Health protocol: Huberman, Galpin and Attia, evidence-checked
Contents
There’s a lot of health and longevity advice out there. I picked three people I’ve been listening to for a while, Andrew Huberman, Andy Galpin and Peter Attia, because they mostly back up what they say with research. Not always, though, so I’ve checked their claims against other researchers too. I use this page to keep my own routine simple.
How to use this
One protocol built from Huberman, Galpin and Attia, with each claim checked against researchers outside their network. Each topic gives the action first, then where the sources differ, an evidence grade and links.
- Core voices: Andrew Huberman, Andy Galpin, Peter Attia.
- Check voices: Stuart Phillips and Martin Gibala (exercise, protein), Kevin Hall (nutrition). They publish the trials and reviews the core three cite, and they sit outside that podcast circle.
- Scope: built from each person’s distilled material (books, recap episodes, newsletters, recent AMAs), not every episode. Open the linked source before relying on a specific figure.
- Dating: positions move. Where a source has updated (Attia on fasting and Zone 2), the newer position is used and the old one goes in Didn’t hold up.
- Not medical advice: these are personal notes. Clinical items (blood markers, screening) are for a conversation with a GP, not self-directed.
| Grade | What it means |
|---|---|
| A | Randomised trials or meta-analyses in humans, broadly consistent |
| B | Human observational data, or small or short trials: direction probably right, size uncertain |
| C | Mechanism, animal work or a single study: plausible, unproven |
The one-page protocol
The core three agree on almost all of this; where they differ, the column on the right says which way the evidence points.
| Area | Do this | Grade | Where the core three differ |
|---|---|---|---|
| Strength | All major muscle groups, 2+ sessions a week, sets near failure; add heavy 3-5 rep work for strength | A | Dose only: 2 sessions is the minimum, 3-5 the optimiser’s version |
| Cardio | Weekly volume at a conversational pace, plus one hard interval session (e.g. 4 x 4 min) | A | Share of Zone 2: less than Attia’s old 80% if you train under about 3 hours a week |
| Protein | About 1.6 g/kg/day over 3-4 meals (128 g at 80 kg) | A | Core three say 2.2 g/kg; trials show gains flatten near 1.6 |
| Diet | Calories and protein first; mostly minimally processed food | A-B | Meal timing: Huberman treats it as a lever, trials do not |
| Sleep | Fixed wake time, about 8 h in bed, morning outdoor light, dark cool room | A-B | Huberman’s caffeine delay and sleep stack: drop them |
| Caffeine | Cut off about 8 hours before bed; morning timing is your call | B | Huberman’s 90-120 min delay is unsupported |
| Alcohol | Less is better; if drinking, cap the week and stop 3 h before bed | B | Ceiling: Huberman 2 drinks, Attia about 7, UK guideline 14 units |
| Heat and cold | Sauna if available; cold optional and never straight after lifting | B / C | How much cold matters: Attia found no lifespan evidence |
| Supplements | Creatine 3-5 g/day; vitamin D 10 mcg Oct-Mar (UK) | A | Huberman’s wider stacks are grade C |
| Clinical | Know BP, ApoB or non-HDL, HbA1c; test Lp(a) once | A | Attia’s targets are far below guidelines |
Strength training
Action: train every major muscle group at least twice a week, sets taken close to failure. For strength specifically, add heavy sets (roughly 80%+ of your one-rep max, 3 to 5 reps). Evidence: A.
| Source | Position |
|---|---|
| Galpin | Treats strength and muscle size as different adaptations. Strength: the “3 to 5” rule (3-5 days a week, 3-5 exercises, 3-5 reps, 3-5 sets, 3-5 minutes’ rest). Size: 5-30 reps per set, within 1-2 reps of failure. Progress load 3-5% a week, in 5-8 week blocks followed by a lighter week. |
| Huberman | Foundational Fitness Protocol: three resistance and three cardio sessions a week plus one rest day. Alternates monthly between 4-8 reps (3-4 sets, 2-4 min rest) and 8-15 reps (2-3 sets, about 90 s rest). Sessions of 50-60 minutes. |
| Attia | Strength is one of four pillars (Zone 2, VO2 max, strength, stability). Ranks VO2 max then strength as the strongest predictors of lifespan. Tracks grip strength and dead hang (target 2 minutes for men at 40). |
| ACSM 2026 position stand (Phillips co-author) | Built on 137 systematic reviews and 30,000+ participants. The biggest jump is from no training to any training. Minimum: all major muscle groups, 2+ days a week. Strength: heavier loads, 2-3 sets per exercise. Muscle growth tracks weekly volume; training to failure is not required. |
Where they agree: resistance training is non-negotiable, progressive overload drives results, and effort per set matters more than the exact rep scheme.
Where they differ: dose. Galpin and Huberman give full programmes of 3-6 sessions a week. The ACSM review (McMaster summary) says two sessions of any style captures most of the benefit for a non-athlete. Not a contradiction: theirs is the optimiser’s version, ACSM’s is the minimum effective dose.
Cardio: Zone 2, VO2 max and intervals
Action: accumulate weekly aerobic volume at a conversational pace, plus one session a week of hard intervals (e.g. 4 x 4 minutes hard, 4 minutes easy). With less than about 3 hours a week, shift more of the time to higher intensity. Evidence: A that training raises fitness; B that fitness itself extends life (large observational data).
| Source | Position |
|---|---|
| Attia, Outlive era | At least 3 hours of Zone 2 a week (e.g. 4 x 45 min), plus one VO2 max session: 4 minutes at the hardest sustainable pace, 4 minutes recovery, repeated. Roughly 80% Zone 2, 20% hard. |
| Attia, AMA #79, Jan 2026 | Total volume is the main driver of adaptation, provided intensity is at least Zone 2. Zone 2’s value is that it adds volume at low physical cost. The talk test (can speak in full sentences, would rather not) is a workable guide. |
| Huberman | Three cardio sessions a week in the Foundational Fitness Protocol, including one long Zone 2 session of 60-75 minutes and one high-intensity session. |
| Galpin | Splits fitness into nine trainable adaptations, with several kinds of endurance among them; you pick the ones that match your goal rather than defaulting to one zone. |
| Storoschuk, Gibala et al. 2025 | Zone 2 advice rests mainly on observational data from elite endurance athletes doing very high volumes. Evidence does not support Zone 2 as the best intensity for mitochondria; training above Zone 2 matters most when total volume is low. |
Why fitness matters: in 122,007 adults given treadmill tests, the least fit quarter had about 5 times the death rate of the fittest 2.3% over a median 8.4 years (adjusted HR 5.04), with no observed upper limit. Observational, so some of that is genetics and health status.
Where they differ: how much of the week should be Zone 2. Attia’s earlier 80/20 split suits people training 4+ hours a week. Gibala’s review, and Attia’s own 2026 framing, put total volume first and intensity ahead of Zone 2 for anyone short on time.
Protein
Action: aim for about 1.6 g per kg of bodyweight a day, spread over 3-4 meals. Going up to 2.2 g/kg is optional, not required. Total daily intake matters far more than timing. Evidence: A.
Worked example at 80 kg: 1.6 g/kg is 128 g a day; 2.2 g/kg (1 g per lb) is 176 g.
| Source | Position |
|---|---|
| Attia | Long-standing heuristic of 1 g per lb (2.2 g/kg). Also frames it as 1.6-2 g/kg, spaced through the day, with a minimum of 1.6 g/kg rising with age. |
| Galpin | About 1 g per lb. Once total intake is high enough, timing, type and quality matter much less; timing matters mainly for athletes. |
| Huberman | Echoes 1 g per lb. Adds eating protein early in the day (before 10am) for muscle, citing a body-clock gene (BMAL). That rests mainly on animal work: C. |
| Phillips: Morton et al. 2018 | 49 trials, 1,863 people lifting weights: protein beyond about 1.6 g/kg/day produced no further gain in lean mass. |
| Phillips: Nunes et al. 2022 | 74 randomised trials: extra protein gives small additional gains in lean mass and leg strength. Under-65s benefited at 1.6 g/kg and above; over-65s at 1.2-1.59 g/kg. |
Where they agree: the official 0.8 g/kg allowance is too low for anyone who trains or is ageing.
Where they differ: the ceiling. The core three round up to 1 g per lb for simplicity and margin. The trial data says gains flatten around 1.6 g/kg, so the last 48 g a day (at 80 kg) buys little extra muscle. Treat 2.2 g/kg as a ceiling for heavy training, dieting or older age, not as the default.
Wider nutrition and fasting
Action: manage total calories and hit your protein target; build most meals from minimally processed food, since that cuts intake without counting. Use a shorter eating window only if it helps you eat less. Skip multi-day fasts if you want to keep muscle. Evidence: A that time-restricted eating adds nothing beyond the calorie cut; B on ultra-processed food (tightly controlled but small, short trials).
| Source | Position |
|---|---|
| Attia | Follows no named diet; now optimises mainly for energy balance and protein. Concludes the benefits of time-restricted eating come from the calorie deficit, and narrow windows make it hard to eat enough protein. Stopped quarterly 7-10 day water fasts after losing about 20 lb of muscle over three years. |
| Galpin | Calls fasting overrated. Fine if it helps you get quality food and manage calories; otherwise no reason to do it. |
| Huberman | Draws on Satchin Panda’s work: when you eat matters as much as what you eat; keep an 8-12 hour window, skip food for the first hour after waking. Much of the case rests on mouse studies: C. |
| Hall: ultra-processed food trial, 2019 | 20 inpatients, 2 weeks on each diet, matched for calories, sugar, fat and fibre on offer. On the ultra-processed diet they ate about 500 kcal a day more and gained weight. |
| Hall: low-fat vs keto, 2021 | Same design. A low-fat, plant-based diet led people to eat 550-700 kcal a day less than an animal-based keto diet, despite higher insulin. Contradicts the claim that carbs drive overeating via insulin. |
| Liu et al., NEJM 2022 | 139 adults with obesity, 12 months: an 8am-4pm eating window plus calorie restriction was no better than calorie restriction alone for weight, body fat or metabolic markers. |
Where they differ: Huberman’s older material treats meal timing as a lever in its own right. Attia has moved away from that, Galpin never held it, and the 12-month human trial found no added effect. Keep the window for convenience, not for metabolism.
For the full picture, Hall’s book Food Intelligence (2025, with Julia Belluz) is the single best reference.
Sleep and light
Action: fix your wake time first, then give yourself about 8 hours in bed. Get outdoor light within an hour of waking. Keep the bedroom dark and cool, stop food and alcohol about 3 hours before bed, and stop caffeine roughly 8 hours before bed. Evidence: A-B for consistency, light and caffeine cut-off; C for supplement stacks.
| Source | Position |
|---|---|
| Attia | Consistent sleep-wake times, about 8 hours in bed (his own is 10pm-6am, giving roughly 7.5 hours’ sleep), dark cold room, no stimulating content for 2 hours and no food or alcohol for 3 hours before bed. Rates wake-time consistency as the biggest single lever. |
| Huberman | Bright light, ideally sunlight, within 30-60 minutes of waking: about 5 minutes on a clear day, 10 when cloudy, 20-30 on dense overcast. Also: delay caffeine 90-120 minutes after waking, and a sleep stack of magnesium threonate, apigenin and theanine. |
| Antonio et al. 2024 (caffeine review) | Found no basis for the claim that delaying morning caffeine prevents an afternoon crash. Caffeine late in the day does cut sleep: one meta-analysis put the loss at about 45 minutes a night. |
Where they agree: consistency, morning light, a dark cool room, and keeping alcohol and late meals away from bedtime. Attia and Huberman are closely aligned here, and this is the area nearest Huberman’s own research.
Where they differ: Huberman’s add-ons. The 90-minute caffeine delay does not hold up, and the supplement stack has thin human evidence. Neither is needed for the core protocol.
UK note: on a grey UK winter morning you are in the 20-30 minute bracket most days; a walk or commute outdoors covers it.
Alcohol
Action: treat alcohol as a pleasure with a cost, not a health food. Less is better, zero is best. If you drink, cap the week, spread it out, and finish at least 3 hours before bed. Evidence: B (large observational data, consistent in direction).
| Source | Weekly ceiling | In UK units | Notes |
|---|---|---|---|
| Huberman | 0 best; 1-2 drinks “probably fine” | about 2-4 | Cites grey matter thinning above 2 drinks a week, cancer risk and sleep disruption. |
| Attia | about 7 drinks | about 12 | No more than 2 a day, no more than 3 days a week, none within 3 hours of bed. Says alcohol has no health benefit at any dose. |
| UK Chief Medical Officers | 14 units | 14 | Spread over 3+ days. States there is no level of regular drinking that is completely safe. |
Units: a US “drink” (14 g alcohol) is about 1.75 UK units (8 g). A pint of 4% beer is 2.3 units.
Where they agree: no health benefit, real harms, keep it away from sleep.
Where they differ: only the ceiling. Huberman’s is the strictest; Attia’s sits just under the UK guideline.
Cold and heat exposure
Action: if you have access to a sauna, it is the better-supported of the two: several 15-20 minute sessions a week. Cold is optional, for alertness or mood, and should be kept well away from lifting sessions. Neither replaces exercise. Evidence: B for sauna (observational); A that cold straight after lifting blunts muscle gain; C for cold extending life.
| Source | Position |
|---|---|
| Huberman | Uses Susanna Søberg’s figures: about 11 minutes of cold a week (1-2 minutes a session, 2-3 sessions) and 57 minutes of sauna (10-15 minutes a session), ending on cold. Avoid ice baths for 6-8 hours after lifting, or take them before training. |
| Galpin | Cold after training blunts the muscle-building response; separate it from hypertrophy sessions. |
| Attia | After reviewing the literature, found no compelling evidence that cold exposure extends life; good evidence it blunts hypertrophy straight after lifting. More positive on sauna, while noting the data is non-randomised. |
| Laukkanen et al. 2015 | 2,315 Finnish men aged 42-60, followed about 20 years. Versus once a week, 4-7 sauna sessions a week was linked to 40% lower all-cause mortality and 63% lower sudden cardiac death. |
| 2024 meta-analysis on cold and muscle | Cold plunges after resistance training can attenuate muscle growth. |
Where they agree: no cold plunge straight after lifting.
Where they differ: how much to make of cold. Huberman presents it as a tool for metabolism and mood; Attia’s review found no lifespan evidence. The Søberg 11-minute figure comes from observing winter swimmers, not a trial. The sauna data is observational, in Finnish men, so frequent sauna users may simply be healthier and more relaxed in other ways.
Supplements
Action: creatine monohydrate, 3-5 g a day, is the one supplement all three back and the evidence supports. In the UK, add vitamin D at 10 micrograms (400 IU) a day from October to March, per NHS advice. Treat everything else as optional and unproven. Evidence: A for creatine; A (null) for vitamin D and fish oil preventing heart disease or cancer in healthy adults; C for most of Huberman’s stacks.
| Supplement | Who backs it | What the evidence says | Grade |
|---|---|---|---|
| Creatine monohydrate, 3-5 g/day | All three: Galpin calls it the most effective supplement for strength and size; Huberman 5 g; Attia 5 g | Consistent gains in strength and lean mass alongside training; a 2024 meta-analysis suggests memory and processing-speed benefits | A (muscle), B (cognition) |
| Vitamin D | Attia 5,000 IU, while calling most vitamin D studies poor | VITAL (25,871 adults, 2,000 IU, about 5 years): no reduction in heart attacks, strokes or cancer. NHS: 10 micrograms in autumn and winter for bone and muscle health | A (null for heart and cancer) |
| Omega-3 (EPA/DHA) | Attia: about 2 g EPA + 1.5 g DHA, dosed to a blood omega-3 level | VITAL’s 1 g/day arm showed no significant drop in major cardiovascular events in healthy adults; higher-dose EPA has helped in high-risk patients already on statins | B |
| Testosterone herbs (tongkat ali, fadogia agrestis) | Huberman | Tongkat has small trials, mostly in men with low testosterone; fadogia evidence is largely animal | C |
| Sleep stack (magnesium threonate, apigenin, theanine) | Huberman | Thin human data for these doses and combinations | C |
| Pre-workout (alpha-GPC, L-tyrosine) | Huberman | Little outcome data in trained adults | C |
Where they differ: volume and conflicts. Huberman recommends far more products than the other two and has supplement sponsors; McGill’s Office for Science and Society has criticised him for presenting thin evidence as settled. Attia’s list is shorter and he says himself that the vitamin D data is weak. Galpin mostly stops at creatine and protein.
Clinical markers and screening
Action: know your blood pressure, cholesterol (ideally ApoB, or at least non-HDL), blood sugar and Lp(a) once. Discuss the results with a GP, especially with any family history of early heart disease. Evidence: A that ApoB-carrying particles cause atherosclerosis and lower is better; expert opinion for Attia’s very low targets and early drug treatment.
This is Attia’s territory. Huberman and Galpin are not clinicians and rarely go here, so there is little to reconcile; the check is Attia against mainstream guidelines.
| Topic | Attia | Mainstream guidance |
|---|---|---|
| ApoB / LDL | ApoB is the key metric: it counts every particle that can lodge in an artery wall. Proposes ApoB of 20-30 mg/dL, a ceiling of 60, and aggressive treatment from the late 30s or early 40s for anyone above the 20th percentile. April 2026: no safe gamble with high LDL. | Guidelines still treat LDL-C and non-HDL-C as the main targets, with ApoB as an add-on. Treatment thresholds depend on overall risk, not a universal low target. |
| Lp(a) | Measure it; it is genetic and largely fixed. | The European Atherosclerosis Society (2022) and ESC say measure at least once in every adult. HEART UK (2019) targets people with personal or family history of early heart disease. |
| Fitness testing | Tracks VO2 max as a core longevity metric. | Cardiorespiratory fitness strongly predicts mortality (see Cardio); not part of routine NHS checks. |
| Routine checks | Runs a private concierge practice, Early Medical, built around his “Medicine 3.0” approach. | NHS Health Check: every 5 years for ages 40-74 without existing conditions; covers blood pressure, cholesterol and lifestyle risk. |
Where they differ: Attia treats earlier and to far lower targets than guidelines, and practises in a private concierge setting. His case is that atherosclerosis builds over decades, so lifetime exposure matters. The open question is long-term outcome data for very low ApoB in healthy people, which is still limited. Under 40 in the UK, a lipid panel and Lp(a) test usually means asking your GP or paying privately.
Didn’t hold up
Claims from this circle that were reversed or weakened. If you meet one of these in an older episode, skip it.
| Claim | Who made it | What changed |
|---|---|---|
| Regular prolonged fasts (7-10 days) for longevity | Attia, pre-2021 | He stopped after losing about 20 lb of muscle over three years. |
| Eating window has metabolic benefits beyond eating less | Huberman (via Panda); Attia earlier | 12-month NEJM trial found no added benefit; Attia now attributes it to the calorie deficit. |
| 80% of cardio in Zone 2; Zone 2 is the best intensity for mitochondria | Attia, Outlive era | Gibala and colleagues, 2025: not supported; Attia’s 2026 AMA puts total volume first. |
| Delay caffeine 90-120 minutes after waking to avoid a crash | Huberman | 2024 review found no basis for it. |
| Ice bath after lifting for recovery | Common practice | Blunts muscle growth; all three now say keep it away from lifting. |
| Cold exposure as a longevity tool | Huberman’s framing | Attia’s review found no compelling lifespan evidence. |
| Metformin for longevity in non-diabetics | Popular in longevity circles | Attia presented a paper casting doubt on it in his journal club with Huberman. |
| Vitamin D and fish oil prevent heart disease and cancer | Widely implied | VITAL: no effect in healthy adults. |
| More protein than 1.6 g/kg builds meaningfully more muscle | 1 g/lb heuristic | Gains flatten near 1.6 g/kg; extra is margin, not a muscle lever. |
Core source list
About 3 books and a dozen episodes.
| Person | Start with | Then | Note |
|---|---|---|---|
| Huberman | Protocols (book, Sept 2026): read the sleep, light and exercise chapters | Free newsletter PDFs: Toolkit for Sleep, Foundational Fitness Protocol; newsletter sign-up | Strongest on sleep and light. Treat supplement content as grade C; he has supplement sponsors. |
| Galpin | Huberman Lab guest series, episode 2: strength and muscle (2023, part of six) | Perform recap episodes: “Fundamentals of Training for Muscle, Strength & Longevity” (22 Jul 2026) and “Fundamentals of Nutrition, Hydration & Performance Fueling” (5 Aug 2026) | Best single voice on training design. |
| Attia | Outlive (2023) for the framework, with this page’s updates on fasting and Zone 2 | AMA #79 on cardio (Jan 2026); #229 on ApoB | Strongest on clinical risk. |
| Phillips and Gibala | Exercise Science podcast: “Strength Training for Everybody!” (26 May 2026), “Much Ado About Zone 2” (1 Jun 2026), “How Much Protein Do You Need?” (21 Jul 2026) | ACSM 2026 position stand summary | The check on training and protein claims. Gibala discloses equity in Longevity League Ltd. |
| Hall | Food Intelligence (2025, with Julia Belluz) | His 2019 ultra-processed food trial | The check on nutrition claims. |
Most of these voices have commercial ties: sponsors, equity, clinics or books. That is why each claim on this page is graded on its evidence, not on who said it.