The goal of this interventional study is to build a high quality, real world multimodal dataset that combines continuous glucose monitoring (CGM), wearable and fitness data, performance metrics, and saliva and urine omics collected during a prolonged, moderate intensity outdoor gravel-cycling session in adults with type 1 diabetes (T1D).
The main questions it aims to answer are:
* Can we collect and synchronize comprehensive CGM, physiological, performance, and omics data around a single cycling session to enable further artificial intelligence (AI) model development?
* What molecular changes in saliva and urine occur during exercise, and how do they relate to glycemic outcomes?
Participants will:
* Complete a supervised \~75 km gravel-cycling route at their own pace under real-world conditions, without protocolized therapy adjustments.
* Wear a Dexcom G7 starting \~4 days before the ride and continue through the sensor lifespan to capture CGM data.
* Provide saliva and urine immediately before and after the ride for epigenomic and proteomic analyses.
This study will generate an integrated resource that supports the development and validation of AI models for predicting glucose responses to exercise in T1D and will help guide future studies on how prolonged exercise affects glucose control.