Serum or plasma, what is the difference? Investigations to facilitate the sample material selection decision making process for metabolomics studies and beyond.

In analytical chemistry serum as well as plasma are recommended as sample material of choice. However, blood processing for the generation of serum or plasma is rather different. Whether plasma or serum is the preferable sample material is still controversial discussed. We performed in paired samples three UHPLC-mass spectrometry-driven metabolomics studies. In study 1 metabolite profiles of serum vs plasma were compared. 46% out of 216 identified metabolites showed significant different levels (paired Wilcoxon signed-rank test, p < 0.05, FDR <0.01) with only three metabolites (methionine, C2:0- and C3:0-carnitine) showing lower levels in serum. In study 2 comparison of three different serum blood collection tubes revealed that coagulation and associated processes distinctly alter metabolite levels depending on the tube-specific clotting process. Most pronounced differences were found for the dipeptide phenylalanine-phenylalanine (highest levels in silicate containing serum blood collection tubes). In study 3 possible adverse effects of platelets, which still remain in standard plasma even after correct processing, were investigated. No differences in a pattern of 216 metabolites were detected in the comparison of standard and platelet-free plasma (PFP). Our results give novel insights in fundamental differences between serum and plasma, thereby providing valuable information for analytical chemists for decision making to either use serum or plasma before starting complex and time-consuming analytical investigations.