Background Obesity perturbs metabolism, resulting in differential diet–disease associations, but metabolic pathways that differ by obesity remain unclear. We investigated whether cross‐sectional associations between metabolites and diet quality differ by obesity status in 2832 participants in the CARDIA (Coronary Artery Risk Development in Young Adults) cohort study. Methods We used untargeted metabolomics (7255 metabolite peaks) and an a priori diet quality score (APDQS) to (1) compare model accuracy and prediction, stratified by body mass index (BMI); (2) assess modification of APDQS‐metabolite associations by BMI in linear regression models; (3) identify pathways with differential metabolic activity by BMI; and (4) assess the strength and direction of associations with identified/annotated metabolites. Results In BMI‐stratified models, metabolite peaks had higher model accuracy but similar APDQS predictive ability among participants with lower versus higher BMI (BMI<25: Q 2 =0.52; 25≤BMI<30: Q 2 =0.41; BMI≥30: Q 2 =0.21). Of the 7255 metabolite peaks, 180 peaks were associated with APDQS and differed by BMI (false discovery rate<0.1). From pathway enrichment analyses, 6 pathways had differential metabolic activity by BMI (Fisher's exact test <0.05) including primary bile acid biosynthesis; steroid biosynthesis; steroid hormone biosynthesis; caffeine metabolism; valine, leucine, and isoleucine biosynthesis; linoleic acid metabolism. Strength of associations were strongest for those without overweight or obesity. Conclusions Diet‐metabolite associations were best detected among participants who did not have overweight/obesity. We identified associated diet–metabolic pathways that differ by obesity, suggesting potential metabolic‐related mechanisms perturbed for those with obesity.
Hullings et al. (Tue,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: