Differences in metabolite profiles between blood matrices,ages, and sexes among Caucasian individuals and their inter-individual variations

Endobiotic metabolites are associated with biological processes in the body and therefore may serve as biomarkers for disease states or therapeutic efficacy and toxicity. However, information is limited regarding how differences between blood matrices, patient backgrounds, and sample handling affect human metabolite profiles. Our objective was to obtain metabolite profiles from Caucasian individuals, based on different matrices (plasma and serum), subject backgrounds (male/female and young/old), and storage conditions (2 or 10 freeze–thaw cycles). In total, 297 metabolites were detected by LC/MS and GC/MS, and more than 75 % of them were highly represented in all sample groups. The multivariate discriminant analysis (OPLS-DA as a model) singled out the matrix type as the most important variable influencing global metabolic profiles; that is, more than 100 metabolites were significantly different based on the matrix type. The influence of subject backgrounds on global metabolic profiles was consistent between plasma and serum. Age-associated differences were more predominant in females than males, whereas gender-associated differences were more prevalent in young subjects than old individuals were. The relative standard deviation of metabolite levels in subjects with the same background ranked from 0.1 to 1.5. Moreover, the changes of metabolite levels caused by freeze–thaw cycles were limited, and the effect was more prominent in plasma than serum. These data demonstrate the impact of matrix, age, gender, and freeze–thaw cycles on the metabolite profiles and reveal metabolites affected by these factors. Thus, our results provide would useful fundamental information for exploring and qualifying biomarkers for clinical applications.