Machine learning uncovers manganese as a key nutrient associated with reduced risk of steatotic liver disease

Abstract Background Metabolic dysfunction‐associated steatotic liver disease (MASLD) affects approximately 20%–30% of the general population and is linked to high‐caloric western style diet. However, there are little data that specific nutrients might help to prevent steatosis. Methods We analysed the UK Biobank (ID 71300) 24 h‐nutritional assessments and investigated the association between nutrient intake calculated from food questionnaires and hepatic steatosis indicated by imaging or ICD10‐coding. The effect of manganese (Mn) on subgroups with risk single nucleotide polymorphism carriage as well as the effect on metabolomics was investigated. All analyses are corrected for age, sex, body mass index, Townsend index for socioeconomic status, kcal, alcohol, protein intake, fat intake, carbohydrate intake, energy from beverages, diabetes, physical activity and for multiple testing. Results We used a random forest classifier to analyse the feature importance of 63 nutrients and imaging‐proven steatosis in a cohort of over 25 000 UK Biobank participants. Increased dietary Mn intake was associated with a lower likelihood of MRI‐diagnosed steatosis. Subsequently, we conducted a cohort study in over 200 000 UK Biobank participants to explore the relationship between Mn intake and hepatic or cardiometabolic outcomes and found that higher Mn intake was associated with a lower risk of ICD‐10 coded steatosis (OR = .889 [.838–.943], p < .001), independent of other potential confounders. Conclusion Our study provides evidence that higher Mn intake may be associated with lower odds of steatosis in a large population‐based sample. These findings underline the potential role of Mn in the prevention of steatosis, but further research is needed to confirm these findings and to elucidate the underlying mechanisms.