This article is part of the supplement: Proceedings of the Fourth Annual MCBIOS Conference. Computational Frontiers in Biomedicine

Open Access Open Badges Proceedings

Metabonomics evaluations of age-related changes in the urinary compositions of male Sprague Dawley rats and effects of data normalization methods on statistical and quantitative analysis

Laura K Schnackenberg1, Jinchun Sun2, Parvaneh Espandiari3, Ricky D Holland1, Joseph Hanig3 and Richard D Beger1*

Author Affiliations

1 National Center for Toxicological Research (NCTR), Jefferson, AR, 72079, USA

2 University of Arkansas for Medical Sciences, Little Rock, AR, 72205, USA

3 Center for Drug Evaluation and Research (CDER), Silver Spring, MD, 20993, USA

For all author emails, please log on.

BMC Bioinformatics 2007, 8(Suppl 7):S3  doi:10.1186/1471-2105-8-S7-S3

Published: 1 November 2007



Urine from male Sprague-Dawley rats 25, 40, and 80 days old was analyzed by NMR and UPLC/MS. The effects of data normalization procedures on principal component analysis (PCA) and quantitative analysis of NMR-based metabonomics data were investigated. Additionally, the effects of age on the metabolic profiles were examined by both NMR and UPLC/MS analyses.


The data normalization factor was shown to have a great impact on the statistical and quantitative results indicating the need to carefully consider how to best normalize the data within a particular study and when comparing different studies. PCA applied to the data obtained from both NMR and UPLC/MS platforms reveals similar age-related differences. NMR indicated many metabolites associated with the Krebs cycle decrease while citrate and 2-oxoglutarate, also associated with the Krebs cycle, increase in older rats.


This study compared four different normalization methods for the NMR-based metabonomics spectra from an age-related study. It was shown that each method of normalization has a great effect on both the statistical and quantitative analyses. Each normalization method resulted in altered relative positions of significant PCA loadings for each sample spectra but it did not alter which chemical shifts had the highest loadings. The greater the normalization factor was related to age, the greater the separation between age groups was observed in subsequent PCA analyses. The normalization factor that showed the least age dependence was total NMR intensity, which was consistent with UPLC/MS data. Normalization by total intensity attempts to make corrections due to dietary and water intake of the individual animal, which is especially useful in metabonomics evaluations of urine. Additionally, metabonomics evaluations of age-related effects showed decreased concentrations of many Krebs cycle intermediates along with increased levels of oxidized antioxidants in urine of older rats, which is consistent with current theories on aging and its association with diminishing mitochondrial function and increasing levels of reactive oxygen species. Analysis of urine by both NMR and UPLC/MS provides a comprehensive and complementary means of examining metabolic events in aging rats.