Start · Catalogue · Profile · Table
Screening BODY HANDBOOK
Screening · §98
APOE ε4
One gene, three common versions — and one of them is the strongest known genetic driver of late-life Alzheimer's. About one in four people carry a single copy of APOE ε4; about one in fifty carry two, and at two copies the lifetime risk of Alzheimer's approaches a near-certain genetic disease. The same variant also nudges cardiovascular risk upward through how it handles cholesterol. The real question isn't whether the gene matters — it does — but whether you should find out your status, and what changes the day after you do.
Decide · Once Evidence Strong Chapter Screening

If you're a carrier, the dementia-prevention playbook stops being abstract — exercise, sleep, lipid control, and treating sleep apnea each get a sharper edge, and the cardiovascular thresholds you should treat to drop. If you're not, you get a useful piece of calibration. The test itself is cheap and one-time. The catch: the answer is probabilistic, the psychology of sitting with it is real, and in the United States it can affect long-term-care, life, and disability insurance — so think before you spit in the tube.

APOE is a tiny lipid-shuttling protein your body uses to move cholesterol around — including inside the brain, where it ferries fats between neurons and the support cells that feed them. Most people inherit two copies of the common ε3 version. About one person in four inherits one copy of ε4; about one in fifty inherits two. A rarer version, ε2, is actively protective Reiman et al. 2020.

The ε4 version of the protein does its job badly in two ways that matter. In the brain, it clears amyloid-β — the sticky protein fragments that build into Alzheimer's plaques — more slowly than ε3 does, and it makes the brain's resident immune cells worse at cleaning up the same fragments Holtzman et al. 2012. In the bloodstream, it leaves more LDL particles circulating, which is the same machinery that drives heart attacks. Two mechanisms, one molecule.

How big the effect actually is

The numbers have been replicated for thirty years. Compared to two copies of the common version, one copy of ε4 roughly triples lifetime Alzheimer's risk; two copies push it up about twelve to fifteen times Farrer et al. 1997. Each copy also shifts the age of onset earlier by seven to nine years. The link is the most settled finding in the genetics of dementia Bellenguez et al. 2022.

The cardiovascular signal is smaller but real. Across more than 121,000 people in pooled analyses, each ε4 allele was tied to slightly higher LDL cholesterol and about 6% higher coronary heart disease risk per copy versus the common version Bennet et al. 2007. Modest at the individual level. Substantial across a lifetime, especially stacked on top of the brain effect.

Women carry an extra burden. In a meta-analysis of nearly sixty thousand people, women with one copy of ε4 between roughly age 65 and 75 had about double the Alzheimer's risk of men with the same genotype in the same decade Neu et al. 2017. The gap is widest in the years just after menopause, then narrows with age.

What it looks like if you carry it and do nothing

For a single-copy carrier living an average life, the curve bends late. The first signs aren't a diagnosis — they're the partner who asks, twice in a year, why the same story got told at dinner. The route home from a place you've driven for twenty years feels wrong for half a minute. A friend from college calls and the name takes longer than it should. Cohort data put this clustering of small failures in the seventh and eighth decades, with full diagnosis trailing by a few years.

For a double-copy carrier — about one in fifty — the onset window pulls forward by roughly a decade Fortea et al. 2024. The mid-60s, not the mid-70s, become the high-risk years. For women, the steepest stretch arrives in the first ten years after menopause; their partners are often the first to flag it. The cardiovascular contribution stays quieter — a heart attack at 62 instead of 70, an LDL number that crept up despite a reasonable diet, a small white-matter lesion picked up on an unrelated scan. None of it is destiny. All of it is more likely than the population average.

How to actually find out

Two routes. A direct-to-consumer kit (23andMe, AncestryDNA with a third-party interpretation tool for the raw data) costs $50–200 and gives you the answer in a few weeks. Or your doctor can order a clinical test for $100–300, often covered if there's a real clinical reason — symptoms, a strong family history, or a decision about an anti-amyloid drug. A genetic counseling session before testing runs another $150–500 and is worth it for most people; it covers what the result means, what it doesn't, and the insurance issue (see below).

What changes if you're a carrier

The general advice for protecting your brain — exercise, sleep, lipid control, treating sleep apnea, social engagement, not getting hit in the head — is the same prevention list everyone gets. What changes for a carrier is the urgency and the threshold. Three pieces have the strongest backing.

The full multidomain protocol. A two-year program of Mediterranean-style diet, structured exercise, cognitive training, and aggressive cardiovascular-risk treatment improved a global cognitive score by about a quarter of a standard deviation in older adults at elevated risk Ngandu et al. 2015. A pre-specified follow-up looked specifically at carriers and non-carriers: the carriers benefited at least as much as everyone else Solomon et al. 2018. The gene doesn't lock you out of the payoff.

Exercise. A six-month home-based walking program — roughly 150 minutes per week — improved cognition in older adults with memory complaints, and the gain was still detectable 18 months later Lautenschlager et al. 2008. Across observational cohorts, the protective slope of physical activity is steeper in ε4 carriers than in non-carriers — the genetic susceptibility makes the same exercise worth more.

Lipids, treated harder. If you're a carrier, ApoB and LDL cholesterol are not just about heart disease — they're a second route to dementia through small-vessel cerebrovascular damage. Pragmatically: get an ApoB measurement, treat to the lower end of the range your doctor uses, and respond well to statins (carriers usually do). A Mediterranean-style pattern with most of the fat from olive oil, nuts, and fish — not saturated fat — drops LDL more in carriers than in non-carriers in head-to-head dietary trials.

Sleep. Deep sleep is when the brain clears the same amyloid that ε4 handles poorly. Fragmented sleep over years tracks with a 1.5-fold higher Alzheimer's risk in cohort data Lim et al. 2013. The single highest-yield move: get tested for obstructive sleep apnea if there's any suggestion of it — loud snoring, gasping awakenings, daytime fatigue despite enough hours in bed. Treat it. Then protect a real 7–8 hour window.

One newer lever sits just outside that list: the shingles vaccine, which recent research has begun tying to lower dementia risk — still-settling evidence, but a low-effort, low-downside one to keep current on.

When not to test — at least not yet

What most people get wrong

"One copy means I'm getting Alzheimer's." No. A single copy is common — roughly one person in four — and most single-copy carriers do not develop Alzheimer's, even in their 80s. The risk is meaningfully higher; it isn't a sentence.

"Two copies of the common version means I'm safe." Also no. About six in ten Alzheimer's cases happen in people without a single ε4 copy. The gene explains a real chunk of risk; it doesn't explain all of it. The lifestyle list applies to everyone.

"There's nothing you can do anyway, so why bother knowing." The lifestyle protocol that helps non-carriers helps carriers at least as much, possibly more — that's what the multidomain trial showed when it split the results by genotype Solomon et al. 2018. The argument against testing is real, but it isn't this one.

Who actually benefits from finding out

The decision to test isn't the same for everyone. Three groups get the most decision value from it.

  • Adults with a strong family history of Alzheimer's or early-onset dementia. The prior probability of carrying ε4 is higher; the answer refines a question you're already asking.
  • People with elevated cardiovascular risk who want sharper targets. A carrier result lowers the LDL and ApoB thresholds your doctor will be willing to treat to, and tightens the case for an early statin.
  • Patients facing an anti-amyloid drug decision. If you or a parent is being considered for one of the new Alzheimer's antibodies, genotype is now standard of care — it determines the risk of a specific brain-imaging side effect and changes how the drug is dosed and monitored.

For women, the calculus has a sharper time component. The decade after menopause is when ε4-related risk runs hottest; that window is also when other decisions — hormone therapy timing, retirement planning, cardiovascular treatment intensity — are being made. Finding out at 45 instead of 65 leaves room to act.

Adults under 30 with no family history get the least: a long lead time, identical lifestyle advice either way, and a decision they could make better with more information later.

Where this goes wrong in practice

The impulse test. Someone orders a kit on a Sunday night, gets the result alone three weeks later, and spends six months catastrophizing. The disclosure-study literature is consistent: counseled testing produces modest, manageable distress; uncounseled testing produces a much wider distribution, with a real tail of harm.

The one-way insurance door. A carrier finds out, then a year later tries to buy long-term-care insurance and gets declined or rated up. The U.S. law that protects you from health-insurance and employment discrimination doesn't cover this. The order of operations matters.

The fatalist response. A carrier reads the result, decides nothing matters, and stops the gym membership. This is exactly backwards — the protective effect of the lifestyle list is at least as large in carriers as in non-carriers.

The non-carrier's free pass. The opposite mistake: the common-version result gets read as permission to ignore sleep, lipids, and exercise. Most Alzheimer's happens in non-carriers. The license isn't there.

What changes if you act on it

A 50-year-old carrier who runs the full prevention playbook from the day of the result doesn't get a clean escape — the gene still raises the baseline. What they get is a delayed onset window: probably into the 80s instead of the mid-70s, sometimes much further. The cardiovascular curve bends harder, faster, because statins work especially well in carriers and the doctor now has a reason to push the targets. Sleep apnea, if it was there, gets treated and stops feeding the silent vascular drip.

The texture across decades looks like this. In the first year, a routine: known LDL number trending down, sleep window enforced, weekly cardio reliable, hearing checked. By year five, the partner notices that the cognitive sharpness hasn't softened in the way it did for friends of the same age. By year ten, the friends without the protocol are starting to ask whether they should be doing something — and you tell them yes. By the 70s, the years that statistically would have been the dimmest are still bright; the executive function that holds up a financial life is still intact.

For non-carriers, the payoff is calibration. The dementia-prevention list still applies, but the cardiovascular thresholds are normal and the relentless sleep optimization can relax to a steady, sane baseline. Worry retires from a meaningful slot in the week.

Related, worth knowing about

  • ApoB testing — the single best cardiovascular risk number, and the one to push hardest on if you're a carrier.
  • Obstructive sleep apnea screening — the highest-leverage modifiable sleep contributor to dementia risk.
  • Hearing loss in midlife — the largest single modifiable dementia risk factor across the population Lancet Commission 2024.
  • Anti-amyloid antibody drugs (lecanemab, donanemab) — a separate decision for symptomatic Alzheimer's where genotype now drives dosing and monitoring.
  • Polygenic risk scores — a more granular but harder-to-interpret alternative to single-gene testing.
  • Plasma p-tau biomarkers — a newer, dynamic blood test that catches early Alzheimer's pathology directly rather than predicting it from genetics.
Related in the handbook
Helps
  • — If your APOE status worries you, the shingles vaccine is one of the few proven moves that nudges dementia risk down.
  • — For a carrier, treating sleep apnea is one of the dementia-prevention levers that gets a sharper edge.
Related
  • — Two e4 copies raise the bleeding risk of anti-amyloid drugs, a key factor if treatment's on the table.
  • — ε4 also raises cardiovascular risk through cholesterol, and carriers should treat to tighter targets — so know your ApoB and Lp(a).
  • — This is one specific result a genetic test returns — think about the insurance angle before you spit.
  • — If you carry ε4, the dementia playbook sharpens: a hearing test, and a hearing aid if you need one, is part of it.
  • — If you carry e4, clearing brain-fogging meds is a free, sensible part of the prevention plan.
·
98