Dynamic FIB-4 trajectory and a multi-state Markov model analysis reveal fibrosis progression and cardiovascular-cerebrovascular risk in MAFLD patients

Metabolic associated fatty liver disease (MAFLD) is becoming increasingly prevalent in China. Identifying patients at high risk of fibrosis is crucial for clinical management. This study aimed to assess the dynamic Fibrosis-4 (FIB-4) trajectory and its association with cardiovascular-cerebrovascular events (CCEs) and explore the progression of fibrosis in patients with MAFLD.

This retrospective analysis included adult patients with at least three records who underwent physical examination between January 2017 and December 2023. A latent class mixed model (LCMM) was used to fit the FIB-4 trajectory. The multistate Markov model was applied to assess transitions in advanced liver fibrosis risk states. A total of 12,839 patients were analyzed, with 5777 MAFLD patients included in the final study. During a median follow-up of 46.2 months, CCEs occurred in 221 patients (3.83%). Based on the LCMM model, three trajectory groups were identified: Class 1 (low-stable), Class 2 (high-stable), and Class 3 (high-rising). In the multivariable-adjusted models, both class 2 and class 3 were significantly associated with the incidence of CCEs, with aHR (95% CI) of 2.73 (2.01-3.71) and 2.46 (1.25-4.86), respectively. Additionally, class 3 showed a rapid increase in FIB-4 at approximately 30 months. The multi-state Markov model indicated that the probability of high-risk of advanced liver fibrosis alleviating to medium-low risk was significantly higher than the probability of medium-low risk progressing to high risk.

Monitoring FIB-4 trajectory aids in the early identification and intervention of high-risk MAFLD patients, potentially preventing irreversible fibrosis and reducing CCEs risk.