Founding price: Foundation is $69 for the first 250 buyers or until 31 Dec 2026, whichever comes first, then $79. Free creatine tub with Foundation, code FIRST100, one per customer, first 100 Foundation orders.$69 founding price. Free extra creatine tub, one per customer, first 100 Foundation orders.
Public record

What we've said

Every post we have published, quoted as posted, with its sources and caveats, and every correction. Every science summary is listed below with its current grade.

How this works: we do not edit past posts here. If something we said needs a caveat or a fix, we add it next to the post and in Corrections, with a date. Grade changes are also in the grade change log. Last updated: .

Posts

20:36 CEST · X quote post: @hubermanlab on sweet taste

Dr. Scott Kanoski on @hubermanlab: cut very sweet foods and they start tasting too sweet. A 2016 trial backs the taste shift, but a 180-person, 6-month trial found liking for sweets did not change. Both studies: everstackhealth.com/en/university/metabolic-health twitter.com/hubermanlab/status/2109790633884066076

View on X · Related page: /university/metabolic-health

Sources
Caveats
The 2016 trial was small (13 completed the key tests) and found that liking for sweet foods did not change either; only perceived sweetness shifted.

15:36 CEST · X thread: Vitamin D: test results, units and dose

1/ Got a vitamin D result and not sure what it means? We read the guidelines and dose trials. The short version: under 30 nmol/L (12 ng/mL) is a deficiency risk, 50 nmol/L (20 ng/mL) or more is adequate for most people, and higher is not clearly better.

2/ Units trip people up. 1 ng/mL equals 2.5 nmol/L. A "30" in ng/mL is 75 in nmol/L. Check which one your lab used before you change anything.

3/ How much does D3 move it? In a controlled winter trial, about 0.7 nmol/L per extra 40 IU a day, roughly 7 ng/mL per 1,000 IU (Heaney 2003). The rise flattens at high doses, and people with obesity ran about 20 nmol/L lower (Ekwaru 2014).

4/ More is not better. In a 3-year trial, 4,000 and 10,000 IU a day lowered forearm bone density more than 400 IU, with more mild high calcium at 10,000 (Burt 2019). One 500,000 IU dose a year increased falls and fractures (Sanders 2010).

5/ Daily beats big boluses. In a 25-trial analysis of immune outcomes, daily or weekly dosing helped and large one-off doses did not. People starting very low gained the most (Martineau 2017).

6/ Our take: test if you have a reason to run low, count every source toward the 4,000 IU adult upper limit, and retest after about 3 months at the same lab. Not medical advice. Limits, funding and DOIs: everstackhealth.com/en/science/vitamin-d-testing-and-dose

View on X · Related page: /science/vitamin-d-testing-and-dose · Grade BB Good human data, but mixed outcomes or surrogate markers. How we grade →

Sources
Caveats
Not medical advice (stated in post 6). Cutoffs differ between guideline bodies. Post 5 was reworded before posting to describe the immune-outcome analysis rather than say vitamin D "cut respiratory infections".

10:36 CEST · X quote post: @bryan_johnson on creatine, water and lean mass

@bryan_johnson raises a fair point: DEXA counts water as lean mass, so short creatine trials can overstate muscle. The strength data still hold across many RCTs. Both sides, cited: everstackhealth.com/en/science/creatine twitter.com/bryan_johnson/status/2107516394418147828

View on X · Related page: /science/creatine · Grade AA Multiple randomized trials or meta-analyses, consistent results. How we grade →

Sources
Caveats
We sell creatine. The grade A is for strength and lean mass alongside resistance training, not for creatine alone.

20:31 CEST · X thread: Quercetin and inflammatory markers

1/ We published a plain-English research summary on quercetin and inflammatory markers.

Grade C. Our claim is structure/function only: supports a healthy inflammatory response. It is not meant to diagnose or treat any condition.

2/ Mohammadi-Sartang 2017 (meta-analysis of 7 RCTs): average CRP drop of 0.33 mg/L at 500 mg/day or more. A small marker change, not a clinical outcome.

3/ Honest null: Knab 2011 (1,002 adults, 500 or 1,000 mg for 12 weeks). CRP and cytokines essentially unchanged vs placebo. Heinz 2010: no drop in upper respiratory infections in the same cohort.

4/ Phytosome absorption (Riva 2019): plasma levels up to ~20x higher than plain quercetin in 12 volunteers. Industry-tied (ingredient maker). It measures absorption, not inflammation outcomes.

5/ Interactions we list on the page: warfarin (case report), P-gp/fexofenadine (+55% exposure), CYP3A4 lab inhibition, quinolones (gyrase binding in lab). Ask a pharmacist if you take prescription drugs.

6/ Full summary with every DOI:

everstackhealth.com/en/science/quercetin-inflammatory-markers

Waitlist (no payment today): everstackhealth.com/en/products/quercetin

View on X · Related page: /science/quercetin-inflammatory-markers · Grade CC Small or few human trials. Promising, not proven. How we grade →

Sources
Caveats
The CRP change is a marker, not a health outcome, and the largest trial found no change. The absorption study was tied to the ingredient maker. Post 6 links the Quercetin Phytosome product page, which is waitlist only (no payment). That product was not tested in these trials, and nothing in the thread shows that it changes inflammation.

15:36 CEST · X quote post: @foundmyfitness on coffee and visceral fat

Big observational signal from @foundmyfitness: heavier coffee drinkers carried less visceral fat and more lean mass. Observational, so it shows association, not cause. Our cited visceral fat explainer covers what trials show actually moves it: everstackhealth.com/en/university/visceral-fat twitter.com/foundmyfitness/status/2103167999611985960

View on X · Related page: /university/visceral-fat

Sources
No source linked in the post.
Caveats
Observational association, as the post says. Sources are on the linked explainer.

10:30 CEST · X thread: Berberine and glucose metabolism

We published a plain-English research summary on berberine and glucose metabolism.

Grade C, on purpose: meta-analyses show a metabolic signal, but most trials were small and low quality. Drug-interaction risk is real (CYP enzymes).

1/2

Read it, including limits and what it does not mean: everstackhealth.com/en/science/berberine-glucose-metabolism

Sources include Dong 2012, Lan 2015, Lee 2006, Yin 2008, Guo 2012 (DOIs on the page).

2/2

View on X · Related page: /science/berberine-glucose-metabolism · Grade CC Small or few human trials. Promising, not proven. How we grade →

Sources
Caveats
Lee 2006 is cell and rodent research. On 18 October 2026 the independent grader graded the popular claim "lower blood sugar" B, based on a 2023 meta-analysis. That is a claim people hear, not one we make; the grade for the whole summary stays C. Berberine interacts with drugs processed by CYP enzymes, including cyclosporine.

22:39 CEST · X reply: @foundmyfitness on blueberries and blood-vessel function

@foundmyfitness The dose comparison is the most useful part: one cup improved FMD where half a cup or placebo did not. The gain came from adults with metabolic syndrome over six months, so it is a signal for that group, not yet proof for healthy adults.

View on X · Related page: /university/metabolic-health

Sources
Caveats
The result came from adults with metabolic syndrome, so it does not show the same effect in healthy adults (stated in the reply).

22:35 CEST · X reply: @davidasinclair on menopause and hormone therapy

@davidasinclair Worth reading the HRT part closely: the paper says the step at the final period looks markedly attenuated in women on hormone therapy. That is an association in cross-sectional lab data, not a trial showing delayed change.

View on X

Sources
No source linked in the post.
Caveats
Refers to the paper in the original post; we did not link it. The reply is itself a caveat: cross-sectional data show an association, not a delayed change.

22:31 CEST · X reply: @bryan_johnson on jet lag and light timing

@bryan_johnson Which markers moved most? Circadian research points to light timing as the strongest lever after a time-zone shift: morning light after flying east, evening light after flying west. Meal timing is a secondary cue.

View on X

Sources
No source linked in the post.
Caveats
No source was linked in the reply. Light timing as the main cue for shifting the body clock is standard circadian research, but we have no published summary on it yet.

20:33 CEST · X quote post: @bryan_johnson on weekly training basics

@bryan_johnson lists the weekly basics: heavy lifting, zone 2, HIIT, sleep. The research behind lifting and VO2 max for lifespan is strong. Our cited training page: everstackhealth.com/en/training twitter.com/bryan_johnson/status/2107147209959727153

View on X · Related page: /training

Sources
Caveats
We called the lifespan research behind lifting and VO2 max "strong". More precisely: the link between fitness and lower mortality is large and consistent in observational cohorts (our summary grades it B), but trials have not settled whether improving fitness lowers mortality. Quoting a public figure is not an endorsement either way.
Correction
Wording clarified here on 18 October 2026: "strong" refers to observational association (grade B), not proof that training extends life.

16:54 CEST · X thread: L-theanine (Spanish)

Posted in Spanish

1/2 La L-teanina, el aminoácido del té, se vende para la calma sin somnolencia. En adultos sanos, 200 mg al día redujeron el estrés y mejoraron el sueño autoevaluado en un ensayo pequeño financiado por la industria.

2/2 En personas con ansiedad generalizada no superó al placebo. Grado C. Fuentes y DOIs: everstackhealth.com/es/science/l-theanine-stress-sleep-trials

View on X · Related page: /science/l-theanine-stress-sleep-trials · Grade CC Small or few human trials. Promising, not proven. How we grade →

Sources
Caveats
Same evidence and limits as the English thread above.

16:52 CEST · X post: Magnesium L-threonate (Spanish)

Posted in Spanish

El L-treonato de magnesio se vende como "el magnesio del cerebro". Hay tres ensayos en humanos, pequeños, cortos y todos con vínculos con la industria. Nuestra nota: C. Resumen con DOIs: everstackhealth.com/es/science/magnesium-l-threonate-human-trials

View on X · Related page: /science/magnesium-l-threonate-human-trials · Grade CC Small or few human trials. Promising, not proven. How we grade →

Sources
Caveats
Same evidence and limits as the English thread above.

16:51 CEST · X thread: L-theanine: stress and sleep trials, including the null one

1/ L-theanine is the "calm without drowsiness" amino acid from tea. We read the human trials, including the one that did not work. Here is what holds up.

2/ Best case: a 4-week crossover in 30 healthy adults. 200 mg a day lowered self-rated stress, anxiety and sleep-quality scores versus placebo. Funded by an L-theanine ingredient maker. (Hidese 2019, doi.org/10.3390/nu11102362)

3/ In a lab stress test with 12 people, it blunted the heart rate rise during mental arithmetic. Small, single session. (Kimura 2007)

4/ The honest null: in 46 people with generalized anxiety disorder, 450 to 900 mg a day did not beat placebo for anxiety or insomnia. The only publicly funded trial we found was the most negative. (Sarris 2019)

5/ Our read: grade C. A consistent small signal for everyday stress and self-rated sleep in healthy adults. Not a treatment for anxiety or insomnia. Avoid with sedatives, in pregnancy, and for kids.

6/ Full summary with every DOI and the funding ties: everstackhealth.com/en/science/l-theanine-stress-sleep-trials

View on X · Related page: /science/l-theanine-stress-sleep-trials · Grade CC Small or few human trials. Promising, not proven. How we grade →

Sources
Caveats
The positive trials are small and the best one was funded by an ingredient maker. The safety advice in post 5 (sedatives, pregnancy, children) is caution, not a trial result.

16:49 CEST · X thread: Magnesium L-threonate: the three human trials

1/ Magnesium L-threonate gets sold as "the brain magnesium." We read every human trial we could find. There are three.

2/ 80 adults with sleep complaints, 1 g a day for 21 days: sleep quality held steady while placebo declined, with better ring-measured deep and REM scores. Three of the six authors work at an ingredient supplier.

3/ 44 adults aged 50 to 70, 12 weeks: a composite cognition score improved vs placebo. Run by the product developer. Sleep effects could not be judged because placebo responses were strong.

4/ 109 healthy adults, 30 days: memory scores improved, but the capsule also had phosphatidylserine and vitamins C and D, and the ingredient maker paid for it.

5/ Our grade: C. Promising, small, short, never independently replicated. Our only claim: supports relaxation and restful sleep. Full summary with DOIs: everstackhealth.com/en/science/magnesium-l-threonate-human-trials

View on X · Related page: /science/magnesium-l-threonate-human-trials · Grade CC Small or few human trials. Promising, not proven. How we grade →

Sources
Caveats
The 2022 trial (post 4) tested a capsule that also contained phosphatidylserine and vitamins C and D, so it cannot show what magnesium L-threonate does alone. All three trials have industry ties. Post 5 names the only claim we would use for this ingredient; we do not sell magnesium L-threonate.

Corrections

  • · Grade change: NR ratings card: raising blood NAD+, C to B. Decided by the independent grader. Human trials consistently raise blood NAD+. The NR add-on label stays C, because its claim mentions cellular NAD+ and trials did not show a rise in muscle. grade change log
  • · Grade change: Omega-3: a higher Omega-3 Index and longer life, C to D. Decided by the independent grader. The longevity link is observational, and large trials of omega-3 did not lower deaths. grade change log
  • · Wording: Training quote post of 2026-10-15. We called the lifespan research behind lifting and VO2 max "strong". It is a large, consistent observational association (grade B); trials have not settled cause. The post stays up as written. View on X
  • · Grade change: Berberine (Metabolic add-on), B to C. Most trials were small, short and low quality, and the drug-interaction risk is real. grade change log
  • · Reworded before posting: Vitamin D thread, post 5. Changed from "cut respiratory infections" to a description of the immune-outcome analysis before it was posted. View on X
  • · Reworded before posting: Quercetin thread, post 1. The disclaimer was reworded to "not meant to diagnose or treat any condition" before it was posted. View on X

Science summaries