Factors Associated With Glycemic Variability in Hospitalized Patients With Type 2 Diabetes Mellitus and Heart Failure

Patients with concomitant heart failure (HF) and type 2 diabetes mellitus (T2DM) are at high risk for adverse clinical outcomes. Glycemic variability (GV) has emerged as a crucial metric for assessing dysglycemia. However, its determinants in this specific patient population remain poorly characterized. This study aimed to investigate factors associated with GV in hospitalized patients with HF and T2DM.

A total of 150 patients hospitalized with T2DM and HF were enrolled. Clinical and laboratory data were collected, and multiple linear regression analysis was performed to identify independent factors associated with four GV indices: standard deviation of blood glucose (SDBG), coefficient of variation (CV), mean of daily differences (MODD), and mean amplitude of glycemic excursions (MAGE).

Multivariate analysis revealed that C-peptide level was significantly negatively associated with all four GV indices (SDBG: β = -0.219, p < 0.001; log(CV): β = -0.080, p < 0.001; MODD: β = -0.176, p < 0.001; MAGE: β = -0.284, p < 0.001). Age showed significant positive associations with SDBG (β = 0.020, p < 0.001), log(CV) (β = 0.009, p < 0.001), and MODD (β = 0.027, p < 0.001). Diabetes duration was significantly positively associated with SDBG (β = 0.028, p < 0.001), log(CV) (β = 0.011, p < 0.001), and MODD (β = 0.029, p < 0.001). Glycated hemoglobin (HbA1c) was significantly positively associated with SDBG (β = 0.062, p = 0.002), MODD (β = 0.125, p < 0.001), and MAGE (β = 0.196, p < 0.001). Log-transformed N-terminal pro-B-type natriuretic peptide (NT-proBNP) (log(NT-proBNP)) was significantly positively associated with all GV indices (SDBG: β = 0.085, p = 0.002; log(CV): β = 0.046, p = 0.002; MODD: β = 0.090, p = 0.009; MAGE: β = 0.162, p = 0.003). Additionally, family history of diabetes was positively associated with SDBG (β = 0.184, p = 0.001), log(CV) (β = 0.088, p = 0.004), and MODD (β = 0.175, p = 0.012). A history of cardiovascular disease was positively associated with MAGE (β = 0.265, p = 0.024). Body mass index (BMI) was negatively associated with MODD (β = -0.036, p < 0.001) but positively associated with MAGE (β = 0.037, p = 0.007). The regression models explained 34.9% to 49.6% of the variance across the different GV indices.

Glycemic variability in hospitalized patients with T2DM and HF is influenced by multiple clinical and metabolic factors. C-peptide level, age, diabetes duration, HF severity (reflected by NT-proBNP), and overall glycemic control are primary factors associated with GV. These findings suggest that clinical management should adopt individualized strategies that account for the heterogeneity and distinct characteristics of different GV indices.