Format

Send to

Choose Destination
BMC Nephrol. 2015 Jul 7;16:98. doi: 10.1186/s12882-015-0100-y.

Cross-sectional examination of metabolites and metabolic phenotypes in uremia.

Author information

1
Division of Nephrology, Massachusetts General Hospital (MGH), 165 Cambridge Street, Suite 302, Boston, MA, 02114, USA. skalim@mgh.harvard.edu.
2
Broad Institute, Cambridge, MA, USA. clary@broadinstitute.org.
3
Division of Nephrology, Massachusetts General Hospital (MGH), 165 Cambridge Street, Suite 302, Boston, MA, 02114, USA. joseph.deferio@downstate.edu.
4
Division of Nephrology, Massachusetts General Hospital (MGH), 165 Cambridge Street, Suite 302, Boston, MA, 02114, USA. gortiz@partners.org.
5
Division of Nephrology, Massachusetts General Hospital (MGH), 165 Cambridge Street, Suite 302, Boston, MA, 02114, USA. amoffet2@illinois.edu.
6
Broad Institute, Cambridge, MA, USA. rgerszten@partners.org.
7
Cardiology Division, MGH, Boston, MA, USA. rgerszten@partners.org.
8
Cardiovascular Research Center, MGH, Boston, MA, USA. rgerszten@partners.org.
9
Division of Nephrology, Massachusetts General Hospital (MGH), 165 Cambridge Street, Suite 302, Boston, MA, 02114, USA. rthadhani@partners.org.
10
Division of Nephrology, Massachusetts General Hospital (MGH), 165 Cambridge Street, Suite 302, Boston, MA, 02114, USA. eprhee@partners.org.
11
Broad Institute, Cambridge, MA, USA. eprhee@partners.org.

Abstract

BACKGROUND:

Although metabolomic approaches have begun to document numerous changes that arise in end stage renal disease (ESRD), how these alterations relate to established metabolic phenotypes in uremia is unknown.

METHODS:

In 200 incident hemodialysis patients we used partial least squares discriminant analysis to identify which among 166 metabolites could best discriminate individuals with or without diabetes, and across tertiles of body mass index, serum albumin, total cholesterol, and systolic blood pressure.

RESULTS:

Our data do not recapitulate metabolomic signatures of diabetes and obesity identified among individuals with normal renal function (e.g. elevations in branched chain and aromatic amino acids) and highlight several potential markers of diabetes status specific to ESRD, including xanthosine-5-phosphate and vanillylmandelic acid. Further, our data identify significant associations between elevated tryptophan and long-chain acylcarnitine levels and both decreased total cholesterol and systolic blood pressure in ESRD. Higher tryptophan levels were also associated with higher serum albumin levels, but this may reflect tryptophan's significant albumin binding. Finally, an examination of the uremic retention solutes captured by our platform in relation to 24 clinical phenotypes provides a framework for investigating mechanisms of uremic toxicity.

CONCLUSIONS:

In sum, these studies leveraging metabolomic and metabolic phenotype data acquired in a well-characterized ESRD cohort demonstrate striking differences from metabolomics studies in the general population, and may provide clues to novel functional pathways in the ESRD population.

PMID:
26149577
PMCID:
PMC4491861
DOI:
10.1186/s12882-015-0100-y
[Indexed for MEDLINE]
Free PMC Article

Supplemental Content

Full text links

Icon for BioMed Central Icon for PubMed Central
Loading ...
Support Center