Researchers from Innotomy Consulting Data Science Center and Lata Research Center in collaboration with M&H Research Center in the United States have developed DiaBeats, an algorithm that delivers highly accurate and specific screening for prediabetes based on electrocardiogram (ECG) results. The work was published in the BMJ Innovations journal, signaling a notable advancement in how clinicians might approach early detection of glucose-related disorders. DiaBeats represents a significant step forward in translating routine heart rhythm data into meaningful insights about metabolic health, particularly in the early, often asymptomatic stages of disease.
The study aggregated clinical data from 1,262 participants, each contributing a digital ECG alongside relevant health information. The data were partitioned into three distinct cohorts: one to train the algorithm, another to validate its performance, and a third for independent testing to gauge real-world applicability. Within the participant pool, the prevalence of type 2 diabetes and prediabetes stood at 30% and 14%, respectively. In an independent validation set, DiaBeats demonstrated a predictive accuracy of 97.1% for detecting either diabetes or prediabetes, highlighting its potential as a robust screening tool that could complement existing methods.
Early detection of type 2 diabetes and prediabetes is crucial because timely intervention can reduce the risk of progression and associated complications. Current standard screening approaches—such as the glucose tolerance test and measurements of glycosylated hemoglobin (HbA1c)—can be invasive, time-consuming, or impractical for large-scale population screening. Importantly, diabetes and prediabetes are known to induce structural and functional changes in the heart even before symptoms emerge. These cardiac alterations can be observed through ECG analysis, and the DiaBeats algorithm could broaden access to early diagnosis by leveraging a noninvasive, inexpensive modality that many patients already undergo as part of routine care. The research team emphasizes that while ECG-based screening should not replace conventional diagnostic tests, it can function as a powerful triage tool to identify individuals who would most benefit from follow-up testing and lifestyle interventions, potentially reducing the burden on healthcare systems. The implications for public health are meaningful, offering a scalable approach to improve early detection in diverse populations across Canada and the United States (BMJ Innovations).