the starting dose too, so a whole course can now be given at home., on 17 October 2026: [reading 2026-10-17] 'The drug needs infusions' stopped being true on 2026-07-13, when the companies announced the United States approval of the autoinjector as a starting dose. Scans, price and the lack of uptake figures stand, so the stage stays at clinical use.; estimate.why on An antibody clears amyloid from the brain and slows the decline of Alzheimer's disease, (nothing; a dated note added at the end of the reasoning; no year changed) to [Note 2026-10-17, from the weekly reading of that date: the first sentence above names infusions as a bar to wide use and says the under-the-skin form was approved in 2025 for maintenance only. Since 2026-07-13 it is also approved as a starting dose (eisai-2026, the companies' own release), so that bar no longer holds. The frequent brain scans, the price and the absence of any published uptake figure stand. The years are unchanged; John may want to weigh whether the earliest and likely wide-use years (2028, 2035) should move at the next review.], on 17 October 2026: [reading 2026-10-17] A dated note that one of the estimate's stated bars to wide use has been removed; no year changed.

Added under the weekly reading of 2026-10-17. It removes the entry's stated condition that the drug needs infusions, but not the brain scans, the price or the missing uptake figures, so the entry stays at clinical use. Whether it moves the estimate's earliest or likely wide-use year is for John.

Not yet verified against a source: The regulator's own notice of the approval was not found or opened; the date and the regimen are the companies'.

The record is data/papers/eisai-2026.toml in the public repository.

Twenty-five overweight or obese post-menopausal women with prediabetes took either 250 mg a day of nicotinamide mononucleotide (NMN, a building block of the cell's energy-carrying molecule NAD) or a placebo for ten weeks in a double-blind trial. Muscle insulin sensitivity, measured directly with an insulin clamp, rose by 25 percent in the NMN group and not at all on placebo; blood sugar, blood pressure, blood fats, liver insulin sensitivity and liver fat did not change. The first controlled trial to show an NAD precursor doing anything measurable in people, on one tissue and one measurement.

Mihoko Yoshino, Jun Yoshino, Brandon D. Kayser, Gary J. Patti, Michael P. Franczyk, Kathryn F. Mills, Miriam Sindelar, Terri Pietka, Bruce W. Patterson, Shin-Ichiro Imai, Samuel Klein. Science 372(6547), 1224–1229 (2021). Read it.

Bears on: A supplement that raises a cell's NAD improves an age-related function in a controlled trial.

What it settled: the rung "first human" of A supplement that raises a cell's NAD improves an age-related function in a controlled trial.

Intended leaf: an NAD precursor with a measured effect in people, under energy and fuel sensing. Thirteen women were randomised to NMN and twelve to placebo. Online 2021-04-22; the print issue is dated 2021-06-11. Registered as NCT03151239.

The record is data/papers/yoshino-2021.toml in the public repository.

Adult male cynomolgus monkeys were given metformin daily for 40 months and compared with untreated monkeys of the same and younger ages on several biological age clocks built from gene activity, DNA marks, blood proteins and metabolites across many tissues. The clocks ran slower in the treated animals, with the brain reading about six years younger; the authors tie the brain effect partly to the protective Nrf2 pathway. A controlled animal study on measurements, not on lifespan or disease, in a small group.

Yuanhan Yang, Xiaoyong Lu, Ning Liu, Shuai Ma, Hui Zhang, Zhiyi Zhang, Kuan Yang, Mengmeng Jiang, Zikai Zheng, Yicheng Qiao, Qinchao Hu, Ying Huang, Yiyuan Zhang, Muzhao Xiong, Lixiao Liu, Xiaoyu Jiang, Pradeep Reddy, Xueda Dong, Fanshu Xu, Qiaoran Wang, Qian Zhao, Jinghui Lei, Shuhui Sun, Ying Jing, Jingyi Li, Yusheng Cai, Yanling Fan, Kaowen Yan, Yaobin Jing, Amin Haghani, Mengen Xing, Xuan Zhang, Guodong Zhu, Weihong Song, Steve Horvath, Concepcion Rodriguez Esteban, Moshi Song, Si Wang, Guoguang Zhao, Wei Li, Juan Carlos Izpisua Belmonte, Jing Qu, Weiqi Zhang, Guang-Hui Liu. Cell 187(22), 6358–6378.e29 (2024). Read it.

Bears on: Metformin delays age-related disease in people who do not have diabetes.

What it moved: nothing. It was read against the treatments it bears on and no stage or estimate changed. Logging it anyway is how the record shows what has been considered.

Intended leaves: a drug makes people live longer in good health (metformin), under energy and fuel sensing; and a clock validated as a trial measure, under measuring ageing. Online 2024-09-12; the print issue is dated 2024-10-31.

Not yet verified against a source: The group sizes (12 treated elderly males, 16 elderly untreated, 18 young or middle-aged) are from news coverage, not the paper's own text, which was behind the publisher's access wall; the dose of 20 milligrams per kilogram a day is supported by a published commentary but not checked in the paper itself.

The record is data/papers/yang-2024.toml in the public repository.

The CALERIE trial had randomised 220 healthy adults without obesity to cut their calories by a quarter or eat as usual for two years; those assigned to restriction managed 11.9 percent on average. In the 197 with blood samples, a clock that reads the pace of ageing from DNA methylation (DunedinPACE) ran 2 to 3 percent slower in the restricted group at one and two years, while two clocks that read accumulated age (PhenoAge and GrimAge) did not move. The first randomised trial in people to show an intervention slowing a biological age clock, by an amount the authors call small.

R. Waziry, C. P. Ryan, D. L. Corcoran, K. M. Huffman, M. S. Kobor, M. Kothari, G. H. Graf, V. B. Kraus, W. E. Kraus, D. T. S. Lin, C. F. Pieper, M. E. Ramaker, M. Bhapkar, S. K. Das, L. Ferrucci, W. J. Hastings, M. Kebbe, D. C. Parker, S. B. Racette, I. Shalev, B. Schilling, D. W. Belsky. Nature Aging 3(3), 248–257 (2023). Read it.

Bears on: A treatment slows a biological age clock in a controlled trial.

What it moved: nothing. It was read against the treatments it bears on and no stage or estimate changed. Logging it anyway is how the record shows what has been considered.

Intended leaves: a clock validated to predict outcomes and used as a trial measure, under measuring ageing; and the fuel-sensing branch, since calorie restriction is the intervention the rapamycin and metformin leaves imitate. A correction appeared on 2023-05-09 (Nature Aging 3(6), 753) and was not read here. Online 2023-02-09; the print issue is March 2023. Ruled by John on 2026-10-17: this paper does not settle the rung of the clock-in-a-controlled-trial leaf, because it calls its own clock analysis post hoc and the leaf's test asks for one planned in advance; settled_stage stays false.

The record is data/papers/waziry-2023.toml in the public repository.

In a phase 2 trial, 183 people with moderate to severe knee osteoarthritis pain had a single injection into the knee of placebo or one of three doses of UBX0101, a drug meant to kill worn-out cells in the joint. At twelve weeks there was no statistically significant difference in pain between any dose and placebo (p-values 0.52, 0.81 and 0.99 against placebo); the drug was well tolerated. The company said it would not take the drug into pivotal trials and would narrow its work to eye and nerve diseases. The first controlled trial of a cell-clearing drug on a clinical outcome, and it failed.

UNITY Biotechnology, Inc.. Company press release, South San Francisco, 2020-08-17, filed the same day as exhibit 99.1 to a Form 8-K with the US Securities and Exchange Commission. Read it.

Bears on: A drug that clears aged cells relieves knee osteoarthritis.

What it moved: nothing. It was read against the treatments it bears on and no stage or estimate changed. Logging it anyway is how the record shows what has been considered.

Intended leaf: a drug clears aged cells from a person, under worn-out cells; the example leaf's description and verify list already refer to this failure, so the master can clear that verify line against this record. The release's own page on unitybiotechnology.com and the wire copy on GlobeNewswire are no longer served; the wire address was https://www.globenewswire.com/news-release/2020/08/17/2079116/0/en/. Doses 0.5, 2.0 and 4.0 mg, 45 to 46 people per arm; the primary outcome was change in the WOMAC-A pain score at week 12.

The record is data/papers/unity-2020.toml in the public repository.

Eighteen people aged 50 to 90 with mild to moderate Alzheimer's disease were given four weekly infusions of plasma from male donors aged 18 to 30 at Stanford: nine in a double-blind crossover against salt water, and nine more in an open-label plasma-only group. The infusions were safe and well tolerated, with no related serious adverse events. Cognition, measured only as an exploratory outcome, did not change; caregivers rated day-to-day function better, which the authors treated as unconfirmed. A safety trial of the young-blood idea in people, not a test of whether it works.

Sharon J. Sha, Gayle K. Deutsch, Lu Tian, Kara Richardson, Maria Coburn, Jennifer L. Gaudioso, Tatiana Marcal, Ethan Solomon, Athanasia Boumis, Anthony Bet, Maarten Mennes, Erik van Oort, Christian F. Beckmann, Steven P. Braithwaite, Sam Jackson, Karoly Nikolich, Darby Stephens, Geoffrey A. Kerchner, Tony Wyss-Coray. JAMA Neurology 76(1), 35–40 (2019). Read it.

What it moved: nothing. It was read against the treatments it bears on and no stage or estimate changed. Logging it anyway is how the record shows what has been considered.

Intended leaf: plasma exchange or dilution with a measured effect in people, under inflammation and the body's messaging. The trial had no primary measure of cognition, so "no effect" should not be read as a failed efficacy trial. Online 2018-10-29; the print issue is dated 2019-01-01, which is why the id carries 2019. Registered as NCT02256306.

The record is data/papers/sha-2019.toml in the public repository.

Using national death records from 1990 to 2019 for the eight longest-lived countries plus Hong Kong and the United States, the authors show that gains in life expectancy have slowed since 1990: life expectancy at birth rose 6.5 years on average over the period, and in every population but Hong Kong and South Korea the yearly gain has fallen below 0.2 years. They conclude that survival to 100 is unlikely to exceed 15 percent for women and 5 percent for men, and that radical life extension (defined as three years gained per decade, the twentieth-century pace) is implausible this century unless biological ageing itself is markedly slowed. The demographers' case against the headline, and the baseline the headline has to beat.

S. Jay Olshansky, Bradley J. Willcox, Lloyd Demetrius, Hiram Beltrán-Sánchez. Nature Aging 4(11), 1635–1642 (2024). Read it.

What it moved: nothing. It was read against the treatments it bears on and no stage or estimate changed. Logging it anyway is how the record shows what has been considered.

Bears on the headline question itself rather than any one branch, so listed against all eight. Its own escape clause, that the trend holds unless ageing is slowed, is exactly what the branches track. Online 2024-10-07; the print issue is November 2024.

The record is data/papers/olshansky-2024.toml in the public repository.

Mice were given the four Yamanaka genes that turn an adult cell back into an embryonic-like stem cell, but switched on for only two days in every seven so that cells reset part of the way without losing their identity. In mice bred to age fast (the LAKI progeria model) the cycling treatment reduced the signs of ageing in several tissues and lengthened life; in normally aged mice it improved the repair of muscle and pancreas after injury. The first demonstration that partial reprogramming can be done in a living animal without killing it with tumours.

Alejandro Ocampo, Pradeep Reddy, Paloma Martinez-Redondo, Aida Platero-Luengo, Fumiyuki Hatanaka, Tomoaki Hishida, Mo Li, David Lam, Masakazu Kurita, Ergin Beyret, Toshikazu Araoka, Eric Vazquez-Ferrer, David Donoso, Jose Luis Roman, Jinna Xu, Concepcion Rodriguez Esteban, Gabriel Nuñez, Estrella Nuñez Delicado, Josep M. Campistol, Isabel Guillen, Pedro Guillen, Juan Carlos Izpisua Belmonte. Cell 167(7), 1719–1733.e12 (2016). Read it.

Bears on: Partially resetting cells reverses ageing in a living animal.

What it settled: the rung "shown in animals" of Partially resetting cells reverses ageing in a living animal.

Intended leaf: partially resetting cells reverses ageing in a living animal, under the cell's settings; one of the five leaves shared with The Upgrade Tree. The paper itself gives no percentage for the lifespan gain, only "a dramatic increase in median and maximal lifespan" and a survival curve; the widely quoted figure of about 30 percent longer comes from the Salk Institute's news release of 2016-12-15, not from the paper, so the record does not carry it.

The record is data/papers/ocampo-2016.toml in the public repository.

A position paper from a consortium of ageing researchers setting out what a measure of biological age would have to do before a trial could use it in place of waiting for disease and death: be practical to take, track age, predict function and death better than a birth date does, rest on a mechanism, and move when an intervention works, then pass laboratory and clinical validation. It states that no such measure has yet been validated as a stand-in outcome for a trial. The field's own account of how far the measuring branch has to go.

Mahdi Moqri, Chiara Herzog, Jesse R. Poganik, Biomarkers of Aging Consortium, Jamie Justice, Daniel W. Belsky, Albert Higgins-Chen, Alexey Moskalev, Georg Fuellen, Alan A. Cohen, Ivan Bautmans, Martin Widschwendter, Jingzhong Ding, Alexander Fleming, Joan Mannick, Jing-Dong Jackie Han, Alex Zhavoronkov, Nir Barzilai, Matt Kaeberlein, Steven Cummings, Brian K. Kennedy, Luigi Ferrucci, Steve Horvath, Eric Verdin, Andrea B. Maier, Michael P. Snyder, Vittorio Sebastiano, Vadim N. Gladyshev. Cell 186(18), 3758–3775 (2023). Read it.

Bears on: A regulator accepts a marker of ageing as proof that a treatment slowed it.

What it moved: nothing. It was read against the treatments it bears on and no stage or estimate changed. Logging it anyway is how the record shows what has been considered.

Intended leaf: a regulator accepts a marker of ageing as a trial endpoint, under measuring ageing; this paper is the criteria such a marker must meet, not an acceptance. Cell prints the consortium fourth in the author list, as given here; some indexes move it first or drop it.

The record is data/papers/moqri-2023.toml in the public repository.

The PEARL trial gave healthy adults aged 50 to 85 a weekly 5 mg or 10 mg dose of compounded rapamycin, or placebo, for 48 weeks in a double-blind trial run remotely; 114 completed (40, 35 and 39 in the three groups) and 11 dropped out. The primary outcome, deep abdominal fat on a body scan, did not change. Women on the higher dose gained lean mass and reported less pain, and the lower-dose group reported better emotional well-being and general health; serious adverse events were 1, 2 and 3 in the three groups. Mid-trial the team found the compounded drug reached about a third of the blood level of the commercial product, so the real doses were about two-thirds lower than intended. The first year-long placebo-controlled trial of rapamycin in healthy people, run by the company that sells it but paid for mainly by a crowdfunding appeal, and it missed its primary outcome.

Mauricio Moel, Girish Harinath, Virginia Lee, Andy Nyquist, Stefanie L. Morgan, Anar Isman, Sajad Zalzala. Aging (Albany NY) 17(4), 908–936 (2025). Read it.

Bears on: A rapamycin-type drug improves an age-related measure in older people.

What it moved: nothing. It was read against the treatments it bears on and no stage or estimate changed. Logging it anyway is how the record shows what has been considered.

Intended leaf: a drug makes people live longer in good health (rapamycin), under energy and fuel sensing; a first human on a measurement, not a healthy-years result. All seven authors are employees and shareholders of AgelessRx. Registered as NCT04488601 on 2020-07-28. The paper gives no single randomised total in its text; 125 is the completers plus the dropouts.

The record is data/papers/moel-2025.toml in the public repository.

Half of the blood plasma of old mice was replaced with salt water containing 5 percent albumin, with no young blood involved. A single exchange matched or exceeded what young blood had done in earlier work: better muscle repair, less fat and scarring in the liver, and more new neurons in the memory centre of the brain. In a small human arm, serum taken after one ordinary therapeutic plasma exchange no longer suppressed the growth of progenitor cells the way old serum does. The authors' reading is that diluting what accumulates in old blood matters more than adding anything young; the abstract gives no percentages.

Melod Mehdipour, Colin Skinner, Nathan Wong, Michael Lieb, Chao Liu, Jessy Etienne, Cameron Kato, Dobri Kiprov, Michael J. Conboy, Irina M. Conboy. Aging (Albany NY) 12(10), 8790–8819 (2020). Read it.

Bears on: Replacing the liquid part of a person's blood slows an age-related disease.

What it settled: the rung "shown in animals" of Replacing the liquid part of a person's blood slows an age-related disease.

Intended leaf: plasma exchange or dilution with a measured effect in people, under inflammation and the body's messaging. The animal founding paper is conboy-2005 from the same group; the young-plasma trial in Alzheimer's is sha-2019.

The record is data/papers/mehdipour-2020.toml in the public repository.

Thirty healthy adults aged 55 to 79 were enrolled, and 24 completed, a crossover in which each took 1,000 mg a day of nicotinamide riboside (a form of vitamin B3 that the body turns into NAD) and a placebo for six weeks each. Levels of NAD in white blood cells rose by about 60 percent on the supplement; blood pressure trended lower by about 4 points but not significantly after correction for multiple comparisons; no serious side effects. The first trial to show the supplement raises NAD in older people, with no effect on any health outcome tested.

Christopher R. Martens, Blair A. Denman, Melissa R. Mazzo, Michael L. Armstrong, Nichole Reisdorph, Matthew B. McQueen, Michel Chonchol, Douglas R. Seals. Nature Communications 9, 1286 (2018). Read it.

Bears on: A supplement that raises a cell's NAD improves an age-related function in a controlled trial.

What it moved: nothing. It was read against the treatments it bears on and no stage or estimate changed. Logging it anyway is how the record shows what has been considered.

Intended leaf: an NAD precursor with a measured effect in people, under energy and fuel sensing. The NAD rise was measured in blood mononuclear cells; the paper says the molecules are not detectable in plasma, so "blood NAD" would overstate it.

The record is data/papers/martens-2018.toml in the public repository.

Volunteers aged 65 and over took RAD001 (everolimus, a relative of rapamycin that damps the cell's main fuel-sensing pathway) or a placebo for six weeks, then had a flu vaccine. At doses that were reasonably well tolerated, the antibody response to the vaccine was about 20 percent stronger, and fewer of their T cells carried the exhaustion marker PD-1. The first controlled trial in people to show that a drug acting on the ageing pathways found in animals improves an age-related function, here the immune system's.

Joan B. Mannick, Giuseppe Del Giudice, Maria Lattanzi, Nicholas M. Valiante, Jens Praestgaard, Baisong Huang, Michael A. Lonetto, Holden T. Maecker, John Kovarik, Simon Carson, David J. Glass, Lloyd B. Klickstein. Science Translational Medicine 6(268), 268ra179 (2014). Read it.

Bears on: A rapamycin-type drug improves an age-related measure in older people.

What it settled: the rung "first human" of A rapamycin-type drug improves an age-related measure in older people.

Intended leaves: immune rejuvenation with a measured effect in people, under inflammation and the body's messaging; and a drug makes people live longer in good health (rapamycin and relatives), under energy and fuel sensing. The same group's 2018 follow-up (not logged here) reported fewer infections in older people on a related drug.

The record is data/papers/mannick-2014.toml in the public repository.

A 77-year-old Japanese woman with a wet form of age-related macular degeneration had skin cells turned back into stem cells, grown into a sheet of the retina's supporting layer 1.3 by 3 millimetres across, and transplanted under her retina on 2014-09-12. A year later the sheet was intact, her sight was unchanged rather than worse, and there was no rejection and no serious complication without any immune-suppressing drugs. The first person to receive tissue made from their own reprogrammed cells; a second patient was not treated because of DNA changes found in his cells.

Michiko Mandai, Akira Watanabe, Yasuo Kurimoto, Yasuhiko Hirami, Chikako Morinaga, Takashi Daimon, Masashi Fujihara, Hiroshi Akimaru, Noriko Sakai, Yumiko Shibata, Motoki Terada, Yui Nomiya, Shigeki Tanishima, Masahiro Nakamura, Hiroyuki Kamao, Sunao Sugita, Akishi Onishi, Tomoko Ito, Kanako Fujita, Shin Kawamata, Masahiro J. Go, Chikara Shinohara, Ken-ichiro Hata, Masanori Sawada, Midori Yamamoto, Sachiko Ohta, Yasuo Ohara, Kenichi Yoshida, Junko Kuwahara, Yuko Kitano, Naoki Amano, Masafumi Umekage, Fumiyo Kitaoka, Azusa Tanaka, Chihiro Okada, Naoko Takasu, Seishi Ogawa, Shinya Yamanaka, Masayo Takahashi. New England Journal of Medicine 376(11), 1038–1046 (2017). Read it.

Bears on: Cells grown from stem cells restore a lost organ function in people.

What it moved: nothing. It was read against the treatments it bears on and no stage or estimate changed. Logging it anyway is how the record shows what has been considered.

Intended leaf: a stem cell therapy restores a lost function in a controlled trial, under replacing lost cells and parts; this is a first-human safety record, not a restored function or a controlled trial. Thirty-nine authors; Mandai, Watanabe and Kurimoto contributed equally. Registered as UMIN000011929.

The record is data/papers/mandai-2017.toml in the public repository.

Three of the four Yamanaka reprogramming genes (Oct4, Sox2 and Klf4, leaving out the cancer-linked c-Myc) were delivered by a gene-therapy virus into the eyes of mice and switched on with an antibiotic. The treatment made crushed optic nerves regrow, restored sight lost to a glaucoma model and to ordinary ageing in twelve-month-old mice, and turned back a DNA-methylation measure of the cells' age; the effect needed two enzymes that strip methylation marks, and switching the genes on for up to fifteen months caused no tumours. The animal result behind the first reprogramming therapy to be dosed in a person (life-biosciences-2026).

Yuancheng Lu, Benedikt Brommer, Xiao Tian, Anitha Krishnan, Margarita Meer, Chen Wang, Daniel L. Vera, Qiurui Zeng, Doudou Yu, Michael S. Bonkowski, Jae-Hyun Yang, Songlin Zhou, Emma M. Hoffmann, Margarete M. Karg, Michael B. Schultz, Alice E. Kane, Noah Davidsohn, Ekaterina Korobkina, Karolina Chwalek, Luis A. Rajman, George M. Church, Konrad Hochedlinger, Vadim N. Gladyshev, Steve Horvath, Morgan E. Levine, Meredith S. Gregory-Ksander, Bruce R. Ksander, Zhigang He, David A. Sinclair. Nature 588(7836), 124–129 (2020). Read it.

Bears on: Partially resetting cells reverses ageing in a living animal; A person receives cell-resetting genes in a registered trial.

What it moved: nothing. It was read against the treatments it bears on and no stage or estimate changed. Logging it anyway is how the record shows what has been considered.

Intended leaves: partially resetting cells reverses ageing in a living animal (the cell's settings; shared with The Upgrade Tree), and a first person dosed with a reprogramming therapy, whose company trial uses the same three genes. The paper gives the age reversal as a direction of change against controls in its figures, not as a single number of years. Online 2020-12-02; the print issue is dated 2020-12-03.

The record is data/papers/lu-2020.toml in the public repository.

The same five authors' update ten years on, raising the count from nine hallmarks to twelve by adding the failure of the cell's recycling of its own worn parts (disabled macroautophagy), low-grade inflammation that never switches off, and an unbalanced community of gut microbes (dysbiosis). It restates the three tests a hallmark must pass: it shows up with age, making it worse speeds ageing, and treating it slows, stops or reverses ageing.

Carlos López-Otín, Maria A. Blasco, Linda Partridge, Manuel Serrano, Guido Kroemer. Cell 186(2), 243–278 (2023). Read it.

What it moved: nothing. It was read against the treatments it bears on and no stage or estimate changed. Logging it anyway is how the record shows what has been considered.

Bears on every branch; the twelve hallmarks named here are the ones data/branches.toml groups, and the validator refuses a grouping that drops or doubles one. Online 2023-01-03; the print issue is dated 2023-01-19.

The record is data/papers/lopez-otin-2023.toml in the public repository.

A review that proposed nine "hallmarks" of ageing, the kinds of damage and malfunction that build up with age in most animals: unstable DNA, shortening chromosome ends, drifting gene settings, misfolded proteins, mis-set fuel sensing, failing power plants in the cell, worn-out cells that will not die, exhausted stem cells, and garbled signalling between cells. It argued that each should get worse with age, speed ageing when made worse, and slow ageing when treated. The starting map for the field, and for this site's branches.

Carlos López-Otín, Maria A. Blasco, Linda Partridge, Manuel Serrano, Guido Kroemer. Cell 153(6), 1194–1217 (2013). Read it.

What it moved: nothing. It was read against the treatments it bears on and no stage or estimate changed. Logging it anyway is how the record shows what has been considered.

Bears on every branch: the eight branches group these hallmarks and the three added in 2023 (lopez-otin-2023). The mapping from each hallmark to a branch is in docs/record-schema.md. Cell prints the title in capitals as given; the issue date is 2013-06-06 and no earlier online date is recorded.

The record is data/papers/lopez-otin-2013.toml in the public repository.

The SELECT trial randomised 17,604 adults aged 45 and over with existing heart or blood-vessel disease and a body mass index of 27 or more, but no diabetes, to a weekly injection of semaglutide 2.4 mg or placebo, and followed them for a mean of 39.8 months. Heart attack, stroke or death from cardiovascular causes came to 6.5 percent on the drug against 8.0 percent on placebo, a hazard ratio of 0.80 (95 percent confidence interval 0.72 to 0.90). The first large controlled trial to show that a GLP-1 drug prevents hard outcomes in people without diabetes; it was not designed to measure ageing, and 16.6 percent stopped the drug for side effects against 8.2 percent on placebo.

A. Michael Lincoff, Kirstine Brown-Frandsen, Helen M. Colhoun, John Deanfield, Scott S. Emerson, Sille Esbjerg, Soren Hardt-Lindberg, G. Kees Hovingh, Steven E. Kahn, Robert F. Kushner, Ildiko Lingvay, Tugce K. Oral, Marie M. Michelsen, Jorge Plutzky, Christoffer W. Tornøe, Donna H. Ryan, for the SELECT Trial Investigators. New England Journal of Medicine 389(24), 2221–2232 (2023). Read it.

Bears on: A weight-loss drug is approved to prevent heart attacks and strokes in people without diabetes.

What it moved: nothing. It was read against the treatments it bears on and no stage or estimate changed. Logging it anyway is how the record shows what has been considered.

Intended leaf: a GLP-1 drug with an ageing endpoint, under energy and fuel sensing; this is the result that made such a trial thinkable, not that trial. Sixteen named authors write for the SELECT Trial Investigators, who are listed as a group. Funded by Novo Nordisk; registered as NCT03574597. Online 2023-11-11; print issue 2023-12-14.

The record is data/papers/lincoff-2023.toml in the public repository.

The company says a first participant has been given ER-100, a one-time injection into one eye of a gene therapy that switches on three of the four Yamanaka reprogramming genes (OCT4, SOX2 and KLF4) for 56 days under the control of an oral antibiotic, in a phase 1 safety trial for open-angle glaucoma and a sudden loss of blood supply to the optic nerve. The registry entry plans about 18 adults in sequential dose groups and measures safety to days 56 and 112. This is a first person dosed, not a result: the release reports no measurement of any kind.

Life Biosciences, Inc.. Company press release, Boston, 2026-06-09; trial registered as NCT07290244. Read it.

Bears on: A person receives cell-resetting genes in a registered trial.

What it settled: the rung "first human" of A person receives cell-resetting genes in a registered trial.

Intended leaf: a first person dosed with a reprogramming therapy (the cell's settings branch). The animal result behind it is the 2020 optic nerve paper, lu-2020. The company's earlier release of 2026-01-28 announced the FDA's clearance to start the trial and called ER-100 the first reprogramming therapy cleared for human trials. The registry entry (last updated 2026-05-19) says the trial started on 2026-03-02 and gives the doses; the release itself gives no site, dose or participant count.

The record is data/papers/life-biosciences-2026.toml in the public repository.

The paper that set out the Dog Aging Project: the case that pet dogs, which age ten times faster than people, share our homes and get our diseases and our veterinary care, are the right animal for testing whether a drug slows ageing within a human researcher's working life, and the plan for a long-term study of tens of thousands of dogs with a rapamycin trial inside it. A rationale paper with no results.

Matt Kaeberlein, Kate E. Creevy, Daniel E. L. Promislow. Mammalian Genome 27(7–8), 279–288 (2016). Read it.

Bears on: A drug makes people live longer in good health.

What it moved: nothing. It was read against the treatments it bears on and no stage or estimate changed. Logging it anyway is how the record shows what has been considered.

Intended leaf: a drug makes people live longer in good health (rapamycin), under energy and fuel sensing; the dog trial is the step between harrison-2009 in mice and any trial in people. The project's founding paper proper, with the study's design and first enrolment, is creevy-2022. Online 2016-05-03; the print issue is August 2016.

The record is data/papers/kaeberlein-2016.toml in the public repository.

Fourteen people with stable scarring of the lungs (idiopathic pulmonary fibrosis) at two centres took dasatinib 100 mg and quercetin 1,250 mg a day for three days a week over three weeks, with no placebo group. All fourteen stayed in; their six-minute walking distance, walking speed and chair-stand time improved, while lung function, blood chemistry and frailty did not change, and the effect on blood markers of worn-out cells was inconclusive. The first trial of a drug meant to kill such cells in people, on physical function rather than on the cells themselves.

Jamie N. Justice, Anoop M. Nambiar, Tamar Tchkonia, Nathan K. LeBrasseur, Rodolfo Pascual, Shahrukh K. Hashmi, Larissa Prata, Michal M. Masternak, Stephen B. Kritchevsky, Nicolas Musi, James L. Kirkland. EBioMedicine 40, 554–563 (2019). Read it.

Bears on: A drug clears aged cells from a person.

What it moved: nothing. It was read against the treatments it bears on and no stage or estimate changed. Logging it anyway is how the record shows what has been considered.

Intended leaf: a drug clears aged cells from a person, under worn-out cells; as the leaf's notes already say, this trial did not measure the cells, so hickson-2019 is the stage record and this is the first human dosing. Online 2019-01-05; the print issue is February 2019. Registered as NCT02874989.

The record is data/papers/justice-2019.toml in the public repository.

From about 8,000 samples across 51 healthy tissues and cell types, a predictor of a person's age was built from the chemical methyl marks on 353 sites in their DNA. In samples it had not been trained on it matched birth-certificate age with a correlation of 0.96 and a median error of 3.6 years, across tissues as different as blood, brain and skin; 20 kinds of cancer tissue read on average 36 years older than the patient. The first multi-tissue "epigenetic clock", and the tool every later claim of age reversal in people is measured with.

Steve Horvath. Genome Biology 14(10), R115 (2013). Read it.

Bears on: A clock read from blood predicts a person's death and disease better than their age does.

What it moved: nothing. It was read against the treatments it bears on and no stage or estimate changed. Logging it anyway is how the record shows what has been considered.

Intended leaf: a clock validated to predict outcomes in a large cohort, under measuring ageing; this is the clock, and the validation against outcomes came in later clocks (PhenoAge, GrimAge, DunedinPACE) and in moqri-2023's criteria. The training-set error is 2.9 years; 3.6 years is the test-set figure. An erratum appeared in Genome Biology 16, 96 (2015) and was not read here.

The record is data/papers/horvath-2013.toml in the public repository.

Nine people with diabetic kidney disease took a three-day course of two drugs that kill senescent cells (dasatinib, a cancer drug, and quercetin, a plant compound). Eleven days later, samples of their fat and skin held fewer such cells than before, and the blood carried less of the inflammatory signal those cells give off. A first demonstration, in people, on a measurement; not a clinical outcome, and with no control group.

LaTonya J. Hickson, Larissa G. P. Langhi Prata, Shane A. Bobart, Tamara K. Evans, Nino Giorgadze, Shahrukh K. Hashmi, Sandra M. Herrmann, Michael D. Jensen, Qingyi Jia, Kyra L. Jordan, Todd A. Kellogg, Sundeep Khosla, Daniel M. Koerber, Anthony B. Lagnado, Donna K. Lawson, Nathan K. LeBrasseur, Lilach O. Lerman, Kathleen M. McDonald, Travis J. McKenzie, João F. Passos, Robert J. Pignolo, Tamar Pirtskhalava, Ishran M. Saadiq, Kalli K. Schaefer, Stephen C. Textor, Stella G. Victorelli, Tammie L. Volkman, Ailing Xue, Mark A. Wentworth, Erin O. Wissler Gerdes, Yi Zhu, Tamara Tchkonia, James L. Kirkland. EBioMedicine 47, 446–456 (2019). Read it.

Bears on: A drug clears aged cells from a person.

What it settled: the rung "first human" of A drug clears aged cells from a person.

Checked against the open-access text on 2026-10-16: 33 authors as listed; nine people; dasatinib 100 mg and quercetin 1,000 mg daily for three days; fat and skin samples on study days 0 and 14, so eleven days after the last dose. A corrigendum of 2020-01-23 (EBioMedicine 52, 102595, https://doi.org/10.1016/j.ebiom.2019.12.004) re-analysed the raw data with five added statistical authors and says some conclusions changed; the detail is in its appendix, not read here, so the finding above should be read with that in mind.

The record is data/papers/hickson-2019.toml in the public repository.

Mice of a deliberately mixed genetic background were fed rapamycin, a drug that damps the cell's main fuel-sensing pathway, from 600 days of age, the equivalent of about sixty human years, at three laboratories at once (The Jackson Laboratory, the University of Michigan and the University of Texas Health Science Center). Across 1,901 mice the age by which nine in ten had died rose by 14 percent in females and 9 percent in males, and mean lifespan by 13 and 9 percent; the pattern of diseases at death did not change. The first drug shown to lengthen life in a mammal when started in old age, and the result the rapamycin trials in dogs and people grow from.

David E. Harrison, Randy Strong, Zelton Dave Sharp, James F. Nelson, Clinton M. Astle, Kevin Flurkey, Nancy L. Nadon, J. Erby Wilkinson, Krystyna Frenkel, Christy S. Carter, Marco Pahor, Martin A. Javors, Elizabeth Fernandez, Richard A. Miller. Nature 460(7253), 392–395 (2009). Read it.

Bears on: A drug makes people live longer in good health.

What it settled: the rung "shown in animals" of A drug makes people live longer in good health.

Intended leaf: a drug makes people live longer in good health (rapamycin), under energy and fuel sensing; this is its animal record. The paper attaches no percentage to median lifespan: the 14 and 9 percent figures are for the age at 90 percent mortality, and the 13 and 9 percent for mean lifespan, so a line reading "median lifespan up 14 percent" would mislabel the figure. Online 2009-07-08; the print issue is dated 2009-07-16.

The record is data/papers/harrison-2009.toml in the public repository.

On 2022-01-07 a 57-year-old man with heart failure who could not get a human heart received the heart of a pig carrying ten gene edits (four pig genes removed, six human genes added) at the University of Maryland. The heart worked for seven weeks, then thickened and failed suddenly from day 49; life support was withdrawn on day 60, and the damage did not look like ordinary rejection. The first pig organ to keep a living person alive, for two months.

Bartley P. Griffith, Corbin E. Goerlich, Avneesh K. Singh, Martine Rothblatt, Christine L. Lau, Aakash Shah, Marc Lorber, Alison Grazioli, Kapil K. Saharia, Susie N. Hong, Susan M. Joseph, David Ayares, Muhammad M. Mohiuddin. New England Journal of Medicine 387(1), 35–44 (2022). Read it.

Bears on: An organ from a gene-edited pig keeps a person alive.

What it moved: nothing. It was read against the treatments it bears on and no stage or estimate changed. Logging it anyway is how the record shows what has been considered.

Intended leaf: a pig organ keeps a person alive, under replacing lost cells and parts; one of the five leaves shared with The Upgrade Tree. Online 2022-06-22; the print issue is dated 2022-07-07. Later pig kidney transplants into living people (2024 onward) are for the leaf session to cite from their own reports.

The record is data/papers/griffith-2022.toml in the public repository.

David Gobel, co-founder of the Methuselah Foundation, opens the foreword by saying he coined the phrase "longevity escape velocity" while working with Aubrey de Grey and Peter Diamandis on the history of prizes. He gives no year, and no dated document from 2004 or earlier in which he uses the phrase has been found; the earliest dated publication of the idea is de Grey's 2004 essay, which uses the words "actuarial escape velocity" instead.

David Gobel. Let's Talk Longevity, Substack newsletter, 2022-05-12. Read it.

What it moved: nothing. It was read against the treatments it bears on and no stage or estimate changed. Logging it anyway is how the record shows what has been considered.

Bears on the headline question itself, so listed against all eight branches. Secondary sources (Wikipedia; a 2018-09-05 Lifespan.io interview whose interviewer calls Gobel the first to put the concept forward) repeat the attribution without a dated primary source. Kept so that the site can say plainly who is credited with the phrase and on what evidence.

The record is data/papers/gobel-2022.toml in the public repository.

Ten healthy men aged 51 to 65 were recruited, and nine analysed, in a one-year trial (TRIIM) of growth hormone with DHEA and metformin, aimed at regrowing the thymus, the gland that trains immune cells and turns to fat with age. The thymus did regain working tissue on MRI in seven of nine men, and four DNA-methylation clocks read about 1.5 years younger than at the start, which is about 2.5 years against the one year of ageing expected with no treatment. Nine men, no placebo group, and the clocks were the outcome; the paper most often cited for age reversal in people, and the one that shows how thin that claim is.

Gregory M. Fahy, Robert T. Brooke, James P. Watson, Zinaida Good, Shreyas S. Vasanawala, Holden Maecker, Michael D. Leipold, David T. S. Lin, Michael S. Kobor, Steve Horvath. Aging Cell 18(6), e13028 (2019). Read it.

Bears on: A person's biological age, read from the marks on their DNA, falls after a treatment.

What it settled: the rung "first human" of A person's biological age, read from the marks on their DNA, falls after a treatment.

Intended leaves: immune rejuvenation with a measured effect in people, under inflammation and the body's messaging; and a clock as a trial measure, under measuring ageing. The 2.5-year figure often quoted is relative to expected ageing; the absolute change from baseline is about 1.5 years. Why one of the ten recruited men is missing from the analysis is not stated in the main text. Online 2019-09-08; the print issue is December 2019.

The record is data/papers/fahy-2019.toml in the public repository.

Five babies and young children at Boston Children's Hospital whose hearts had failed after surgery and who were on a heart-lung machine had mitochondria, the cell's power plants, taken from a small piece of their own abdominal muscle and injected straight into the heart muscle, about a hundred million in ten small injections. Four of the five regained enough heart function to come off the machine within three to six days; two of the five later died of other organ failure. The first report of mitochondria moved from one tissue to another in people, in a handful of emergencies with no comparison group.

Sitaram M. Emani, Breanna L. Piekarski, David Harrild, Pedro J. del Nido, James D. McCully. The Journal of Thoracic and Cardiovascular Surgery 154(1), 286–289 (2017). Read it.

Bears on: Fresh mitochondria are transplanted into a person's failing tissue.

What it settled: the rung "first human" of Fresh mitochondria are transplanted into a person's failing tissue.

Intended leaf: mitochondrial transfer with a measured effect in people, under energy and fuel sensing. The paper has no abstract; the figures are from its text. Online 2017-02-15; the print issue is July 2017 and the journal gives no day.

The record is data/papers/emani-2017.toml in the public repository.

An editorial welcoming the news that the TAME metformin trial's 75 million dollar budget had been found: nearly half pledged through the American Federation for Aging Research by the philanthropist Paul Glenn, the National Institute on Aging having twice turned down applications for the rest, and the remainder, "very recently", from an anonymous donor. De Grey, no admirer of metformin as a treatment, argues the trial is worth the money because it would establish ageing as something a regulator lets a drug be tested against. Seven years on, the trial has not started (afar-2026).

Aubrey D. N. J. de Grey. Rejuvenation Research 22(5), 375–376 (2019). Read it.

Bears on: Metformin delays age-related disease in people who do not have diabetes; A regulator recognises ageing itself as a condition a drug can be approved to treat.

What it moved: nothing. It was read against the treatments it bears on and no stage or estimate changed. Logging it anyway is how the record shows what has been considered.

Intended leaves: the TAME trial enrolling, under energy and fuel sensing, and a regulator accepting ageing as something a drug can be tested against, under measuring ageing. Kept because it is a dated, published statement that the money was in hand, against which the later record (afar-2026) can be read.

Not yet verified against a source: The quoted lines and the funding figures are from a Fight Aging! post of 2019-10-22 reproducing only the opening of the editorial, not from the editorial itself; the issue month and pages are confirmed (Crossref, PubMed).

The record is data/papers/de-grey-2019.toml in the public repository.

An essay arguing that once treatments cut the yearly risk of death at a given age fast enough, by roughly a tenth a year in the age range where most people die, remaining life expectancy stops falling with age and people pull away from death the way a rocket pulls away from the ground. De Grey names that rate "actuarial escape velocity". The paper does not use the phrase "longevity escape velocity"; that wording came later.

Aubrey D. N. J. de Grey. PLoS Biology 2(6), e187 (2004). Read it.

What it moved: nothing. It was read against the treatments it bears on and no stage or estimate changed. Logging it anyway is how the record shows what has been considered.

Bears on the headline question itself rather than any one branch, so it is listed against all eight. Early PLoS Biology also carried print page numbers (723–726); the article number is the publisher's citation. The companion claims log holds de Grey's dated predictions; this record is only the paper that framed the question.

The record is data/papers/de-grey-2004.toml in the public repository.

The Dog Aging Project's description of itself as of its first data release: more than 30,000 pet dogs and their owners enrolled, more than 100 staff at more than 20 institutions, genome sequencing planned for at least 10,000 dogs, a closely studied group of 1,000, and a placebo-controlled trial of rapamycin in 350 middle-aged large dogs (TRIAD, the Test of Rapamycin In Ageing Dogs) at seven US veterinary teaching hospitals, into which enrolment began in June 2021. The largest study of ageing in any animal that lives with people; the rapamycin result is still to come.

The Dog Aging Project Consortium (Kate E. Creevy, Joshua M. Akey, Matt Kaeberlein, Daniel E. L. Promislow and members). Nature 602(7895), 51–57 (2022). Read it.

Bears on: A drug makes people live longer in good health.

What it moved: nothing. It was read against the treatments it bears on and no stage or estimate changed. Logging it anyway is how the record shows what has been considered.

Intended leaf: a drug makes people live longer in good health (rapamycin), under energy and fuel sensing. Nature prints the author as the consortium with a member list that begins with the four names given; indexes list those four and the consortium. The enrolment figures are as of December 2020. Online 2022-02-02; the print issue is February 2022.

The record is data/papers/creevy-2022.toml in the public repository.

Old and young mice were surgically joined so that they shared one bloodstream. The old mice's muscle repair cells regained a young animal's ability to multiply and mend damage, their liver cells divided more, and young serum alone did the same to old muscle cells in a dish. The paper that showed something in young blood, or something missing from old blood, drives part of ageing in tissue; the abstract gives no sample sizes or effect sizes.

Irina M. Conboy, Michael J. Conboy, Amy J. Wagers, Eric R. Girma, Irving L. Weissman, Thomas A. Rando. Nature 433(7027), 760–764 (2005). Read it.

Bears on: Young blood rejuvenates an old animal.

What it settled: the rung "shown in animals" of Young blood rejuvenates an old animal.

Intended leaf: plasma exchange or dilution with a measured effect in people, under inflammation and the body's messaging; this is the animal result the field traces that leaf to. The later dilution paper (mehdipour-2020) and the young-plasma trial in Alzheimer's (sha-2019) follow from it.

The record is data/papers/conboy-2005.toml in the public repository.

Mice aged one year and two years were given a single injection of a gene-therapy virus (AAV9) carrying the gene for telomerase, the enzyme that rebuilds the protective caps on chromosome ends. Median lifespan rose by 24 percent in the one-year-old group and 13 percent in the two-year-old group, with better insulin sensitivity, bone and coordination, and no more cancer than in untreated littermates; an inactive version of the gene did nothing. The animal record for lengthening chromosome ends in a living adult without a rise in cancer.

Bruno Bernardes de Jesus, Elsa Vera, Kerstin Schneeberger, Agueda M. Tejera, Eduard Ayuso, Fatima Bosch, Maria A. Blasco. EMBO Molecular Medicine 4(8), 691–704 (2012). Read it.

Bears on: A gene therapy lengthens the ends of chromosomes in an old animal, with no rise in cancer.

What it settled: the rung "shown in animals" of A gene therapy lengthens the ends of chromosomes in an old animal, with no rise in cancer.

Intended leaf: telomerase therapy in a living animal, under damaged DNA. The AAV9 serotype is given in the full text; the abstract says only a wide-tropism AAV. Online 2012-05-15; the print issue is August 2012.

The record is data/papers/bernardes-de-jesus-2012.toml in the public repository.

The design paper for the Targeting Aging with Metformin (TAME) trial: the authors plan to enrol 3,000 people aged 65 to 79 at about 14 US centres and measure the time to a first new occurrence of any of heart attack and stroke, cancer, dementia or death, a composite chosen so that a trial can test whether a drug slows ageing as a whole rather than one disease, with function and frailty measured alongside. The paper reviews the observational and animal evidence for metformin and says the trial has planning money from the American Federation for Aging Research. The template for how an ageing trial could be run and judged; as of 2026-10-16 the trial has not been shown to have started (see afar-2026).

Nir Barzilai, Jill P. Crandall, Stephen B. Kritchevsky, Mark A. Espeland. Cell Metabolism 23(6), 1060–1065 (2016). Read it.

Bears on: Metformin delays age-related disease in people who do not have diabetes; A regulator accepts a marker of ageing as proof that a treatment slowed it.

What it moved: nothing. It was read against the treatments it bears on and no stage or estimate changed. Logging it anyway is how the record shows what has been considered.

Intended leaves: the TAME trial enrolling, and a drug makes people live longer in good health (metformin), under energy and fuel sensing; and a regulator accepting an ageing marker or a composite as a trial endpoint, under measuring ageing. The paper does not give the six-year duration or describe the 2015 meeting with the FDA; those are from AFAR's pages and news reports, logged under afar-2026.

The record is data/papers/barzilai-2016.toml in the public repository.

In UK general-practice records from 2000 onward, 78,241 people with type 2 diabetes who started on metformin were matched to 90,463 people without diabetes. After adjustment, the people without diabetes had a median survival 15 percent shorter than the metformin group (survival time ratio 0.85, 95 percent confidence interval 0.81 to 0.90), while 12,222 people started on a sulphonylurea had a survival 38 percent shorter. An observational comparison, open to the objection that people who start metformin are healthier than those who start the other drug; it is the result most often cited as the reason to test metformin against ageing.

C. A. Bannister, S. E. Holden, S. Jenkins-Jones, C. Ll. Morgan, J. P. Halcox, G. Schernthaner, J. Mukherjee, C. J. Currie. Diabetes, Obesity and Metabolism 16(11), 1165–1173 (2014). Read it.

Bears on: Metformin delays age-related disease in people who do not have diabetes.

What it moved: nothing. It was read against the treatments it bears on and no stage or estimate changed. Logging it anyway is how the record shows what has been considered.

Intended leaf: a drug makes people live longer in good health (metformin), under energy and fuel sensing. The paper prints the ratio with metformin as the reference, so the figure is 0.85 for the people without diabetes, not a 15 percent gain for metformin users; the gain is the same comparison read the other way and the paper does not print it. Online 2014-07-31; the print issue is November 2014.

The record is data/papers/bannister-2014.toml in the public repository.

The same genetic switch as in the 2011 paper, this time in normally ageing mice of two genetic backgrounds, with the cell-killing drug given twice a week from twelve months of age until death. Median lifespan rose by 27 percent in the mixed-background mice and 24 percent in the inbred strain, with increases of 17 to 35 percent by sex; kidney, heart and fat deteriorated more slowly, cataracts came later, and tumours came later though not less often. The result that turned clearing worn-out cells from a curiosity into a drug programme.

Darren J. Baker, Bennett G. Childs, Matej Durik, Melinde E. Wijers, Cynthia J. Sieben, Jian Zhong, Rachel A. Saltness, Karthik B. Jeganathan, Grace Casaclang Verzosa, Abdulmohammad Pezeshki, Khashayarsha Khazaie, Jordan D. Miller, Jan M. van Deursen. Nature 530(7589), 184–189 (2016). Read it.

Bears on: Clearing aged cells makes old animals live longer.

What it settled: the rung "shown in animals" of Clearing aged cells makes old animals live longer.

Intended leaf: a drug clears aged cells from a person (animal record for lifespan, where baker-2011 is the record for healthy life). The percentages are in the results text, not the abstract, which gives none. Online 2016-02-03; the print issue is dated 2016-02-11.

The record is data/papers/baker-2016.toml in the public repository.

In mice bred to age fast (the BubR1 progeroid strain), a genetic switch was added that kills cells carrying the senescence marker p16 whenever a drug is given. Clearing those cells throughout life delayed the age-related decline of fat tissue, muscle and the eye, and clearing them late in life slowed decline already under way. A gain in healthy life in a fast-ageing mouse, not in lifespan or in a normally ageing animal; the first direct evidence that worn-out cells drive ageing rather than merely mark it.

Darren J. Baker, Tobias Wijshake, Tamar Tchkonia, Nathan K. LeBrasseur, Bennett G. Childs, Bart van de Sluis, James L. Kirkland, Jan M. van Deursen. Nature 479(7372), 232–236 (2011). Read it.

Bears on: A drug clears aged cells from a person; A drug that clears aged cells beats a sham treatment in a controlled trial.

What it settled: the rung "shown in animals" of A drug clears aged cells from a person; the rung "shown in animals" of A drug that clears aged cells beats a sham treatment in a controlled trial.

Intended leaf: a drug clears aged cells from a person (its shown-in-animals rung, which the leaf's earlier_stage entry already cites by this paper's link). Online 2011-11-02; the print issue is November 2011 and its exact day was not confirmed from the indexed records.

The record is data/papers/baker-2011.toml in the public repository.

Seven children and young adults aged 4 to 19 with spina bifida and failing bladders had cells taken from their own bladders, grown on a scaffold shaped like a bladder for about seven weeks, and implanted to enlarge the organ. Over a mean follow-up of 46 months the best results came from the composite scaffold wrapped in abdominal tissue: pressure at capacity fell by 56 percent, capacity rose 1.58-fold and the bladder stretched 2.79 times better, with kidneys preserved and no stones. The first organ-like tissue grown from a person's own cells to work in people for years.

Anthony Atala, Stuart B Bauer, Shay Soker, James J Yoo, Alan B Retik. The Lancet 367(9518), 1241–1246 (2006). Read it.

Bears on: A printed or lab-grown organ works in a person for a year.

What it settled: the rung "first human" of A printed or lab-grown organ works in a person for a year.

Intended leaf: a lab-grown organ works in a person for a year, under replacing lost cells and parts; one of the five leaves shared with The Upgrade Tree. A bladder is the simplest organ to grow, and the approach has not reached a solid organ. Only the issue date, 2006-04-15, is on record; no online-first date was confirmed.

The record is data/papers/atala-2006.toml in the public repository.

Ten years after its design paper, the sponsor's own page still describes TAME in the future tense: a six-year trial at 14 US institutions that "will engage" over 3,000 people aged 65 to 79, with Wake Forest as coordinating centre, under the heading "Raising Funds to Launch" and a call for donors. The published funding history runs: planning money from AFAR (2016); nearly half of a stated 75 million dollar budget pledged through AFAR by the philanthropist Paul Glenn, two refusals by the National Institute on Aging, and a reported anonymous pledge for the remaining 40 million in September 2019 that no later source confirms; a claim in April 2022 that the Hevolution Foundation would fund a third, which the foundation did not confirm; about 5 million dollars set aside by the National Institute on Aging and cost estimates of 45 to 70 million dollars in April 2024. No primary source says enrolment has begun, and no registration for the trial was found on ClinicalTrials.gov on 2026-10-16.

American Federation for Aging Research. AFAR's own trial page, afar.org/tame-trial, as archived on 2026-07-04. Read it.

Bears on: Metformin delays age-related disease in people who do not have diabetes.

What it moved: nothing. It was read against the treatments it bears on and no stage or estimate changed. Logging it anyway is how the record shows what has been considered.

Intended leaf: the TAME trial enrolling, under energy and fuel sensing. This record exists so the leaf can say, with sources, that the field's flagship ageing trial has been announced as imminent since 2015 and has not started. The weekly reading should check AFAR's page and the trial registry; the day a registration or a first-participant notice appears, that becomes a new record. AFAR's 2024 news post gives the age range as 65 to 80; the trial page says 65 to 79.

Not yet verified against a source: The two National Institute on Aging refusals are from Aubrey de Grey's 2019 editorial (de-grey-2019), seen only as excerpted in a 2019-10-22 Fight Aging! post, not from the editorial itself; the Paul Glenn gift through AFAR is also reported in a Lifespan.io article of 2023-02-08. The AFAR page returned 'not found' in the web archive from 2026-10-01 and still did on 2026-10-17; it is absent from AFAR's site map and no replacement page was found, but AFAR has not said why it was removed.

The record is data/papers/afar-2026.toml in the public repository.