Tom Dayspring on Brain Cholesterol, APOE4, and What Lipid Drugs Do (or Don't Do) to the Brain

This is our independent, journal-club summary of Peter Attia’s conversation with lipidologist Tom Dayspring on brain cholesterol and what lipid drugs do (or do not do) to the brain (Episode 395). We distill what was said, separate solid evidence from opinion, and flag where it gets murky. You can check every point against the episode yourself.

Why this one matters for carriers

If you carry APOE4, you have probably worried at some point that lowering your LDL “too far” might starve your brain. This conversation takes that fear apart piece by piece, and the central message is reassuring: the brain runs its own, almost entirely separate cholesterol economy. What floats in your blood has very little to do with what your neurons have to work with.

The episode also walks through why APOE4 specifically causes trouble inside the brain, and it ends with a genuinely interesting (but still early) story about a new drug class and APOE4 carriers.

The core ideas, plainly

The brain and the body keep separate books. ApoB particles (LDL and its relatives), which carry most of the cholesterol in your blood, are too big to cross the blood-brain barrier. So the brain makes its own cholesterol on site. Dayspring’s framing: how much cholesterol is stored in the brain “has zero to do with what is floating in the plasma.” This is established physiology, not opinion.

Who makes what inside the brain. Dayspring described a division of labor between two brain cell types. Oligodendrocytes produce roughly 70% of brain cholesterol, mostly to build the myelin that insulates the wiring between neurons. Astrocytes are the cells that wrap cholesterol in ApoE and ship it out to neurons, because once the brain reaches adult size (around age 10) neurons largely stop making their own cholesterol to save energy. So the high-volume myelin producers and the delivery service are not the same cell.

The numbers on brain cholesterol. Dayspring put the brain’s cholesterol content at roughly 20 to 25 grams, against about 3 to 5 grams in the liver, out of roughly 140 grams total in the body. So the brain holds something like 20 times more than the liver. The brain also hoards it: the half-life of brain cholesterol was given as about five years, versus a few days in the rest of the body.

Children prove the point. A two-year-old’s brain is growing faster than it ever will again, yet Dayspring noted a toddler’s measured LDL cholesterol might be around 30 mg/dL. Low circulating LDL and a furiously building brain coexist just fine, because the brain is manufacturing its own supply. This is the cleanest argument against the “low LDL starves the brain” worry.

How APOE4 causes harm in the brain. Inside the brain, the cholesterol carrier is ApoE, not ApoB. Astrocytes build the ApoE-wrapped particles and ship cholesterol to neurons. If you make the E4 version of the protein, those particles bind neuron receptors less effectively, cholesterol delivery into the neuron is disrupted, and (per Dayspring’s account) the resulting imbalance in the neuron’s membrane cholesterol tips amyloid precursor protein toward the more toxic amyloid-beta 42. This is a coherent mechanistic story, but treat the cholesterol-to-amyloid chain as the speaker’s synthesis of the biology, not a settled, proven pathway.

The APOE4 risk numbers, in context. Using E3/E3 as the reference, the episode cited roughly 2x (sometimes up to 3x) higher Alzheimer’s risk for one copy of E4, and about 8 to 12-fold for two copies. Notably, Attia pointed out that the old “20 to 25 times” figure for E4/E4 has come down considerably. Population frequencies he gave: about 55% E3/E3, 20 to 25% E3/E4, and 1 to 2% E4/E4. These are relative multiples. The speaker did not put an absolute baseline risk on them in this conversation, and the multiples sit on top of an age-dependent baseline. Most carriers never develop the disease.

Statins and the brain. This is the part most carriers will care about. Both lipophilic (fat-soluble) and hydrophilic (water-soluble) statins reach the brain at steady state, so Dayspring argued you do not need to pick one based on solubility for brain reasons. On outcomes, the hosts walked through their own review of the randomized trial literature: Attia recalled an AMA in which every study he looked at for mild cognitive impairment, Alzheimer’s, or dementia showed neutrality or improvement, and Dayspring said a few trials show reduced incidence, with none showing harm. Treat this as the hosts’ reading of the trial evidence rather than a flat statement of settled fact. They flagged the important caveat themselves: in these trials, dementia was a secondary outcome, not the main question the trial was built to answer. So the signal is encouraging and consistent, but it is not the same as a trial designed to test it.

Two biomarkers worth knowing about. Plasma desmosterol tracks brain cholesterol production and correlates highly with spinal-fluid levels; a study cited found low desmosterol associated with higher cognitive impairment. By contrast, Dayspring noted, the other production-marker sterol (lithosterol) mostly reflects cells outside the brain, about 95% of it, so it is not a useful brain marker. And 24S-hydroxycholesterol rises in plasma when the brain is dumping excess cholesterol, so it can serve as a brain-health marker. The honest practical note from the episode: desmosterol is measurable commercially, but 24S-hydroxycholesterol largely is not, outside research.

The carrier takeaways

• The fear that lowering LDL harms the brain is not supported by the physiology or by the statin trials discussed. If anything, the trial signal the hosts reviewed leans neutral-to-protective for cognition.
• The reason APOE4 matters for the brain is local: it is about how the E4 protein moves cholesterol inside the brain, not about your blood LDL number.
• The over-suppression worry is real but speculative. Dayspring offered the hypothesis that statin “brain fog” in a minority of people reflects too-rapid suppression of brain cholesterol production, which reverses when the drug stops. He was explicit that no trial has tested this.

Fair warning

A few places where enthusiasm runs ahead of the evidence, and where interests are worth naming:

• Ezetimibe for the brain rests on a breakdown product, ezetimibe glucuronide, that crosses the barrier in small amounts. The mechanism Dayspring described from animal work is that it interferes with an enzyme (hexokinase) and the sugar-coating of brain proteins, which he tied to less brain inflammation. He layered the anecdotal belief of named neurologists on top of that. Dayspring was candid: do not hold your breath for a randomized trial. This is hypothesis, not established benefit.
• Omega-3s (EPA/DHA). The brain data here are observational. Dayspring conceded there is “no level one evidence” for the brain claims. On the heart side he pointed to a trial where EPA outperformed combined EPA plus DHA for residual risk in insulin-resistant, high-risk people who already had ApoB controlled. For the brain, he treated an omega-3 index of 8 to 9% as roughly the right target, but specifically called the push to 10% guesswork. The episode also leaned on work by Bill Harris, who is closely tied to omega-3 index testing.
• Obicetrapib (the CETP inhibitor). This is the most exciting and the most preliminary thread. (CETP is a blood protein that moves cholesterol from HDL onto LDL; inhibiting it lowers LDL and raises HDL.) The biological rationale Dayspring offered: people born with low CETP function appear to have less Alzheimer’s or cognitive impairment, so blocking CETP with a drug might mimic that protected state. The Broadway trial was designed to lower ApoB, but secondary Alzheimer’s biomarkers (phosphorylated p-tau, the amyloid 40-42 ratio, and other ratios of these markers) reportedly moved in a favorable direction. That is a biomarker signal in a short study, not a cognitive-outcome result. The drug is not FDA approved. And the trials are run and funded by the company (New Amsterdam Pharma), with trialists Dayspring named, Michael Davidson and John Castellan, described as friends of the host. That is a conflict worth holding in mind. Dayspring himself called it “a wonderful story on paper right now.”
• The hosts and guest run a clinical practice that uses these drugs, so the framing naturally favors aggressive lipid management.

Bottom line for a carrier

The strongest, most established message here is the reassuring one: your brain makes and keeps its own cholesterol, and lowering blood LDL has not been shown to harm cognition, with the statin trial evidence the hosts reviewed leaning neutral-to-protective. The APOE4-specific brain mechanism, the ezetimibe and omega-3 brain claims, and the obicetrapib biomarker story are all interesting and biologically plausible, but they remain emerging, observational, anecdotal, or company-funded. Treat them as reasons for optimism and for asking good questions, not as settled guidance.

This is general information, not medical advice, and it is our distillation rather than the guest’s exact words. Listen to the episode yourself and talk to your own doctor about what it means for you.