Metabolomic Profiles Associated with BMI, Waist Circumference, and Diabetes and Inflammation Biomarkers in Women

Objective This study was undertaken to identify metabolites associated with BMI and waist circumference (WC) in women and to determine whether these metabolites are associated with biomarkers of metabolic health. Methods Untargeted metabolomic analysis was done on serum from 1,534 women. Metabolites associated with BMI and WC were identified using linear regression with a Bonferroni-corrected P value. Clustered blocks of these metabolites were then defined whose association with the anthropometric measures could be represented by a single metabolite. The association of these representative metabolites with biomarkers for diabetes and inflammation was then determined. Results About one-third of 781 metabolites included in the analyses were associated with BMI and/or WC. Associations were found for some novel metabolites, including several sphingolipids, nucleotides, and modified fatty acids. Among metabolites most strongly inversely associated with BMI, the choline-containing plasmalogen (O-16:0/18:1) (beta = -0.30, P = 6.62 x 10(-32)) was also inversely associated with c-peptide and positively associated with adiponectin. Adjustment for BMI attenuated the metabolite-biomarker associations more for hemoglobin A1c (> 100%) and c-peptide (58.8% to > 100%) than for C-reactive protein (10.5%-40.0%) and adiponectin (7.0%-30.4%). Conclusions These results add to the list of metabolites associated with adiposity and indicate that some may influence processes that contribute to the development of obesity-related diseases.