Abstract
AIMS: To evaluate the short-term metabolic effects of a low-carbohydrate diet (LCD) compared with a control diet (CON) in hospitalized patients with insulin-deficient diabetes, with particular focus on glycemic variability (GV). METHODS: This non-randomized clinical trial included 359 inpatients with fasting C-peptide levels ≤1.0 ng/mL. Participants chose either an LCD or CON during hospitalization. GV was assessed using capillary and continuous glucose monitoring, with coefficient of variation (CV) as primary outcome. Secondary measures included other GV metrics, antidiabetic medication use, ketone levels, and adverse events. Subgroup analysis was conducted based on baseline C-peptide levels. RESULTS: LCD led to a greater reduction in CV (diet-by-time interaction, p = 0.03), as well as improvements in MAGE, SD, TIR, and iAUC compared with CON. LCD patients also showed reduced requirements for oral agents and insulin injection at discharge. Benefits were more pronounced in individuals with lower C-peptide levels (P for interaction = 0.002). No increase in adverse events was observed. Ketone levels increased in the LCD group and negatively correlated with discharge blood glucose and CV. CONCLUSION: Short-term LCD intervention improved GV in insulin-deficient diabetes, especially in patients with more significant β-cell dysfunction, with blood ketones negatively correlated with blood glucose and GV. CLINICAL TRIAL REGISTRATION: ChiCTR2000038006, https://www.chictr.org.cn/showproj.html?proj=60712.