Alzheimer's risk by 85, against the general population
| Genotype | Develop Alzheimer's by 85 | Against the general population |
|---|---|---|
| General population (all genotypes) | 11 to 14 in 100 | The reference |
| 4/4 (two copies) | 51 to 60 in 100 | About 40 to 46 more in 100 |
| 3/4 (one copy) | 23 to 30 in 100 | About 12 to 16 more in 100 |
| 3/3 (no APOE4) | 8 to 10 in 100 | About 3 to 4 fewer in 100 |
Source: Genin, 2011. 7,351 people with Alzheimer's and 10,132 without, all of European ancestry, combined with US incidence data. Each range runs from men to women, and each difference compares men with men and women with women. These figures count Alzheimer's in people who live to 85.
For two copies, only 2 to 4 in 100 have Alzheimer's by 65, against 51 to 60 in 100 by 85 (Genin, 2011, Table 3A). Almost all of the risk builds between 65 and 85.
What can you act on?
I carry two copies of APOE4, and I built Phoenix because I needed it. The genotype is fixed. The numbers below are not.
- A structured program beats going it alone, for carriers too. In the US POINTER trial, 2,111 adults aged 60 to 79 at higher risk followed a two-year lifestyle program: exercise, healthy eating, mental activity, social activity and heart-health checks. The structured version, with more intensity and accountability, improved thinking scores more than the self-guided version, and APOE4 carriers benefited as much as non-carriers (Baker, 2025).
- Lifestyle moves the absolute numbers. In 196,383 UK adults over 60 followed for about 8 years, among those at high genetic risk, 1.13% with a healthy lifestyle developed dementia, against 1.78% with an unhealthy one. Healthy meant not smoking, regular physical activity, a healthy diet and moderate alcohol (Lourida, 2019).
- The Lancet Commission's 14 modifiable risk factors include high LDL cholesterol, high blood pressure, diabetes, physical inactivity, obesity, smoking, excess alcohol, hearing loss, vision loss, depression, social isolation, head injury, air pollution and less education (Livingston, 2024).
- Not every study agrees, so measure. In the Rotterdam Study, a healthy risk profile went with lower dementia risk at low and intermediate genetic risk, but not at high genetic risk (Licher, 2019). That is the reason Phoenix tracks your numbers: you see whether what you do moves them.
The blood markers Phoenix tracks for carriers, each with its own page:
- ApoB and LDL-C: high LDL cholesterol is on the Lancet list, and LDL measured before 65 links most strongly to dementia years later (Iwagami, 2021).
- HbA1c, fasting glucose and fasting insulin: in people without diabetes, an average glucose of 115 against 100 mg/dL went with an 18% higher dementia rate (Crane, 2013).
- hs-CRP: chronic inflammation, a CRP of 8 mg/L or higher on repeat tests, went with earlier Alzheimer's in carriers and not in non-carriers (Tao, 2018).
- Homocysteine: above 11 µmol/L is one of the causes of cognitive decline, by expert consensus (Smith, 2018).
- Vitamin D: a genetic study says deficiency is part of the cause of dementia (Navale, 2022).
- Omega-3 index: people with the highest red-cell DHA had 49% less Alzheimer's (Sala-Vila, 2022).
- Triglycerides and Lp(a): heart markers that complete the picture.
Why do studies give different numbers?
You will see figures from about 30% to about 60% for two copies. They answer different questions.
- Alzheimer's or any dementia. Some studies count Alzheimer's only, others every cause of dementia.
- "If you live that long" or real life. Some figures assume you reach the age in the table. Others count people who die of something else first, which lowers the number.
- Who was studied. People who volunteer at memory centers have higher rates than people drawn from the community (Qian, 2017).
- Ancestry. APOE4's effect differs across populations (Belloy, 2023).
Three more reference points:
- The largest US community study. It counted dementia from any cause up to age 95, including deaths from other causes, in 15,043 US adults followed for a median of 23 years: 59 in 100 people with two copies developed dementia after 55, against 48 with one copy and 39 with none (Fang, 2025).
- Framingham and Rotterdam. For people who started at 60 to 64, dementia by age 80 to 85 reached 6 to 7% with no APOE4, 16 to 17% with one copy and 35 to 38% with two. The Generation Study, a prevention trial for carriers, told its volunteers that the lifetime chance of mild cognitive impairment or dementia was 30 to 55% for 4/4, 20 to 25% for 3/4 and 10 to 15% for 3/3 (Qian, 2017).
- Your other genes matter. In the Rotterdam Study, among 4/4 carriers, Alzheimer's risk by 85 differed by 27 percentage points between the top and bottom third of a genetic score built from 23 other Alzheimer's variants. That is a 7 to 10 year difference in age at onset (van der Lee, 2018).
Is amyloid the same as Alzheimer's dementia?
No. Brain changes come early in 4/4 carriers; dementia is not certain.
- The brain changes. In 3,297 donated brains and 10,039 living people, nearly all 4/4 carriers had Alzheimer's brain changes. By 65, nearly all had abnormal amyloid in their spinal fluid, and 75% had a positive amyloid scan. The authors call 4/4 a distinct genetic form of Alzheimer's (Fortea, 2024).
- The symptoms. In the same study's brain-bank group, 4/4 carriers who developed symptoms did so at 65.6 on average, mild cognitive impairment followed at 71.8 and dementia at 73.6, about 7 to 10 years earlier than 3/3 (Fortea, 2024).
- The community picture. In the US community study, the median age at dementia diagnosis was 79 for two copies, 81 for one and 82 for none (Fang, 2025). And lifetime dementia for two copies runs from about 35% to about 60%, not close to 100% (Qian, 2017; Fang, 2025).
Amyloid is the first step, years before symptoms. Those years are the window to work on everything above.
Do sex and ancestry change the numbers?
Yes, both.
- Women. By 85, Alzheimer's reached 60% for women with 4/4 against 51% for men, and 30% against 23% for 3/4 (Genin, 2011). In the US community study, lifetime dementia for two copies was 64% for White women and 60% for White men (Fang, 2025).
- Ancestry. In 68,756 people, one copy of APOE4 raised Alzheimer's risk most in people of East Asian ancestry, then White, and less in Black and Hispanic people. Women with 3/4 carried extra risk between 60 and 70 (Belloy, 2023).
APOE4 among Phoenix members
Of the 1,720 Phoenix members who have recorded their genotype (2 October 2026, free and paid accounts), 1,009 carry two copies of APOE4, 604 carry 3/4 and 79 carry 2/4. Fewer than 30 carry no APOE4 at all. Almost 6 in 10 Phoenix members sit in the 4/4 column of the tables above: these numbers describe the people Phoenix is built for.
When should you talk to your doctor?
- If you notice changes in memory or thinking, get them assessed. Do not put them down to your genes.
- If your genotype came from a consumer test, a doctor or genetic counselor can confirm it and explain what it means for your parents, siblings and children.
- For each blood marker, the line to talk to your doctor is on its page above.