Plasma multi-omics signatures of the EAT-Lancet diet and associations with health outcomes

Background:
The EAT-Lancet diet aims to optimize human health and environmental sustainability, but the biological changes associated with this dietary pattern remain incompletely characterized. We aimed to identify plasma proteomic and metabolomic signatures associated with adherence to the EAT-Lancet diet and to examine their associations with long-term health outcomes.

Methods:
We analyzed 204,938 UK Biobank participants with at least one 24-h dietary recall and baseline covariate data. Plasma proteins and metabolites associated with the EAT-Lancet diet score were screened using multivariable regression, and diet-related omics signatures were derived using least absolute shrinkage and selection operator regression. Associations with 45 incident health outcomes and life expectancy were examined using multivariable Cox and flexible parametric survival models. Mediation analyses were performed to elucidate underlying biological pathways.

Results:
Here we show that diet-related omics signatures consist of 163 proteins and 37 metabolites. These signatures show stronger associations with multiple chronic disease outcomes than the self-reported diet score and are associated with larger estimated differences in life expectancy. Mediation analyses support candidate pathways involving immune-inflammatory regulation and lipid-related metabolism.

Conclusions:
Plasma multi-omics signatures provide complementary biological information on EAT-Lancet diet adherence and disease risk. These signatures should be interpreted as baseline biological correlates and candidate intermediates rather than validated causal mechanisms or replacements for dietary assessment.