The Effect of Blood Lipid Profiles on Chronic Kidney Disease in a Prospective Cohort: Based on a Regression Discontinuity Design

Abstract

ObjectivePrevious studies on the association between lipid profiles and chronic kidney disease (CKD) have yielded inconsistent results and no defined thresholds for blood lipids. MethodsA prospective cohort study including 32,351 subjects who completed baseline and follow-up surveys over 5 years was conducted. Restricted cubic splines and Cox models were used to examine the association between the lipid profiles and CKD. A regression discontinuity design was used to determine the cutoff value of lipid profiles that was significantly associated with increased the risk of CKD. ResultsOver a median follow-up time of 2.2 (0.5, 4.2) years, 648 (2.00%) subjects developed CKD. The lipid profiles that were significantly and linearly related to CKD included total cholesterol (TC), triglycerides (TG), high-density lipoprotein cholesterol (HDL-C), TC/HDL-C, and TG/HDL-C, whereas lowdensity lipoprotein cholesterol (LDL-C) and LDL-C/HDL-C were nonlinearly correlated with CKD. TC, TG, TC/HDL-C, and TG/HDL-C showed an upward jump at the cutoff value, increasing the risk of CKD by 0.90%, 1.50%, 2.30%, and 1.60%, respectively, whereas HDL-C showed a downward jump at the cutoff value, reducing this risk by 1.0%. Female and participants with dyslipidemia had a higher risk of CKD, while the cutoff values for the different characteristics of the population were different. ConclusionThere was a significant association between lipid profiles and CKD in a prospective cohort from Northwest China, while TG, TC/HDL-C, and TG/HDL-C showed a stronger risk association. The specific cutoff values of lipid profiles may provide a clinical reference for screening or diagnosing CKD risk.