All studies
Metabolism·June 2, 2026·1 min read

Same Meal, Different Bodies: What PREDICT Taught Us About Food

The largest study of its kind fed a thousand people identical meals — and found their bodies responded up to ten times differently. What you ate explained almost none of it.

For decades, nutrition science ran on averages. PREDICT — the largest study of its kind — did something radical instead: it fed around a thousand people the exact same meals under controlled conditions, then used continuous monitors to watch how each body actually responded.

~1,000
people, identical meals
10×
difference in response between people
3.6%
of the difference explained by the meal itself
r = 0.77
accuracy once many signals are combined

The same meal is not the same meal

Two people can eat an identical muffin — one barely moves, the other spikes for hours. PREDICT found up to a tenfold difference between people in how their blood sugar and blood fat reacted to the very same food.

Same meal. Two bodies.
Steady responderSpiker
701001301601900306090120 minmg/dL

Illustrative blood-sugar response to one identical meal — a steady responder versus a spiker.

So what decides your response?

This is the finding that rewired the field. The researchers measured everything — gut microbiome, blood markers, sleep, activity, genetics, and the meal itself — then asked what actually explained the differences between people. The meal's own nutrition came near the bottom.

What explains why two people differ?
Gut microbiome7.5%
Body basics (age, sex, BMI, blood)7.1%
What the meal contained“what you eat”3.6%
Context (sleep, activity, timing)2.1%
Genetics“your genes”0.8%

Share of the person-to-person variation in blood-sugar response explained by each factor (PREDICT 1).

Your gut microbiome mattered more than what was on the plate. Your genes barely mattered at all. That is the quiet death of "tell me your DNA and I'll tell you what to eat."

After eating beats fasting

A normal check-up measures you fasting. But the differences between people show up far more clearly after a meal — exactly the window a continuous monitor can see, and a standard blood test cannot.

After-meal responses vary far more than fasting tests
Blood fat after eating103%
Blood sugar after eating68%
Insulin after eating59%
Fasting blood fat50%
Fasting blood sugar10%

Variation between people (coefficient of variation). A bigger bar means people differ more.

Why this is the whole idea behind HealthOS

No single number defines you, and no generic guideline fits you. When PREDICT combined many signals into one model, it predicted a person's glucose response with an accuracy (r = 0.77) that no single factor came close to. That is HealthOS in one sentence: measure many things continuously, learn your patterns, and act on what is true for you.

References

For informational purposes only. For medical advice or diagnosis, consult a professional.

Written by

HealthOS Research

Keep reading