// learn.shawon.ch / evidence-desk / bryan-johnson SCORECARD
← Evidence desk

Evidence desk · Source scorecard

Bryan Johnson

The verdict

Not a scientist — a tech entrepreneur (he founded Braintree, which bought Venmo, and sold it to PayPal for ~$800M) who now runs Project Blueprint: a radical, ~$2M/year n-of-1 experiment on himself, measuring 100+ biomarkers and publishing the data, wrapped in the "Don't Die" brand (a Netflix film, summits, and a quasi-religion) and a supplement-and-food company that sells the protocol. He has no medical or science training; he runs Blueprint with a hired medical team.

The split is unusually clean. His core is genuinely evidence-based — a whole-food diet, daily Zone 2 and strength work, mild calorie restriction, and fanatically regular sleep — and, to real credit, he abandons interventions when his own data show nothing (he quit rapamycin and young-plasma transfusions publicly). The failure mode is just as consistent: he treats surrogates as if they were proven lifespan — epigenetic "aging-speed" scores become "31 years younger" — bolts on unproven, expensive, sometimes risky bets (offshore gene therapy, mega-stacks), and sells the whole thing. So the rule: copy the boring core, ignore the scoreboard and the catalog, and treat anything exotic as an experiment with a sample size of one.

What's inside

  1. The pattern at a glance
  2. Who he actually is
  3. The core
  4. The exotic interventions
  5. Measurement & the clock problem
  6. Risks & the results question
  7. The business
  8. Credibility
  9. For you

The pattern at a glance

His claims sort by how far they sit from the basics. The disciplined fundamentals — sleep, diet, training — track the evidence and sit right. The things that make Blueprint distinctive — the clock-based "age reversal" scoreboard, the mega-stack, offshore gene therapy — drift left, because they ride on surrogate markers and animal data, and each comes with a product or a price tag.

outruns the evidence tracks the evidence Regular sleep schedule Whole-food diet · Zone 2 · strength Drops what his own data debunk Early time-restricted eating Big supplement stack Young-plasma transfusions Offshore gene therapy "31 years younger" clock score
The fundamentals track the data; the distinctive, surrogate-driven interventions outrun it.

Verdicts: Solid (mainstream-backed), Mixed (real but weaker/contested than implied), Overstated (more confidence than the data support), Safety flag (a caution worth heeding). The thread through all of it: he leans on surrogate markers (epigenetic clocks, NAD, follistatin, muscle %) and animal data as if they were hard human outcomes — on an n-of-1 with no control. Where he wins is exactly where the evidence is most boring.

Who he actually is

The honest finding: the money and the transparency are real; the scientific authority is borrowed. He's a businessman running a self-experiment with a medical team, not a researcher — and he doesn't claim a degree he lacks.

Background & credentials
ClaimRealityVerdict
Tech entrepreneur — founded Braintree (acquired Venmo), sold to PayPal (2013, ~$800M)True and verifiable. That fortune is what funds Blueprint and lets him run the experiment at all.Solid
No medical or scientific trainingCorrect — his background is business/tech, not biology or medicine. He runs Blueprint with hired physicians and doesn't claim a science degree.Solid — not a clinician
Runs Blueprint as a rigorous "measure everything" n-of-1 (~$2M/year)Real and unusually transparent — he publishes protocols and raw data. But an n-of-1 with no control group can't, by design, prove what works.Mixed
"Don't Die" — a measurement-driven anti-aging missionAccurate. Began as a personal protocol (~2021); now a Netflix documentary, a summit series, a commercial brand, and an explicitly quasi-religious movement.Solid

The core

His strongest territory by far — and the part worth actually copying. It's the most disciplined version of the boring basics, and it's where the human evidence genuinely lives.

The evidence-based foundation
ClaimEvidenceVerdict
Whole-food, plant-forward diet; no added sugar or saltSquarely mainstream — the least exotic and best-supported part of the whole protocol.Solid
~1 hr/day exercise: Zone 2, strength, some intensityVO₂ max and strength are among the strongest mortality predictors; combined aerobic + resistance training is ideal. Textbook.Solid
Fanatically regular sleep on a fixed scheduleSleep regularity predicts mortality even better than duration (UK Biobank, n≈61,000). Genuinely well-supported and under-appreciated.Solid
Mild caloric restriction (~10–20%)CALERIE — the human CR trial — improved cardiometabolic markers and slowed DunedinPACE. The best human data tying a Blueprint behavior to an aging marker (still a surrogate, ≤2 yr).Solid — surrogate
Early time-restricted eating (last meal ~11am)Early TRE beats longer eating windows for weight, glucose, and blood pressure — but the benefit largely tracks the calorie deficit, not magic in the timing.Mixed

The exotic interventions

Everything past the basics. The honest pattern is animal-and-mechanism promise that hasn't translated to humans — and, to his credit, two interventions he ran hard and then publicly dropped when his own numbers showed nothing. That self-correction is the best thing on this page.

Past the basics
ItemEvidenceVerdict
Large supplement stack (100+ pills, later ~40 — NMN, spermidine, CaAKG, CoQ10…)Multivitamins show no mortality benefit in healthy adults, and most "longevity" compounds (NMN/NR, spermidine, CaAKG) are preclinical with no human hard-outcome data. Cutting the stack and gating additions on biomarkers is the defensible move.Overstated
Rapamycin (~5 years) — stopped in 2024Strong mouse lifespan data; no proven human longevity benefit, and he hit the known side effects (lipids, glucose, infections). He quit publicly when his data turned — exactly the right behavior, and the same shape as Attia walking back metformin.Solid — updated
Young-plasma exchange (with his son & father) — stoppedParabiosis is mouse-only; the FDA says young-donor plasma has "no proven clinical benefit" for aging. His own panel detected nothing and he discontinued. (His later total plasma exchange is a different, still-speculative therapy — don't conflate them.)Solid — updated
Follistatin gene therapy (offshore, ~$25k)The most speculative item: animal-only longevity data, done in an unregulated overseas setting, with surrogate "results" (follistatin +160%, muscle +7%) that diet and training could plausibly produce. Aging researcher Matt Kaeberlein: "there is absolutely no reason to trust these people. They are trying to sell you something."Safety flag

Measurement & the clock problem

The engine of the whole project — and its central scientific error. The disciplined-measurement ethos is admirable; the mistake is treating biological-age scores as if they were proven lifespan. A clock that predicts mortality across a population does not prove that moving your score changes how long you live.

Biomarkers & "biological age"
ClaimEvidenceVerdict
Tracks 100+ biomarkers; "let the data decide, not my mind"The measure-then-act discipline is genuinely good, and some tracking is sensible. But more testing isn't free — it invites overdiagnosis and acting on noisy single readings (the same screening caution that applies to Attia).Mixed
DunedinPACE ~0.66 — "aging at two-thirds speed," "slowest measured"DunedinPACE is among the better-validated clocks and predicts mortality in cohorts — but no clock is an FDA-accepted surrogate endpoint, and lowering your score is not proven to extend life.Mixed
"I reduced my pace of aging by 31 years" / the age-reversal scoreboardA category error. Morgan Levine (who built the PhenoAge clock): a rate measured at a single timepoint can't be converted into "31 years," and clocks aren't denominated in years at all.Overstated
"I have the biological age of an 18-year-old [organ]"Cherry-picked surrogates: different clocks disagree with each other and are noisy. Predicts ≠ causes, and it's still n-of-1.Overstated

Risks & the results question

Two honest worries the marketing skates past: the protocol may be too aggressive in places, and as an uncontrolled experiment it can't actually show the exotic parts work.

Where to be careful
ConcernWhat's knownVerdict
~5–6% body fatBody fat vs mortality is J-shaped; very low body fat is linked to hormonal disruption, lower bone density, and higher mortality in older men. Critics suggest ~8–10% would be safer.Safety flag
~31–40% calorie deficit for his activity levelBeyond the ~25% range where CR's benefits cluster; chronic under-fueling risks mood, libido, bone, and nutrient status.Safety flag
Do the exotic parts actually work?Much of his improvement is almost certainly the basics — weight loss, clean diet, exercise, sleep — not the add-ons. As an n-of-1 with no control, the protocol simply can't separate them.Mixed
Heavy metals in Blueprint productsA 2024 California Prop 65 notice alleged lead/cadmium in some products; independent testing flagged heavy metals in the cocoa and matcha (categories that are industry-wide offenders), while the olive oil tested non-detect. Blueprint says it tests every product. Awkward for a purity-branded line — with real industry-wide context.Mixed

The business

Blueprint began as a personal protocol and is now a commercial supplement-and-food company — so nearly every recommendation has a product attached, and the protocol is the marketing.

Credibility

The fair summary: transparent and self-correcting, but non-clinical and surrogate-driven — with a storefront attached.

The pattern — how to use him

Copy the core, skip the scoreboard and the catalog. The boring foundation — whole-food diet, daily Zone 2 + strength, religiously regular sleep, mild (not extreme) calorie restriction — is the real, evidence-based product, and he runs the most disciplined version of it. Treat the epigenetic-clock scores, the mega-stack, and anything exotic (gene therapy, plasma) as an experiment with a sample size of one, and run them through surrogate vs hard outcomes, external validity (mouse → human), and is it strong enough to act? His best habit is the one worth stealing: when the data say an intervention does nothing, stop doing it.

For you

Of everyone on this desk, Johnson is the purest "measure, then act" case — which fits you, since you already run on biomarkers and a Life OS. The whole job here is taking his discipline without his excess.

His position → your call
His positionYour call
Whole-food diet, daily Zone 2 + strength, fixed-schedule sleepAdopt the discipline — you mostly do. The regular-sleep point is the cheapest high-value habit to lock in if it isn't already; it outranks most of your supplement queue for effort-to-payoff.
Mild caloric restrictionYou're already in a deliberate deficit on Zepbound — but his ~31–40% is a cautionary tale, not a target. Your ~1 lb/week with a >1.75 lb/wk stop-trigger is the disciplined version; don't out-Bryan Bryan.
Drive body fat to ~5–6%Hard no — and instructive. Your plan stops at 14–15% partly to protect your face and hormones. His very-low-fat physique is a downside to avoid, not a goal to chase.
Run a large supplement stackResist stack-creep. Your queue (creatine, psyllium, whey, D3…) is being re-derived one item at a time against your markers — that's the right model; his 100-pill maximalism is the anti-pattern.
Rapamycin / gene therapy / plasmaNot now / never offshore. Same call as your own filters and the Attia rapamycin row: animal-data bets, wrong trade mid-cut — and unregulated overseas gene therapy is a flat no.
Chase an epigenetic-clock "age reversal" scoreDon't buy in. Track the markers that actually drive your risk — ApoB, HOMA-IR, visceral fat — actionable numbers with hard-outcome links, not a noisy clock score you'd be tempted to optimize for its own sake.
"Measure everything; let the data decide"Take the good half. Disciplined tracking of decision-relevant markers is smart and very you. Measuring everything and acting on single noisy readings is overdiagnosis — let your filters gate what's worth measuring.

Check it yourself

Re-run any row through surrogate vs hard outcomes, the evidence ladder, external validity, observational studies, correlation vs causation, and is it strong enough to act?

Sources · Background & business: Fortune (Braintree/PayPal), the Blueprint protocol, finances & CEO (summarizing NYT) · Core: caloric restriction CALERIE (Nature Aging), sleep regularity (UK Biobank), early TRE · Rapamycin: his "I stopped rapamycin" post, mouse data (NIA ITP, Nature 2009), human PEARL trial · Young plasma: "no benefits detected" (Fortune), FDA warning · Gene therapy & clocks critique: Kaeberlein, Levine on the "31 years" claim, DunedinPACE validation (eLife) · Risks: low-body-fat & calorie-deficit critique (NOVOS) · Product heavy metals: California Prop 65 notice · Scientist skepticism: Northeastern News. An evidence assessment of public claims, not a personal judgment — and not medical advice; confirm anything touching your medications with your prescriber.

Learn · Shawon Chowdhury · a source scorecard, kept rough on purpose · not medical advice