WJ. Sun, LL. Xing, SF. Zhao and HM. Ying (2026). The effects of ketogenic diet and calorie-restricted diet on metabolic dysfunction-associated steatotic liver disease: a retrospective study. Frontiers in nutrition, 13, 1790674. https://doi.org/10.3389/fnut.2026.1790674

PubMed 42163963

[Abstract]
OBJECTIVE: Metabolic dysfunction-associated steatotic liver disease (MASLD) is a prevalent condition with limited established dietary therapies. Lifestyle modifications, particularly dietary adjustments, represent the primary therapeutic approach for MASLD. This study investigated the effects of ketogenic diet (KD) and calorie-restricted diet (CRD) on hepatic fat accumulation and metabolic parameters in MASLD patients.

METHODS: This retrospective cohort study analyzed 102 MASLD patients completing 12-week nutritional management (KD or CRD). FibroTouch, anthropometric, body composition, and laboratory parameters were assessed at baseline and week 12.

RESULTS: After 12 weeks, the KD group achieved a significantly greater reduction in CAP (ΔCAP) compared to the CRD group (median ΔCAP: 62.0 dB/m (IQR: 48.0-78.5) vs. 36.0 dB/m (IQR: 26.0-52.0); p < 0.001, Cohen's d = 0.93). A clinically meaningful reduction (≥40 dB/m) was achieved in 43 patients (84.3%) in the KD group, compared with 23 patients (45.1%) in the CRD group (p < 0.001). The proportion of patients with moderate-to-severe steatosis decreased more significantly in the KD group (68.6% vs. 39.2%, p < 0.05). No significant between-group difference was observed in liver stiffness measurement (LSM), although both groups showed significant within-group improvements. The KD group demonstrated significantly greater improvements in BMI, fat mass, body fat percentage, visceral fat grade, and waist-to-height ratio (all p < 0.05, Cohen's d range: 0.44-1.07). However, the KD group also exhibited a significantly greater reduction in muscle mass compared to the CRD group (median Δ: 3.0 kg vs. 1.8 kg; p < 0.001, Cohen's d = 0.59). Changes in body weight and waist-to-hip ratio were similar between groups. Among blood biomarkers, only fasting insulin showed a significantly greater reduction in the KD group (median Δ: 6.4 μIU/mL (IQR: 3.4-13.1) vs. 4.7 μIU/mL (IQR: 1.4-9.2); p < 0.05, Cohen's d = 0.37). No significant between-group differences were observed in liver enzymes, lipid profile, fasting glucose, HOMA-IR, or inflammatory markers (all p > 0.05).

CONCLUSION: KD significantly reduced hepatic steatosis and fasting insulin versus CRD, with comparable weight loss but greater muscle loss warranting caution. These findings support KD as a promising MASLD intervention, requiring confirmation in larger trials.