Type 2 diabetes subtypes for precision medicine: methodological challenges and alternative prediction‑based approaches – published on 25/07/2026

Tim Mori, Christian Herder, Pedro Cardoso, John M. Dennis, Oliver Kuß

The classification of type 2 diabetes into data-driven subtypes—severe insulin-deficient diabetes (SIDD), severe insulin-resistant diabetes (SIRD), moderate obesity-related diabetes (MOD) and moderate age-related diabetes (MARD)—has attracted growing interest as a framework for precision medicine. However, as subtyping research continues to mature, its methodological foundations deserve careful scrutiny. In this issue, Mori et al (https://doi.org/10.1007/s00125-026-06802-6) address this gap by examining key statistical and conceptual challenges that are sometimes overlooked in the literature. The authors also discuss individualised prediction models, which use routinely available clinical features to guide treatment selection without requiring a discrete subtype label, as a complementary approach. Together, these perspectives offer a timely and balanced assessment of where data-driven subtyping stands today—and what will be needed to translate precision medicine in type 2 diabetes from an aspirational goal into clinical reality. The figures from this review are available as a downloadable slideset.