Study On The Influencing Factors Of Hs-cTnT,NT-proBNP And Their Relationship With Kidney Injury/cardiovascular Disease Among CKD Non-dialysis Patients

Study on the influencing factors of hs-cTnT, NT-proBNP and their relationship with kidney injury/cardiovascular disease among CKD non-dialysis patientsObjectiveTo analyze the influencing factors of high-sensitivity cardiac troponin T (hs-cTnT), N-terminal pro-brain natriuretic peptide (NT-proBNP) and their relationship with kidney injury as well as cardiovascular disease (CVD) among non-dialysis chronic kidney disease(CKD) patients. And to further explore their value of evaluating and predicting cardiovascular disease in these patients.MethodsCross-sectional study. We recruited 577 non-dialysis CKD patients who were hospitalized in the nephrology department of Zhongshan Hospital Fudan University. Those who had a heart failure history or CVD/infection event in three months were excluded. Grouping the patients based on quartiles of hs-cTnT and NT-proBNP, we used one-way ANOVA, rank-sum test or Chi-square test for comparisons among groups according to variables’ characteristics. Univariate and multivariate linear regression analysis were applied for assessing the influencing factors of hs-cTnT, NT-proBNP and their relationship with kidney injury/cardiovascular disease. Their relationship with CVD was analyzed with binary logistic regression and linear regression. And receiver operating characteristic (ROC) curve was utilized to further explore the two markers’ value of predicting CVD. To establish a method combining hs-cTnT and NT-proBNP together for CVD evaluation, we joined ROC curve and binary logistic analysis to conduct ROC curve analysis of multivariate variables.Results1.577 patients were recruited, male 58.9%, with age 51.7±16.1 years old. The primary cause for renal disease were glomerulonephritis 61.5%, diabetic nephropathy 12.0%, hypertensive renal injury 5.2%, et al. The proportions of CKD 1-5 stages were 20.5%,20.8%,22.9%,13.9%, and 22.0% respectively.2. Left ventricular mass index (LVMI) was 48.87±16.02 g/m2.7, with the proportion of left ventricular hypertrophy (LVH) 43.7%, concentric remodeling 8.7%, eccentric hypertrophy 33.3%, concentric hypertrophy 10.4%. Left ventricular ejection fraction was (65.68±5.40)%, with the porportion of LVEF greater than 50% 98.8%, between 45 and 50% 0.7%, less than 45% 0.5%.9.4% patients had left ventricular diastolic dysfunction.3. Median level of hs-cTnT was 0.013 (0.007-0.029) ng/mL, with 1.7% undetectable, 46.4% greater than 99th percentile of the general population. Median level of NT-proBNP was 170.2 (53.9～792.4) pg/mL. Both went up from CKD 1 to CKD 5 stage, with significant difference between groups.4. Multivariate linear analysis displayed that higher Ln hs-cTnT was significantly associated with older age, male, diabetes, higher Cys C, urine albumin-to-creatinine ratio(UACR), LVMI and lower estimated glomerular filtration rate (eGFR)(P<0.05).5. Hs-cTnT had negative association with eGFR (regression coefficients-0.019; 95%CI,-0.021～-0.017; P<0.05), with statistical significance in CKD 1-5 stages, and positive association with UACR (regression coefficients 0.00010; 95%CI, 0.00007-0.00013; P<0.05), with statistical significance in CKD 2-4 stages.6. Multivariate linear analysis displayed that higher Ln NT-proBNP was associated with older age, higher β2 microglobulin, Cys C, UACR, LVMI and lower hemoglobin, serum albumin, eGFR, left ventricular ejection fraction (LVEF) (P<0.05).7. NT-proBNP had negative association with eGFR (regression coefficients -0.034; 95%CI,-0.037～-0.031; P<0.05), with statistical significance in CKD 1-5 stages, and positive association with UACR (regression coefficients 0.00020; 95%CI, 0.00014-0.00026; P<0.05), with statistical significance in CKD 1-4 stages.8. Compared with the lowest quartile of hs-cTnT, the highest quartile was approximately six times as likely to have LVH (OR,6.418; 95% CI,3.073～13.408, P<0.05) in the fully adjusted model. Ln cTnT level had a more modest association with LVEF(OR,-1.117; 95 CI,-5.839～-0.594; P<0.05). The highest quartile was approximately 30 times as likely to have left ventricular diastolic dysfunction compared with the lowest one in univariate analysis, which lost statistical significance in the fully adjusted model. From the perspective of varied CKD stages, hs-cTnT had statistically significant association with LVH in CKD 2/3/5 stages, with LVEF in CKD 3/5 stages, with left ventricular diastolic dysfunction in CKD 5 stage.9. Compared with the lowest quartile of NT-proBNP, the highest quartile was approximately five times as likely to have LVH (OR,4.839; 95%CI,1.552-15.086; P<0.05) in the fully adjusted model. Ln NT-proBNP level had a more modest association with LVEF (OR,-0.990; 95%CI,-1.426～-0.554; P<0.05). The highest quartile was approximately 6 times as likely to have left ventricular diastolic dysfunction compared with the lowest one in univariate analysis, which lost statistical significance in the fully adjusted model. From the perspective of varied CKD stages, NT-proBNP had statistically significant association with LVH in CKD 1/3/4/5 stages, with LVEF in CKD 1/3/5 stages, with left ventricular diastolic dysfunction in CKD 3-4 stages.10. When evaluated as a screening test, the aera under the curve (AUC) of ROC curves for hs-cTnT and NT-proBNP evaluating LVH, LVEF<50%, left ventricular diastolic dysfunction were 0.718,0.788,0.736 and 0.733,0.859,0.696, respectively(P<0.05).11. With a specificity of 90% as a diagnostic criterion, from CKD 1 stage to CKD 5 stage, the diagnostic values of cTnT and NT-proBNP to evaluate LVH, LVEF<50%, left ventricular diastolic dysfunction ascended.Conclusion1. Among CKD non-dialysis patients, inflencing factors of serum hs-cTnT were older age, male, renal injury, concomitant diabtes and CVD. Inflencing factors of serum NT-proBNP were older age, anemia, hypoproteinemia, renal injury and concomitant CVD.2. In CKD non-dialysis population, hs-cTnT and NT-proBNP were valuable for evaluating LVH, left ventricular systolic dysfunction and left ventricular diastolic dysfunction. It will be of great importance to monitor these two markers regularly as to screen patients with high cardiovascular risk and provide early preventive intervention.