Discordance Between FIB-4 and BAST Fibrosis Risk Classifications in Obese Patients with MASLD: Results from the OBREDI-TR Study

<b>Background/Objectives</b>: Non-invasive fibrosis scores are widely used for risk stratification in metabolic dysfunction-associated steatotic liver disease (MASLD); however, their performance in obese individuals remains controversial. The Fibrosis-4 (FIB-4) index is commonly recommended as a first-line tool, yet it may underestimate fibrosis risk in severe obesity. The BAST score, which incorporates metabolic and anthropometric parameters, has been proposed as an alternative. This study aimed to characterize both the degree and direction of discordance between FIB-4 and BAST in obese patients with MASLD. <b>Methods</b>: This predefined secondary analysis included 2950 adults with obesity (BMI ≥ 30 kg/m²) and MASLD from the multicenter OBREDI-TR cohort. Fibrosis risk categories were assigned using standard cut-offs for FIB-4 and BAST, and agreement was assessed using weighted Cohen's kappa. Associations among discordance patterns, obesity class, and the visceral adiposity index (VAI) were evaluated using chi-square tests and general linear models. <b>Results</b>: Overall agreement between FIB-4 and BAST was very poor (κ = 0.041, <i>p</i> < 0.001). Discordance was observed in 22.3% of patients and increased markedly with obesity severity. In class III obesity, discordance was predominantly driven by low-risk classification according to FIB-4 despite high-risk classification according to BAST. Patients with this discordant pattern exhibited significantly higher VAI values than concordant cases (<i>p</i> < 0.001), independently of the study center. <b>Conclusions</b>: In obese patients with MASLD, particularly those with morbid obesity, FIB-4 frequently classifies patients as low risk, while BAST identifies elevated fibrosis risk. This systematic discordance suggests that FIB-4 may underestimate fibrosis burden in the context of severe obesity and visceral adiposity, supporting the need for a phenotype-oriented, multimodal approach to fibrosis risk assessment.