What question did this study set out to answer?

This study aims to quantify how physical activity characteristics and initial metabolic states affect glycemic outcomes in individuals with Type 1 Diabetes (T1D).

June 7, 2026

1064-OR: Independent Impact of Physical Activity Characteristics and Initial Metabolic State on Glycemic Outcomes in the DBLG1 System: A Linear Mixed-Model Analysis

Key Points

This study aims to quantify how physical activity characteristics and initial metabolic states affect glycemic outcomes in individuals with Type 1 Diabetes (T1D).
Analyzed 20,959 physical activity sessions from 1,720 adult users of the DBLG1 System.
Used linear mixed-effects models (LMMs) to predict time below range (TBR) and time above range (TAR).
Evaluated fixed effects including initial glycemia, insulin-on-board, PA duration, and declared intensity.
Higher initial insulin-on-board increased TBR by 0.58 percentage points (p<0.001).
Greater initial glycemia decreased TBR (-0.41 pp per 10 mg/dL) but increased TAR (+4.9 pp per 10 mg/dL).
Short, high-intensity sessions were linked to a 4.2 pp increase in TBR, while long-duration sessions raised TAR by 5.1 pp.

Abstract

Introduction and Objective: Exercise-induced hypoglycemia and hyperglycemia remain challenging for people with T1D. Variability in outcomes is often driven by physical activity (PA) characteristics and metabolic states. This study quantified the independent impact of declared intensity, duration, initial Insulin-On-Board (IOB) and initial glycemia on glycemic outcomes using Linear Mixed Effects Models (LMM). Methods: We analyzed 20,959 PA sessions from 1,720 adult users of the DBLG1 System. LMMs predicted Time Below Range (TBR 70 mg/dL) and Time Above Range (TAR 180 mg/dL) during activity. Fixed effects included initial glycemia, initial IOB, PA duration (30 min, 30-60 min, 60 min), declared intensity (low, moderate, high), and the intensity-duration interaction. Patient ID was included as a random effect to account for repeated measures. Results: Initial metabolic conditions were the strongest predictors. Each unit of IOB at start increased TBR by 0.58 percentage points (pp) (p0.001) and rescue carbohydrate intake by 0.82g. Conversely, higher initial glycemia protected against hypoglycemia (-0.41 pp TBR per 10 mg/dL) but significantly increased TAR (+4.9 pp per 10 mg/dL). A significant intensity-duration interaction was observed: short, high-intensity sessions (30 min) were associated with the highest hypoglycemia risk (+4.2 pp absolute TBR), while long-duration sessions (60 min) were associated with increased hyperglycemia (+5.1 pp absolute TAR). This suggests a more specific adaptation for short, intense effort for prolonged activity. Conclusion: While the system effectively personalizes treatment, this analysis identifies potential improvements. High IOB and short, intense exercise are a risk factor for hypoglycemia, while long-duration activities are a risk factor of hyperglycemia. Future algorithms could incorporate more specific adaptation to intensity-duration combination and dynamic IOB-aware adjustments to improve the glycemic control. Disclosure A. Adenis: Employee; Current; Diabeloop SA. S. Lachal: Employee; Current; Diabeloop SA. P. Gauthier: Employee; Current; Diabeloop SA. P. Gimenez: Employee; Current; Diabeloop SA. C. Desir: Employee; Current; Diabeloop SA. T. Le Roux-Mallouf: None. E. Huneker: Employee; Current; Diabeloop SA. P.Y. Benhamou: Employee; Current; Diabeloop SA. Advisory Panel; Ended; Eli Lilly and Company, Novo Nordisk.

Mark Helpful

Bookmark

Relay