Endocrinology

GLP-1 RAs outperform DPP-4 inhibitors for kidney, CV, and mortality outcomes in T2D without CKD

This emulated target trial matched 24,510 new GLP-1 RA users to DPP-4 inhibitor users with T2D but no baseline CKD. GLP-1 RAs reduced composite kidney outcomes and incident CKD versus DPP-4 inhibitors (both HR 0.85). Risks of dialysis, acute kidney injury, major adverse cardiovascular events, and all-cause mortality were all significantly lower with GLP-1 RAs. Benefits were consistent across subgroups, supporting GLP-1 RAs over DPP-4 inhibitors when kidney–heart protection is prioritized. 1

Incretin-based weight loss often includes substantial lean mass and skeletal muscle loss

This systematic review of 36 RCTs assessed body composition with liraglutide, semaglutide, tirzepatide, or dulaglutide in adults with elevated BMI. Incretin therapies produced greater total weight, fat mass, and visceral fat loss than placebo or lifestyle comparators. Across incretin arms, a median 34.9% of weight loss came from muscle-related indices, frequently exceeding prespecified lean-mass benchmarks. Few trials prioritized body composition, and none reported objective physical function, limiting understanding of clinical consequences. 2

Glycaemic variability emerges as a key driver of cardiomyocyte dysfunction and myocardial injury risk

This multimodal study combined human genetic data, mouse models, and in vitro systems to dissect myocardial risk in diabetes. Traditional cardiovascular risk markers and mean hyperglycaemia did not fully explain excess post-MI mortality in diabetes. Experimental models showed glycaemic variability caused greater cardiomyocyte dysfunction and myocardial injury susceptibility than sustained hyperglycaemia alone. Patient plasma assays suggested glycaemic variability markers may predict subclinical cardiac injury. 3

Updated CGM-derived GMI improves agreement with HbA1c across glycaemic ranges

Investigators analyzed 26,647 paired average glucose and HbA1c measurements from 18,860 individuals to derive an updated GMI (uGMI). The HbA1c–GMI regression slope improved from 1.4 to 1.0 with uGMI, effectively eliminating proportional bias. Absolute bias at HbA1c <5.5% and >9.0% dropped from >0.4% to ≤0.1% (NGSP units). Performance was similar across Abbott Freestyle Libre and Dexcom systems, indicating device independence. 4