An open-label, randomized controlled trial to assess a ketogenic diet in critically ill patients with sepsis.

Patients with sepsis experience metabolic and immunologic dysfunction that may be amplified by standard carbohydrate-based nutrition. A ketogenic diet (KD) may offer an immunologically advantageous alternative, although clinical evidence is limited. We conducted a single-center, open-label, randomized controlled trial to assess whether a KD could induce stable ketosis in critically ill patients with sepsis. Secondary outcomes included assessment of feasibility and safety of KD, as well as explorative analysis of clinical and immunological characteristics. Forty critically ill adults were randomized to either a ketogenic or standard high-carbohydrate diet. Stable ketosis was achieved in all KD patients, with significant increases in β-hydroxybutyrate levels compared with controls [mean difference 1.4 milimoles per liter; 95% confidence interval (CI): 1.0 to 1.8; P < 0.001). No major adverse events or harmful metabolic side effects (acidosis, dysglycemia, or dyslipidemia) were observed. After day 4, none of the patients in the KD group required insulin treatment, whereas in the control group, insulin dependency ranged between 35% and 60% (P = 0.009). There were no differences in 30-day survival, but ventilation-free [incidence rate ratio (IRR) 1.7; 95% CI: 1.5 to 2.1; P < 0.001], vasopressor-free (IRR 1.7; 95% CI: 1.5 to 2.0; P < 0.001), dialysis-free (IRR 1.5; 95% CI: 1.3 to 1.8; P < 0.001), and intensive care unit-free days (IRR 1.7; 95% CI: 1.4 to 2.1; P < 0.001) were higher in the ketogenic group. Next-generation sequencing of CD4+/CD8+ T cells and protein analyses showed reduced immune dysregulation, with decreased gene expression of T-cell activation and signaling markers and lower pro-inflammatory cytokine secretion. This trial demonstrated the safe induction of a stable ketogenic state in sepsis, warranting larger trials to investigate potential benefits in sepsis-related organ dysfunction.