AI Algorithm Predicts Diabetic Kidney Disease through Blood Tests

Notable advancements have been made in devising treatments for kidney disease in diabetic patients. Yet, evaluating an individual's risk for kidney disease based solely on clinical factors can be challenging. Consequently, identifying who is most susceptible to developing diabetic kidney disease is a vital clinical need. Now, scientists have created a computational method that predicts the likelihood of a person with type 2 diabetes developing kidney disease, a common yet severe diabetes complication. This could aid physicians in preventing or improving the management of kidney disease in type 2 diabetes patients.

The new algorithm developed by researchers from Sanford Burnham Prebys (La Jolla, CA, USA) and the Chinese University of Hong Kong (CUHK, Hong Kong) relies on measuring a process known as DNA methylation, which is the accumulation of subtle changes in the DNA. DNA methylation can provide essential insights into gene activation and deactivation and can be easily measured via blood tests.

Utilizing comprehensive data from over 1,200 type 2 diabetes patients registered in the Hong Kong Diabetes Register, the researchers constructed their model which they also tested on an independent group of 326 Native Americans with type 2 diabetes. This confirmed the model's predictive power for kidney disease across diverse populations. The researchers are presently fine-tuning their model and extending its application to address other health and disease-related inquiries, such as why some cancer patients do not respond favorably to certain treatments.

“This study provides a glimpse into the powerful future of predictive diagnostics,” said co-senior author Kevin Yip, Ph.D., a professor and director of Bioinformatics at Sanford Burnham Prebys. “Our team has demonstrated that by combining clinical data with cutting-edge technology, it’s possible to develop computational models to help clinicians optimize the treatment of type 2 diabetes to prevent kidney disease.”

“Our computational model can use methylation markers from a blood sample to predict both current kidney function and how the kidneys will function years in the future, which means it could be easily implemented alongside current methods for evaluating a patient’s risk for kidney disease,” added Yip.

Related Links:
Sanford Burnham Prebys
CUHK