Pulse pressure mediation and hypertension modulation: the impact of TyG index on renal function deterioration in CKD patients under the CKM syndrome framework

Figures

• 

  Figure 1. Directed Acyclic Graph (DAG) The total effect of TyG on RFD is estimated by adjusting only for pre-specified baseline confounders, which closes all backdoor paths while fully retaining all potential mediated pathways to capture the complete causal effect. For mechanism analysis of the core hypothesized pathway (TyG → pulse pressure [PP] → RFD), the controlled direct effect is estimated by additional adjustment for parallel mediators (body mass index, diabetes mellitus, low-density lipoprotein cholesterol [LDL], high-density lipoprotein cholesterol [HDL]). This adjustment blocks competing indirect pathways from these independent parallel mediators, eliminates interference from overlapping mediated effects, and thus isolates and accurately quantifies the specific, independent mediating effect of PP. As a validated surrogate of insulin resistance, elevated TyG index increases PP by promoting arterial stiffness and endothelial dysfunction; elevated PP further impairs renal autoregulation, accelerates glomerular injury, and ultimately drives estimated glomerular filtration rate decline and end-stage renal disease progression. The proportion of RFD risk mediated by PP is calculated via…

• 

  Figure 2. Distribution and association of TyG index with eGFR changes in the CHARLS cohort (a) Number distribution of participants with eGFR decline (red) and non-decline (green) after quartile stratification of the 2011 TyG index (Q1: low, Q2, Q3, Q4: high) in the CHARLS cohort. The proportions of eGFR decline in Q1 to Q4 were 29.3%, 28.5%, 34.9%, and 43.1%, respectively, showing a stepwise increasing trend with higher TyG quartiles. (b) Scatter plot and linear fitted line illustrating the relationship between the TyG index and ΔeGFR (2011–2015) in the CHARLS cohort. Observed data points are distributed around the fitted line without obvious abnormal deviations, indicating a linear association trend.

• 

  Figure 3. Distribution of regression coefficients and confidence intervals in the multiple linear regression model (CHARLS cohort) This figure presents the regression coefficients and their 90%, 95%, and 99% confidence intervals for each variable in the multiple linear regression model, with ΔeGFR as the dependent variable and the 2011 TyG index as the core independent variable. Covariates included demographic factors, lifestyle factors, disease history, and medical-related factors. The key result is that the confidence intervals of the TyG index did not include 0 at all three confidence levels, with a negative regression coefficient (β=-0.736) and statistical significance (P<0.001). An increase in the TyG index was significantly associated with a decrease in ΔeGFR, and this association remained stable across the three confidence levels.

• 

  Figure 4. Distribution and association of TyG index with ESRD risk in the CJFH cohort (a) ESRD incidence rates across TyG tertiles (T1: low, T2: medium, T3: high) in the CJFH cohort. The incidence rates were 23.48% (T1), 26.52% (T2), and 45.45% (T3), respectively, showing a clear stepwise upward trend with increasing TyG levels. (b) Kaplan-Meier survival curves for ESRD-free survival stratified by TyG tertiles in the CJFH cohort. The survival probability of the high TyG group (T3) declined most steeply, and the Log-rank test confirmed significant differences in survival curves between groups (P = 0.0001).

• 

  Figure 5. Moderating effect of hypertension. Regression coefficients (90% CI) of the TyG index on eGFR decline stratified by baseline hypertension status in the CHARLS cohort (hypertension group: n=2196; non-hypertension group: n=2279). The TyG coefficient was negative in both groups, with a larger absolute value in the hypertension group. Bootstrap test (1000 repetitions) showed that the difference in coefficients between the two groups was statistically significant (P = 0.033), indicating that hypertension amplifies the adverse effect of the TyG index.

• 

  Figure 6. Nonlinear relationship between TyG index and renal function deterioration (a) Nonlinear relationship between the TyG index and ΔeGFR in the CHARLS cohort (adjusted for baseline covariates) using a quadratic regression model (ΔeGFR = β1×TyG + β2×TyG2). The linear term coefficient (β₁=3.293, P<0.01) and quadratic term coefficient (β2=-0.227, P<0.01) indicated an inverted U-shaped relationship, with an inflection point at TyG=7.26. Only 11 participants (0.2% of the total sample) were below this inflection point, and 99.8% of samples were on the right side, where ΔeGFR continued to decline with increasing TyG, indicating the overall association was dominated by a negative linear trend. **P<0.01. (b) Dose-response relationship between the TyG index and ESRD risk in the CJFH cohort using restricted cubic spline (RCS) analysis. The relative hazard ratio (HR) showed a continuous upward trend as the TyG index increased from 5 to 15, with no plateau or downward inflection point. The nonlinear test yielded P = 0.742, confirming no significant nonlinear relationship, and the association was mainly characterized by a linear trend without an obvious “risk threshold” in the TyG range of 5–15.…