Harnessing dysglycaemia heterogeneity for precision type 2 diabetes prevention

Type 2 diabetes (T2D) is a heterogeneous, progressive metabolic disease and a major contributor to cardiovascular disease worldwide. Current prevention strategies, centred on population-level lifestyle recommendations, have had limited success, highlighting the need for more comprehensive approaches that account for interindividual variation in intervention response.

Although discrete T2D subtypes using genomics or clinical variables have been proposed for targeted diabetes prevention and care strategies, emerging evidence supports a more dynamic view of heterogeneity as a continuous spectrum of metabolic dysfunction. Continuous glucose monitoring and multiomics profiling have shown promises in detecting early metabolic dysfunction in individuals with normal glycemia.

In this review, we synthesize advances in modelling T2D phenotypic continuum and highlight how emerging technologies, including continuous glucose monitoring and multiomics profiling, can detect early metabolic dysregulation. We also discuss methodological challenges, such as the need for standardized deep phenotyping, robust analytic frameworks, and inclusion of diverse populations to ensure equity in translation. Finally, we outline future directions for translating these insights into scalable, mechanism-informed interventions that address glycaemic progression from subclinical changes to more advanced forms of the disease.