What HbA1c should an APOE4 carrier aim for?

The HbA1c bands

The line rule. "Under 5.3" excludes 5.3, and "5.7 or higher" includes 5.7. So 5.3 is Monitor and 5.7 is Talk to your doctor.

Confidence in the optimal line: low to moderate. The direction is solid: risk climbs with blood sugar well below the prediabetes line. The exact number is my choice.

What does HbA1c measure?

HbA1c is the share of your hemoglobin that has sugar stuck to it. A red blood cell lives about 120 days, so HbA1c is a running average of your blood sugar, weighted toward the most recent weeks (ADA, 2026).

Fasting glucose is one morning. HbA1c is your last three months, which makes it the steadiest blood-sugar number you can test.

Why does blood sugar matter for an APOE4 brain?

Higher blood sugar goes with a faster-aging brain, well below the diabetes line.

• In 5,189 older English adults followed for about 8 years, every 1 mmol/mol of extra HbA1c went with faster decline in memory, executive function and overall cognition. People with prediabetes declined faster than people with normal blood sugar, and people with diabetes faster still (Zheng, 2018).
• In 2,067 older adults without diabetes, an average blood sugar of 115 mg/dL against 100 mg/dL went with an 18% higher rate of dementia (hazard ratio 1.18) (Crane, 2013).

What APOE4 changes. The carrier studies point in two directions:

• In 636 people with type 2 diabetes, a higher HbA1c amplified the cognitive and smell problems that came with APOE4 (interaction odds ratio 2.69) (Wang, 2025).
• A 2025 meta-analysis of 170 studies found that diabetes raised Alzheimer's risk only in people without APOE4 (Huang, 2025).
• The mechanism is real: in mice and in neurons, APOE4 traps insulin receptors inside the cell and blunts insulin signalling, worse with age and a high-fat diet (Zhao, 2017).

No study gives carriers their own HbA1c number. But blood sugar is a risk you can measure and move, and the brain data favour lower across the normal range. That is my reason to aim low.

What do the guidelines say?

The American Diabetes Association's 2026 Standards of Care set the lines (ADA, 2026):

• Prediabetes: 5.7 to 6.4% (39 to 47 mmol/mol).
• Diabetes: 6.5% (48 mmol/mol) or higher, confirmed by a second abnormal result unless the diagnosis is obvious.

The same guideline says "risk is continuous, extending below the lower limit of the range." A 5.6% is not safe and 5.7% is not a cliff. These lines diagnose diabetes. They were not written for APOE4 carriers.

Why did Phoenix choose under 5.3%?

I chose under 5.3% for three reasons:

1. It sits inside ARIC's reference band, 5.0 to under 5.5%. Every band above it carried more diabetes risk, starting with nearly double at 5.5 to under 6.0% (hazard ratio 1.86) (Selvin, 2010).
2. The ADA itself says risk keeps falling below 5.7%, so the lab's "normal" line is too loose for someone who wants to protect their brain.
3. The app and the Blood Work Blueprint already used 5.3, and no evidence argues for moving it.

Why not lower? In ARIC, people under 5.0% had the lowest diabetes risk (hazard ratio 0.52), but death from any cause followed a J-shaped curve (Selvin, 2010). So Phoenix sets no lower line and does not push you below 5.0.

Know the noise. HbA1c varies about 1.8% within one person from test to test (Chai, 2017). At 5.3%, two tests on the same true blood sugar can differ by up to about a quarter of a point. A 5.2 and a 5.4 can be the same you. Read the trend across two or more tests, not one decimal.

How do you get HbA1c tested?

Ask for "hemoglobin A1c" (HbA1c, or A1C). No fasting is needed, and the Blood Work Blueprint lists it with the rest of the APOE4 panel.

US labs report a percentage; the UK and most of Europe report mmol/mol (5.3% is 34, 5.7% is 39, 6.5% is 48; ADA, 2026).

Some things move HbA1c without changing your blood sugar, because they change how long your red cells live: anemia, iron deficiency, recent blood loss or transfusion, hemolysis, kidney failure, pregnancy and some hemoglobin variants (ADA, 2026).

What moves HbA1c?

Ranked by the size of the drop in the best trial data. The biggest drops come from people with type 2 diabetes, who have the most room to fall. Near 5.3 to 5.7%, expect changes in tenths of a point.

Lifestyle beats medicine for prevention. In the Diabetes Prevention Program, 3,234 people with raised blood sugar who aimed for 7% weight loss and 150 minutes of activity a week cut new diabetes by 58%; metformin cut it by 31% (Knowler, 2002).

At a normal HbA1c, exercise works on glucose and insulin first: in 24 trials of inactive adults without diabetes, combined aerobic and strength exercise lowered both but left HbA1c unchanged (Silva, 2024).

One lever pushes it up: statins. Across 13 trials and 91,140 people, statins raised the rate of new diabetes by 9% (odds ratio 1.09): one extra case for every 255 people treated for 4 years. The authors judged the heart benefit far larger (Sattar, 2010). 80 Phoenix members log a statin. If you take one for ApoB, watch your HbA1c, and do not stop the statin because of this page.

What Phoenix members use. Among Phoenix members who log their stack (counted 2 Oct 2026): magnesium (156 members), a statin (80), berberine (38), metformin, semaglutide or tirzepatide (38) and psyllium (34). Magnesium also lowers glucose in trials of people with or at risk of diabetes (Veronese, 2021).

How often should you retest?

The ADA's 2026 intervals are for screening (ADA, 2026):

• HbA1c 5.7% or higher: test at least once a year.
• Normal result: repeat at least every 3 years, sooner if your risk changes, for example after weight gain. Routine testing starts at age 35.

If you are working on it: retest about 3 months after a change. Because HbA1c is weighted toward recent weeks, the ADA notes that meaningful changes can show in under 120 days.

What Phoenix members do: among the 88 members with two or more HbA1c results, the median time from first to latest was 295 days.

My numbers

I carry APOE4/4. Here is my HbA1c over 18 months:

Cutting liquid calories and refined carbs probably moved my HbA1c more than anything else. I also took psyllium before meals and went through Mediterranean and keto phases.

I went from the prediabetes range into Monitor, and I hold myself to the same line as you. One carrier's result, with dates.

What Phoenix members' results show

Where members stand. Of 179 Phoenix members who uploaded an HbA1c result (counted 2 Oct 2026), 79 (44%) are under 5.3%, 82 (46%) are in Monitor and 18 (10%) are at 5.7% or higher. Among the 101 with two copies of APOE4, 49% are under 5.3%.

Who moved. Of the 88 who uploaded two or more results, 14 (16%) moved to a better band, 57 (65%) stayed and 17 (19%) moved to a worse one. Phoenix Research Release 001 saw the same split: of 76 repeat testers, 25 lowered HbA1c by 0.1 point or more and 31 raised it. HbA1c is the stubborn marker. Most carriers hold steady, and as many drift up as come down, which is why I keep working on mine.

One member. An APOE3/4 woman in her 50s went from 5.7% to 5.1% in 338 days: from the talk band to optimal. Release 001 credits magnesium glycinate; her stack also included berberine, ezetimibe and vitamin D3.

Members chose their own changes, and there is no placebo group. How Phoenix reads member data

When should you talk to your doctor?

Talk to your doctor if your HbA1c is 5.7% (39 mmol/mol) or higher. That is the ADA's prediabetes line. A result of 6.5% or higher is in the diabetes range and needs a second test to confirm it.

Between 5.3 and under 5.7%, the result is worth working on. Read it next to your fasting glucose and fasting insulin.

Also raise it if your HbA1c and your glucose readings disagree.

If you take metformin, a GLP-1 medicine or a statin, do not change it because of this page.