The participants ate the same meal one week apart (7-day rotating menus, 14 days total) and continuous glucose monitors (CGMs) were used to track their glucose responses.

The participants were exposed to 4 diets:

Minimally processed, plant-based, low-fat diet^[1]
Minimally processed, animal-based, very-low carbohydrate (ketogenic) diet^[1]
Highly ultraprocessed food diet^[2]
Minimally processed food diet^[2]

The glucose responses of each participant to the same meal on separate occasions were inconsistent and highly variable according to the following results:

There was a weak-to-moderate correlation (r = 0.45) between glucose responses to the same meal.
The difference between the 2 glucose responses was within −29 to 32 mg/dL 95% of the time.
Approximately 80% (72%–83%) of the variation was estimated as due to differences within the participants (biological factors or external variables) or measurement error.

Variability in the glucose responses to identical meals, as measured using continuous glucose monitors

Although baseline glucose, carbohydrate content of meals, energy intake, snack intake, and exercise were identified as potential predictors of individual variability in the postmeal glucose responses to duplicate meals, they explained less than 33% of the variability. None of the results were meaningfully changed after adjusting for these predictors.

Continuous glucose monitors reveal variable glucose responses to the same meals Original paper