Trans-ethnic Mendelian randomization study reveals causal relationships between cardio-metabolic factors and chronic kidney disease
Abstract BACKGROUND The chronic kidney disease (CKD) public health burden is substantial and has not declined as expected with current interventions on disease treatments. A large number of clinical, biological, and behavioural risk factors have been associated with CKD. However, it is unclear which of them are causal. OBJECTIVE To systematically test whether previously reported risk factors for CKD are causally related to the disease in European and East Asian ancestries. DESIGN Two-sample Mendelian randomization (MR) and non-linear MR analyses. PARTICIPANTS 53,703 CKD cases and 960,624 controls of European ancestry from CKDGen, UK Biobank and HUNT, and 13,480 CKD cases and 238,118 controls of East Asian ancestry from Biobank Japan, China Kadoorie Biobank and Japan-Kidney-Biobank/ToMMo. MEASURES Systematic literature mining of PubMed studies identified 45 clinical risk factors and biomarkers with robustly associated genetic variants, including phenotypes related to blood pressure, diabetes, glucose, insulin, lipids, obesity, smoking, sleep disorders, nephrolithiasis, uric acid, coronary artery disease, bone mineral density, homocysteine, C-reactive protein, micro-nutrients and thyroid function, which were selected as exposures. The outcome was CKD (defined by clinical diagnosis or by estimated glomerular filtration rate (eGFR) RESULTS Eight risk factors showed evidence of causal effects on CKD in European ancestry, including body mass index (BMI), hypertension, systolic blood pressure, high density lipoprotein cholesterol, apolipoprotein A-I, lipoprotein A, type 2 diabetes (T2D) and nephrolithiasis. In East Asian ancestry, BMI, T2D and nephrolithiasis showed evidence of causal effects on CKD. Hypertension showed reliable evidence of a strong causal effect on CKD in Europeans but in contrast appeared to show a null effect in East Asians, suggesting the possibility of different causal risk factors in Europeans and East Asians. Although liability to T2D showed consistent effects on CKD, the effect of glycemic traits on CKD was weak, suggesting T2D may have glucose-independent mechanisms to influence CKD. Nonlinear MR indicated a threshold relationship between genetically predicted BMI and CKD, with increased risk at BMI above 25 kg/m2. LIMITATION Due to the unbalanced distribution of data between ancestries, we could only test 17 of the 45 risk factors in East Asian participants. CONCLUSIONS Eight CKD-associated risk factors showed evidence of causal effects on the disease in over 1.2 million European and East Asian ancestries. These risk factors were predominantly related to cardio-metabolic health, which supports the shared causal link between cardio-metabolic health and kidney function. This study provides evidence of potential intervention targets for primary prevention of CKD, which could help reduce the global burden of CKD and its cardio-metabolic co-morbidities. Research in context Evidence before this study Chronic kidney disease (CKD) has a major effect on global health, both as a direct cause of morbidity and mortality, and as an important complication for cardio-metabolic diseases. However, even with the existing interventions, the burden of CKD has not declined as expected over the last 30 years. Existing epidemiological studies of CKD have mainly focused on disease treatment in patients from specific populations and estimated association rather than causality. A systematic assessment of the causal determinants of CKD in different populations is urgently needed, to help promote a shift from treatment of CKD patients to prevention of the disease in high-risk groups. The use of genetic data and the latest Mendelian randomization (MR) methodologies offers a cost-effective way to evaluate the potential intervention targets for prevention of CKD in high-risk groups. Added value of this study In this study, we systematically constructed a causal atlas of 45 risk factors on CKD in European and East Asian ancestries using MR. To maximise power of these analyses and accuracy of the findings, we collected and harmonised CKD genetic association data from six large-scale biobanks (in over 1.1 million Europeans and 250,000 East Asians). By applying a comprehensive MR framework, including linear two-sample MR, bidirectional MR, multivariable MR and non-linear MR approaches, we identified eight risk factors with reliable evidence of causal effects on CKD in European ancestry studies, including body mass index (BMI), hypertension, systolic blood pressure, high density lipoprotein cholesterol, apolipoprotein A-I, lipoprotein A, type 2 diabetes (T2D) and nephrolithiasis. In East Asian studies, BMI, T2D and nephrolithiasis also showed causal effects on CKD. Among other factors, hypertension showed reliable evidence of a strong causal effect on CKD in Europeans but in contrast appeared to show a null effect in East Asians. This MR finding together with previous literature evidence opens up the possibility that hypertension could play different causal roles on CKD across ancestries. For diabetes and glycemic phenotypes, our MR and sensitivity analyses suggested the causal role of liability of T2D on CKD but suggested weak effects of glycemic phenotypes on CKD. This aligns with the recent trial of SGLT2 inhibitors on kidney disease, which implies T2D may have glucose-independent mechanisms to influence CKD. For lipids phenotypes, we found good evidence to support the role of high-density lipoprotein cholesterol on CKD and further suggested the effects of two lipids targets: circulating CETP level and lipoprotein A concentration. For body weight, our study quantified a threshold relationship between BMI and CKD, with increased risk at BMI above 25 kg/m2. The causal relationship between nephrolithiasis and CKD were reported in previous studies, but our study confirmed the causal links between the two for the first time. Implication of all the available evidence This study makes a significant advance in comprehensively prioritising intervention targets for CKD in over 1.2 million participants. Our study presents causal evidence from both European and East Asian population samples, widening the generalisability of the causal atlas. Importantly, the prioritised risk factors are predominantly related to cardio-metabolic health, which supports the shared causal link between cardio-metabolic health and kidney function. Clinically, the high-quality evidence from this study highlights the value of exploring these causal factors in the general population and prioritizes drug targets and life-style interventions for CKD primary prevention, which could help reduce the global burden of CKD and its cardio-metabolic co-morbidities.