Abstract Aims Continuous glucose monitoring (CGM) enables scalable digital phenotyping by capturing detailed daily glucose profiles and has been shown to improve glycaemic control in individuals with type 1 diabetes (T1D). We hypothesized that clustering CGM-derived metrics could identify distinct glycemic phenotypes associated with increased insulin resistance, systemic inflammation, and risk of metabolic derangements, with potential implications for optimization of clinical care. Materials and methods This cross-sectional study applied a hierarchical clustering approach to stratify 75 individuals with T1D based on CGM metrics obtained over 14 days (Libre ProIQ). Clustering was performed using Ward’s minimum variance linkage and Euclidean distance on six CGM metrics (average glucose, coefficient of variance, time above range, time below range, time in range, and frequency of low glucose events). The resulting groups, poorly controlled diabetes (PCD) and moderately controlled diabetes (MCD), were compared for insulin resistance markers (estimated glucose disposal rate, eGDR), metabolic dysfunction-associated steatotic liver disease (MASLD) risk markers (fatty liver index, FLI; hepatic steatosis index, HSI), and inflammation and endotoxaemia markers. Results Compared to MCD, subjects in the PCD group had significantly higher insulin resistance (eGDR: 4.4 vs. 7.3 mg/kg/min, p = 0.001), elevated liver enzymes (ALT: 26 vs. 20 U/L, p = 0.005; AST: 28 vs. 23 U/L, p = 0.049), increased MASLD risk markers (HSI: 37 vs. 35, p = 0.033; FLI: 29 vs. 17, p = 0.034), and higher systemic inflammation (CRP: 1.82 vs. 0.81 mg/L, p = 0.020; lipopolysaccharide-binding protein (LBP): 11.7 vs. 8.4 EU/mL, p = 0.024). Time above range correlated positively with liver and inflammatory markers, while time in range correlated negatively. Mediation analysis suggested that FLI partially mediated the effect of poor glycemic control on inflammation. Conclusions In this cross-sectional cohort of adults with T1D, poorer CGM-derived glycaemic control was associated with higher systemic inflammation and higher surrogate-marker-defined MASLD risk.
Fedulovs et al. (Mon,) studied this question.