Background

End stage renal disease (ESRD) is characterized by various metabolic disturbances linked to adverse outcomes, but the nature of these associations is poorly understood or even counterintuitive [1]. For example, low serum albumin has consistently been linked to cardiovascular mortality in dialysis [2, 3], yet the association is not simply a result of inadequate nutrition, as intradialytic parenteral nutrition does not always improve survival in malnourished ESRD patients [4], and alternative pathways including inflammation are known to contribute [5]. Dyslipidemia is also common in ESRD, including atherogenic increases in carbamylated and oxidized-LDL cholesterol. Clinical trials, however, have failed to demonstrate a survival benefit with statin therapy in ESRD, despite substantial reductions in LDL cholesterol [6, 7]. Indeed, several studies have identified an inverse association between cholesterol levels and uremic cardiovascular risk [8, 9]. Similar paradoxical associations, sometimes described as examples of ‘reverse epidemiology’ in ESRD, have been noted with body mass index (BMI) and blood pressure [8, 10–12] and survival on hemodialysis.

Metabolomic approaches enable high resolution interrogation of metabolic phenotypes. For example, studies have begun to highlight specific metabolites associated with obesity and diabetes risk in the general population, particularly branched-chain and aromatic amino acids, but also short chain acylcarnitines, the glutamate/glutamine ratio, and bile acids [13–23]. Ultimately, these findings may lead to a more refined understanding of underlying disease mechanisms. Whether these and other associations identified in the burgeoning metabolomics literature extend to the ESRD population is unknown. While initial applications of metabolomics to nephrology research have generated a broad view of uremia [24] and the immediate effects of the hemodialysis procedure [25, 26], no studies have explicitly examined the cross-sectional relationship between uremic metabolite alterations and metabolic or other clinical phenotypes captured in a given ESRD study population.

We recently performed liquid chromatography-mass spectrometry (LC-MS) based metabolite profiling on plasma obtained from incident hemodialysis subjects in the Accelerated Mortality on Renal Replacement (ArMORR) cohort, and highlighted baseline levels of long-chain acylcarnitines as markers of future cardiovascular death in a longitudinal study [27]. Using this data set, we now leverage the rich clinical phenotyping available in ArMORR to examine cross-sectional associations between plasma metabolite levels and select metabolic phenotypes. Specifically, we sought to identify associations between metabolites and metabolic phenotype data in this well-characterized ESRD cohort, hypothesizing that these findings would differ from metabolomic associations derived from the general population. The aim of such work is to gain new insights into the functional metabolic pathways that underlie the poorly understood associations between clinical outcomes in ESRD and variables such as BMI, albumin, cholesterol, and blood pressure. Further, we examined the correlation between the subset of metabolites measured by our platform that are known uremic retention solutes and numerous clinical variables captured in ArMORR. Taken together, these results provide a new perspective on the metabolic disturbances that accompany and arise in ESRD.

Methods

Study population

The Accelerated Mortality on Renal Replacement (ArMORR) study is a prospective cohort study of 10,044 incident hemodialysis patients in any of 1,056 U.S. centers operated by Fresenius Medical Care North America between June 2004 and August 2005. Full details have been previously published [27–31]. All subjects underwent 1 year of follow-up except for those who died (15.2 %), voluntarily discontinued dialysis (5 %), underwent kidney transplantation (3 %), recovered renal function (4 %), or transferred to a dialysis unit outside the Fresenius system (12 %). Clinical data were prospectively collected by physicians at the point-of-care and included demographics, coexisting conditions, and routine laboratory studies. Plasma samples drawn at the beginning of a dialysis session within 14 days of initiation of outpatient hemodialysis, and that would otherwise have been discarded after routine clinical testing, were saved and stored in liquid nitrogen. We previously evaluated the associations between metabolite profile and mortality in the ArMORR cohort in a study of 100 randomly selected patients who died within the first year of dialysis from a cardiovascular cause, and 100 patients who were alive 1 year after starting dialysis, frequency matched for age, sex, and race [27]. We included this entire group of individuals in the current study. This study was approved by the Massachusetts General Hospital IRB, which waived the need for informed consent because all personal identifiers were removed from the blood samples and clinical data before transfer to the investigators.

Metabolite profiling

We applied two distinct LC-MS based methods to distinct plasma aliquots for each study subject. Amino acids, amino acid derivatives, urea cycle intermediates, nucleotides and other positively charged polar metabolites were profiled as previously described [14]. Briefly, 10 μL of plasma were extracted with 90 μL of 74.9:24.9:0.2 vol/vol/vol acetonitrile/methanol/formic acid containing valine-d8 (Sigma-Aldrich; St Louis, MO). After centrifugation, supernatants underwent hydrophobic interaction chromatography using a 150 × 2.1 mm Atlantis hydrophobic interaction chromatography column (Waters, Milford, MA), and MS data were acquired on a 4000 QTRAP triple quadrupole mass spectrometer (AB SCIEX, Foster City, CA) using ESI and MRM in the positive ion mode. Organic acids, sugars, bile acids, and other negatively charged polar metabolites were profiled as previously described [32]. Briefly, 30 μL of plasma were extracted with the addition of four volumes of 80:20 vol/vol methanol/water containing isotope-labeled inosine-15 N4, supernatants underwent chromatography on a 150 × 2.0 mm Luna NH2 column (Phenomenex; Torrance, CA), and MS data were acquired using a 5500 QTRAP triple quadrupole mass spectrometer (AB SCIEX; Foster City, CA) using ESI and MRM in the negative ion mode.

Statistical analyses

We first examined the association between plasma metabolites and pre-selected metabolic phenotypes: diabetes status, BMI, serum albumin, total cholesterol, and systolic blood pressure (SBP). We focused on these phenotypes because of their strong links to metabolism and their established, but incompletely understood, associations with ESRD mortality [33–39]. To investigate differences in metabolite profiles across these phenotypes, continuous outcomes measures (BMI, albumin, total cholesterol, and SBP) were divided into tertiles. Diabetes status was classified as “yes” or “no”.

Because of the number of metabolites measured by our LC-MS platform (166 total, Additional file 1: Table S1), many with significant inter-correlations, we performed partial least squares discriminant analysis (PLS-DA) to visualize the linear components that discriminate individuals across the categories (tertiles or class) of metabolic phenotypes. PLS-DA was performed on log transformed and auto-scaled (mean-centered and divided by the standard deviation of each metabolite) data using MetaboAnalyst 2.0 [40, 41]. This program performs PLS regression using the plsr function provided by the pls package in R, and classification and cross validation using the corresponding wrapper function offered by the caret package. Variable Importance in Projection (VIP) scores generated by this program estimate the importance of each variable in the projection used within the PLS model. A variable with a VIP score greater than 1 can be considered important in a given model. The metabolites with the highest VIP scores were further analyzed by comparing their levels across classes using Mann-U-Whitney and Kruskal-Wallis tests as appropriate.

Finally, heatmaps were created to represent Pearson correlation (r) and P-value matrices between available clinical variables in the cohort and the subset of metabolites measured by our platform that have previously been described as uremic retention solutes (http://www.uremic-toxins.org) [27, 42–45].

To account for multiple testing, we used a Bonferroni corrected significance threshold of P < 0.0003 (0.05/166) for comparisons of individual metabolites across phenotype class. For the correlation matrices, we used an adjusted significance threshold of P < 0.00007 (0.05/[29 metabolites* 24 clinical variables]). A sensitivity analysis was performed to gauge the influence of case–control status (i.e. whether or not a given individual died of a cardiovascular cause within one year of starting dialysis) on the associations described herein. Stratified analysis by case status did not alter the statistical significance of the models described. Therefore, we did not stratify models by mortality status and analyzed the entire group of 200 individuals together as a cohort. All analyses were performed using SAS software version 9.1.3 (SAS Institute) and MetaboAnalyst 2.0 software (www.metaboanalyst.ca).

Results and discussion

Cohort characteristics

As shown in Table 1, the mean age of the study population was 69.5 years and 69 % of subjects were white. There was an equal representation of males and females and nearly half of the individuals had a history of diabetes or had diabetes listed as their cause of ESRD (49 %). The mean BMI was 26.5 kg/m2 (SD ±7.6 kg/m2) and the mean SBP was 144 mmHg (±27 mmHg). A minority of individuals reported a lipid disorder (12 %) and the median total cholesterol level was 162 mg/dL (quartile1-quartile3, 127–188 mg/dL). The median serum albumin level was 3.6 g/dl (3.2-3.8 g/dL).

Table 1 Baseline characteristics of the study sample
Full size table

Examination of metabolite profiles and select metabolic phenotypes

The PLS-DA approach allowed us to visualize and extract the metabolites that best separated individuals (Figs. 1,2,3,4 and 5; left panels) across phenotype tertile or class (Table 2). Because four of the clinical phenotypes we studied (BMI, serum albumin, total cholesterol, and SBP) are continuous measures, we recognize that tertile cut-offs do not demarcate distinct physiologic or pathophysiologic processes. Thus, plot overlap across tertiles was expected. Variable importance in projection (VIP) provided a score for each metabolite, ranking the metabolites according to their predictive power in the PLS model; the fifteen metabolites with the highest VIP scores for each plot are shown in Figs. 1, 2, 3, 4 and 5 (right panels), and the levels of these metabolites across tertiles (or class) of the phenotypes with corresponding test statistics are shown in Tables 3-7.

Fig. 1
figure 1

Comparison of metabolite profiles and diabetes status. Study subjects were grouped by diabetes status (yes/no). Left: Partial least squares discriminant analysis (PLS-DA) score plot for the study population separated by phenotype class. Oval outlines denote 95 % confidence intervals. Right: The variable importance in projection (VIP) scores for the 15 metabolites with the greatest score. The colored boxes on the right indicate the direction of metabolite alterations across phenotype class. Abbreviations: XMP, Xanthosine-5-phosphate; VMA, vanillylmandelic acid; 3-PG, 3-phosphoglycerate; mevalonate-5P, mevalonate-5-phosphate; PEP, phosphoenolpyruvate; ribose-5P, ribose-5-phosphate and ribulose-5-phosphate

Full size image
Fig. 2
figure 2

Comparison of metabolite profiles across tertile of body mass index (BMI). Study subjects were grouped by BMI tertile (tertile 1 = lowest BMI). Left: Partial least squares discriminant analysis (PLS-DA) score plot for the study population separated by phenotype class. Oval outlines denote 95 % confidence intervals. Right: The variable importance in projection (VIP) scores for the 15 metabolites with the greatest score. The colored boxes on the right indicate the direction of metabolite alterations across phenotype tertile. Abbreviations: BAIBA, β-aminoisobutyric acid; SDMA, symmetric dimethylarginine; 3-PG, 3-phosphoglycerate; b-OH-butyrate, β-hydroxybutryate; p-creatine, phosphocreatine; XMP, Xanthosine-5-phosphate

Full size image
Fig. 3
figure 3

Comparison of metabolite profiles across tertile of serum albumin. Study subjects were grouped by albumin tertile (tertile 1 = lowest). Left: Partial least squares discriminant analysis (PLS-DA) score plot for the study population separated by phenotype class. Oval outlines denote 95 % confidence intervals. Right: The variable importance in projection (VIP) scores for the 15 metabolites with the greatest score. The colored boxes on the right indicate the direction of metabolite alterations across phenotype tertile. Abbreviations: BAIBA, β-aminoisobutyric acid; SDMA, symmetric dimethylarginine; GCDCAs, glycodeoxycholate and glycochenodeoxycholate; ADP, adenosine diphosphate

Full size image
Fig. 4
figure 4

Comparison of metabolite profiles across tertile of total cholesterol. Study subjects were grouped by cholesterol tertile (tertile 1 = lowest). Left: Partial least squares discriminant analysis (PLS-DA) score plot for the study population separated by phenotype class. Oval outlines denote 95 % confidence intervals. Right: The variable importance in projection (VIP) scores for the 15 metabolites with the greatest score. The colored boxes on the right indicate the direction of metabolite alterations across phenotype tertile. Abbreviations: GMP, guanosine monophosphate; AMP, adenosine monophosphate; p-creatine, phosphocreatine

Full size image
Fig. 5
figure 5

Comparison of metabolite profiles across tertile of systolic blood pressure. Study subjects were grouped by systolic blood pressure tertile (tertile 1 = lowest). Left: Partial least squares discriminant analysis (PLS-DA) score plot for the study population separated by phenotype class. Oval outlines denote 95 % confidence intervals. Right: The variable importance in projection (VIP) scores for the 15 metabolites with the greatest score. The colored boxes on the right indicate the direction of metabolite alterations across phenotype tertile. Abbreviations: ADMA, asymmetric dimethylarginine

Full size image
Table 2 Primary metabolic phenotypes by tertiles
Full size table
Table 3 Cross sectional analysis of diabetes status and select metabolites
Full size table

Values are median peak area for the metabolites (quartile 1, quartile 3)*P-value significant at the Bonferroni adjusted level of 3.0 × 10−4

For individuals with or without diabetes, glucose was one of the strongest discriminating metabolites, with higher levels among the individuals with diabetes (Fig. 1). Xanthosine-5-phosphate (XMP) and vanillylmandelic acid (VMA), however, had even higher VIP scores than glucose, with XMP levels higher and VMA levels lower among the individuals with diabetes (P = 5.0 × 10−7 and 1.5 × 10−6 for XMP and VMA, respectively). Other significant differences included higher levels of 3-phosphoglyceric acid (P = 4.9 × 10−5), methionine (P = 4.5 × 10−5), mevalonate-5-phosphate (P = 2.8 × 10−4), and ornithine (P = 3.2 × 10−4) among individuals with diabetes. Tyrosine, an aromatic amino acid, had a nominal association with diabetes status (P = 1.0 × 10−3). There were no significant associations between tyrosine levels, or any aromatic or branched-chain amino acid levels, with BMI (Table 4). In fact, none of these metabolites had even a nominal association with BMI in univariate comparisons (data not shown). Overall, the ability of metabolite profiles to discriminate individuals across tertiles of BMI was poor (Fig. 2). Among the top metabolites by VIP score, only 20:4 carnitine (arachidonyl carnitine), β-hydroxybutyrate, and benzoate had nominal P-values in univariate analysis (Table 4).

Table 4 Cross sectional analysis of BMI tertiles and select metabolites
Full size table

Values are median peak area for the metabolites (quartile 1, quartile 3)*P-value significant at the Bonferroni adjusted level of 3.0 × 10−4

Table 5 Cross sectional analysis of albumin tertiles and select metabolites
Full size table

Values are median peak area for the metabolites (quartile 1, quartile 3)*P-value significant at the Bonferroni adjusted level of 3.0 × 10−4

Thus, our data from an ESRD population do not recapitulate observations derived in the general population linking diabetes and obesity with elevations in branched chain and aromatic amino acids (except for the nominal association with tyrosine noted above), short chain acylcarnitines, the glutamate/glutamine ratio, or bile acids [2]. One potential explanation is that kidney function (and dysfunction) has a direct effect on select metabolites. The interaction between kidney function and amino acid metabolism, in particular, has been closely examined. For example, the kidney is known to make a substantial contribution to whole body tyrosine appearance via intra-organ phenylalanine hydroxylation [46]. Further, the metabolic acidosis that results from renal failure leads to increased leucine oxidation [47], and indeed, generalized muscle catabolism. The significant associations we did identify with diabetes status in ESRD, e.g. XMP (a purine breakdown product) and VMA (a catecholamine metabolite), have not been reported in the general population. To what extent these alterations reflect distinct pathophysiologic processes in uremia warrant further investigation.

For individuals in different tertiles of serum albumin, PLS-DA demonstrated moderate separation across groups (Fig. 3). Tryptophan and its downstream metabolite kynurenine were significantly higher in individuals with higher serum albumin levels (P = 1.8 × 10−4 and 2.5 × 10−4, respectively). Two additional tryptophan metabolites, indoxyl sulfate and indole lactate, had a trend for higher levels among individuals with higher serum albumin levels. Given recent observations that ascribe a functional role for tryptophan metabolism, specifically through indoleamine-2, 3-dioxygenase (IDO), in modulating the immune system and vascular tone [48, 49], it is an appealing link between metabolism, nutritional status, inflammation, and cardiovascular risk in ESRD. However, these data should be interpreted with caution. Because tryptophan and its catabolites are hydrophobic, their positive correlation with albumin may reflect their significant protein-binding [50] – our LC-MS method measures total, not free, plasma metabolite levels. A similar mechanism may underlie the trend for association between thyroxine and albumin levels. By contrast, creatine, the metabolite with the third highest VIP score, had a trend for higher levels among individuals with lower serum albumin levels. In a study that examined spent media from cultured muscle cells treated with mitochondrial respiratory chain inhibitors, as well as plasma obtained from individuals with respiratory chain diseases, creatine levels were reproducibly elevated [51]. In cell culture, extracellular creatine was inversely correlated with the intracellular phosphocreatine/creatine ratio, suggesting that elevated plasma creatine may signal a low energetic state in tissues using the creatine phosphate shuttle. Interestingly, elevated plasma creatine and decreased plasma phosphocreatine levels both have a nominal association with increased 1-year cardiovascular mortality among incident dialysis patients in ArMORR [27]. These observations raise the possibility that impaired mitochondrial respiration in muscle is linked to the pathogenesis of hypoalbuminemia in dialysis.

For individuals across tertiles of total cholesterol and SBP, PLS-DA again demonstrated moderate separation across tertiles (Figs. 4 and 5). For both analyses, long-chain acylcarnitines were among the metabolites with the highest VIP scores, with 18:2 carnitine (linoleylcarnitine, P = 9.0 × 10−7), 18:1 carnitine (oleoylcarnitine, P = 6.0 × 10−7), and 18:0 carnitine (stearoylcarnitine, P = 7.7 × 10−5) all significantly higher in individuals with lower total cholesterol levels (Table 6,) and linoleylcarnitine (P = 5.5 × 10−5) and oleoylcarnitine (P = 8.7 × 10−6) significantly higher in individuals with lower SBP (Table 7); select short-chain and medium-chain acylcarnitines were also significantly elevated among individuals with lower SBP. Like albumin, low cholesterol and blood pressure are established predictors of increased mortality in ESRD [8, 9, 11, 12]. These findings are consistent with our observation that long-chain acylcarnitines are associated with 1-year cardiovascular mortality in ESRD [27], and motivate further interest in long-chain acylcarnitines as markers, or even mediators, of altered lipid metabolism and cardiovascular function in uremia. Tryptophan, one of the strongest discriminators of albumin tertile, was also a top hit in regards to total cholesterol and SBP, with higher tryptophan significantly associated with lower levels for both. Finally, higher arginine levels were significantly associated with higher total cholesterol and higher deoxyadenosine levels were significantly associated with lower SBP.

Table 6 Cross sectional analysis of total cholesterol tertiles and select metabolites
Full size table

Values are median peak area for the metabolites (quartile 1, quartile 3)*P-value significant at the Bonferroni adjusted level of 3.0 × 10−4

Table 7 Cross sectional analysis of systolic blood pressure tertiles and select metabolites
Full size table

Relation of uremic solutes with twenty-four clinical phenotypes

To examine the specificity of the long-chain acylcarnitine-phenotype associations described above, we looked more broadly at the correlation between all of the metabolites captured on our platform that are uremic retention solutes and 24 phenotypes captured in ArMORR. Correlations between previously reported uremic retention solutes measured by our LC-MS platform, including long-chain acylcarnitines, and a broad array of clinical phenotypes measured in ArMORR are shown in Fig. 6 (full results are shown in Additional file 1: Tables S2 and S3). The correlations for glucose, as measured by the mass spectrometer, provide an anchor for interpreting this data – LC-MS glucose had a strong positive correlation with glucose measured by the clinical laboratory (r = 0.80, P <10−15), and a modest negative correlation with serum sodium (r = −0.21, P = 7.0 × 10−3) that did not reach statistical significance, but is well recognized in the clinic (i.e. dilutional hyponatremia due to hyperglycemia). Aside from the correlation between the two glucose measures, the strongest correlations identified by this broader examination of uremic metabolites were indeed between long-chain acylcarnitines and lipid traits (total and LDL cholesterol) and blood pressure. Among lipid phenotypes, long-chain acylcarnitines such as oleoylcarnitine were more strongly correlated with total (r = −0.45, P = 2.6 × 10−9) and LDL cholesterol (r = −0.43., P = 1.7 × 10−8) than with HDL cholesterol (r = −0.15, P = 5.7 × 10−2) or triglycerides (r = −0.27, P = 5.0 × 10−4). Similarly, oleoylcarnitine was more strongly correlated with SBP (r = −0.38, P = 3.4 × 10−8) than diastolic blood pressure (r = −0.25, P = 3.0 × 10−3). Long-chain acylcarnitines were not significantly correlated with any non-lipid or non-blood pressure phenotype.

Fig. 6
figure 6

Relation of uremic solutes with 24 clinical phenotypes in the ArMORR cohort. Correlation coefficient (r, left) and P values (right) generated from Pearson correlation analyses of metabolites and available clinical variables in the ArMORR cohort. Squares outlined in red on the P value matrix represent values significant at the Bonferroni adjusted level of P < 7.4 × 10−5. Abbreviations: ADMA, asymmetric dimethylarginine; TMAO, trimethylamine-N-oxide; SDMA, symmetric dimethylarginine; BMI, body mass index; URR, urea reduction ratio; SBP, systolic blood pressure; DBP, diastolic blood pressure; LDL, low density lipoprotein; HDL, high density lipoprotein; BNP, b-type natriuretic peptide, TSAT, transferrin saturation; Alk Phos, alkaline phosphatase; PTH, parathyroid hormone; WBC, white blood cell

Full size image

Among the other uremic retention solutes included in our analysis, only kynurenic and anthranilic acid (both tryptophan metabolites downstream of IDO) had more than one significant correlation with a clinical variable (Fig. 6). Kynurenic acid was significantly correlated with diastolic blood pressure (r = 0.32, P = 5.0 × 10−6), bicarbonate (r = −0.31, P = 1.1 × 10−5), and phosphorous (r = 0.38, P = 5.7 × 10−8), whereas anthranilic acid was significantly correlated with alkaline phosphatase (r = 0.29, P = 6.3 × 10−5) and ferritin (r = 0.35, P = 9.7 × 10−7), a frequently used marker of inflammation in hemodialysis. Notably, the other metabolites that had a nominal association with ferritin included two other tryptophan metabolites, kynurenic acid (r = 0.16, P = 2.6 × 10−2) and quinolinate (r = 0.20, P = 6.3 × 10−3), and malondialydehyde (r = 0.24, P = 9.0 × 10−4), an oxidative stress marker. These findings reinforce a potential link between the IDO pathway and uremic inflammation. Sorbitol (r = −0.34, P = 3.4 × 10−5), cytidine (r = −0.40, P = 1.1 × 10−6), and uracil (r = −0.42, P = 1.7 × 10−7) were significantly correlated with plasma glucose; urate (r = −0.35, P = 5.2 × 10−6) and malondialdehyde (r = 0.38, P = 5.7 × 10−7) were significantly correlated with triglycerides; hippurate (r = 0.31, P = 7.0 × 10−6) and oxalate (r = 0.36, P = 1.6 × 10−7) were significantly correlated with phosphorous; and indoxyl sulfate (r = −0.28, P = 6.8 × 10−5) was significantly correlated with bicarbonate. Indoxyl sulfate, which has been found to have several proinflammatory effects in model systems [52], had no correlation with ferritin (nor transferrin saturation, nor white blood cell count), but had a trend for a negative correlation with BNP (r = −0.44, P = 4.0 × 10−4).

Although significant work has been devoted to understanding the in vitro effects of uremic retention solutes, less is known about how many of these metabolites relate to clinical traits in humans. Thus, in addition to demonstrating the specificity of select long-chain acylcarnitine associations, our examination of uremic metabolites against clinical phenotypes provides a resource for future uremic toxin research.

Limitations

Our study has several limitations. First, by examining cross-sectional relationships between metabolite levels and various clinical phenotypes, our study is unable to address causation. Instead, by applying parsimonious selection procedures and conservative adjusted significance thresholds, we seek to highlight notable associations for future investigations. Second, study subjects were selected in the context of a case–control study of 1-year cardiovascular mortality, raising the possibility of confounding by case status. However, as noted in the Methods, the statistical significance of the findings are not changed when analyzed stratified by case status. Nevertheless, we acknowledge that our sample is not a random selection of incident dialysis patients, potentially limiting the generalizability of our findings. Third, residual confounding from other sources is likely to influence the associations described, as multivariate adjustments were not pursued. However, because the relationship between various clinical phenotypes and metabolite levels in uremia is unknown, we did not want to obscure novel biological associations by statistical adjustment. Thus, we present these raw associations as a framework for interpreting and appropriately adjusting select findings in future metabolomics studies in ESRD. Finally, our results require replication in an independent sample, ideally selected randomly and including both incident and prevalent dialysis patients.

Conclusions

In this report, we describe the small molecule alterations that accompany various metabolic phenotypes in ESRD. For phenotypes like diabetes and obesity that have been examined using metabolomics in the general population, our data reinforce the notion that unique pathophysiologic processes arise at ESRD onset. Our finding that long-chain acylcarnitines are strongly and inversely correlated with cholesterol and blood pressure corroborate their potential value as markers of cardiovascular risk in ESRD, whereas the association of tryptophan levels across several phenotypes is of interest given recent studies that assign a functional role for tryptophan metabolism in inflammation and vascular biology. Ultimately, we hope that the breadth of the data presented herein serves as a springboard for investigating mechanisms of uremic toxicity and identifying therapeutic targets to improve the unacceptably high morbidity and mortality attributable to ESRD.