Document resource
Background Energy-restricted diets are a cornerstone treatment for metabolic dysfunction-associated steatotic liver disease (MASLD), yet the circulating proteomic changes accompanying hepatic steatosis improvement remain poorly defined. This study aimed to identify reproducible diet-responsive circulating proteins and determine their independent associations with liver fat content.Methods Proteomic profiling was performed in two independent energy-restricted clinical trials in adults with MASLD ( IDDF2026-ABS-0382 Figure 1. Study design and analytical framework of the dual-trial discovery replication proteomics study): a discovery trial (n=23, structured caloric restriction) and a replication trial (n=24, low-calorie diet in adults with type 2 diabetes). Fasting blood samples collected at baseline, week 12, and week 24 were analysed using data-independent acquisition proteomics on the Orbitrap Astral platform, yielding 865 proteins shared across studies. Time-responsive proteins were identified using linear mixed-effects models under a predefined discovery-replication framework (discovery false discovery rate [FDR]-adjusted P<0.05; replication P<0.05 with consistent effect direction). All 865 proteins were independently tested for association with liver fat (controlled attenuation parameter [CAP] and magnetic resonance imaging proton density fat fraction [MRI-PDFF]), adjusting for age, sex, BMI, timepoint, and intervention arm using the same framework. Pathway enrichment analysis was performed on validated proteins.Results Both trials achieved significant liver fat reduction over 24 weeks (discovery: CAP from 344.2 to 265.9 dB/m; replication: MRI-PDFF from 16.5% to 5.6%), accompanied by reductions in alanine aminotransferase, BMI, and triglycerides. Among 865 proteins, 21 were validated as diet-responsive (19 at week 12, 6 at week 24, 4 at both timepoints; IDDF2026-ABS-0382 Figure 2. Volcano plots of diet-responsive proteins at week 12 (A) and week 24 (B), Venn diagram of validated proteins at both timepoints (C), and temporal trajectories of 21 validated proteins across the 24-week intervention (D)). Three proteins, ADH4, ADH1B, and SPP2, were independently validated as both time-responsive and associated with liver fat content (discovery FDR-adjusted P=0.044, 0.046, and 0.044, respectively). ADH4 and ADH1B showed positive associations with liver fat, while SPP2 showed an inverse association (IDDF2026-ABS-0382 Figure 3. Forest plot of standardised effect sizes (Z = β/SE) for three liver fat-associated proteins (ADH4, ADH1B, and SPP2) in both trials). Pathway enrichment revealed convergence on small molecule metabolic process (10/21 proteins, 3.18-fold enrichment, FDR-adjusted P=0.040) and enzyme regulator activity (11/21 proteins, 3.35-fold enrichment, FDR-adjusted P=0.009).Conclusions Energy-restricted diets induce a reproducible circulating proteomic signature associated with hepatic steatosis improvement in MASLD. ADH4, ADH1B, and SPP2 emerged as robust diet-responsive proteins independently associated with liver fat content, converging on small-molecule metabolism pathways. These hepatocyte-derived proteins may serve as non-invasive biomarkers for monitoring dietary intervention response in MASLD.Abstract IDDF2026-ABS-0382 Figure 1Abstract IDDF2026-ABS-0382 Figure 2Abstract IDDF2026-ABS-0382 Figure 3