Abstract
BACKGROUND: Metabolic dysfunction-associated steatotic liver disease (MASLD) is a major global health concern. Ultra-processed foods (UPFs) have been implicated in metabolic disorders, yet prospective evidence linking UPF intake to MASLD remains limited. This study examined the association between UPF consumption and incident MASLD in an Iranian adult population. METHODS: We analyzed data from 5,058 adults (20-70 years) in the Tehran-based Monitoring of Metabolic Diseases Risk Factors (MMRT) cohort. Dietary intake was assessed using a validated 125-item food frequency questionnaire, with UPFs classified according to the NOVA system. MASLD was diagnosed using transient elastography (controlled attenuation parameter ≥ 238 dB/m) alongside metabolic criteria. Multivariable logistic regression models estimated odds ratios (ORs) and 95% confidence intervals (CIs) for MASLD incidence across quartiles of UPF intake. Restricted cubic spline models assessed dose-response relationships, and mediation analysis evaluated the role of adiposity. RESULTS: Over a median follow-up of 5 years, 562 incident MASLD cases were identified. In the fully adjusted model, participants in the highest versus the lowest UPF quartile had higher odds of MASLD (OR = 2.02; 95% CI: 1.23-3.32; p for trend < 0.01). Restricted cubic spline analysis revealed a non-linear association, with risk increasing up to 125 g/day and plateauing thereafter. Mediation analysis suggested that adiposity accounted for 21.1% of the association. Subgroup analyses showed consistent associations across demographic and metabolic strata (all p for interaction > 0.05). CONCLUSIONS: Higher UPF consumption was associated with increased MASLD risk, partly mediated by adiposity. These findings underscore the need for dietary strategies aimed at reducing UPF intake to mitigate liver-related metabolic disease risk.