Combinative predictive effect of left ventricular mass index, ratio of HDL and CRP for progression of chronic kidney disease in non-dialysis patient.

Abstract

This current study scrutinized the association amongleft ventricular mass index (LVMI), ratio ofhigh-density lipoprotein (HDL) and C-reactive protein (CRP), and renal function. Furthermore, we examined the predictive effects of left ventricular mass index and HDL/CRP on progression of non-dialysis chronic kidney disease. We enrolled adult patients with chronic kidney disease (CKD) who were not receiving dialysis and obtained follow-up data on them. We extracted and compared data between different groups. To investigate the relationship between left ventricular mass index (LVMI), high-density lipoprotein (HDL)/C-reactive protein (CRP) levels, and CKD, we employed linear regression analysis, Kaplan-Meier analysis, and Cox proportional hazards regression analysis. Our study enrolled a total of 2351 patients. Compared with those in the non-progression group, subjects in the CKD progression group had lower ln(HDL/CRP) levels (-1.56 ± 1.78 vs. -1.14 ± 1.77, P < 0.001) but higher left ventricular mass index (LVMI) values (115.45 ± 29.8 vs. 102.8 ± 26.31g/m2, P < 0.001). Moreover, after adjusting for demographic factors, ln(HDL/CRP) was found to be positively associated with estimated glomerular filtration rate (eGFR) (B = 1.18, P < 0.001), while LVMI was negatively associated with eGFR (B = -0.15, P < 0.001). In the end, we found that both LVH (HR = 1.53, 95% CI 1.15 to 2.05, P = 0.004) and lower ln(HDL/CRP) (HR = 1.46, 95% CI 1.08 to 1.96, P = 0.013) independently predicted CKD progression. Notably, the combined predictive power of these variables was stronger than either variable alone (HR = 1.98, 95% CI 1.5 to 2.62, P < 0.001). Our study findings indicate that in pre-dialysis patients, both HDL/CRP and LVMI are associated with basic renal function and are independently correlated with CKD progression. These variables may serve as predictors for CKD progression, and their combined predictive power is stronger than that of either variable alone.