THE INTERPLAY BETWEEN DIET AND THE GUT MICROBIOME IN DETERMINING PLASMA C-REACTIVE PROTEIN LEVELS IN HEALTHY INDIVIDUALS

Background: Diet and the gut microbiome are both established modulators of inflammation. However, their combined effect in shaping systemic inflammation remains unknown. Methods: We analyzed data from two cohorts: 284 men from the Men’s Lifestyle Validation Study and 208 women from the Mind-Body Study. Participants provided stool and blood samples along with dietary information via food frequency questionnaires in 2012-2014. Gut microbiomes were sequenced using shotgun metagenomics and taxonomically profiled with MetaPhlAn4. We assessed the variation in serum high-sensitivity C-reactive protein (CRP) levels with 33 food groups, 315 microbial species, and a combination of the two using random forest models with 10-fold cross-validation. We used principal coordination analysis (PCoA) to determine the first or second principal coordinates axis (PCo1 or PCo2) for interaction analysis between food groups and microbial species. We also examined potential interactions between individual food groups and top species that were strongly correlated to PCo1 and PCo2. Results: In the random forest model, the 33 food groups alone accounted for 8.3% of CRP variation, predominantly by processed meats, French fries, and red meats; the 315 microbial species accounted for 5.8% of CRP variation, predominantly by Flavonifractor plautii, Sellimonas intestinalis, and Eggerthellaceae unclassified SGB14322. Incorporating both food groups and microbial species did not increase the explained variation (6.4%). Interaction analysis revealed overall interactions between dietary components and microbial profiles in relation to CRP, evidenced by significant interactions between food PCo1 and species PCo1 (P_interaction=0.008), food PCo1 and species PCo2 (P_interaction=0.004), and food PCo2 and species PCo1 (P_interaction=0.002). Food PCo1 and PCo2 were positively correlated with CRP (Spearman r=0.18 and 0.14, respectively), whereas species PCo1 showed a negative relationship to CRP (Spearman r=-0.15). Among the individual foods and species, we identified a positive interaction between Clostridium fessum and fish intake, and between Enterocloster citroniae and green leafy vegetable intake in lowering CRP levels. Conclusion: Although the gut microbiome may not add much explanatory power for the influence of diet on systemic inflammation in a healthy population, both overall microbial profiles and specific species may interact with dietary components to influence inflammation.