Early Precise Identification and Intervention Strategies for Individuals at High Risk of Prediabetes

Overview

Prediabetes significantly increases the risk of developing diabetes, cardiovascular and cerebrovascular diseases, tumors, and dementia. Early identification and intervention have become a leading focus in current diabetes prevention and control research. Currently, prediabetes screening primarily relies on methods such as fasting blood glucose, oral glucose tolerance tests, and glycated hemoglobin. These approaches suffer from limitations including single-point assessment, static nature, cumbersome procedures, poor reproducibility, delayed diagnosis, and limited accuracy. Continuous glucose monitoring (CGM) technology offers advantages such as ease of use, dynamic continuous monitoring, and round-the-clock surveillance. It comprehensively captures glucose fluctuation patterns, enabling identification of occult hyperglycemia and glucose variability. Integrating artificial intelligence (AI) to perform deep analysis on CGM-generated big data holds promise for pioneering new pathways toward earlier and more precise identification of prediabetes. This project aims to establish a prospective prediabetes cohort integrating multidimensional data-including CGM parameters, body composition analysis, clinical indicators, and biomarkers-to develop novel diagnostic models for prediabetes. Building upon this foundation, we will construct an AI-driven prediabetes intervention management platform with intelligent decision support. This platform will generate personalized intervention strategies based on risk stratification, providing scientific evidence and practical support for advancing diabetes prevention and enabling precision management.

In recent years, the global prevalence of prediabetes has surged rapidly, with approximately 9% of adults exhibiting impaired glucose tolerance and 6% showing impaired fasting glucose. More concerning is the rising prevalence of prediabetes in China. Cross-sectional studies indicate that the prevalence of prediabetes in China increased from 35.7% to 38.1% between 2013 and 2018. Pre-diabetes not only significantly elevates diabetes risk (with annual conversion rates as high as 5-10%) but is also closely associated with multiple adverse clinical outcomes including all-cause mortality, cardiovascular and cerebrovascular diseases, microvascular complications, retinopathy, tumors, and dementia. Therefore, timely identification of pre-diabetic individuals and effective early intervention are critical measures for curbing the diabetes epidemic and mitigating related complications. Existing screening methods for prediabetes exhibit notable limitations. Current domestic and international guidelines primarily recommend fasting plasma glucose (FPG), oral glucose tolerance test (OGTT), and glycated hemoglobin (HbA1c) as diagnostic criteria for prediabetes. However, these methods suffer from cumbersome procedures, delayed diagnosis, and inadequate detection of occult hyperglycemia, making them ill-suited for large-scale screening and precise diagnosis of prediabetes. Therefore, developing a more efficient, precise, and convenient early identification system for prediabetes is urgently needed. Among novel glucose monitoring technologies, CGM demonstrates significant advantages. CGM provides continuous, comprehensive, and reliable 24-hour glucose information, detects occult hyperglycemia and hypoglycemia, and reveals trends and characteristics of glucose fluctuations. This facilitates a shift from HbA1c-based glucose management to a comprehensive glucose management model, while being less susceptible to interference from factors such as anemia and chronic kidney disease. Our research group previously applied CGM technology to type 2 diabetes patients, securing the Chinese invention patent "Method for Optimizing Blood Glucose Management and Monitoring in Type 2 Diabetes Using Continuous Glucose Monitoring Technology (CN 109637677A)". However, research on applying CGM to prediabetes screening is still in its infancy. Currently, there is a lack of early identification and stratified intervention systems for high-risk prediabetes populations, as well as scalable personalized intervention strategies. Prediabetes is associated with visceral fat accumulation. Body composition analysis offers a non-invasive and convenient method for detecting body fat distribution, muscle mass percentage, and other parameters, demonstrating significant potential in prediabetes risk assessment. However, the independent predictive value of each relevant parameter and the optimal cutoff points require further investigation. Furthermore, metabolomics and proteomics research has identified multiple potential biomarkers for prediabetes. However, most biomarker detection relies on high-end platforms like mass spectrometry or nuclear magnetic resonance, facing challenges such as high testing costs, insufficient standardization, lack of validation in large prospective cohorts, and undefined diagnostic cut-off points, which limit their clinical translation. This project will employ novel intelligent monitoring technologies such as CGM and body composition analysis. By integrating multidimensional data including clinical information, laboratory indicators, and biomarkers, and leveraging artificial intelligence and big data technologies, it aims to establish a comprehensive "screen-assess-prevent" system for high-risk populations. This system will cover early diagnosis, precise classification, and personalized intervention, thereby advancing diabetes prevention efforts, improving early detection rates of prediabetes, effectively prevent diabetes and related complications, and provide scientific evidence and practical pathways for proactive chronic disease health management.