MAGELLAN LONGEVITY

Thomas DeLauer's Claims, Checked Against the Papers

Ninety-one claims from fifteen videos — the largest sample in this series. He is not a physician or a dietitian and does not claim to be, and I expected this to be the weakest record here. It is not. The physiology is largely right. The supplement aisle is where it falls apart.

The short version

DeLauer's mechanistic claims — ketone biology, insulin and lipolysis, glycogen and water, contraction-mediated glucose uptake, gluconeogenesis being demand-driven rather than supply-driven — are mostly correct, sometimes more correct than what you hear from credentialed people. He accurately cites the sleep-restriction study showing dieters who sleep 5.5 hours lose 55% less fat than those sleeping 8.5. He gets the keto-adapted ultra-runner glycogen data right. He is right that around 90% of US adults are not in optimal metabolic health.

Two things break the record. First, precise timing claims about processes nobody can measure. Second, and much more consequential, supplement claims that are flatly reversed.

Three claims that are backwards

Claims where the evidence points the other way.
ClaimThe record
Krill oil has "a higher omega-3 content, but also a higher DHA content" than fish oil Krill oil capsules generally contain less EPA and DHA per capsule than fish oil. Krill's claim is the phospholipid delivery form, not the quantity. Reversed.
Soy isoflavone supplements increase estrogen-receptor breast cancers, "which is the worst form" Two errors in one sentence. Prospective cohort data associate soy intake with neutral-to-reduced breast cancer risk and better recurrence outcomes. And ER-positive disease is the most treatable common subtype, not the worst. This one could change a real decision for a woman after menopause.
"Yohimbine and rauwolscine are alpha-2 adrenergic receptor agonists" They are antagonists — blocking the receptor is the entire basis of the lipolytic effect. He then goes on to describe antagonism correctly, contradicting his own label mid-explanation.

He also dismisses the erythritol and xylitol cardiovascular findings as "very, very flawed… reverse causation." That is not a fair characterisation. The xylitol work included an interventional arm: healthy volunteers drank a xylitol-sweetened beverage and showed measurably enhanced platelet reactivity afterward, alongside prothrombotic effects in animal models DOI. You can argue about magnitude and dose. You cannot wave it away as an observational artifact.

Where I was wrong about him

Two corrections that go in his favour, and they matter more than another criticism would.

Calcium supplements. He says calcium supplements — as opposed to dietary calcium — raise cardiovascular risk, and our first pass marked that as overstated. Checking the source properly: a BMJ meta-analysis of randomised trials found calcium supplements without co-administered vitamin D associated with increased myocardial infarction, hazard ratio 1.31 in patient-level data and relative risk 1.27 at trial level, both statistically significant DOI. Later analyses are mixed and absolute risk is modest, but he is describing a genuine randomised-trial signal, and he correctly distinguishes supplemental from dietary calcium — a distinction many better-credentialed people blur.

Selenium. He cites the JNCI analysis showing that men with high baseline selenium who supplemented had a 91% increase in high-grade prostate cancer DOI. That is exactly what the paper reports. He is one of the few voices in this space getting selenium right — and notably, he gets it right in the same week's content that Dan Buettner gets it backwards.

The autophagy clock

His signature format is the precise timing chart: autophagy "starts to ramp up" at 16–20 hours; gut stem cells begin proliferating "around the 20-hour mark"; there is a "tipping point after 36 hours."

The problem is not that these numbers are wrong. It is that they cannot currently be right or wrong, because human autophagy cannot be measured non-invasively at hourly resolution in a living person. The figures come from rodent work and indirect markers. The underlying biology is real — the ketone/NLRP3 inflammasome mechanism he describes is solid, and prolonged fasting genuinely does drive stem-cell-based regeneration in the source paper DOI. What does not exist is the clock.

He also claims BDNF increases "three- to five-fold" after 48 hours of fasting. Human BDNF responses to fasting are small and inconsistent; the large multiples are rodent findings.

And one title claim — "36 hours of eating sardines is better than a 3-day fast for autophagy and visceral fat" — is never substantiated anywhere in the episode it headlines. No study compares those two things.

The protein study he inverts

He describes an AJCN study where "the low-protein group ate 441 more calories" and the high-protein group "lost about 13 lbs more fat."

The actual study raised protein from 15% to 30% of calories and found spontaneous intake fell by 441 kcal/day in the high-protein condition, producing 4.9 kg of weight loss and 3.7 kg of fat loss over 12 weeks DOI. So: the direction of the 441 is flipped, and 3.7 kg is 8 lb, not 13. The underlying protein-leverage point he is making is sound and well-supported. The numbers wrapped around it are not.

Similarly, his guest Jason Fung presents high IGF-1 as one-directionally raising mortality. The dose-response meta-analysis found a U-shape — both low and high IGF-1 associate with increased mortality DOI. The one-directional version is the version that supports a fasting narrative.

The pattern worth naming

DeLauer's channel runs heavy sponsor integration, and the weakest claims cluster in exactly the categories where products are sold: which magnesium form, why you need K2 with your D, methylcobalamin over cyanocobalamin, the C15:0 "first new deficiency in 75 years" framing that originates with the company selling C15:0. The omega-6 claim that seed oils "skyrocket inflammation" is contradicted by randomised trials in which linoleic acid does not raise inflammatory markers DOI.

To his credit, he has said this about himself. On another podcast he described "the echo chamber that I was living in" around keto and fasting, and conceded he had been seeing it "almost myopically." That is a more honest self-assessment than anyone else in this series has offered.

What I think this adds up to

If you watch him for how fasting and ketosis work, you are getting reasonable information from someone who reads the papers. If you watch him for what to buy, you are getting the least reliable content in this series, in the area where being wrong costs you money and occasionally more than money.

The specific thing I would not want repeated in my clinic: the soy claim. Telling postmenopausal women that soy raises their risk of the "worst" breast cancer is wrong on the epidemiology and wrong on the oncology, and it is the sort of thing that quietly removes a food from someone's diet for a decade.

How this was checked

Fifteen of DeLauer's most-viewed videos were transcribed and each factual claim pulled verbatim with a timestamp. Claims by guests — Dominic D'Agostino, Jason Fung — are labelled as theirs. One source video had unusable captions and was excluded; one is an auto-translation from Dutch and its quotes are marked as translations. Each citation was verified against the PubMed record, including retraction status: one paper cited in this corpus, on apple cider vinegar, has since been retracted. Article metadata retrieved from PubMed. Full methodology →

Evidence appraisal, not medical advice. No physician–patient relationship is created by reading this.