Background <p>The dietary index for gut microbiota (DI-GM) is a newly proposed metric for assessing diet quality linked to gut microbiota. However, prospective evidence is scarce on the associations between DI-GM and adverse liver outcomes.</p> Methods <p>The DI-GM was calculated by averaging the intakes of 12 foods and nutrients. Elastic net regression was performed to identify metabolites associated with DI-GM and metabolic signature reflecting higher adherence to DI-GM was constructed. Cox proportional hazards regression and mediation analyses were employed to explore the potential associations and mechanisms.</p> Results <p>This prospective cohort study included 168,456 participants from the UK Biobank. Compared to participants with DI-GM scores of 0–3, those scoring ≥ 6 presented 22% lower risk of MASLD (<i>HR =</i> 0.78, 95% <i>CI</i> = 0.68–0.90). Metabolic signature for DI-GM and dietary index beneficial to gut microbiota (BDI-GM) were also inversely correlated with MASLD. Similar inverse correlations between DI-GM and BDI-GM and the risks of other chronic liver diseases were identified. Furthermore, phenotypic age, body mass index, metabolic score, inflammatory score, and metabolic signature significantly mediated the relationship between DI-GM and MASLD. No significant interactions were observed between DI-GM and polygenic risk score of hepatic steatosis, and the associations between DI-GM and adverse liver outcomes persisted regardless of genetic risk.</p> Conclusions <p>Higher adherence to DI-GM significantly correlates with reduced risks of MASLD and other chronic liver diseases, independent of genetic susceptibility. And the apparent mediating effects of five indices highlight the role of aging, obesity, metabolic disorders, inflammation, and metabolomic alterations in the association between DI-GM and MASLD. Further research is warranted to evaluate the utility of metabolic signatures in metabolic profile monitoring and risk stratification.</p> Impact and implications <p>This large-scale cohort study first demonstrates that higher adherence to a gut microbiota-beneficial diet (DI-GM) is associated with a lower risk of MASLD and other chronic liver diseases, independent of genetic susceptibility. The estimated population attributable fractions, while derived from observational data and requiring cautious interpretation, suggest that a substantial portion of liver disease cases in the study population might be linked to suboptimal DI-GM adherence. These findings underscore the importance of integrating gut microbiome health into public health strategies for liver disease prevention, offering a practical approach to reduce disease burden at both individual and population levels. The DI-GM-associated metabolic signature represents a candidate objective biomarker meriting evaluation in future studies for its potential in early risk assessment. Mediation analyses further reveal that a diet promoting healthy gut microbiota may reduce MASLD risk by maintaining gut microbiota homeostasis, decelerating biological aging, ameliorating obesity, attenuating metabolic disorders, alleviating inflammation, and altering metabolome. Collectively, this study generates important hypotheses and provides a rationale for future interventional research to determine whether promoting DI-GM-aligned diets can effectively reduce liver disease risk at the population level.</p> Graphical abstract <p></p>