Digital signatures of asthma: mechanisms and trajectories from multimodal wearable data

Failing to detect and respond to deteriorating asthma costs lives. Each year, patients experience avoidable attacks and unnecessary admissions because warning signs are missed. The days after hospital discharge are especially high-risk. Current monitoring of asthma is inadequate, relying heavily on self-report and lacking objective, timely signals. Wearable sensors allow continuous capture of physiology in real-world settings. Beyond single measures such as heart rate, respiratory rate and cough frequency, wearables can reveal complex physiological patterns, known as digital signatures. Continuous multimodal physiology from wearables has not yet been applied in adult asthma care, leaving a critical gap. This Fellowship will define digital signatures of asthma through multimodal wearable data integrated with clinical outcomes. I will generate a unique dataset linking high-dimensional physiology with clinical trajectories, enabling discovery through machine learning across rest, activity and sleep transitions, and providing mechanistic insight into altered physiological coordination. I will examine how digital signatures vary across asthma phenotypes and recovery trajectories and evaluate their potential to serve as continuous markers of lung physiology, supporting personalised monitoring. The discoveries made in this Fellowship will lay the foundations to minimise the high-risk period following discharge, reduce readmissions, and pave the way for proactive monitoring.