Continuous glucose monitoring reveals distinct clinical implications of glucose levels at the start and end of hemodialysis in people with type 2 diabetes

BACKGROUND: The clinical significance of glucose levels at the start and end of hemodialysis (HD) sessions in people with type 2 diabetes remains insufficiently explored. METHODS: We retrospectively analyzed 116 individuals with type 2 diabetes undergoing HD (age 61 ± 12 years; 69% male; HbA1c 6.5 ± 1.3%; glycated albumin 21.1 ± 6.9%). Continuous glucose monitoring (CGM) data were collected from two dialysis institutions to evaluate sensor glucose levels at the start and end of HD sessions and their associations with CGM metrics and hypoglycemia. RESULTS: (β = 0.360, P < 0.0001). HD-induced hypoglycemia occurred in 25 participants (21.6%), including 14 (12.1%) during HD and 22 (19.0%) after HD. Lower end-of-HD glucose was significantly associated with all-cause hypoglycemia (OR 0.97; 0.95-0.98), HD-induced hypoglycemia (OR 0.96; 0.94-0.98), hypoglycemia during HD (OR 0.93; 0.90-0.97), and post-dialysis hypoglycemia (OR 0.95; 0.93-0.98; all P < 0.0001). An end-of-HD glucose level < 90 mg/dL was linked to increased all-cause (76.9% vs. 17.8%) and HD-induced (61.5% vs. 10.0%) hypoglycemia, demonstrating clinically meaningful discrimination for hypoglycemia risk stratification. CONCLUSION: Glucose levels at the start and end of HD sessions demonstrate distinct clinical relevance. End-of-HD glucose level serves as a useful marker for hypoglycemia risk, highlighting the potential importance of tailored monitoring strategies to support glycemic risk assessment in this high-risk population.