Your Brain Is a Vascular Organ — PESA, Dementia, and the APOE4 Question

There's a strange division in modern medicine that I want to start with.

When a patient walks into a cardiologist's office, the cardiologist is focused on a specific question. Is this person going to have a heart attack? What's their LDL? What's their blood pressure? Do they have plaque in their coronaries?

When that same patient walks into a neurologist's office decades later, complaining of memory problems, the neurologist is focused on a different question. Is this Alzheimer's? Is it vascular dementia? Is there a stroke? What does the amyloid PET show?

Those are two completely different conversations. They happen in two different rooms, with two different specialties, twenty or thirty years apart. And yet, increasingly, the data are telling us that what those two doctors are looking at is the same disease, in two different organs, at two different points in time.

The thing the cardiologist saw in the carotid artery at age fifty — that plaque, that hypertension, that subtle vascular damage — is the same thing the neurologist will be looking at, on a brain MRI, at age seventy-five. The cardiologist's patient and the neurologist's patient are, statistically, the same person. The vascular tree connects the heart to the brain. The disease that damages it doesn't stop at the neck.

So today we're going to talk about what subclinical atherosclerosis is doing to the organ most people don't think of as vascular: the brain. And we're going to spend the second half on a specific genetic question. About a quarter of the population carries at least one copy of a gene variant called APOE4, which dramatically increases the risk of Alzheimer's disease. What most people who carry that variant have never been told is that APOE4 also dramatically increases the risk of cardiovascular disease — and that suggests a question about whether some of the Alzheimer's risk in APOE4 carriers is, in fact, a vascular risk that could be lowered the same way we lower cardiovascular risk.

The brain is a vascular organ

I want to start with a basic physiologic fact most people underappreciate. The brain weighs about three pounds. It is two percent of your body weight. And it consumes about twenty percent of the body's resting energy supply. The brain is, by a wide margin, the most metabolically expensive organ in the body. Per gram of tissue, it requires about ten times the energy a muscle does at rest.

That energy demand has to be met every second. The brain has essentially no energy storage. Unlike a muscle, which has glycogen, or a liver, which has glycogen and lipid reserves, the brain has effectively no on-board fuel. It depends on a continuous, real-time delivery of glucose and oxygen from the bloodstream. Interrupt that delivery for thirty seconds and the patient loses consciousness. Interrupt it for ten minutes and you have permanent damage.

The vehicle for that delivery is a remarkable vascular network. The brain receives its blood supply from four major arteries — the two carotids and the two vertebrals — and that flow gets distributed through a microvascular bed of approximately four hundred miles of capillaries, packed into the cranium. Each of those capillaries has a specialized blood-brain barrier that selectively controls what gets in and out of the neural environment. And the entire system is regulated by a phenomenon called neurovascular coupling — when a brain region is active, the local blood vessels dilate within milliseconds to deliver more substrate. The healthier the vasculature, the tighter that coupling is.

So the brain is not just near the vasculature. The brain is, in a real sense, the vasculature's most demanding customer. And anything that compromises the vascular tree — atherosclerosis, microvascular dysfunction, blood-brain barrier damage, chronic hypoperfusion — has cascading consequences for neural function.

What we've come to understand over the last twenty years is that the major dementing illnesses are mixed. The clean pathological categories we used to teach in medical school — Alzheimer's is amyloid and tau, vascular dementia is small-vessel disease — turn out to be a simplification. In the great majority of older adults with dementia, the postmortem brain shows both the proteinopathy of Alzheimer's and the vascular damage of small-vessel disease. They co-occur because they share underlying causes: hypertension, diabetes, smoking, inflammation, dyslipidemia. These are not just heart disease risk factors. They are brain disease risk factors.

This is the conceptual setup for everything that follows. Cardiovascular health and brain health are not in two different categories. They are, mechanistically, much closer to being the same category.

What PESA found in the brain

The PESA study is a Spanish cohort designed to follow asymptomatic middle-aged adults over time and look at their vascular tree using imaging. Most of what gets discussed is the cardiovascular findings — the prevalence of plaque, the relationship to LDL, the diet pattern, the sleep data. But the PESA design included something most cardiovascular studies don't: they imaged the brain at the same time they imaged the vasculature.

In 2021, Marta Cortés-Canteli and the PESA team published a paper in the Journal of the American College of Cardiology asking a specific question. In middle-aged people who have subclinical atherosclerosis but no symptoms and no known cognitive disease, is there evidence that their brains are already being affected?

The study population was five hundred and forty-seven PESA participants, all of whom had at least some evidence of subclinical atherosclerosis on imaging. Mean age fifty. Eighty-two percent men. None of them had cognitive symptoms.

What the researchers did was put each participant into a PET scanner with a radioactive tracer called FDG — fluorodeoxyglucose. FDG is a glucose analog. It gets taken up by metabolically active tissue, so a brain region that's metabolically healthy lights up on the scan and a region that's metabolically depressed appears dim. This technique — FDG-PET of the brain — is, by the way, the standard imaging tool we use to diagnose Alzheimer's disease at early stages, because Alzheimer's produces a very characteristic pattern of cortical hypometabolism in specific brain regions before the patient is symptomatic.

Here's what they found. First, total brain glucose uptake was inversely associated with cardiovascular risk. The higher the thirty-year Framingham Risk Score, the lower the brain's metabolic activity, with hypertension the single strongest individual risk factor driving that association.

Second — and this is the finding I think is most important — when they looked at carotid plaque burden specifically, that plaque burden was independently associated with lower brain FDG uptake, even after adjusting for the cardiovascular risk score. Having plaque in your carotids was statistically associated with a less metabolically active brain, beyond what could be explained by your risk-factor profile alone.

Third, they mapped exactly which brain regions were showing the hypometabolism. The answer was the parietotemporal cortex — the angular gyrus, the supramarginal gyrus, the inferior and middle temporal gyri — and the cingulate gyrus. These are the regions most affected in early Alzheimer's disease. Not random regions. The exact regions you would see hypometabolism in if you were watching the disease develop, twenty or thirty years earlier than usual.

Let me say that again, because it matters. In asymptomatic, cognitively normal, middle-aged adults whose only abnormality was the kind of subclinical plaque PESA picks up on routine imaging, the brain regions that go hypometabolic in early Alzheimer's were already, on average, metabolically depressed. The vascular footprint of the disease is detectable in the cortex of asymptomatic forty-year-olds.

The longitudinal follow-up

Cross-sectional data shows a snapshot. Longitudinal data shows what happens over time. In 2023, the PESA team published a longitudinal follow-up in The Lancet Healthy Longevity: three hundred and seventy participants, two FDG-PET scans about four point seven years apart on average, with APOE genotype adjusted for directly in every model.

One. Persistently high cardiovascular risk was associated with a four-point-three percent decline in cortical FDG uptake over the follow-up period. Persistently low risk was associated with a one-point-five percent decline. Sustained high CV risk produced about three times the rate of cortical metabolic decline as low risk — approximately point-nine percent per year of cortical metabolism lost.

Two. Progression of carotid plaque volume independently predicted additional decline in the Alzheimer's-vulnerable regions, above and beyond what could be explained by the risk-factor profile.

Three. About twenty percent of this association was mediated by plasma neurofilament light chain (NfL), a circulating biomarker of axonal injury and neurodegeneration. This is an important detail. NfL doesn't just measure hypometabolism — it measures actual neurons dying. So a significant chunk of the FDG signal isn't a reversible energy deficit. It reflects irreversible neuronal injury accumulating in the brains of middle-aged adults with vascular disease.

Four. The signal held after adjusting for APOE genotype, meaning the vascular contribution to brain metabolic decline is independent of the genetic contribution. APOE4 adds risk. Plaque adds risk. They are additive, not redundant.

What this study did, in my view, is take the cross-sectional finding — vascular disease leaves a footprint on the brain — and convert it into a longitudinal mechanism. Vascular disease drives ongoing brain decline in the regions most relevant to Alzheimer's, and a measurable fraction of that decline is real neuronal death, not just functional energy deprivation.

The APOE4 question

Now I want to spend a few minutes on a specific genetic question, because I think it's the most interesting open question in this whole field — and one that people with the relevant genotype have generally not had explained to them.

About twenty-five percent of the population carries at least one copy of the APOE4 allele. Carrying one copy roughly triples the lifetime risk of Alzheimer's disease. Carrying two copies — homozygous APOE4 — multiplies that risk by approximately twelve. APOE4 is the strongest common genetic risk factor we know for late-onset Alzheimer's.

What gets less attention is that APOE4 is also a cardiovascular risk factor. APOE4 carriers have higher LDL cholesterol throughout life, because the APOE4 protein is less efficient than the APOE3 protein at clearing remnant lipoproteins from circulation. As a result, APOE4 carriers have more subclinical atherosclerosis. The PESA group itself documented this: APOE4 carriers in the cohort had more carotid plaque, more femoral plaque, and more coronary calcification than APOE3 homozygotes — and most of that effect was statistically explained by their higher lifetime LDL exposure.

So APOE4 is doing two things to the people who carry it. It's adding cardiovascular risk through higher LDL. And it's adding Alzheimer's risk through mechanisms that probably include several pathways — lipid handling in the brain, blood-brain barrier function, amyloid clearance.

Here's the question I think is the most interesting. Are these two effects connected? Specifically, if we lower the LDL of an APOE4 carrier to non-carrier levels — if we eliminate the cardiovascular penalty their genotype gives them — do we also lower their dementia risk?

We don't have a definitive randomized-trial answer to that question yet. But we have some intriguing pieces.

The first is a Mendelian randomization study published in 2021 analyzing the relationship between plasma APOE isoforms and ischemic heart disease. The researchers used genetic instruments to ask whether the APOE4 effect on heart disease is mediated by ApoB — whether the vascular damage of APOE4 is essentially LDL-mediated and would disappear if you controlled LDL. The answer was yes. Once they adjusted for ApoB, the association between APOE4 and ischemic heart disease essentially disappeared. The cardiovascular risk attributable to APOE4 is mediated through the higher LDL it produces. Lower the LDL, and that arm of the risk goes away.

The second piece is a 2024 study in Neurology that asked whether statin initiation in older adults reduced the incidence of Alzheimer's disease, and whether that effect was modified by APOE genotype. The finding was that statin initiation was associated with reduced Alzheimer's incidence specifically in APOE4 carriers. That benefit was not seen in non-carriers. The interpretation the authors offered is that the vascular contribution to Alzheimer's risk is larger, on average, in APOE4 carriers — because they're already starting with more vascular insult.

Now I want to be very honest here. We do not have a randomized controlled trial showing that aggressive LDL lowering in APOE4 carriers prevents dementia. That trial would have to be enormous, would have to follow people for decades, and may never be done.

But here is the position I want to put on the table. We know APOE4 carriers carry double risk — genetic and vascular. We know the vascular leg of that risk is, by Mendelian randomization, fully explained by their higher LDL. We know the PESA data show vascular disease in midlife is associated with a measurable cortical hypometabolism fingerprint in the regions Alzheimer's targets. We know that fingerprint is partially mediated by actual neuronal injury, not just functional change. And we know that lowering LDL — aggressively, early, and sustained — is one of the safest and best-evidence-based interventions in all of medicine.

The synthesis I would offer is this. An APOE4 carrier who treats their LDL as if their life depends on it — meaning aggressive lowering from young adulthood into the range PESA suggests is actually protective, fifty to sixty milligrams per deciliter or lower — is, at a minimum, depriving the genotype of one of its two routes of attack. The cardiovascular leg gets disarmed. Whether that also reduces the dementia risk by some amount is a biologically plausible hypothesis, supported by converging evidence. It is not yet proof.

If you are an APOE4 carrier, the most important thing you can do is take that hypothesis seriously. Get your ApoB measured. Drive it as low as you can tolerate. And do all the other things — sleep, exercise, the Mediterranean diet, blood pressure control — that converge on protecting the vascular tree your brain depends on. The available evidence does not yet prove that this lowers dementia risk. But the available evidence makes it the most defensible thing to do.

Key takeaways

• The brain is a vascular organ. The same disease that damages the carotid artery damages the cortex. There is no clean separation between cardiology and the medicine that prevents dementia — the intervention list is largely the same.
• The vascular footprint of Alzheimer's is detectable in midlife. The PESA data show asymptomatic adults with subclinical carotid plaque already have, on average, hypometabolism in the brain regions Alzheimer's takes down decades later — and a measurable fraction of that reflects actual neuronal injury.
• For APOE4 carriers, one of the two routes of attack is modifiable. The Alzheimer's risk attributable to APOE4 is, in important part, a vascular risk — and vascular risk is among the best-treated targets in all of medicine. Lower the LDL aggressively and early, and you have at minimum eliminated the cardiovascular contribution to the genotype's lifetime burden.