Disease progression Mendelian randomization of risk factors for kidney function decline in chronic kidney disease and the general population

Introduction With limited treatment options, identifying risk factors for chronic kidney disease (CKD) disease progression is an important aspect of CKD management. Novel approaches for correcting case-only genome-wide association studies (GWAS) for index event bias and the availability of large GWAS conducted in people with CKD may allow us to identify such risk factors. Methods We applied Mendelian randomization (MR) using European ancestry data from CKDGen consortium to estimate the causal effect of 74 putative risk factors for CKD progression on annual decline in estimated glomerular filtration rate (eGFR) among people with CKD (N=26,653) and in the general population (N=343,339). Additionally, we examined the effect of the same risk factors on incident CKD (N=480,698). We accounted for index event bias using the Dudbridge et.al. and SlopeHunter methods in CKD case-only analyses. Results Type-2 diabetes (T2D) and high thyroid-stimulating hormone (TSH) increased annual decline in eGFR in CKD population of 0.102 (0.035,0.168) and 0.353 (0.108,0.598) ml/min/1.73 m2 per 1 log-odds higher exposure respectively. These effects were consistent after index event bias adjustment. In unadjusted analysis, pulse pressure (PP) (CKD population: 0.221 (0.041,0.402); general population: 0.059 (0.028,0.091) ml/min/1.73 m2 per SD increase); and serum uric acid level (CKD population: 0.252 (0.081,0.423); general population: 0.041 (0.007,0.075) ml/min/1.73 m2 per SD increase) also increased annual eGFR decline. Conversely, educational attainment decreased annual eGFR decline (CKD population: -0.051 (-0.102,-0.001); general population: -0.016 (-0.025,-0.007)). Conclusion We applied novel methodological approaches to correct for index event bias in CKD case-only GWAS, enabling investigation of putative risk factors for CKD progression. Cardiometabolic traits showed suggestive effects on kidney function decline and there was evidence of a causal effect of T2D and high TSH on annual decline in eGFR. These findings highlight potential targets to slow CKD progression.