Loading exact experiment records
Metabolic Observatory
Connecting
Public n-of-1 experiment · live from the VPS

What happens when meals meet a minute-by-minute body?

A continuously updating, evidence-aware study of glucose, food, movement, heart rate, sleep and body measurements. Every chart is generated from the experiment’s source records—not a curated screenshot.

Current glucose—mg/dL
Mean glucose—mg/dL · complete window
Tight range—70–140 mg/dL · descriptive
Variability—coefficient of variation
GMI—6-day estimate · not laboratory A1C
Experiment volume—exact source records
01 / Live metabolic trace

Glucose, meals and movement

GlucoseMealWorkout
02 / Daybook

Six days, side by side

Mean, range and activity reveal context that a single “average glucose” obscures.
03 / Metabolic clock

Glucose by hour

Median and 10th–90th percentile envelope, IST.
Current state of evidence

A strong signal is not the same thing as a strong conclusion.

—minute-level glucose readings
—timestamped meal exposures
6 daystoo short for causal or stable personal claims
Meal response laboratory

Every meal, one common measurement frame

Baseline: median −15→0 min. Response: exact CGM 0→180 min. Overlapping meals and workouts are flagged.

Select a meal

Response ranking

Carbohydrate and excursion

Meal categories

Observed response families

Descriptive only: categories overlap and exposures are not randomized.
Continuous physiology

Heart rate with workout context

Exact source samples; dense during recorded walks.
Activity

Steps, active minutes and mean glucose

Body measurements

Weight and composition

Short-term scale movement is dominated by hydration and conditions.
Sleep windows

Measured sleep architecture and overnight glucose

Only nights exposed by Health Connect are shown; missing nights are not imputed.
Independent analysis

Findings, hypotheses and next experiments

Language is intentionally graded: observation ≠ association ≠ causal inference.
Exploratory association matrix

Meals, macros and response

Meals without another logged intake inside 2h; Spearman ρ, unadjusted p and BH q remain exploratory.
Recommendations

Highest-value next steps

Methodology

How each number is made

Evidence base

Primary references

Analysis catalogue

What this observatory measures

Implemented views and the next-stage causal roadmap, separated by domain.
Data quality ledger

What exists, what is missing, what is never invented

The source databases remain full-resolution; dashboard metrics are computed from exact records.
Live pipelines

Source → VPS → public analysis

Known limitations

Interpretation boundaries

    Accessible records

    Daily and meal data tables

    Keyboard and screen-reader alternative to the canvas charts.