Abstract
Introduction & Objectives
Healthy diets have been linked to greater cardiometabolic health, however, underlying biological mechanisms are still elusive. We therefore aimed to identify metabolomic signatures of dietary habits and investigate their relationships with established cardiometabolic biomarkers.
Material & methods
The analyses were conducted among 2571 participants of the Cooperative Health Research in South Tyrol (CHRIS) study. Information on dietary intake assessed through a semi-quantitative food frequency questionnaire, targeted metabolomics data (175 serum metabolites) and cardiometabolic biomarkers (cLDL, cHDL, triglycerides, cholesterol, CRP, Hba1c) were collected at baseline visit. We derived dietary patterns through principal component analysis and identified two main patterns, i.e. western and prudent-style diet. Moreover, several pre-selected dietary indices were derived (Mediterranean Diet Index, Alternate Healthy Eating Index 2010, overall, healthy and unhealthy plant based dietary indices). Associations between dietary variables and metabolites were determined through multiple regression analysis, and relationships between metabolites and cardiometabolic biomarkers using Spearman correlations.
Results
The study sample comprised of 53.8% females with mean age=46y (SD=16.7) and BMI=25.9 (SD=4.6). Overall, 38 metabolites, mainly phospholipids, were associated with multiple dietary variables, whereas 21 were specifically related to one. Directions of the associations were heterogeneous within metabolite classes. Several diet related metabolites were significantly correlated with cardiometabolic biomarkers. E.g. glutamate and four phosphatidylcholines (PC aa C38:3/C38:4/C40:4/C40:5) showed positive relationships with both, adherence to an unhealthy diet and cardiometabolic biomarkers (Spearman ρ>0.17, adjusted p-values<0.05).
Conclusions
Multiple metabolites relate to specific dietary habits and highlight the potential to identify pathways linking diet to disease development.