Francisco Gonzalez-Lima on Alzheimer's as a Vascular and Metabolic Disease: What It Means for Carriers

This is our independent, journal-club summary of Peter Attia’s conversation with Francisco Gonzalez-Lima, Ph.D., a professor of neuroscience, pharmacology and toxicology at UT Austin (Episode 38). We distill what was said, separate solid evidence from opinion, and flag where it gets murky. You can check it against the episode yourself.

The central argument is provocative and worth your attention, but it is largely one researcher’s framework, not settled science. Read it as a hypothesis with some supporting data, not a consensus position. Throughout, we separate what is established from what is the speaker’s opinion or extrapolation.

Why this one matters for carriers

Gonzalez-Lima never singles out APOE4 carriers. His argument is about late-onset Alzheimer’s in general. But the levers he lands on, blood flow, blood pressure, blood sugar, and metabolic health, are squarely vascular and metabolic, and those are exactly the levers an APOE4 carrier has reason to take seriously given the elevated vascular and Alzheimer’s risk that comes with the gene. So the connection to your situation is ours to draw, not his.

What he is actually claiming

Gonzalez-Lima makes a hard break from the dominant amyloid story. His view: the Alzheimer’s that Alois Alzheimer described in 1907 (a 51-year-old patient) is a different disease from the common late-onset dementia we see in older people. He argues the amyloid-beta hypothesis, as applied to late-onset disease, is simply wrong, and he says it in the strongest possible terms.

His alternative: late-onset Alzheimer’s is fundamentally the brain running out of energy. Two upstream problems feed it. First, chronically reduced blood flow to the brain (vascular hypoperfusion). Second, mitochondria failing to use oxygen to make energy, specifically a downregulation of the enzyme cytochrome c oxidase, also called complex IV, the last step in the cellular machinery that turns oxygen into usable energy. In his telling, amyloid is downstream, a marker of cells already in trouble rather than the driver.

This is a real scientific debate, but it is contested. The amyloid framework remains mainstream, and a single lab’s reframing is not the same as established fact. Treat the whole structure as the speaker’s hypothesis.

The numbers worth holding onto

A few figures came up that have context behind them:

• Cerebral blood flow declines roughly 20% between ages 22 and 60, or about half a percent per year (Gonzalez-Lima, citing the broader literature). Established as a general trend; the precise figure is an approximation.
• Cytochrome oxidase can be inhibited by up to 30% in animals without obvious behavioral change, yet memory tests still show impairment. In his sodium azide work, the animals ate and behaved normally but failed cognitive tasks. Push inhibition past about 40% in a whole organism and it dies (this is why cyanide, a complex IV blocker, kills in minutes). This comes from his own animal work and the toxicology literature; the Alzheimer’s relevance is extrapolation.
• Blood pressure targets: he points to recent trials suggesting best outcomes below 120 systolic / 80 diastolic, versus older targets around 135/90. This is his characterization of recent trial results, which he uses to support his vascular framing; he names no specific trial.
• The rate of disease versus the rate of aging: Attia notes US human longevity is increasing only about 0.4% per year, and argues Alzheimer’s is rising faster than that, implying a real increase rather than just more diagnosis. This is Attia’s framing of the epidemiology, not a settled quantitative finding.
• Amyloid versus symptoms: he argues there is no consistent correlation between the amount of amyloid deposition and memory loss or disease progression in late-onset cases, and that pathologists blinded to the clinical picture often cannot tell a demented brain from a cognitively normal one of the same age. This is his strong claim; it is contested and central to his thesis, so do not treat it as established.

One part of his framework deserves a direct flag: cholesterol. He argues that high total cholesterol does not cause the vascular damage and that lowering it will not make a major difference to this process. For an APOE4 carrier, that is the wrong place to follow him. ApoB-containing lipoproteins are an established causal driver of atherosclerosis, so his dismissal here is his opinion, runs against a large body of evidence, and is not where the field stands.

Why the mitochondrial story matters, and why he thinks it is reversible

A key plank of his argument is that the cytochrome oxidase deficit he found in fresh-frozen Alzheimer’s brains was not a loss of the protein itself but a loss of its working, active state. He stresses that this enzyme is inducible and works “on demand,” so the shutdown is, at least temporarily, not permanent. That is precisely why he believes intervention is possible. This reversibility claim is his interpretation and the basis for his treatment ideas, not an established fact about Alzheimer’s; grade it as his hypothesis.

What a carrier can actually do

The practical thread, if his framework holds, is that prevention beats treatment, and the levers are largely cardiovascular and metabolic. He puts it plainly: anything that helps the heart and circulation helps the brain, because the brain is the most energy-hungry organ (about 2% of body weight, roughly 20% of energy use) and is unusually dependent on aerobic metabolism with little anaerobic backup.

The prevention ideas he raised, none of them APOE4-specific:

• Manage the standard vascular risks: blood pressure, atherosclerosis, blood sugar, lipoproteins, smoking, and avoid head trauma.
• Ketogenic diet or exogenous ketones. His argument: glucose transport into the aging brain becomes impaired, but ketone uptake is not, so ketones can feed neurons when glucose delivery falls short. He says he hopes supplemental ketones work as well as a ketogenic diet, but that this will have to be settled by experiment. Mechanistically plausible and emerging, not proven to prevent dementia.
• His personal routine: a 14 to 16 hour fast once a week, often Friday to Saturday, sometimes with a morning workout to burn through circulating glucose. This is his personal practice and opinion, not a tested protocol.

The honest takeaway: the lifestyle levers he names overlap almost entirely with cardiovascular prevention, which is well supported. The specific metabolic add-ons (ketones, fasting cadence) are reasonable hypotheses, not established dementia prevention.

Fair warning: where this gets speculative, and where the conflicts are

Much of the back half is about two interventions Gonzalez-Lima has personally worked on for years. That is a relevant conflict of interest, and he holds grants (including NIH funding) tied to these approaches.

Methylene blue. His claim is that at low concentration it acts as an electron donor that can bypass bottlenecks in the energy-production chain and support mitochondrial respiration. He describes the dose as biphasic, meaning the effect flips with dose: helpful low, harmful high. At high doses it competes with oxygen, turns into a pro-oxidant, and can induce methemoglobinemia, the very blood condition it treats at correct doses. He says his group showed, in a blinded placebo-controlled study in healthy and older people, improved memory retrieval after a single acute dose, with imaging changes. The study he is describing is published: Rodriguez et al., Radiology, 2016, a randomized, double-blind, placebo-controlled fMRI trial in 26 adults (ages 22 to 62), which found that low-dose methylene blue raised fMRI activity during sustained-attention and short-term-memory tasks and was associated with a roughly 7 percent increase in correct responses during memory retrieval. So the effect is real and now quantified, but it is one small, single-dose study in mostly healthy adults, and the leap to preventing or treating dementia is unproven.

Several cautions he raised himself matter. Purity is a serious problem. He puts the chemical/laboratory grade (he names the Sigma product) at about 15% impurities, and describes industrial grade as even worse than 15%; the impurities he lists include lead, mercury, and cadmium. Only pharmaceutical/USP grade should ever go near a person. The only standing FDA-recognized use is methemoglobinemia. He also notes it turns urine blue.

He is sharply critical of the company that ran a methylene blue derivative (LMTM) as an “anti-tau” agent, arguing they had the wrong hypothesis and the wrong (too high) doses. In recapping that trial it was Attia, not Gonzalez-Lima, who put the apparent benefit at “about 15%” and who recalled it appearing only in patients who received LMTM on its own; neither specified precisely what that figure refers to, so we leave it imprecise. Gonzalez-Lima thinks the most likely explanation is that combining the agent with standard Alzheimer’s drugs muddied the result, and he frames the company’s approach as undermining the drug’s potential. Note: subgroup findings like this are exactly what statisticians warn against reading too much into, and Attia says so.

Transcranial near-infrared light. He describes shining 1,064 nm light through the forehead to activate cytochrome oxidase in the prefrontal cortex. He says only about 1 to 2% of the light reaches the surface of the cerebral cortex when delivered this way, and the effect is local, not systemic. He is explicit that clinical benefit in atrophied Alzheimer’s brains is unproven. He does not combine it with methylene blue, because the dye would absorb the light. The penetration figure and the mechanism are his; clinical benefit is, by his own admission, untested.

He also makes strong, contested clinical claims that deserve a skeptical eye. On the cholinesterase-inhibitor drugs (a common class of Alzheimer’s medication), he points to UK longitudinal data and argues patients die sooner on them, and recounts the UK banning then reinstating those compounds; he gives no magnitude for the mortality claim, so treat it as an unquantified assertion. He discusses memantine separately, arguing the FDA approved it originally only for severe disease and later for moderate-to-severe disease, and that, by calming overexcited neurons, it can “rescue” a cell that is already functionally incompetent, which he sees as counterproductive. He also floats methylene blue as a rescue agent for stroke and traumatic brain injury based on animal models, and it is in that emergency context that he cites a dose of about 1 mg/kg as protective. These are his positions and animal-model extrapolations, not established human treatment.

The recurring caveat he names himself: there is little commercial incentive to run the trials, because methylene blue is off-patent. That is an honest point about why the evidence is thin, but it does not make the thin evidence strong.

Bottom line for a carrier

The defensible message here is one you have heard before, dressed in new mechanism: protect your vascular and metabolic health, because the brain pays first when energy delivery fails. The blood pressure, blood sugar, and cardiovascular targets are well grounded. The ketone and fasting ideas are plausible and emerging. The methylene blue and near-infrared light work is genuinely interesting but early, conflicted, and not ready for self-experimentation, especially given the purity and dosing risks. And be wary of his dismissal of cholesterol and lipid-lowering. For an APOE4 carrier, ApoB management is established cardiovascular and likely brain-protective ground, and walking away from it on the strength of one framework would be a mistake.

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 doctor before acting on anything here.