What fasting glucose should an APOE4 carrier aim for?

The fasting glucose bands

The line rule. A "between" band includes both ends, and "100 or higher" includes 100. So 89 is Optimal, 90 is Monitor, 100 is Talk to your doctor, and 69 is below the range. In mmol/L: 4.9 is Optimal, 5.0 is Monitor, 5.6 is Talk to your doctor.

Confidence in the optimal band: moderate. It is anchored to the lab range and to where diabetes risk starts rising, and it is wide enough not to flip on test noise.

What does fasting glucose tell you?

Fasting glucose is your blood sugar on one morning, after a night without food.

Risk starts rising inside the normal range. In 13,163 men aged 26 to 45, all under 100 mg/dL, diabetes risk climbed step by step from 87 mg/dL against men under 81 (Tirosh, 2005). In 2,067 older adults without diabetes, an average blood sugar of 115 mg/dL against 100 went with an 18% higher dementia rate (hazard ratio 1.18) (Crane, 2013).

What it misses. One morning is noisy. Your fasting glucose varies about 5.7% from day to day (Chai, 2017). At 83 mg/dL, that means any single morning can read about 9 mg/dL above or below your true average. HbA1c, your three-month average, is the steadier number: in 11,092 adults without diabetes, HbA1c predicted heart disease and death more strongly than fasting glucose (Selvin, 2010). Phoenix reads the two together, with fasting insulin.

What does APOE4 change?

Nothing gives carriers a separate glucose number.

• In 3,695 Finnish adults followed for 11 years, fasting glucose did not predict cognition, and insulin resistance predicted decline the same way with or without APOE4 (Ekblad, 2017).
• A 2025 meta-analysis of 170 studies found that diabetes raised Alzheimer's risk only in non-carriers (Huang, 2025).
• 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).

For a carrier, fasting glucose is what it is for anyone: an early warning of where blood sugar is heading. The brain-linked numbers are your average blood sugar (HbA1c) and insulin resistance (fasting insulin).

What do the guidelines say?

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

• Impaired fasting glucose (prediabetes): 100 to 125 mg/dL (5.6 to 6.9 mmol/L).
• Diabetes: 126 mg/dL (7.0 mmol/L) or higher, confirmed by a second abnormal result unless the diagnosis is obvious.

The ADA adds that "risk is continuous, extending below the lower limit of the range." The World Health Organization still starts impaired fasting glucose at 110 mg/dL; the ADA moved its line down to 100 in 2003 (ADA, 2026). None of these lines was written for APOE4 carriers.

Why did Phoenix choose 70 to 89?

I chose 70 to 89 because a target should not flip on test noise.

1. The top line, 89. Diabetes risk began rising from about 87 mg/dL in the young-men study (Tirosh, 2005), so upper-normal results of 90 to 99 get their own Monitor band.
2. The width. One person's fasting glucose swings about 9 mg/dL either side of their true average. Phoenix's old app band, 80 to 85, was narrower than that swing, so the same person moved in and out of "optimal" with no real change.
3. No 80 floor. Results of 70 to 79 are normal, and nothing shows harm there. The old floor labelled them "not optimal" with no evidence.

The band is wider than before, and still tighter than your lab at the top end, where it counts.

How do you get an accurate fasting glucose?

Fast for at least 8 hours with no calories: water is fine (ADA, 2026). Test in the morning, and pair it with HbA1c and fasting insulin on the same draw.

Three things to know:

• The day matters. The ADA lists food, stress, recent illness and activity as short-term factors, plus time of day, medicines, alcohol and smoking (ADA, 2026). Test on an ordinary morning, not the day after a fever.
• Handling matters. Glucose keeps falling in a blood sample unless the plasma is separated straight away or the tube is kept on ice (ADA, 2026). A sample that sat around reads low.
• Units. US labs report mg/dL; most other countries report mmol/L (divide mg/dL by 18).

What moves fasting glucose?

Ranked by the size of the drop in the best trial data. The higher you start, the more it falls.

What the table shows:

• Lifestyle is the strongest 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).
• Add intervals and weights. In prediabetes, plain aerobic exercise alone (5.2 mg/dL lower) was too small to be sure of (Bennasar-Veny, 2023).
• Berberine fades. Its effect shrank in trials that ran longer than 90 days, used more than 2 g a day, or enrolled people over 60 (Liang, 2019).

One lever pushes it up: statins. Across 13 trials and 91,140 people, statins raised new diabetes by 9%: one extra case for every 255 people treated for 4 years, against a much larger heart benefit (Sattar, 2010). If you take one, watch your glucose and HbA1c, and do not stop it because of this page.

What Phoenix members use. Magnesium glycinate with lower glucose is the largest repeated pattern in Phoenix Research Release 001: 12 members. Among Phoenix members who log their stack (counted 2 Oct 2026): magnesium (156 members), berberine (38), metformin, semaglutide or tirzepatide (38) and psyllium (34).

How often should you retest?

The ADA's 2026 screening intervals cover fasting glucose and HbA1c together (ADA, 2026):

• Prediabetes (100 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.

Every time you test: take fasting glucose with HbA1c and fasting insulin. Because one morning can be 9 mg/dL off, never act on a single Monitor result: repeat it, or average two. Averaging repeat tests makes the reading far more consistent (Chai, 2017).

What Phoenix members do: of the 212 Phoenix members who uploaded blood work (counted 2 Oct 2026), 152 (72%) have tested more than once.

My numbers

I carry APOE4/4. In December 2024 my fasting glucose was 104 mg/dL (5.8 mmol/L), in the prediabetes range.

Since then my HbA1c has fallen from 5.9% to 5.4%. That is why I never read glucose alone: one morning's number does not tell the story. I also use a continuous glucose monitor at times. One carrier's result.

What Phoenix members' results show

Where members stand. Of 66 Phoenix members who uploaded a confirmed fasting glucose result (counted 2 Oct 2026), 39 (59%) are in the 70 to 89 band and 21 (32%) are in Monitor. The median is 88 mg/dL.

The repeated pattern. Magnesium glycinate with lower glucose showed up in 12 members in Phoenix Research Release 001, the largest repeated pattern in the report. Magnesium also lowers glucose in placebo-controlled trials (Veronese, 2021).

One member. An APOE4/4 woman in her 60s went from 98.7 to 85 mg/dL in 70 days: from Monitor to optimal. She took magnesium glycinate and started tirzepatide in the same period, and Release 001 credits the two together.

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 fasting glucose is 100 mg/dL (5.6 mmol/L) or higher. That is the ADA's line for impaired fasting glucose. A result of 126 mg/dL (7.0 mmol/L) or higher is in the diabetes range and needs a second test to confirm it.

Mention a result under 70 mg/dL (3.9 mmol/L), especially if you felt unwell that morning.

Between 90 and 99, the result is worth working on. Repeat it before you read much into it, and read it next to your HbA1c and fasting insulin.

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