Long Oral Presentation Nutrition Society of Australia 50th Annual Scientific Meeting 2026

Beyond weight loss: Body composition partially mediates the associations of Mediterranean diet adherence with metabolic dysfunction-associated steatotic liver disease, hospitalisation, and mortality (146748)

Elena S George 1 , Ayesha Sualeheen 1 , Sze-Yen Tan 1 , Nicole Kiss 1 , Robin M Daly 1 , Shelley E Keating 2 , Jacob George 3 , Stuart K Roberts 4 5 , Ekavi N Georgousopoulou 6 7
  1. Institute for Physical Activity and Nutrition, School of Exercise and Nutrition Sciences, Deakin University, Geelong, VIC, Australia
  2. School of Human Movement and Nutrition Sciences, The University of Queensland, Brisbane, QLD, Australia
  3. Storr Liver Centre, Westmead Institute for Medical Research, Westmead Hospital and University of Sydney, Westmead, NSW, Australia
  4. Department of Gastroenterology, Alfred Health, Melbourne, VIC, Australia
  5. Central Clinical School, Department of Medicine, Monash University, Melbourne, VIC, Australia
  6. School of Medicine Sydney, University of Notre Dame Australia, Sydney, NSW, Australia
  7. Discipline of Nutrition and Dietetics, Faculty of Health, University of Canberra, Canberra, NSW, Australia

Metabolic dysfunction-associated steatotic liver disease (MASLD) affects >30% of adults globally, with adiposity being an intrinsic feature contributing to liver and cardiovascular disease (CVD) outcomes (1, 2). Although the Mediterranean diet (MedDiet) is recommended for MASLD management, the extent to which its associations with liver and CVD outcomes are mediated by body weight and body composition remains unclear. Therefore, this study investigated whether body weight and body composition mediated the relationships between MedDiet adherence and MASLD, liver, and CVD outcomes. The study included 119,536 UK Biobank participants with at least one 24-hour dietary recall and complete sociodemographic and clinical data. MedDiet adherence was assessed using the Mediterranean Diet Score (MedDietScore), and MASLD was identified using the Fatty Liver Index (FLI) (3, 4). Mediators included body weight and body composition parameters including whole-body fat mass, fat percentage and fat to fat-free mass ratio, measured through bioimpedance analysis. Longitudinal analyses were conducted using hospitalisation and mortality outcomes using linked health records, with liver- and CVD-related events identified using ICD-10 codes (5). Cross-sectional associations were estimated using regression coefficients (β), and longitudinal associations using Cox proportional hazards models, reported as hazard ratios (HRs). Time-on-study was used as the underlying time scale. Mediation analysis was performed to quantify the direct and indirect effects of MedDiet adherence on outcomes variables, stratified by sex. Overall body weight and all body composition parameters were inversely associated with MedDietScore in both sexes. Body weight significantly mediated the association between MedDiet and MASLD, with a greater indirect association observed in females (β = −0.061, 95% CI −0.066, −0.056) than in males (β = −0.051, 95% CI −0.055, −0.047). Significant indirect associations of MedDiet with MASLD through all body composition parameters were observed in both sexes, while direct associations remained significant after accounting for mediators. Longitudinally, body weight and body composition parameters partially mediated associations between MedDiet and liver-related hospitalisation (3–5% lower hazard), CVD-related hospitalisation and mortality (2–3%), and all-cause mortality (1–2%), with indirect effects corresponding to 1–5% reductions in hazard across mediators. However, the direct association remained significant only in males for liver-related hospitalisation (16–18%) and only in females for CVD-related hospitalisation (8%). For CVD mortality and all-cause mortality, the direct association remained significant in both sexes, except for fat mass in females for CVD mortality. In this UK Biobank cohort, higher MedDiet adherence was associated with lower MASLD risk and reduced liver-related and CVD-related hospitalisation and mortality outcomes. These associations were partly mediated by lower body weight and favourable body composition, with residual direct effects and sex differences suggesting pathways beyond adiposity. These findings support MedDiet as an optimal dietary strategy for MASLD management in non-Mediterranean populations, with potential long-term clinical benefits.

  1. Younossi ZM, Kalligeros M & Henry L (2025) Clin Mol Hepatol 31, S32–S50.
  2. Targher G, Byrne CD & Tilg H (2024) Gut 73, 691.
  3. Panagiotakos DB, Pitsavos C, Arvaniti F et al. (2007) Prev Med 44, 335–340.
  4. Bedogni G, Bellentani S, Miglioli L et al. (2006) BMC Gastroenterol 6, 33.
  5. George ES, Sualeheen A, Freer C et al. (2026) Br J Nutr, 1–11.