Use of synthetic data for digital biomarkers from wearables for glycaemic event monitoring and analysis of the impact of diabetes treatment on cardiovascular complications

Type 2 Diabetes (T2D) is one of the most prevalent chronic diseases worldwide, affecting around 6% of the global population and contributing to over one million deaths annually. Its rising burden, partially linked to social, environmental, and lifestyle factors, represents a critical healthcare challenge. This use case explores how synthetic data can support the development of digital biomarkers from wearable devices for real-time glycaemic event monitoring and improve risk prediction for cardiovascular complications in T2D patients with overweight or obesity. 

The Challenge 

T2D management is complicated by glycaemic variability, frequent comorbidities, and the long-term risk of cardiovascular disease. Continuous glucose monitoring (CGM) and wearable devices provide rich, real-time data, but the availability and accessibility of such data are limited by privacy restrictions and bias in existing datasets. At the same time, conventional clinical trials often require large control arms that are difficult to establish for T2D and related comorbidity subgroups. New approaches are needed to optimise algorithms for early detection of glycaemic events, refine treatment strategies, and reduce the risk of diabetes-related cardiovascular complications. 

Our Research Questions 

1. How can continuous glucose monitoring (CGM) data be utilized to develop algorithms that predict glycaemic events and optimize glycaemic control in patients with type 2 diabetes (T2D)?  
2. How can real-world clinical and monitoring data and SD enhance the optimization of algorithms to improve risk prediction for cardiovascular complications in patients with T2D and overweight or obesity?
3. What role do comorbidities, such as chronic kidney disease and heart disease, play in shaping clinical outcomes for T2D patients, and how can these insights inform regulatory guidelines?
4. Can synthetic data help uncover the interplay between glycaemic variability and long-term complications in T2D? 

Our Approach 

• Develop and optimise algorithms to extract biomarkers from non-invasive wearables (e.g. CGM, smartwatches) for real-time monitoring and early detection of hyperglycaemia and hypoglycaemia. 
• Use synthetic data to correct bias and support comparative studies of treatment effects, particularly in patients living with overweight or obesity and cardiovascular risk.  
• Utilize synthetic data to reinforce an external control arm in clinical trials investigating the effectiveness and safety of new therapies for T2D, addressing the challenge of randomized trials with varied standard-of-care control arms. 

Envisioned Impact 

The type 2 diabetes use case is expected to demonstrate the value of synthetic data by enabling real-time monitoring of glycaemic events through digital biomarkers embedded in wearable devices. This will improve patient quality of life by supporting timely intervention and reducing the risk of chronic complications, while also promoting healthcare system sustainability. At the same time, synthetic data will strengthen cardiovascular risk prediction models and refine treatment recommendations for diabetic and prediabetic individuals, especially those with obesity and comorbidities. These contributions are anticipated to enhance patient outcomes, reduce cardiovascular complications, and inform regulatory decision-making, ultimately fostering a more personalised and effective approach to diabetes care. 

Use Case Leadership