Ageing

1 On vitalism

Vitalism: the 19th-century doctrine of élan vital — Bergson, Driesch — claiming living things contain a non-material vital force absent from inert matter.

The 21st-century version drops the metaphysics and keeps the affect: raw milk, sun on the skin (and sometimes elsewhere), seed-oil panic, anti-pharma scepticism, the “Don’t Die” / “Blueprint” / “MAHA” axis, founder-led biohacking stacks, carnivore diets, Liver Kings, cold plunges, mouth-taping, nicotine pouches as nootropics, testosterone optimisation as personality.

There is also a literal modern Vitalist movement — capital V, self-identified, rallying under the banner of anti-death, with foundations, pop-up cities, and magazine profiles to match. They cite Nick Bostrom’s Fable of the Dragon-Tyrant (death as a dragon we’ve told ourselves is natural because we can’t bear to admit we’re being farmed), Bryan Johnson’s Don’t Die liturgy, and the transhumanist reframe that mortality is a moral failure rather than a natural condition.

So! We reboot élan vital for the supplement-stack era, with the “non-material vital force” redrawn as the sum total of biomarkers we can measure and interventions we can stack. Both Vitalisms share the same underlying intuition — that life is a positive thing being suppressed, and that the suppression can be lifted. Nineteenth-century vitalists thought the suppressor was inert matter; their 21st-century inheritors accuse Big Pharma, seed oils, or evolutionary pressures we can now engineer around.

There’s signal in that noise. Sleep, sun, movement, low-stress social context, and unprocessed food all matter — often more than the supplement debates would suggest, and the medical-industrial complex was slow to take some of those seriously. Genomic medicine and precision medical interventions are likely to be a huge deal in the future, drastically shifting the landscape of what’s possible. But right now that landscape is crowded with shady looksmaxxxers, peptide bros, seed oil truthers and founders looking for (non-oil) seed capital for technological interventions of variable plausibility. The trouble is that the modern vitalist community is structurally epistemically permissive. That is to say, full of shit. It flatters n=1 anecdotes, treats RCTs as colonial impositions, and turns every founder-led open-label trial into a Galileo-vs-the-Pope drama. That posture is commercially convenient. Cashed-up older people rarely want to die, so selling them through small trials with marginal benefits and low reproducibility is a lucrative enough business to prop up bad science, and maybe even some good science, indefinitely.

Phenomenologically, extending my life is not just hard to do but actuarially difficult to measure, and the incentives in the miracle supplement economy are not ideal, since we are ultimately trying to monetise the fear of death, which circumstance (death, I mean) is definitionally one in which the customer cannot demand a refund.

Still, I don’t want to die either. God help me, I will try this stuff.

2 “My biological age is…”

What we actually want from any intervention is more healthy years lived — the economist’s formalisation is the quality-adjusted life-year (QALY). To plug it into an EV calculation we want to know two things — how many extra years, and at what rate any age-related decline gets pushed back.

Neither is easy to measure directly. The gold-standard hard endpoint is all-cause mortality — counting deaths from any cause, because a death is a death and the count is hard to argue with. A proper RCT with that endpoint needs to run ~15–25 years in a middle-aged cohort before the survival curves diverge clearly enough to tell — by which time funders, subjects, and drug patents have mostly aged out, by which I mean: died. So the field leans on proxies: epigenetic clocks, lipid panels, inflammatory markers, performance tests. The headline family are the epigenetic age clocks — scores built from DNA-methylation patterns, calibrated against chronological age, and claimed to estimate “biological age”. Popular versions include Dunedin PACE (Belsky et al. 2022), PhenoAge (Levine et al. 2018), and GrimAge (Lu et al. 2019). When someone says “I lowered my biological age by 12 years,” it’s one of these scores that moved.

Most intervention studies compare treated and control groups at a fixed timepoint — a snapshot — rather than tracking individuals over long periods. Even when the clock is built to estimate rate (Dunedin PACE is one), a snapshot comparison at a single age can only show that the two groups’ biological-age scores differ, not whether the treated group is now ageing more slowly. So most “intervention X slows biological ageing” results can’t cleanly distinguish two very different things:

1. Baseline shift. The treated group ends up healthier than controls at that moment. A one-off gain — useful, finite.
2. Rate change. The treated group is now ageing more slowly than controls. A gain that compounds across every remaining year.

Both can buy us extra healthy years, but it is the second that matches the way “anti-ageing” claims are usually phrased. Keshavarz and Ehninger (2025) tally this across the post-hallmarks-of-ageing literature and find that under 5% of treatment effects are clearly rate effects; more than 90% are “not clear”. That includes the epigenetic clocks themselves — Dunedin PACE, the one CALERIE 2 used, among them. This doesn’t undo the EV arithmetic for any particular intervention; a baseline shift can still buy healthier years. But it should pull confidence down on any claim phrased as “slows ageing” relative to one phrased as “lowers cardiovascular events at 60”.

Add the usual other suspects — sex- and strain-specific animal results, small founder-led open-label trials unlikely ever to be reported as nulls, dirty supercentenarian registries (Newman 2024) — and the practical consequence is that the credences that look low here (15–20% for plausible mechanisms without hard-endpoint data) aren’t not so much pessimism as they are IMO reasonable priors about the replication rate in this biased literature, that nonetheless come out in favour of trying cheap interventions that don’t look terrible.

3 A decision rule for supplements and interventions

My willingness to pay at the margin is something like $100 per expected day of healthy life. The exact number is not so important — I just need to fix a concrete price to plug into a finite, non-trivial trade-off. I picked $100 because it’s roughly the rate at which I’d trade surplus earnings for a day of healthy life; higher numbers turn the calculation into “work more to live longer”, which is not the deal I want — I’d end up funding a supplement regimen by eating the time I’d otherwise spend exercising or, you know, actually living my life. Although obviously I would like to spend infinity moneys on living longer, it doesn’t make sense to pay too much since we can easily end up spending more on the pursuit of longevity than we would gain from it.

For each candidate I try to account for:

1. Money cost, annualized. A $50/month pill is $600/year.
2. Time cost. Time and money are separate budgets — the supplement dollar budget below is additional to exercise’s time budget, not netted against it. Supplement time costs themselves are usually negligible: a pill is ~10 s/day; a five-day fasting-mimicking cycle four times a year is ~20 days of mildly underpowered functioning while still at work. Exercise’s own time cost (~180 hours/year of cardio plus ~2 hours/week of resistance training) is treated as its own intervention below, where the return justifies the time. The interaction to watch is when one budget eats the other: work hours funding supplements come out of leisure-or-exercise; a supplement’s attention-load can erode the exercise habit; an exercise expansion can come out of paid hours.
3. Opportunity cost / what it precludes. Dollars on a hip peptide are dollars not on a trainer. Hours working to pay for supplements are hours not lifting. Interventions that actively subtract from others (e.g. metformin blunting exercise adaptations) carry that subtraction as a cost.
4. Expected benefit. Credence-that-it-works × expected healthspan gain if it does. Plausible mechanism plus one positive small human RCT gets me to maybe ~20% credence; mouse-only evidence, ~10%; strong cross-species evidence plus hard human endpoints, 60–70%+.
5. Downside risk. Side-effects, drug interactions, adulteration, black-swan cancer-risk tails.

Putting that together:

• \(\text{credence} \times \text{expected days} \times \$100/\text{day} − \text{costs} \gg 0\) → I should take it.
• Marginal → take with a pre-committed re-evaluation point (“one year then re-read the literature”, “if my biomarker X doesn’t move”).
• If costs dominate → skip.

Three notes about my reasoning:

• Weak evidence + plausible mechanism + cheap + safe = take now. The question isn’t “have we proved it works” but “if it’s confirmed later, will I wish I’d been taking it all along?” At $100/day WTP, a $600/year supplement breaks even at ~six days/year of expected healthspan — still a low bar for anything with a plausible story, though not a trivial one. A 20%-credence intervention needs to plausibly deliver ~30 days/year of healthspan if it works; a 10%-credence one, ~60 days/year. Skipping plausible cheap interventions until hard data lands leaves a lot of expected value on the table, in exchange for a slightly cleaner epistemic conscience. Note that this only works at the margin — if the number of plausible cheap interventions grows large, then… well see the next point.
• “Cheap” in dollars isn’t cheap overall. A six-item regimen costs attention, compliance effort, interaction risk, and also makes you into the kind of person who talks about their “stack” on Reddit. A prescription drug that blunts exercise adaptations is expensive no matter how cheap the tablet is. The question is total marginal cost, including what it crowds out.
• Waiting is itself expensive. Per the methodological caveat above, the natural reflex of “wait for the definitive trial” has a price tag attached. A healthspan RCT needs decades to read out, more or less by definition: the more an intervention actually extends life, the longer the trial has to run before the mortality curves diverge. By the time clean evidence lands, many of us won’t be around to use it, and anything the intervention would have bought us in the meantime is forfeit. So “wait and see” is itself a bet, and its premium is paid in years of our own life. This asymmetry is what makes “take now and pre-commit to re-evaluating” the right call for cheap, plausible, safe interventions in a way it wouldn’t be for, say, a chemotherapy decision.

The base rate for “pop a pill and live longer” claims surviving replication is not great. Most molecules that extended lifespan in worms or flies failed to do so in the NIH Interventions Testing Program’s (ITP) genetically heterogeneous mice (Nadon et al. 2017; Strong et al. 2013), and the few that did — rapamycin, acarbose, 17α-estradiol — aren’t things to self-prescribe from a podcast. But “bad replication rate” means “low credence”, not “zero credence”, and for cheap interventions that’s often enough.

4 Boring stuff that probably works

The largest effect sizes are in lifestyle. In EV terms, these dominate every supplement decision by at least an order of magnitude, and their time cost is the main reason to be stingy about supplement attention:

• Cardiorespiratory fitness. In a 122,000-patient treadmill cohort, elite fitness was associated with roughly 5× lower mortality than the low-fit group, with no observed ceiling (Mandsager et al. 2018). Call this ~30 min/day ≈ 180 hours/year (for me, mostly bike commuting). That time comes out of leisure or paid work — both material — but the mortality arithmetic is so lopsided that it clears the bar by orders of magnitude. This is the highest-EV intervention in the whole document, and everything that follows is marginal relative to it. That’s what makes “crowds out exercise” show up as a real cost elsewhere (metformin’s adaptation tax, attention-heavy stacks, work hours funding supplements): each such item is trading the best deal on the menu for an uncertain one.
• Resistance training. Preserves muscle mass and bone density in a way no supplement substitutes for. Budget another ~2 hours/week.
• Sleep — 7–9 h, consolidated, non-alcoholic. Free, but costs sleep-hygiene discipline.
• Not smoking, not drinking to excess, not getting too chubby, eating plants.

Everything below is likely a marginal add to this baseline. If a supplement regimen would plausibly eat enough attention or money to erode the exercise habit, the regimen is net-negative regardless of its direct effect.

5 Sirtuin stuff, and the Sinclair regimen

Sinclair (2021) recommends supplements targeting the sirtuin pathway — resveratrol, NAD⁺ precursors (NMN / NR), metformin — plus fasting. In 2026, as far as I can tell, most of this has aged poorly: the evidence hasn’t kept up with the enthusiasm. See also David Sinclair: Learnings Since Lifespan + What’s Next.

6 SGLT2 inhibitors and GLP-1 agonists

Two drug classes from the metabolic-disease side of the clinic have accumulated longevity interest since the Sinclair-era attention peaked, on rather different kinds of evidence.

Canagliflozin (an SGLT2 inhibitor, widely prescribed for type-2 diabetes) extended median lifespan by ~14% in male — but not female — ITP mice (Miller et al. 2020). Sex-specific results are a recurring feature of the ITP and make interpretation harder. The mechanism is plausible: better glucose handling, mild caloric loss via glucose excreted in the urine, heart and kidney benefits. Human cardiovascular and renal outcome trials in diabetics are apparently as clean as these things get; for non-diabetics the longevity case is mostly extrapolation, and off-label access is non-trivial.

EV is hard to pin down outside diabetes: direct evidence in non-diabetics is thin, access requires a prescription, and the mouse result is male-only. I’d file it under “re-evaluate when we get a decent randomised controlled trial in non-diabetics”.

Semaglutide and the GLP-1 class have moved faster. The SELECT trial randomised ~17,000 non-diabetic adults with obesity and established cardiovascular disease to semaglutide or placebo and reported a ~20% reduction in major adverse cardiovascular events (“MACE” — the standard composite endpoint of cardiovascular death, non-fatal heart attack, and non-fatal stroke) over a median 40 months (Lincoff et al. 2023). That is a hard endpoint in a non-diabetic population 🎉— the kind of evidence most longevity interventions lack. For people in a SELECT-like risk stratum (i.e. overweight with established cardiovascular disease) the EV maths looks favourable.

For metabolically healthy, lean people the picture is murkier. Most of the effect is plausibly down to weight loss and its cardiovascular consequences, which cheaper interventions (exercise, diet) can deliver with less side-effect tail risk. Dollar costs are non-trivial (hundreds a month, prescription-only), and sarcopenia (loss of muscle) during rapid weight loss is an apparent concern — relevant to anyone already using resistance training as the main longevity lever. The direct longevity case outside obesity-plus-CVD is open.

EV: we should take it seriously when we struggle with weight management; otherwise hold until evidence outside obesity-with-cardiovascular-disease materialises. If it works as a general ageing intervention the story is probably through weight, glucose, and cardiovascular endpoints that the boring-stuff section already covers — which means the interesting question is whether it beats exercise in the populations who can’t or won’t train.

7 Fasting and caloric restriction

Caloric restriction is apparently the oldest and most robust lifespan intervention in laboratory animals. The translational question is whether it works in humans without the costs of semi-starvation — and unlike most longevity interventions, fasting is primarily time-and-discomfort-costly rather than dollar-costly, so the EV maths is different.

Two useful data points:

1. The CALERIE 2 RCT — two years of ~12% caloric restriction in healthy non-obese adults — slowed epigenetic ageing on the Dunedin PACE clock by ~2–3%, which (via other studies) apparently maps to roughly a 10–15% reduction in mortality risk (Waziry et al. 2023). PhenoAge and GrimAge didn’t move. Long-term mortality data aren’t in.
2. Five-day fasting-mimicking-diet (FMD) cycles, repeated a few times per year, reduced biological-age markers and several disease-risk markers in a small human trial (Brandhorst et al. 2024).

Simple time-restricted eating (16:8 and friends) hasn’t shown clear benefits beyond calorie-matched controls in the head-to-head RCTs I’ve seen summarised (Lowe et al. 2020; Liu et al. 2022; Lin et al. 2023). The conclusion is that it’s useful as a weight-management tool if it suits the appetite; not anti-ageing magic.

A rough ordering of the three:

• Periodic FMD cycles (5 days × ~3–4/year). Direct cost: groceries. Time cost: ~20 days/year of mildly underpowered functioning. Credence of a healthspan benefit ~30–40%; expected gain if it works, non-trivial. EV looks clearly positive for anyone who can tolerate it. I’d do it, with appetite/social tolerance as the gating factor, not the evidence. Re-evaluation trigger: a big negative RCT, or personal biomarkers going the wrong way.
• Daily severe caloric restriction (CALERIE-style). The lifelong time and social cost is large — arguably the largest of anything considered here, because it’s every meal forever. Benefit is small in absolute terms, with obvious downsides for muscle/bone if not paired with resistance training and adequate protein. EV is marginal at best, and the precluded-by cost (social meals, time, enjoyment) is high. Most people shouldn’t bother.
• Time-restricted eating. A weight-management tool. If skipping breakfast keeps us from snacking, fine; we probably shouldn’t bill it as longevity. Which is annoying because I want to go all-in on this.

8 Things that seem worth taking at the cheap end

These clear the EV bar under the framework above — cheap enough in money, time, and interaction cost that even modest credence of an effect makes them obvious takes:

• Vitamin D₃ if plausibly deficient (low-sun, older, covered-up, dark-skinned). ~$50/year. Chowdhury et al.’s meta-analysis of 22 RCTs reports ~11% lower all-cause mortality on vitamin D₃, concentrated in people who started deficient (Chowdhury et al. 2014). I’d test 25(OH)D once, supplement if low, then re-test. This is closer to “correcting a deficiency” than “taking a supplement”.
• Creatine monohydrate, 3–5 g/day. ~$100/year. Decades of safety data, a 22-RCT meta-analysis reporting ~1.4 kg extra lean mass plus strength gains on top of resistance training in older adults (Chilibeck et al. 2017), and plausible bonus cognitive and bone effects. I take this.
• Omega-3 (EPA+DHA) if oily fish isn’t a regular feature of the diet. ~$150/year. Modest but apparently consistent cardiovascular-mortality signal across meta-analyses. I’d take it, or eat two servings of fatty fish a week.
• Adequate protein — ~1.2–1.6 g/kg for people over 50 defending against sarcopenia. Not a supplement; a diet-design choice. The 0.8 g/kg default was set for sufficiency in young adults, not healthy ageing.
• GlyNAC (glycine + NAC). ~$100–200/year. Plausible mechanism (topping up glutathione, a key cellular antioxidant), one positive RCT in older adults (Kumar et al. 2023). Cheap, safe, plausible but only one supporting trial. I’d take it and re-evaluate after a replication or a clear negative trial.
• Taurine, 1–3 g/day. ~$50/year. Singh et al.’s cross-species evidence looks strong, and taurine apparently extended both lifespan and healthspan in their mouse and monkey work (Singh et al. 2023); the human correlational data is weaker than the hype, and some have challenged whether taurine levels actually decline with human age. The side-effect profile is benign in people without existing issues. Even at 10–15% credence of a human benefit the EV comes out positive by a wide margin. I’d take a moderate dose and re-evaluate when a proper human RCT comes out.
• NR (bulk 300 mg/day). Already covered above. Clears the bar cleanly at the ~$90/year bulk-supplement price.
• CoQ10, conditional on age or statin use. ~$100–400/year depending on form (generic ubiquinone at the cheap end; ubiquinol more bioavailable and more expensive, especially in older adults) and dose. The unusual item on this list is that it actually has hard-endpoint RCT data. Q-SYMBIO gave 300 mg/day to ~420 adults with moderate-to-severe heart failure for two years and reported all-cause mortality cut roughly in half (18% → 10%) against a composite MACE hazard ratio of 0.50 (Mortensen et al. 2014) — a treatment effect larger than most cardiology interventions achieve on top of standard care. KiSel-10 gave CoQ10 plus selenium to healthy Swedes aged 70–88 for four years and reported reduced cardiovascular mortality apparently persisting at 12-year follow-up (Alehagen et al. 2013); the combined-supplement design muddies attribution, and Sweden has a baseline selenium-deficient population, so the effect may not generalise cleanly. Statins blunt endogenous CoQ10 synthesis, and a 2025 meta-analysis finds CoQ10 reduces statin-associated muscle pain (Kovacic, Habicht, and Eckert 2025), which matters mainly because it can keep someone tolerating a drug with real downstream benefit. EV is a clear take in heart failure, in statin users with muscle symptoms, and plausibly in older adults more generally; markedly less compelling in a healthy, training-age, non-statin-using person, where direct evidence thins out and the mechanism is reaching.

Borderline cases I’d consider similarly admissible but am less committal on: spermidine (similar to taurine, weaker cross-species data); glucosamine/chondroitin (interesting UK Biobank mortality signal, an obvious take if joints already hurt); L-citrulline at ~6–10 g/day (~$300–600/year in bulk powder — pills are impractical at that dose). The mouse story is accumulating: citrulline apparently declines with age, extends lifespan in C. elegans, and attenuates age-related glucose and inflammatory signs in mice (Rajcic et al. 2023; Xie et al. 2025). Human meta-analyses show small blood-pressure reductions (more reliably at ≥6 g/day) and a ~0.9-point improvement in flow-mediated dilation — a blood-vessel-function test (Barkhidarian et al. 2019), but no hard-endpoint data yet. The mechanism is plausible (nitric-oxide precursor, plus effects on how immune cells handle energy). For a training-focused person with normal BP, most of the surrogate effect is probably already being moved by exercise; for a hypertensive older adult who won’t or can’t train, EV starts to look competitive. Below NR and taurine in priority, above nothing.

Firmly in the skip tier by this framework:

• Resveratrol (evidence of absence, not just absence of evidence).
• Metformin / berberine for a training-focused non-diabetic (the exercise-adaptation tax dominates the direct benefit).
• The Bryan-Johnson-style 60-item stack — even if each individual line item is marginally positive, the interaction risk, compliance cost, and opportunity cost crowd out the exercise lever that matters most. Stacks are overpriced in ways the per-bottle cost hides.
• Anything whose main evidence is a single founder-led open-label trial.

I remain interested in reports that fasting benefits the immune system. Fasting apparently does this in mice (Brandhorst et al. 2024); the human equivalent is less certain.

9 Blood plasma dilution

Blood boys, etc.

Do we even need blood boys?

See this Open Philanthropy summary:

The challenge: Given the central role ageing plays in disease progression, a better understanding of the ageing process could lead to improvements in a broad range of health outcomes.

The research: Dr. Irina Conboy, a professor in the Department of Bioengineering at UC Berkeley, has made significant strides in the scientific understanding of how blood factors—hormones, proteins, and other molecules circulating in the bloodstream—influence ageing. Though Dr. Conboy’s work had already gained recognition by the time of our first grant in 2017, we believed it was still neglected relative to its potential in a field that had received much private investment but little public research funding.

With support from our team and other funders, Dr. Conboy:

• Developed micro-apheresis for mice, an innovative technique that allows filtration of blood to remove small molecules.
• Identified 10 novel biomarkers of ageing.
• Uncovered a major mechanism for rejuvenation through modulation of the TLR4 receptor, which appears to play a key role in age-related inflammation.
• Revealed (Kim et al. 2022) that diluting old blood (with saline and purified albumin (Yilmaz et al. 2011)) — rather than adding young blood, which has garnered provocative headlines—can have rejuvenating effects on muscle, liver, and brain tissue.

The impact: Dr. Conboy’s research offers a new perspective on how factors in blood affect ageing. Her findings suggest that identifying and counteracting pro-ageing factors in blood could potentially slow or reverse certain effects of ageing, opening new avenues for improving human health and longevity.

10 What biomarkers are cheap to track to assess the effectiveness of ageing interventions?

TODO

11 Collagen

Controversial! Who doesn’t want to look young forever?

Two different questions often get conflated. Does eating more collagen help cosmetic ageing — wrinkles, skin elasticity, hydration? And does it help connective-tissue outcomes — tendons, joints, soft-tissue recovery? I thought that collagen did not survive the gut. Turns out I was wrong: small proline-containing peptides (notably prolyl-hydroxyproline) appear to survive digestion and show up in plasma (Iwai et al. 2005; Shigemura et al. 2009), and pre-exercise gelatin plus vitamin C apparently boosts tendon collagen synthesis (Shaw et al. 2017; DePhillipo et al. 2018). So the tendon / soft-tissue-recovery usage has at least a plausible story.

The cosmetic-ageing story is weak. Meta-analyses of oral collagen peptides for skin elasticity, hydration, and wrinkles report small positive effects on short-term surrogate endpoints over 8–12 weeks (de Miranda, Weimer, and Rossi 2021; Pu et al. 2023), but the evidence base is dominated by small, industry-funded trials using proprietary endpoints, rarely running beyond three months (e.g. Czajka et al. 2018). I can’t find any trial showing that collagen peptides prevent or reverse visible wrinkle formation on a clinically meaningful timescale.

The interventions that actually dominate cosmetic-ageing prevention — daily sunscreen, topical retinoids, not smoking, not getting sun-baked — are cheap or already baked into the baseline, with evidence better than the collagen literature by an order of magnitude.

EV for the cosmetic indication at ~$600/year looks marginal-to-negative once funding bias is priced in. For tendon or soft-tissue recovery the picture tilts positive: a short course of ~15 g collagen/gelatin plus vitamin C an hour before rehab exercise, per the Shaw protocol, is cheap and low-risk. Outside that specific use case, I’m not currently enthused.

12 Will I have a nice planet to spend my long healthspan on?

I love the way you are thinking. Not necessarily. Get to work on saving it, now that you are going to be around for a while.

13 Incoming

• John Wentworth, Gears of Aging is some interesting deductive work on how aging might actually work.

• Regional diversity in longevity trends in Western Europe / interactive map visualization

• The Human Lifespan Probably has an Upper Limit … and Why That’s Good

  […] we can imagine a future in which we slow aging; but today, there’s no sign of such an elixir. And so we are faced with a simple mathematical problem; as long as aging remains inexorable, there is likely a natural upper limit to the human lifespan.

  And that’s probably a good thing. You see, if the elixir of life were actually discovered, the risk is that the spoils of longevity would be hoarded by the rich. Thankfully, the opposite is true today. Indeed, one of the most stunning features of longer life expectancy is that it makes the human lifespan more equal.

• The Old, Old, Very Old Man: Thomas Parr and the Longevity Trade

  As the story goes, Old Tom Parr was relatively healthy for being 152 until a visit to noxious, polluted London in 1635 cut his long life short. Katherine Harvey investigates the early modern claims surrounding this supercentarian and the fraudulent longevity business that became his namesake in the 19th century.

• Don’t Die: The Man Who Wants to Live Forever about Bryan Johnson.

• Longevity and the Mind

• How many extra days of life do we get from taking statins?

• A review of Eric Topol’s Super Agers:

  The book is broad and shallow, as if he’s trying to show off how many topics he’s familiar with. Too much of the book consists of long lists of research that Topol finds interesting, but for which I see little connection with aging. He usually doesn’t say enough about the research for me to figure out why I should consider it promising.

  He mostly seems to be saying that the number of new research ideas ought to impress us. I care more about the quality of the most promising research than about the quantity of research.

  The book is mostly correct and up-to-date, but I’m unclear what kind of reader would get much out of it.

14 References