Proteomic resolution of the MASLD cardiometabolic spectrum identifies sex-driven endotypes and predicts systemic mortality

Abstract

Background: Metabolic dysfunction-associated steatotic liver disease (MASLD) is a heterogeneous condition where overlapping cardiometabolic risk factors challenge accurate prognostic prediction. Objectives: We aimed to define the molecular heterogeneity underlying MASLD clinical presentation, resolving disease patterns and identifying sub-phenotypes that account for prognostic differences in the general population. Design: We applied latent class analysis (LCA) to 48 806 UK Biobank participants with MRI-based proton density fat fraction (MRI-PDFF) data to define phenotypic subtypes over a >15-year follow-up. Large-scale proteomic profiling characterised the molecular basis of the identified prognostic patterns. This led to the derivation of a four-protein score (PS4), which was robustly validated for all-cause mortality prediction in an expanded cohort with/ without MRI-PDFF data (n=115 971). We used mediation analysis to assess the potential contribution of PS4 proteins to the observed class-outcome associations. Results: The LCA identified clinically distinct MASLD subtypes, revealing pronounced sexual dimorphism in disease severity and identifying high-risk prognostic subgroups. The derived PS4 score significantly outperformed established non-invasive scores in predicting all-cause mortality across the expanded cohort. The PS4 components, linked to hepatic lipid/one-carbon metabolism (FABP1, FTCD) and immune/vascular integrity (ADGRG1, GAST) pathways, were found to account for a significant fraction (10-20%) of the estimated effect of the patient's phenotypic class on mortality outcome. Conclusions: MASLD-driven proteomic shifts reflect biological pathways that connect hepatic metabolic dysfunction with long-term survival. By resolving the MASLD cardiometabolic spectrum, we identified high-risk endotypes that are not captured by traditional liver-centric scores. The PS4 represents a biologically grounded framework for deciphering the systemic consequences of MASLD.