Preanalytical (Mis)Handling\nof Plasma Investigated\nby <sup>1</sup>H NMR Metabolomics

The preanalytical\nhandling of plasma, how it is drawn,\nprocessed,\nand stored, influences its composition. Samples in biobanks often\nlack this information and, consequently, important information about\ntheir quality. Especially metabolite concentrations are affected by\npreanalytical handling, making conclusions from metabolomics studies\nparticularly sensitive to misinterpretations. The perturbed metabolite\nprofile, however, also offers an attractive choice for assessing the\npreanalytical history from the measured data. Here we show that it\nis possible using Orthogonal Projections to Latent Structures Discriminative\nAnalysis to divide plasma NMR data into a multivariate “original\nsample space” suitable for further less biased metabolomics\nanalysis and an orthogonal “preanalytical handling space”\ndescribing the changes occurring from preanalytical mishandling. Apart\nfrom confirming established preanalytical effects on metabolite levels,\ne.g., the consequent changes in glucose, lactate, ornithine, and pyruvate,\nthe sample preparation protocol involved methanol precipitation which\nallowed the observation of reversible changes in short-chain fatty\nacid concentrations as a function of temperature.