Association of polygenic scores with chronic kidney disease phenotypes in a longitudinal study of older adults

Abstract

Risk of chronic kidney disease (CKD) is influenced by environmental and genetic factors and increases sharply in individuals 70 years and older. Polygenic scores (PGS) for kidney disease-related traits have shown promise but require validation in well-characterized cohorts. Here, we assessed the performance of recently developed PGSs for CKD-related traits in a longitudinal cohort of healthy older individuals enrolled in the Australian ASPREE randomized controlled trial of daily low-dose aspirin with CKD risk at baseline and longitudinally. Among 11,813 genotyped participants aged 70 years or more with baseline eGFR measures, we tested associations between PGSs and measured eGFR at baseline, clinical phenotype of CKD, and longitudinal rate of eGFR decline spanning up to six years of follow-up per participant. A PGS for eGFR was associated with baseline eGFR, with a significant decrease of 3.9 mL/min/1.73m2 (95% confidence interval -4.17 to -3.68) per standard deviation (SD) increase of the PGS. This PGS, as well as a PGS for CKD stage 3 were both associated with higher risk of baseline CKD stage 3 in cross-sectional analysis (Odds Ratio 1.75 per SD, 95% confidence interval 1.66-1.85, and Odds Ratio 1.51 per SD, 95% confidence interval 1.43-1.59, respectively). Longitudinally, two separate PGSs for eGFR slope were associated with significant kidney function decline during follow-up. Thus, our study demonstrates that kidney function has a considerable genetic component in older adults, and that new PGSs for kidney disease-related phenotypes may have potential utility for CKD risk prediction in advanced age. Risk of chronic kidney disease (CKD) is influenced by environmental and genetic factors and increases sharply in individuals 70 years and older. Polygenic scores (PGS) for kidney disease-related traits have shown promise but require validation in well-characterized cohorts. Here, we assessed the performance of recently developed PGSs for CKD-related traits in a longitudinal cohort of healthy older individuals enrolled in the Australian ASPREE randomized controlled trial of daily low-dose aspirin with CKD risk at baseline and longitudinally. Among 11,813 genotyped participants aged 70 years or more with baseline eGFR measures, we tested associations between PGSs and measured eGFR at baseline, clinical phenotype of CKD, and longitudinal rate of eGFR decline spanning up to six years of follow-up per participant. A PGS for eGFR was associated with baseline eGFR, with a significant decrease of 3.9 mL/min/1.73m2 (95% confidence interval -4.17 to -3.68) per standard deviation (SD) increase of the PGS. This PGS, as well as a PGS for CKD stage 3 were both associated with higher risk of baseline CKD stage 3 in cross-sectional analysis (Odds Ratio 1.75 per SD, 95% confidence interval 1.66-1.85, and Odds Ratio 1.51 per SD, 95% confidence interval 1.43-1.59, respectively). Longitudinally, two separate PGSs for eGFR slope were associated with significant kidney function decline during follow-up. Thus, our study demonstrates that kidney function has a considerable genetic component in older adults, and that new PGSs for kidney disease-related phenotypes may have potential utility for CKD risk prediction in advanced age. Lay SummaryThe sum of multiple small genetics contributors to a trait such as kidney function can be captured by polygenic risk scores (PGSs), derived from large-scale genome-wide association studies; however, the clinical significance of these scores is still being understood. We tested several recently developed PGSs for kidney traits in 11,813 European Australians, aged >70 years, from the ASPirin in Reducing Events in the Elderly population, with a mean estimated glomerular filtration rate (eGFR) of 72.9 ml/min per 1.73 m2. We validated multiple PGSs for both eGFR and decline in eGFR. At a population level, those with higher PGSs had the lowest mean eGFR of 57 ml/min per 1.73 m2, with significant individual variability. Overall, this study validated PGSs for kidney function in an older population, showing PGSs, which are static across the life span, still impact kidney function in later stages of life. The sum of multiple small genetics contributors to a trait such as kidney function can be captured by polygenic risk scores (PGSs), derived from large-scale genome-wide association studies; however, the clinical significance of these scores is still being understood. We tested several recently developed PGSs for kidney traits in 11,813 European Australians, aged >70 years, from the ASPirin in Reducing Events in the Elderly population, with a mean estimated glomerular filtration rate (eGFR) of 72.9 ml/min per 1.73 m2. We validated multiple PGSs for both eGFR and decline in eGFR. At a population level, those with higher PGSs had the lowest mean eGFR of 57 ml/min per 1.73 m2, with significant individual variability. Overall, this study validated PGSs for kidney function in an older population, showing PGSs, which are static across the life span, still impact kidney function in later stages of life. Chronic kidney disease (CKD) is a significant and growing health care burden worldwide.1Kalantar-Zadeh K. Jafar T.H. Nitsch D. et al.Chronic kidney disease.Lancet. 2021; 398: 786-802Abstract Full Text Full Text PDF PubMed Scopus (320) Google Scholar CKD prevalence increases with age, with peak incidence after the age of 60 years.2Xie Y. Bowe B. Mokdad A.H. et al.Analysis of the Global Burden of Disease study highlights the global, regional, and national trends of chronic kidney disease epidemiology from 1990 to 2016.Kidney Int. 2018; 94: 567-581Abstract Full Text Full Text PDF PubMed Scopus (467) Google Scholar Risk is driven by conventional risk factors, including hypertension, diabetes, obesity, and smoking.2Xie Y. Bowe B. Mokdad A.H. et al.Analysis of the Global Burden of Disease study highlights the global, regional, and national trends of chronic kidney disease epidemiology from 1990 to 2016.Kidney Int. 2018; 94: 567-581Abstract Full Text Full Text PDF PubMed Scopus (467) Google Scholar,3Levey A.S. Coresh J. Chronic kidney disease.Lancet. 2012; 379: 165-180Abstract Full Text Full Text PDF PubMed Scopus (1388) Google Scholar In addition, CKD is strongly influenced by genetic factors, especially in monogenic kidney disorders.4Aung T.T. Bhandari S.K. Chen Q. et al.Autosomal dominant polycystic kidney disease prevalence among a racially diverse United States population, 2002 through 2018.Kidney360. 2021; 2: 2010-2015Crossref PubMed Scopus (6) Google Scholar,5Mallett A. Patel C. Salisbury A. et al.The prevalence and epidemiology of genetic renal disease amongst adults with chronic kidney disease in Australia.Orphanet J Rare Dis. 2014; 9: 98Crossref PubMed Scopus (47) Google Scholar Although rare monogenic variants are associated with severe, rare forms of CKD dominant polycystic kidney disease and the between genetic variants and CKD risk is well in older individuals in CKD prevalence is et of age with risk of kidney in adults with stage chronic kidney disease in Scopus Google Chronic kidney disease and the 2014; PubMed Scopus (6) Google Scholar for kidney function from a but still significant genetic component across the life J. A. A. The of kidney function an older Australian Full Text Full Text PDF PubMed Scopus Google Scholar association of CKD between and the population, but have from B. Y. et genome-wide significant variants for kidney function Int. 2021; Full Text Full Text PDF PubMed Scopus Google Y. D. et of and in the 2021; PubMed Scopus Google A. et eGFR to the renal and in the PubMed Scopus Google Y. et of genetic associated with kidney function from of a PubMed Scopus Google Scholar are to the of genetic to CKD in older risk prediction a clinical for CKD, especially in older of individuals for CKD through genetic decline in kidney may to individuals and disease et of age with risk of kidney in adults with stage chronic kidney disease in Scopus Google Chronic kidney disease and the 2014; PubMed Scopus (6) Google Scholar Polygenic scores are an of risk for a disease or trait and are by the of a PGSs for traits associated with kidney function have recently including from the and phenotypes to kidney estimated glomerular filtration rate rate of decline in eGFR, and such as and B. Y. et genome-wide significant variants for kidney function Int. 2021; Full Text Full Text PDF PubMed Scopus Google Y. D. et of and in the 2021; PubMed Scopus Google et polygenic scores for individuals with risk to monogenic 2018; PubMed Scopus Google Scholar have more and PGSs for kidney function have to and disease that are to kidney et Polygenic risk scores for in the J. et a polygenic risk for is associated with in a chronic kidney disease Chronic Scholar and of these scores in older aged >70 years, has the burden of kidney disease in this older age This study to the performance of recently developed kidney disease PGSs, with to association with the CKD risk at baseline and more decline in kidney function in older of environmental and risk the we a cross-sectional analysis baseline to a PGS for eGFR was associated with these in our and to the performance of a PGS for eGFR with PGSs from several to CKD the we PGSs for the in phenotype to were associated with a rate of eGFR decline our longitudinal of eGFR. In addition, we the PGS for eGFR was associated with that strongly with CKD and disease and to these C. et in chronic kidney from a Global Int. Full Text Full Text PDF PubMed Scopus Google Scholar The study genotyped Australian participants from the ASPirin in Reducing Events in the Elderly a randomized trial of daily low-dose aspirin in older et of aspirin in the healthy J 2018; 379: PubMed Scopus Google et of aspirin and in the healthy J 2018; 379: PubMed Scopus Google et of aspirin in the healthy J 2018; 379: PubMed Scopus Google Scholar The analysis from the which follow-up the to a mean of The ASPREE study of ASPirin in Reducing Events in the Elderly a controlled PubMed Scopus Google et of a trial of low-dose aspirin for in older in from the ASPirin in Reducing Events in the Elderly J PubMed Scopus Google Scholar and baseline et for clinical from the in Reducing Events in the Elderly J PubMed Scopus Google Scholar have ASPREE participants were in with a of in and the United The study was by the in with the in for genetic analysis was from study the genetic we Australian ASPREE participants with a baseline eGFR a for analysis we cross-sectional analysis of the baseline we up participants with longitudinal for analysis of eGFR we participants with stage CKD at baseline or those a kidney were for the cross-sectional association with eGFR as a and the of CKD with CKD being to 3 was Global CKD was as eGFR ml/min per 1.73 or was derived for the of participants had at was as CKD eGFR ml/min per 1.73 a to CKD phenotype as eGFR ml/min per 1.73 m2. The eGFR was the Chronic Disease A.S. et new to glomerular filtration PubMed Scopus Google Scholar and measured at baseline and study was measured at baseline with as a the longitudinal we the eGFR in a to decline in ml/min per 1.73 m2. at baseline was of or or for disease and were as in the trial assessed by as et of aspirin and in the healthy J 2018; 379: PubMed Scopus Google was the standard and were to the were to for and for as well as and variants and and variants with were by D. et of diverse from the 2021; PubMed Scopus Google Scholar that were or with scores were the of population the PGS to we individuals of European genetic by component analysis and the European of the 3 The A for genetic PubMed Scopus Google Scholar and the were with 3 with The of the component was and to with in with 3 with the European population to individuals in this study and The PGS for the eGFR analysis was by et to as J. A. et risk scores for kidney function and associations with and kidney 2021; PubMed Scopus Google Scholar This was with the recently developed PGS for CKD stage to as the of eGFR we PGSs for decline of glomerular filtration to as and B. Y. et genome-wide significant variants for kidney function Int. 2021; Full Text Full Text PDF PubMed Scopus Google A. et and of for kidney function decline derived from a of longitudinal genome-wide association Int. Full Text Full Text PDF PubMed Scopus Google Scholar We a PGS for as A. Y. et association and PubMed Scopus Google Scholar the PGSs and in the A. et and of for kidney function decline derived from a of longitudinal genome-wide association Int. Full Text Full Text PDF PubMed Scopus Google of PGSs for eGFR and decline in multiple PGSs for CKD have including population, of in or per (95% CKD eGFR of 60 ml/min per Y. Bowe B. Mokdad A.H. et al.Analysis of the Global Burden of Disease study highlights the global, regional, and national trends of chronic kidney disease epidemiology from 1990 to 2016.Kidney Int. 2018; 94: 567-581Abstract Full Text Full Text PDF PubMed Scopus (467) Google Scholar eGFR decline during a follow-up with in population and validated in eGFR ml/min per 1.73 eGFR ml/min per 1.73 of eGFR, to for in eGFR by is and of in 3 ml/min per 1.73 per decline and ml/min per 1.73 at follow-up among those with 60 ml/min per 1.73 from and kidney phenotype of eGFR to in A. et and of for kidney function decline derived from a of longitudinal genome-wide association Int. Full Text Full Text PDF PubMed Scopus Google Scholar in eGFR ml/min per at follow-up eGFR at of follow-up eGFR decline between cross-sectional CKD and decline in eGFR by European population, in have in Risk In confidence CKD, chronic kidney eGFR, estimated glomerular filtration and genome-wide association PGS, polygenic of at United and are for in a new Risk In confidence CKD, chronic kidney eGFR, estimated glomerular filtration and genome-wide association PGS, polygenic of at United and are for PGSs were by the B. K. et a for association and J Full Text Full Text PDF PubMed Scopus Google et to the of and PubMed Scopus Google Scholar This function the for in a for participant. We 3 eGFR risk for for those with PGS 95% the for those with PGS between and and for those with PGS The PGS is to have the lowest eGFR, and be to have the risk of CKD the cross-sectional we a to the between and eGFR at and the age, hypertension, diabetes, and of or and the CKD was to the between CKD and the as the the longitudinal a the between the and and eGFR in individual with for and slope as the to eGFR in the were baseline age, and hypertension, diabetes, and or the between rate of decline of eGFR and the an between the PGS and We the function from in Scholar to of the rate of per for individuals estimated to mean mean and mean were to the associations between and of disease and for were the A and for Scholar with the as the was to the between baseline eGFR and of from are the was to the between was to of the participants randomized to aspirin or in the ASPREE had a baseline eGFR and a for A were as have European genetic and individuals had stage Thus, the study population of 11,813 genotyped European the analysis had a baseline this analysis was to a of The baseline of the study are in The mean age of the study population was years and of participants were A of participants the and participants the 60 ml/min per 1.73 The up during which participants aspirin or was the of the follow-up with the follow-up to of the ASPREE participants in the mean is to that by of mean ml/min per 1.73 m2, mean of CKD, with of kidney ASPirin in Reducing Events in the CKD, chronic kidney eGFR, estimated glomerular filtration is to that by of with of kidney in a new ASPirin in Reducing Events in the CKD, chronic kidney eGFR, estimated glomerular filtration At baseline, the mean measured eGFR for the population was 72.9 ml/min per 1.73 ml/min per 1.73 years of of eGFR were with the mean eGFR to ml/min per 1.73 ml/min per 1.73 and were measured at baseline, and were measured at mean was and the was at The in our was associated with eGFR at for decrease in the was a in eGFR of ml/min per 1.73 (95% confidence interval to ml/min per 1.73 a of of The mean eGFR for participants in the and were 72.9 and ml/min per 1.73 m2, with individuals in the and individuals in the The was to be associated with the baseline of (95% to the PGS for was associated with measured at a increase in associated with a in of (95% with a of CKD we was associated with CKD with an of (95% the by this and small of in genome-wide association of our analysis eGFR. We to eGFR and CKD the The was associated with per 95% and with 95% with the had the risk for 95% the as were for to CKD 3 and In we that was still associated with 95% and with 95% were for and and for PGS are in of (95% (95% (95% (95% the confidence CKD, chronic kidney eGFR, estimated glomerular filtration glomerular filtration Global PGS, polygenic in a new of (95% (95% (95% (95% the confidence CKD, chronic kidney eGFR, estimated glomerular filtration Global PGS, polygenic and for for 60 ml/min per 1.73 or 3 60 ml/min per 1.73 and ml/min per 1.73 for age, hypertension, diabetes, and and In the as individuals with are in the the of individuals are in a new for of CKD with PGS by and risk of (95% (95% confidence CKD, chronic kidney eGFR, estimated glomerular filtration glomerular filtration Global PGS, polygenic and for the and risk 60 ml/min per 1.73 or 60 ml/min per 1.73 and ml/min per 1.73 in a new for of CKD with PGS by and risk of (95% (95% confidence CKD, chronic kidney eGFR, estimated glomerular filtration Global PGS, polygenic and for the and risk 60 ml/min per 1.73 or 60 ml/min per 1.73 and ml/min per 1.73 in a new the confidence CKD, chronic kidney eGFR, estimated glomerular filtration glomerular filtration Global PGS, polygenic the confidence CKD, chronic kidney eGFR, estimated glomerular filtration Global PGS, polygenic and for for 60 ml/min per 1.73 or 3 60 ml/min per 1.73 and ml/min per 1.73 for age, hypertension, diabetes, and and In the as individuals with are in the the of individuals are confidence CKD, chronic kidney eGFR, estimated glomerular filtration glomerular filtration Global PGS, polygenic and for the and risk 60 ml/min per 1.73 or 60 ml/min per 1.73 and ml/min per 1.73 confidence CKD, chronic kidney eGFR, estimated glomerular filtration Global PGS, polygenic and for the and risk 60 ml/min per 1.73 or 60 ml/min per 1.73 and ml/min per 1.73 We a analysis to of kidney disease the PGS individuals with of kidney disease associated with with of (95% and of (95% individuals with a of kidney disease was associated with with of (95% and of (95% In a for including individuals and the PGS, had an association in the with of (95% and of (95% a of of in was associated with decline in eGFR In was associated with more decline in eGFR the of the trial as was for the rate of decline for an individual with a in the was ml/min per 1.73 (95% to ml/min per 1.73 with ml/min per 1.73 (95% to ml/min per 1.73 for those in the The estimated mean decline per was ml/min per 1.73 (95% to ml/min per 1.73 the of decline for those in the and were (95% to and (95% to ml/min per 1.73 m2. We to the impact of by baseline A. et and of for kidney function decline derived from a of longitudinal genome-wide association Int. Full Text Full Text PDF PubMed Scopus Google Scholar In a for eGFR at follow-up with baseline eGFR as a the are significant for or We tested in the ASPREE cohort to the of CKD The was associated with of disease per or per the of the with the or of at baseline per of disease the of the ASPREE study with per (95% per (95% ASPirin in Reducing Events in the glomerular filtration PGS, polygenic for disease and are age, and The for baseline is for age, hypertension, diabetes, and and In the as individuals with are in the the of individuals are Odds in a new ASPirin in Reducing Events in the glomerular filtration PGS, polygenic for disease and are age, and The for baseline is for age, hypertension, diabetes, and and In the as individuals with are in the the of individuals are In this we that PGSs for traits are associated with CKD risk in older adults, aged years, the of genetic factors for kidney function in older age. that PGSs may have utility for risk prediction of CKD in older The was associated with baseline eGFR, both and were associated with clinical CKD by CKD individuals with PGSs had higher risk of CKD those with The was associated with or strongly associated with CKD and The PGSs for decline in and the of genetic factors associated with of Overall, our study that PGSs for kidney traits CKD risk prediction in older The of this study is the that higher per is associated with ml/min per 1.73 decrease in per in the ASPREE Although this decrease per is this is for individuals with PGS at age years had a mean eGFR of ml/min per 1.73 at the the study of ml/min per 1.73 m2. This that genetic has the to for CKD, of kidney function was by CKD or CKD stage we of PGS for kidney function in this older The derived from baseline eGFR as a has a higher per in in both and The derived from eGFR as a has a higher per A. et polygenic to chronic kidney disease across PubMed Scopus Google Scholar in in both and per and The which CKD eGFR or ml/min per 1.73 m2, may tested in a A. et polygenic to chronic kidney disease across PubMed Scopus Google Scholar The is associated with risk of CKD in individuals of CKD, by in an older population with of of disease at The has validated in a cohort age, with higher eGFR ml/min per 1.73 with a of for CKD per decrease of J. A. et risk scores for kidney function and associations with and kidney 2021; PubMed Scopus Google Scholar that a PGS is the life and can be measured in the between the and PGS may individuals at significant risk for CKD risk PGSs in disease were to of disease and individuals to risk of et and clinical risk scores and impact age at and prediction of and PubMed Scopus Google Scholar clinical of PGS those at higher risk to and the of PGS may new et analysis of the and polygenic risk of 2021; PubMed Scopus Google Scholar The we with was for and and that can kidney function or which highlights that the PGS is a of risk of conventional risk as with risk factors for CKD, a PGS individuals from kidney and PGS as of a clinical risk prediction of these Disease are with genetic risk from PGS significant in clinical as we with eGFR, with higher monogenic which more to kidney function The and PGS in our population the of genetic Chen et of renal in PubMed Scopus Google A. et al.The a new to renal in dominant polycystic kidney PubMed Scopus Google Scholar In the population, the with of disease or during follow-up or with baseline prevalence of This that the genetic captured by the PGS may be to kidney the that the ASPREE population was of disease at the that can be these in with higher risk with the ASPREE cohort have PGSs for CKD are associated with of kidney and J. et a polygenic risk for is associated with in a chronic kidney disease Chronic Scholar A PGS for CKD has associated with J. et and between kidney function and 94: PubMed Scopus Google Scholar is to in the of PGS and the are eGFR across a of individuals may have a PGS for CKD is more to individuals with disease and We validated a PGS for in the A increase in the PGS was associated with a increase in the of from of to in genome-wide association up to in the study and up to in with A. Y. et association and PubMed Scopus Google et of rate in of with PubMed Scopus Google Q. et analysis to in a the Int. Full Text Full Text PDF PubMed Scopus Google Scholar this with may be more of the of to or controlled this we for this The of our study was the validation of PGS for decline in eGFR eGFR of an of years of follow-up per participant. a significant genetic component associated with more decline in older The is as to our the has validated in to especially in this age The is a more decline phenotype however, both were in our population, and we were to significant between to of decline phenotype in our a rate in later years and the that estimated of decline between the and PGS was to a ml/min per 1.73 per Although the of eGFR decline from genome-wide association is significant genetic the of our analysis that the of between the and PGS was and may to clinical This may have influenced by the health of the ASPREE trial population at In a of was more significant in those with a eGFR, that those with a eGFR be for is at risk of more and be more to decline genetic at older J. K. et in estimated glomerular filtration rate and risk of renal disease and 2014; PubMed Scopus Google Scholar The in which baseline eGFR is in the analysis the of and are associated with eGFR which may be to is for decline in eGFR baseline eGFR as of the but which genetic and has in including genetic of eGFR decline be from genetics to cross-sectional A. K. disease and in Int. Full Text Full Text PDF PubMed Scopus Google Scholar A of this study was validation of multiple PGSs for kidney phenotypes in an healthy population, across both eGFR and eGFR had associations a the stage as the but both are of kidney scores performance for both and in kidney with higher in The is as both are from and from European is with higher of and of in the with of eGFR CKD to of eGFR across that in of European but of or A. et polygenic to chronic kidney disease across PubMed Scopus Google Scholar was for ml/min per 1.73 per of eGFR, the rate in was still validated in this We were to this from a phenotype with a rate of the a our of significance of PGS at older Overall, kidney phenotypes in ASPREE were with in validation of recently developed of this study in the longitudinal and of eGFR in an older population, validated in multiple The ASPREE cohort is a with that participants were from or at the population is to have higher of CKD with ASPREE at age and a of eGFR A of this study a of the cohort and healthy is that were of CKD to and kidney to these phenotypes or to the PGS was associated with or individuals as of that are of the population of older A in PGS is the risk of by population strongly associated with may be more of the in Y. to population in of or PubMed Scopus Google Scholar we this risk by the study to individuals of European a our be to and is more for including of in A. et polygenic to chronic kidney disease across PubMed Scopus Google Scholar The clinical of the eGFR PGSs an as for are The of this study that individuals for CKD, decline in eGFR, at older may be a of PGSs in the This may have in older aged years, in the prevalence of CKD increases validation of the PGSs tested is in the population, and across age and to for clinical PGSs for kidney traits CKD risk prediction and may a risk that is to and have from the of for the ASPREE trial and has from the of and and and for from of this has from the Australian of and and from of this the and phenotype that the of this study have in the European to from the ASPREE clinical trial including can be for the ASPREE This was by an ASPREE the and Australian and of by and from the and the at the of and from the and of and and by and the of the ASPREE cohort was by is by a is by a with of for component analysis of the ASPirin in Reducing Events in the Elderly cohort with the of polygenic scores SD, in estimated glomerular filtration rate (eGFR) across the and of individual for estimated glomerular filtration rate (eGFR) and estimated glomerular filtration rate (eGFR) by polygenic (PGS) and by estimated glomerular filtration rate (eGFR) polygenic (PGS) Polygenic scores for chronic kidney disease (CKD) in for component

Topics & Concepts