In a large cohort of over 11,000 adults with diabetes or prediabetes, systemic inflammatory indices —such as NLR, MLR, SIRI, SII, and AISI—were found to independently predict all-cause and cardiovascular mortality. Notably, the monocyte-to-lymphocyte ratio (MLR) had the strongest association with mortality risk. Leveraging machine learning, especially XGBoost, researchers ranked SIRI as the top predictor for cardiovascular mortality, and the integration of these indices improved risk prediction beyond traditional models. Population-level analysis estimated that optimizing inflammatory profiles could prevent 10–20% of mortality events in this vulnerable population.
To read more about novel risk stratification strategies in diabetes and prediabetes, click here . ##Reference## Zhang Z, Li C, Xiao Y, et al. Integrated machine learning and population attributable fraction analysis of…
