Tryptophan levels associate with incident cardiovascular disease in chronic kidney disease

Figures

• 

  FIGURE 1. Tryptophan metabolites are altered in early CVD in CKD. Tryptophan catabolism to kynurenine and other downstream metabolites represented in oval shapes; enzymes in this pathway: tryptamine dioxygenase (TDO); IDO1, kynureninase (Kyn); kynurenine aminotransferase (KAT); kynurenine monooxygenase (KMO); 3-hydroxyanthranilic acid dioxygenase (3-HAA dioxygenase); metabolites that are decreased in the new-onset CVD are represented in green ovals and enzymes that potentially altered are represented in red boxes. (Adapted from Figure 1 in Badawy et al. [7].)

• 

  FIGURE 2. Tryptophan metabolites by cardiovascular groups. Tryptophan metabolites that are altered by the cardiovascular group are represented as box plots that represent median and interquartile range. The error bars represent the 5–95% range. (A) Tryptophan and (B) kynurenine.

• 

  FIGURE 3. Tryptophan metabolites altered by CKD stage. Tryptophan metabolites that are elevated with increasing CKD stage are represented as bar plots that represent median and interquartile range. The error bars represent the 5–95% range. (A) Kynurenic acid, (B) hydroxykynurenine and (C) kynurenine.

• 

  FIGURE 4. Forest plot representing the relative risk of CVD. Forest plot describes multinomial logistic regression models with dots representing the odds ratio and the error bars representing 95% CI. The odds ratios of incident and old CVD groups compared with the no CVD group (reference) are represented for each model. Model 1 includes unadjusted tryptophan; Model 2 includes unadjusted kynurenine; Model 3 is tryptophan adjusted by kynurenine (kynure); Model 4 includes tryptophan adjusted by kynurenine, total cholesterol, urine protein creatinine ratio and the use of β-blockers and diuretic use; Model 5 includes the tryptophan fully adjusted for kynurenine, age, gender, race, smoking status, body mass index, systolic blood pressure, total cholesterol, serum albumin, serum C-reactive protein, eGFR, urine protein creatinine ratio and use of statins, β-blockers and diuretics.

• 

  FIGURE 5. ROC curve prediction of new CVD in CKD patients. ROC curve using probabilities of group membership in incident cardiovascular group using discriminant analysis controlling for base clinical model including age, gender, race, smoking status, body mass index, systolic blood pressure, serum albumin, serum C-reactive protein, eGFR, total cholesterol, urine protein creatinine ratio and use of statins, β-blockers and diuretics results in an area under the curve (AUC) of 0.766 and addition of tryptophan (TRYP) and kynurenine (KYN) to the base clinical model increases the AUC to 0.823 (P = 0.0482).