Kinetic Modeling of Storage Effects on Biomarkers Related to B Vitamin Status and One-Carbon Metabolism

Steinar Hustad*, Simone Eussen, Oivind Midttun, Arve Ulvik, Puck M. van de Kant, Lars Morkrid, Randi Gislefoss, Per Magne Ueland

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review

82 Citations (Web of Science)

Abstract

Biomarkers and metabolites related to B vitamin function and one-carbon metabolism have been studied as predictors of chronic diseases in studies based on samples stored in biobanks. For most biomarkers, stability data are lacking or fragmentary.Degradation and accumulation kinetics of 32 biomarkers were determined at 23 ?C in serum and plasma (EDTA, heparin, and citrate) collected from 16 individuals and stored for up to 8 days. In frozen serum (-25 ?C), stability was studied cross-sectionally in 650 archival samples stored for up to 29 years. Concentration vs time curves were fitted to monoexponential, biexponential, linear, and nonlinear models.For many biomarkers, stability was highest in EDTA plasma. Storage effects were similar at room temperature and at -25 ?C; notable exceptions were methionine, which could be recovered as methionine sulfoxide, and cystathionine, which decreased in frozen samples. Cobalamin, betaine, dimethylglycine, sarcosine, total homocysteine, total cysteine, tryptophan, asymetric and symmetric dimethyl argenine, creatinine, and methylmalonic acid were essentially stable under all conditions. Most B vitamins (folate and vitamins B2 and B6) were unstable; choline increased markedly, and some amino acids also increased, particularly in serum. The kynurenines showed variable stability. For many biomarkers, degradation (folate and flavin mononucleotide) or accumulation (pyridoxal, riboflavin, choline, amino acids) kinetics at room temperature were non-first order.Data on stability and deterioration kinetics for individual biomarkers are required to optimize procedures for handling serum and plasma, and for addressing preanalytical bias in epidemiological and clinical studies.
Original languageEnglish
Pages (from-to)402-410
JournalClinical Chemistry
Volume58
Issue number2
DOIs
Publication statusPublished - Feb 2012

Cite this