FIB-4 risk categories and cardiovascular events in metabolic dysfunction-associated steatotic liver disease: metabolomic insights from a UK Biobank cohort study

Metabolic dysfunction-associated steatotic liver disease (MASLD) is closely associated with cardiovascular events (CVEs). Although the fibrosis-4 index (FIB-4) is widely used to assess liver fibrosis risk, its associations with CVEs and related metabolic alterations remain unclear. We investigated these associations and evaluated the incremental predictive value of NMR metabolites. This study included UK Biobank participants with MASLD, defined as a Fatty Liver Index ≥60 in the absence of significant alcohol consumption or other liver diseases. FIB-4 risk was categorized as low ( < 1.30), intermediate (1.30–2.67), or high ( > 2.67). The primary outcome comprised heart failure, myocardial infarction, and stroke. Cox proportional hazards models were used to assess associations between FIB-4 categories and CVEs. Within each FIB-4 group, CVE-associated metabolites were screened using nominal P < 0.05 for exploratory analysis, with false discovery rate-corrected results examined in sensitivity analyses. Shared metabolites were assessed using Spearman correlation networks and manual metabolic category enrichment. Cox-LASSO regression with 10-fold cross-validation was used to select a metabolite panel. Discrimination was compared among the Framingham score, Framingham score plus FIB-4, and Framingham score plus FIB-4 and metabolites using AUCs and DeLong tests. Participants with CVEs were older, more often male, and had less favorable cardiometabolic profiles. Compared with the low-risk group, adjusted HRs for CVEs were 1.57 (95% CI, 1.49–1.66) for intermediate risk and 2.31 (95% CI, 2.00–2.66) for high risk. Sixty-four nominally associated metabolites were shared across the three FIB-4 groups, mainly involving cholesterol, lipids, lipoprotein particles, and phospholipids. AUCs were 0.663 for the Framingham score, 0.668 after adding FIB-4, and 0.684 after further adding metabolites. Higher FIB-4 risk was associated with greater CVE risk among participants with MASLD. Shared metabolic features predominantly reflected lipid transport and lipoprotein particle composition, suggesting a common metabolic background rather than a causal pathway.