Short-term postprandial glucose monitoring reveals stable traits from noisy free-living meals
Abstract
Postprandial glucose responses differ greatly between meals, which complicates the identification of individual glucose response traits in free-living conditions. We analysed 50,463 free-living meals from 992 adults without diagnosed diabetes using a multivariate mixed-effects machine-learning framework and evaluated transfer to 4,524 standardized meals excluded from training. This model separated the population-shared from participant-specific patterns across four joint glucose response outcomes. Within-individual variability dominated free-living responses, yet the between-individual variation was concentrated in two dimensions. The dominant axis represented an overall tendency toward larger glucose excursions. The second axis reflected carbohydrate responsiveness with lower late glucose elevations. Dominant-axis scores were reproducible across non-overlapping meal subsets and associated with held-out standardized-meal responses, beyond age, sex, BMI, and glucose mean and variability. Dominant-axis scores estimated from 3 days of monitoring approximated those estimated from 14 days. These findings support short-term monitoring to measure glucose response individuality and identify a candidate coordinate for stratification.
Related
- The Second-Meal Effect Under Free-Living Conditions: A Continuous Glucose Monitoring Study of Meal History and Postprandial Glycemia
- Letter to the Editor: Dietary Antioxidant Minerals (Cr, Mg, Cu, Se, Zn) in Diabetic Children and their Relationship with Fasting and Postprandial Blood Glucose
- Continuous glucose monitoring reveals thyroid feedback-related residual glycemic variability and hypoglycemia during inpatient glucose optimization in euthyroid adults with type 2 diabetes: a CGM-based observational study.
- Continuous glucose monitoring reveals glycemic profile in adrenal insufficiency patients on glucocorticoid replacement
- Postprandial glucose management of people with insulin-treated diabetes
BibTeX
@misc{toumi2026shortterm,
title = {{Short-term postprandial glucose monitoring reveals stable traits from noisy free-living meals}},
author = {M. Toumi and M. Salathe},
howpublished = {medRxiv},
year = {2026},
month = sep,
doi = {10.64898/2026.09.14.26362991},
url = {https://www.medrxiv.org/content/10.64898/2026.09.14.26362991v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - Short-term postprandial glucose monitoring reveals stable traits from noisy free-living meals
AU - M. Toumi
AU - M. Salathe
PY - 2026
DA - 2026/09/14
PB - medRxiv
DO - 10.64898/2026.09.14.26362991
UR - https://www.medrxiv.org/content/10.64898/2026.09.14.26362991v1
AB - Postprandial glucose responses differ greatly between meals, which complicates the identification of individual glucose response traits in free-living conditions. We analysed 50,463 free-living meals from 992 adults without diagnosed diabetes using a multivariate mixed-effects machine-learning framework and evaluated transfer to 4,524 standardized meals excluded from training. This model separated the population-shared from participant-specific patterns across four joint glucose response outcomes. Within-individual variability dominated free-living responses, yet the between-individual variation was concentrated in two dimensions. The dominant axis represented an overall tendency toward larger glucose excursions. The second axis reflected carbohydrate responsiveness with lower late glucose elevations. Dominant-axis scores were reproducible across non-overlapping meal subsets and associated with held-out standardized-meal responses, beyond age, sex, BMI, and glucose mean and variability. Dominant-axis scores estimated from 3 days of monitoring approximated those estimated from 14 days. These findings support short-term monitoring to measure glucose response individuality and identify a candidate coordinate for stratification.
ER -