Glucose Observatory

—

TIRTime in Range: percentage of the day glucose stays between 70-180 mg/dL. The ADA target for non-diabetics is >90%. More actionable than HbA1c because it responds to changes within days.

—

mg/dL

Avg glucose

—

SDStandard Deviation: measures glucose variability. SD <15 mg/dL is excellent, 15-20 is good, >20 indicates elevated metabolic stress. Lower is better.

Variability

—

OptimalOptimal Range: 70-120 mg/dL. A tighter window used in longevity medicine. Standard range is 70-180 mg/dL; optimal is where glucose, energy, and cognition are all stable.

The first CGM finding will appear here once enough meal × glucose data is collected to identify your body’s best and worst foods.

Pending data Dexcom Stelo × MacroFactor cross-reference

Every meal logged in MacroFactor is cross-referenced against CGM glucose readings. Over time, patterns emerge — which foods spike, which don’t, and what context (sleep, stress, timing) changes the curve.

Glucose Response

—

Optimal (70–120)The tighter glucose window used in longevity medicine. Keeping glucose here means stable energy, clear cognition, and minimal metabolic stress.

Longevity medicine target. The tighter window where glucose, energy, and cognition are all stable.

—

In Range (70–180)ADA standard range for non-diabetic individuals. Target >90%. More actionable than HbA1c because it responds to dietary changes within days, not months.

ADA standard for non-diabetic. Target >90%. More actionable than HbA1c because it responds in days.

—

Elevated (>140)

Post-meal spikes above 140 indicate metabolic stress. Reduced by protein-first meal ordering.

Once the broad numbers looked stable, the real experiment began: mapping individual foods to their glucose curves. Every meal became a data point.

Meal Response Table

This table grows over time — every logged meal adds a data point

Meal ↓

Cal ↓

Protein ↓

Carbs ↓

Spike ↓

Grade ↓

Honestly? I don't expect to see much here while I'm eating the way I should be. My glucose tends to stay stable when I'm on. The interesting data will come later — when I reintroduce things, when life gets in the way, when the experiment gets messier. That's when the CGM earns its place.

— Matthew

30-Day Glucose Trend

Avg glucose

TIR %

The trend line tells one story; the individual days tell another. Some days are textbook flat. Others spike and crash. The difference almost always traces back to one variable.

Fiber above 30g/day correlates with lower glucose variability — SDStandard Deviation of glucose readings. Measures how much glucose fluctuates. Lower SD means more stable glucose throughout the day. 14 vs 22 mg/dL on high vs low fiber days.

N=1 · Correlational MacroFactor fiber data × Dexcom Stelo SD — Nutrition Observatory

Cross-Domain Patterns

Glucose doesn’t exist in isolation. Sleep quality, movement, and stress all shape the same curve. These comparisons surface the interactions.

Sleep × Glucose

Same meal, different sleep

Hypothesis: identical meals produce higher spikes after poor sleep. Cortisol from sleep deprivation may impair insulin sensitivity the following day. Testing with paired data.

Comparison chart — populates with data

Movement × Glucose

With vs without post-meal walk

Hypothesis: a 10-minute walk after eating lowers the post-meal glucose peak. Research suggests muscle contraction creates insulin-independent glucose uptake. Testing with paired data.

Walk vs no-walk overlay — populates with data

Stress × Glucose

Glucose and stress level

Hypothesis: high-stress days elevate fasting glucose and widen variability, even with identical food intake. Testing the cortisol-glucose axis with paired data.

Stress correlation — populates with data

The more cross-domain data I layered in, the clearer it became: glucose isn’t just about food. It’s about everything — sleep, movement, stress, timing.

Nocturnal Glucose Patterns

What happens to glucose while you sleep reveals metabolic health that daytime readings can’t capture. The dawn phenomenon, overnight stability, and how sleep architecture shapes nighttime glucose are all visible in CGM data.

Dawn Phenomenon

A natural glucose rise between 4–7 AM driven by cortisol and growth hormone. Normal in healthy individuals. Tracking magnitude over time to establish personal baseline. Currently seeing 10–18 mg/dL pre-dawn rise.

Overnight Stability

Glucose should remain relatively flat during deep sleep (10 PM – 3 AM). Elevated overnight variability often correlates with late-night eating or poor sleep quality. Target: SD <10 mg/dL overnight.

Sleep Architecture × Glucose

Deep sleep stages correlate with the most stable glucose. REM sleep shows slight increases likely tied to cortisol cycling. Nights with >90 min deep sleep show lower morning fasting glucose the next day.

Testing whether sleep quality changes how the same meal affects glucose — this requires repeated meals across different sleep conditions.

Pending data Whoop sleep quality × Dexcom Stelo meal response

“The CGM didn’t change what he eats. It changed the order. Protein first, then everything else — and the spikes just stopped mattering.”

Chronicle · Elena Voss AI narrative voice — Read the chronicle →

Glucose Hypotheses Under Test

H-01

Protein-first meal ordering blunts post-meal glucose spikes

Research-backed → testing with paired CGM data → Nutrition

H-02

Post-meal walking reduces peak glucose

Muscle contraction drives insulin-independent glucose uptake → testing with step data

H-03

Higher fiber intake correlates with lower glucose variability

Testing with MacroFactor fiber logs × CGM variability → Data Explorer

H-04

Poor sleep elevates next-morning fasting glucose

Cortisol pathway → testing with Whoop sleep quality × morning CGM → Sleep

Genomic Context

Why my glucose curve is mine

Genetic variants in FADS2 (fatty acid desaturase) and MTHFR (methylation pathway) affect how individuals metabolize carbohydrates and process glucose differently. Population averages for glycemic index don’t account for these variants — which is exactly why N=1 CGM data matters more than any food label.

Why CGM

I've been curious about glucose for years — specifically whether certain foods respond differently in my body than they might in someone else's. When Dexcom launched the Stelo without a prescription, I didn't need much convincing. It was the first time I could actually get the data without a doctor's visit.

The question I'm interested in isn't whether my glucose is 'good or bad.' It's more specific than that: what does my body actually do with what I eat? Which foods spike me that probably shouldn't? Which ones don't? And what context — sleep, stress, timing, activity — changes the curve?

The CGM doesn't answer those questions on day one. It answers them over months of logged meals and paired data. That's what this page tracks.