CGM: what two weeks of continuous glucose data show, and how to use them without obsessing

What it means

A continuous glucose monitor is a coin-sized sensor with a fine filament under the skin of the upper arm that reads glucose in the interstitial fluid every one to five minutes, for 10–14 days, and sends it to a phone. Readings lag blood glucose by about 10–15 minutes and run a few percent off on single values; the curves — peaks, time to return, overnight baseline — are reliable.

Developed for type 1 diabetes, CGMs are now used by non-diabetics to see post-meal spikes, the dawn rise, the effect of sleep and exercise, and the variability that HbA1c averages away.

Why it matters for longevity

Glucose variability and post-meal peaks predict cardiovascular risk independently of average glucose, and they are invisible on a standard panel. A CGM also shows, within days, how individual the response to a given food is — the finding from the Weizmann PREDICT work that made personalised nutrition a thing. For insulin resistance it is the earliest and most actionable readout available at home.

What the evidence says

Zeevi 2015 (Cell, 800 adults): post-meal glucose responses to identical foods varied enormously between people and were predictable from microbiome and lifestyle data. Hall 2018 (Stanford) found that a quarter of non-diabetics had glucose patterns resembling prediabetes on CGM despite normal HbA1c. Shah 2019 defined normal non-diabetic CGM ranges: median glucose about 95–100 mg/dL, 96% of time between 70–140.

What the CGM has also shown: walking 10–15 minutes after a meal cuts the peak by about a third; eating protein and fibre before carbohydrate roughly halves it; a poor night's sleep raises next-day glucose; and the wide variation between people is real.

How to measure or use it

• A two-week sensor, two or three times a year, is enough; continuous use is for diabetes.
• Non-diabetic targets: fasting 70–90 mg/dL (3.9–5.0 mmol/L), post-meal peak under 140 (7.8), back to baseline within two hours, overnight flat.
• Use it as an experiment log: test your usual meals, the walk-after-meal effect, meal sequencing and sleep, then act on the two or three findings that matter. Pair it with HOMA-IR for the insulin side it cannot see.

Related

• Metabolic health — the full metabolic picture
• Fasting and time-restricted eating — what a CGM shows about meal timing
• The fat-loss protocol — where a CGM fits in a plan