Glucose Responses to Aerobic Exercise Are Influenced by Glucose Levels before the Exercise and Menstrual Cycle Phase in Adults with Type 1 Diabetes on Multiple Daily Insulin Injections

Abstract

Aims:

This study aimed to evaluate how pre-exercise glucose levels and menstrual cycle phase influence glucose responses to aerobic exercise in adults with type 1 diabetes (T1D) treated with multiple daily insulin injections, with the goal of improving personalized exercise management.

Materials and Methods:

We analyzed 51 moderate-intensity, 30-min aerobic sessions (17 male, 34 female) from the TAILOR/1a study. Glucose (plasma and interstitial), heart rate, and anthropometric data were collected. Female participants performed sessions in both the follicular and luteal phases. Glucose trends during exercise were clustered using k-medoids and Dynamic Time Warping. Features from clinical, glucose, and heart rate data were extracted and correlated with cluster assignment.

Results:

Two distinct glucose response patterns emerged: descending and stable. Higher pre-exercise glucose levels were associated with greater glucose decline and increased hypoglycemia risk, particularly in men. Female participants more frequently exhibited stable glucose profiles, particularly during the luteal phase. Fitness and body composition influenced cluster assignment: fitter individuals—particularly women—were more likely to exhibit stable glucose trends. Pre-exercise glucose was the strongest predictor of response. The menstrual cycle phase had a modest but noticeable effect on glucose dynamics.

Conclusions:

Glucose response to exercise in T1D is highly variable and influenced by pre-exercise glycemia, sex, fitness level, and menstrual cycle phase. Women, particularly during the luteal phase, demonstrated more stable glycemic responses. These findings support the need for individualized exercise recommendations and for integrating physiological and behavioral factors into predictive models for automated insulin dosing and exercise guidance in T1D.