Introduction and Objective: While high-dose metformin use often signals advanced disease, the clinical relevance of dose-associated proteomic profile remains unclear. We aimed to identify a dose-dependent proteomic profile and associations with clinical traits and cardiometabolic events.Methods: In a discovery cohort of 1,246 Koreans with type 2 diabetes and eGFR >30 mL/min/1.73 m², we analyzed 5,416 proteins (Olink HT Explore) using multivariable linear regression against metformin doses, adjusting for age, sex, BMI, eGFR, HbA1c, diabetes duration and other drugs. A weighted proteomic score was constructed from the identified proteins and validated in 32,192 UK Biobank participants. Associations between score quartiles (Q1-Q4) and clinical traits or cardiometabolic events were assessed using multivariable regression and Cox model.Results: We identified 17 proteins significantly associated with metformin dose (Bonferroni-adjusted P < 0.05), including GDF15, REG4, and SPINK1. A weighted proteomic score was developed using 16 of these proteins (excluding RBP4, unavailable in the UK Biobank). In the UK Biobank, higher score groups showed a higher proportion of metformin use. However, in a longitudinal subset of the Korean cohort (N=287) with mean follow-up of 7 years, an increase in metformin dose was associated with a reduction in the score (P < 0.0002). The proteomic score was strongly linked to adverse clinical profiles, including higher HbA1c, BMI, and waist-to-hip ratio, and lower eGFR (Q4 vs. Q1; P < 2×10-16). Also, the score predicted cardiometabolic risk; individuals with Q4 (highest) had a significantly higher risk of myocardial infarction (HR 2.40; 95% CI 1.99-2.91) and heart failure (HR 2.19; 1.82-2.63) than those in Q1 (lowest).Conclusion: While the score was derived from metformin dose, it characterizes high-risk traits such as poor glycemic control, renal impairment, and higher cardiometabolic risk. This signature reflects the underlying metabolic milieu rather than metformin effects and may aid risk prediction.
S. Lee: None. M. Kho: None. S. Lee: None. H. Lee: None. H. Lee: None. J. Choi: None. J. Lee: None. J. Park: None. Y. Jo: None. S. Kwak: None.
Source link
Leave a Reply