Lipids, Blood Pressure, and Diabetes Mellitus on Risk of Cardiovascular Diseases in East Asians: A Mendelian Randomization Study

Abstract

Elevated blood pressure, dyslipidemia, and impaired glycemic control are well-established cardiovascular risk factors in Europeans, but there are comparatively few studies focused on East Asian populations. This study evaluated the potential causal relations between traditional cardiovascular risk factors and disease risk in East Asians through a 2-sample Mendelian randomization approach. We collected summary statistics for blood pressure parameters, lipid subsets, and type 2 diabetes mellitus liability from large genome-wide association study meta-analyses conducted in East Asians and Europeans. These were paired with summary statistics for ischemic heart disease (IHD), ischemic stroke (IS), peripheral vascular disease, heart failure (HF) and atrial fibrillation (AF). We performed univariable Mendelian randomization analyses for each exposure-outcome pair, followed by multivariable analyses for the available lipid subsets. The genetically predicted risk factors associated with IHD and AF were similar between East Asians and Europeans. However, in East Asians only genetically predicted elevated blood pressure was significantly associated with IS (odds ratio 1.05, 95% confidence interval 1.04 to 1.06, p <0.0001) and HF (odds ratio 1.05, 95% confidence interval 1.04 to 1.06, p <0.0001), whereas nearly all genetically predicted risk factors were significantly associated with IS and HF in Europeans. In conclusion, this study provides supportive evidence for similar causal relations between traditional cardiovascular risk factors and IHD and AF in both East Asian and European ancestry populations. However, the identified risk factors for IS and HF differed between East Asians and Europeans, potentially highlighting distinct disease etiologies between these populations. Elevated blood pressure, dyslipidemia, and impaired glycemic control are well-established cardiovascular risk factors in Europeans, but there are comparatively few studies focused on East Asian populations. This study evaluated the potential causal relations between traditional cardiovascular risk factors and disease risk in East Asians through a 2-sample Mendelian randomization approach. We collected summary statistics for blood pressure parameters, lipid subsets, and type 2 diabetes mellitus liability from large genome-wide association study meta-analyses conducted in East Asians and Europeans. These were paired with summary statistics for ischemic heart disease (IHD), ischemic stroke (IS), peripheral vascular disease, heart failure (HF) and atrial fibrillation (AF). We performed univariable Mendelian randomization analyses for each exposure-outcome pair, followed by multivariable analyses for the available lipid subsets. The genetically predicted risk factors associated with IHD and AF were similar between East Asians and Europeans. However, in East Asians only genetically predicted elevated blood pressure was significantly associated with IS (odds ratio 1.05, 95% confidence interval 1.04 to 1.06, p <0.0001) and HF (odds ratio 1.05, 95% confidence interval 1.04 to 1.06, p <0.0001), whereas nearly all genetically predicted risk factors were significantly associated with IS and HF in Europeans. In conclusion, this study provides supportive evidence for similar causal relations between traditional cardiovascular risk factors and IHD and AF in both East Asian and European ancestry populations. However, the identified risk factors for IS and HF differed between East Asians and Europeans, potentially highlighting distinct disease etiologies between these populations. Dyslipidemia, hypertension, and type 2 diabetes mellitus (T2DM) have consistently been identified as common modifiable risk factors for cardiovascular diseases. However, most evidence comes from European ancestry populations,1Mihaylova B Emberson J Blackwell L Keech A Simes J Barnes EH Voysey M Gray A Collins R Baigent C Cholesterol Treatment Trialists' (CTT) CollaboratorsThe effects of lowering LDL cholesterol with statin therapy in people at low risk of vascular disease: meta-analysis of individual data from 27 randomised trials.Lancet. 2012; 380: 581-590Abstract Full Text Full Text PDF PubMed Scopus (2105) Google Scholar, 2Blood Pressure Lowering Treatment Trialists' CollaborationPharmacological blood pressure lowering for primary and secondary prevention of cardiovascular disease across different levels of blood pressure: an individual participant-level data meta-analysis [published correction appears in Lancet 2021;397:1884].Lancet. 2021; 397: 1625-1636Abstract Full Text Full Text PDF PubMed Scopus (305) Google Scholar, 3Griffin SJ Leaver JK Irving GJ. Impact of metformin on cardiovascular disease: a meta-analysis of randomised trials among people with type 2 diabetes.Diabetologia. 2017; 60: 1620-1629Crossref PubMed Scopus (274) Google Scholar with comparatively little evidence from subjects of East Asian ancestry despite comprising at least one-fifth of the world population. Several observational studies have demonstrated notable epidemiologic differences in the burden of cardiovascular diseases for East Asians compared with Europeans and North Americans.4Zhao D. Epidemiological features of cardiovascular disease in Asia.JACC Asia. 2021; 1: 1-13Crossref PubMed Scopus (84) Google Scholar, 5Peters SAE Wang X Lam TH Kim HC Ho S Ninomiya T Knuiman M Vaartjes I Bots ML Woodward M Asia Pacific Cohort Studies CollaborationClustering of risk factors and the risk of incident cardiovascular disease in Asian and Caucasian populations: results from the Asia Pacific Cohort Studies Collaboration.BMJ Open. 2018; 8e019335Crossref Scopus (40) Google Scholar, 6Park TH Ko Y Lee SJ Lee KB Lee J Han MK Park JM Cho YJ Hong KS Kim DH Cha JK Oh MS Yu KH Lee BC Yoon BW Lee JS Lee J Bae HJ. Identifying target risk factors using population attributable risks of ischemic stroke by age and sex.J Stroke. 2015; 17: 302-311Crossref PubMed Scopus (35) Google Scholar However there remains a relative scarcity of randomized trials performed in East Asian populations to investigate these observational findings. Mendelian randomization (MR) is a research methodology that can test potentially causal relations between cardiovascular risk factors and diseases. A proportion of the phenotype of an individual is determined by genetic polymorphisms that are randomly inherited at birth. The random allocation of phenotype-determining genetic polymorphisms is analogous to assignment to a treatment group in a randomized control trial (Figure 1). MR leverages this random and non-modifiable allocation of genetic variants to help eliminate the risk of bias because of reverse causality and minimize bias from confounding, both of which typically hinder interpretation of nonrandomized observational studies. The present study therefore investigates the potentially causal relations between traditional cardiovascular risk factors and common cardiovascular diseases in East Asian ancestry individuals and compares results with those obtained in European ancestry individuals. The present study uses a 2-sample MR approach. The exposure variables included low-density lipoprotein cholesterol (LDL-c), triglycerides (TGs), high-density lipoprotein cholesterol (HDL-c), systolic blood pressure (SBP), diastolic blood pressure (DBP) and T2DM liability. Ischemic heart disease (IHD), ischemic stroke (IS), peripheral vascular disease (PVD), heart failure (HF), and atrial fibrillation (AF) were used as the outcome variables. Population homogeneity is a requirement for the samples used in a 2-sample MR analysis, and thus both exposure and outcome data were sourced from either East Asian or European ancestry populations, according to the specified ancestry in the analysis. MR makes 3 further core assumptions which are that the genetic variants should be: (1) strongly associated with the exposure; (2) exclusively associated with the outcome through the exposure; and (3) independent of potential confounders. To address the first criterion, uncorrelated genetic variants were selected that were significantly associated with the exposure subsets at p <5 × 10−8; and to evaluate the latter 2 criteria, sensitivity analyses were performed as described later. Data for the present study are publicly available. Ethical approval and consent were obtained by the original studies. Preparation of this manuscript was based on the Strengthening the Reporting of Observational Studies in Epidemiology using Mendelian Randomization Guidelines.7Skrivankova VW Richmond RC Woolf BAR Davies NM Swanson SA VanderWeele TJ Timpson NJ Higgins JPT Dimou N Langenberg C Loder EW Golub RM Egger M Davey Smith G Richards JB Strengthening the reporting of observational studies in epidemiology using Mendelian randomisation (STROBE-MR): explanation and elaboration.BMJ. 2021; 375: n2233Crossref PubMed Scopus (171) Google Scholar For the East Asian analyses: the data for lipid subsets was extracted from a genome-wide association study (GWAS) meta-analysis performed by the for and using data from which included East Asian KH S S I I S Han S A A A B HC X G A I M S C M Y T SA J Y A G A C M A MR Yu A S X J Yoon C A G Y S M J A J J A M R C AF M S N M L J B A A C L M EW JM JS C MK M M N J RC KS S J J J B M M Wang C M A N X R A M MR C Kim S L R X G L M J X A S Wang T YJ YJ S J M M Smith M C J S N Lee J S Lee M S L Wang Wang JS C G B J X N L T J S S T I T T Kim X T M L C A A C Cho Lee JM G MK G S G R S G N G B X M X M JB NJ J SA RM A L B Kim J M JM T L M RM T N M J KH C J X M G M M I S J M J A M MR N N Timpson NJ C A C M Kim YJ Y NJ Langenberg C J M G JS North M JB S C Smith A C B M Y Y Y Ho Cho JM S M RC G M Kim S M of genetic in genome-wide association studies of [published correction appears in 2021; PubMed Scopus Google Scholar statistics available data were extracted from a meta-analysis of and from East Asian from statistics for are available from the of the original M N T M Y A S Kim YJ L Wang C R Wang J A Y M M M A S T M Han S Lee Lee J ML C J L S J JM Y L R L JS M JB J X X M Kim T T M Y M N of Pressure analyses of blood pressure in populations of East Asian and European 2018; PubMed Scopus Google and T2DM liability data were extracted from a meta-analysis of East Asian ancestry individuals from individuals from statistics available M Kim YJ S Y J M J M M A M ML S M X Y Han S YJ S M G JB Y T T Lee MS Lee L J J Y C J A B RM T Yoon Yu C JM L M Cho Wang M N Park KS S Lee J T Kim of type 2 diabetes in East Asian PubMed Scopus Google Scholar For the European analyses: the and data were extracted from a meta-analysis performed by the using data from to and KH S S I I S Han S A A A B HC X G A I M S C M Y T SA J Y A G A C M A MR Yu A S X J Yoon C A G Y S M J A J J A M R C AF M S N M L J B A A C L M EW JM JS C MK M M N J RC KS S J J J B M M Wang C M A N X R A M MR C Kim S L R X G L M J X A S Wang T YJ YJ S J M M Smith M C J S N Lee J S Lee M S L Wang Wang JS C G B J X N L T J S S T I T T Kim X T M L C A A C Cho Lee JM G MK G S G R S G N G B X M X M JB NJ J SA RM A L B Kim J M JM T L M RM T N M J KH C J X M G M M I S J M J A M MR N N Timpson NJ C A C M Kim YJ Y NJ Langenberg C J M G JS North M JB S C Smith A C B M Y Y Y Ho Cho JM S M RC G M Kim S M of genetic in genome-wide association studies of [published correction appears in 2021; PubMed Scopus Google Scholar statistics available data were extracted from a meta-analysis of and to European statistics for available from the of the original B R G Dimou N I M A Cho JM T R L T S J R J M C Y M N S N J R M C R G A S G N C T M M M A A G M M Collins Davies G J M R S T M I C G A X A A SJ A R MR R A M S J I JS S M M B T L L Y Y A C J J Y A T T S A A M A M R JS S A I Y NJ R N Smith S R JM J A M M S J C A G AF J T J C M Langenberg C L NJ C A I Barnes MR of people associated with blood pressure [published correction appears in 2018; PubMed Scopus Google and T2DM liability data were extracted from a meta-analysis of and to European ancestry individuals statistics available A Y L J Y AF J J association analyses risk variants and for type 2 2018; PubMed Scopus Google Scholar polymorphisms that were associated with each exposure at p <5 × were was performed East Asian or European ancestry of the and the variants with the p were selected for further analysis. analyses of data were used for all East Asian S M Y J M S A T M A S Y T N S S S M A Y C T I T G M Y M Y Y A Y A of genetic for 2021; PubMed Scopus Google Scholar S M Y J M S A T M A S Y T N S S S M A Y C T I T G M Y M Y Y A Y A of genetic for 2021; PubMed Scopus Google Scholar peripheral disease S M Y J M S A T M A S Y T N S S S M A Y C T I T G M Y M Y Y A Y A of genetic for 2021; PubMed Scopus Google Scholar HF S M Y J M S A T M A S Y T N S S S M A Y C T I T G M Y M Y Y A Y A of genetic for 2021; PubMed Scopus Google Scholar and A Y S M M M T M S M T M Y of genetic associated with atrial fibrillation in the 2017; PubMed Scopus Google Scholar For each this and IS and and and and AF and were by and of Data for and are available S M Y J M S A T M A S Y T N S S S M A Y C T I T G M Y M Y Y A Y A of genetic for 2021; PubMed Scopus Google Scholar Data for AF is available A Y S M M M T M S M T M Y of genetic associated with atrial fibrillation in the 2017; PubMed Scopus Google Scholar For the of reporting results is as as and as For the European ancestry analyses: IHD data were extracted from a meta-analysis of studies disease and with of of European ancestry statistics available M A C S T L A M A S I N C C S J SJ Kim X Y N L R M Smith A X M L G M Han J S T L L ML N T MS A A M J S Wang L KS Collins R J I R T A C A C Kim JS T A M T S R AF J J R S R NJ M A genome-wide association meta-analysis of 2015; PubMed Scopus Google IS data were extracted from a meta-analysis IS and of European ancestry statistics available R G M M Y A L S M T S C A I C J M S J A G MR X J M C M Langenberg C Lee JM R J A TH Ninomiya T T S R A C T G S S C M NJ S S T A S M Smith C SA JS N J C I A J J G SJ S M Y S M for and in Epidemiology of Pressure of the genome-wide association study of subjects associated with stroke and stroke [published correction appears in 2018; PubMed Scopus Google data were extracted from the using the which included and statistics available Davies NM Davey Smith G a for in J 2018; PubMed Scopus Google HF data were extracted from a meta-analysis of studies and of European ancestry statistics available S A C G G JB A N A C J ML J MR L X X J S M G T C I M S JS S R M L A Langenberg C L B RC J A KB T A T ML M J Smith S ML M A G C S B A Wang X NJ R Yu B NJ J C A SA N Smith association and Mendelian randomisation the of heart PubMed Scopus Google and AF data were extracted from a of studies and of European ancestry statistics available JB TJ S G I MR M M R A C X M S B Lee S J J atrial fibrillation 2018; PubMed Scopus Google Scholar the methodology and phenotype can in the original The primary MR was an meta-analysis of the for each genetic are as of the outcome in genetically predicted lipid variables and in variables. A p was at p for all primary analyses based on exposure and outcome and p for multivariable analyses on univariable because 3 lipid subsets were included in each 2 to to To evaluate for of the MR assumptions because of performed sensitivity analyses and and makes different assumptions and thus between provides confidence in the The that at least the variables are that the most common causal is with the uses the of which that the of effects from the genetic variants to the outcome is independent of the of the association between genetic variants and this the from the of the genetic variants typically compared with is to and and was used to test for genetic analyses were performed for each exposure-outcome 2 and and are in to MR was performed to the effects of the lipid subsets on the outcome were independent of each 2 and 1). that were available in all lipid subsets and outcome summary were variants associated with at least lipid at p <5 × were from to p of association with East or European ancestry of the and multivariable MR was randomisation for the of in genetically predicted elevated low-density lipoprotein triglycerides and high-density lipoprotein cholesterol for the lipid heart disease stroke vascular disease failure fibrillation lipoprotein lipoprotein lipoprotein lipoprotein in a analyses were performed using R with the R for In the a association was between nearly all traditional cardiovascular risk factors and IHD in East Asians ratio 95% confidence interval to p 95% to p 95% to p 95% to 1.06, p 95% to p (Figure The association between T2DM and IHD in East Asians on 95% to p In the European all genetically predicted risk were significantly associated with IHD For both the East Asian and European multivariable MR for all 3 lipid subsets in the demonstrated a similar association between genetically predicted and with risk of but there was association between genetically predicted and IHD risk in the East Asian 1). In East a association was between the genetically predicted and IS on univariable 1.05, 95% 1.04 to 1.06, p 95% to p <0.0001), but there was association between of the genetically predicted lipid or T2DM on IS risk (Figure This is in to the European analysis, in which were identified between all risk and were between genetically predicted T2DM and with risk of 95% to p 95% to p 95% to p 95% to 1.06, p but genetically predicted or with risk 95% to p 95% to p (Figure multivariable there a association between both and with In for the European only genetically predicted T2DM liability was significantly associated with risk on univariable analysis. to the results for in East Asians a association was between the genetically predicted and HF on univariable 1.05, 95% 1.04 to 1.06, p 95% to p <0.0001), but association between of the genetically predicted lipid subsets or T2DM liability on HF risk (Figure This the European analysis, in which were identified between all risk factors and In East genetically predicted was associated with risk of AF 1.06, 95% to 95% to with association between of the genetically predicted lipid subsets or T2DM liability on AF (Figure In the European analysis, genetically predicted was associated with risk of but the lipid subsets, and T2DM liability were pressure, lipid levels and T2DM are to modifiable risks by both genetic and The present study data and the MR to evaluate the association between traditional cardiovascular risk factors and diseases in East Asian ancestry individuals. We similar risk for IHD and AF between East Asian and European ancestry populations. However, for East in to Europeans, only genetically predicted elevated was significantly associated with IS and The risk of cardiovascular diseases because of dyslipidemia, hypertension, and T2DM been in European populations through large observational and randomized B Emberson J Blackwell L Keech A Simes J Barnes EH Voysey M Gray A Collins R Baigent C Cholesterol Treatment Trialists' (CTT) CollaboratorsThe effects of lowering LDL cholesterol with statin therapy in people at low risk of vascular disease: meta-analysis of individual data from 27 randomised trials.Lancet. 2012; 380: 581-590Abstract Full Text Full Text PDF PubMed Scopus (2105) Google Scholar, 2Blood Pressure Lowering Treatment Trialists' CollaborationPharmacological blood pressure lowering for primary and secondary prevention of cardiovascular disease across different levels of blood pressure: an individual participant-level data meta-analysis [published correction appears in Lancet 2021;397:1884].Lancet. 2021; 397: 1625-1636Abstract Full Text Full Text PDF PubMed Scopus (305) Google Scholar, 3Griffin SJ Leaver JK Irving GJ. Impact of metformin on cardiovascular disease: a meta-analysis of randomised trials among people with type 2 diabetes.Diabetologia. 2017; 60: 1620-1629Crossref PubMed Scopus (274) Google Scholar and to in The present study this for IHD and between East Asian and European ancestry populations for the association between traditional cardiovascular risk factors and these However, the present study evidence for a causal of genetically predicted elevated but or as risk factors for IS and HF in East studies in East Asians have consistently demonstrated that elevated to a risk stroke compared with and diabetes risk factors for stroke in East Stroke. PubMed Scopus Google Scholar evidence demonstrated that East Asians are with hypertension, and East Asian populations a Europeans which is of stroke S Y S Y A M J and blood pressure and risks of incident stroke and disease in the the blood pressure PubMed Scopus Google Scholar the of East Asians to elevated was by Park S Park S Wang on in Asian Asian 2018; PubMed Scopus Google Scholar in a on for Asian populations. The that the lipid subsets were significantly associated with IHD and but or AF in East Asians was in the of the results from European ancestry individuals. This to have been The of Elevated Cholesterol in the of which to a in risk of and heart disease but in risk of IS in the A Y T N S M A Y prevention of cardiovascular disease with in a randomised Full Text Full Text PDF PubMed Scopus Google Scholar In the but Treatment in which with IS and were randomized to or the group of but the of stroke or ischemic was significantly different from the control N Y T T S T N C Y M S M M S Y T M The statin treatment stroke a 2015; Full Text Full Text PDF PubMed Scopus Google Scholar study outcome of or The association between genetically predicted and nearly all the in the European analyses is in of studies which have the causal association between and cardiovascular elevated high-density lipoprotein cholesterol cardiovascular [published J A M A A in the a for 2021; PubMed Scopus Google Scholar This the in the present study compared with these results were consistently by the sensitivity which is similar to the by and Davey S Davey Smith G Mendelian randomization high-density lipoprotein in of 2017; Full Text Full Text PDF PubMed Scopus Google Scholar that demonstrated a association in MR between and disease on but sensitivity ancestry been to the of T2DM on risk of The present study an association between genetically predicted T2DM and elevated risk of in East Asians and a association with IHD that for studies on the between T2DM and cardiovascular diseases in East Asians have been in findings. A study of in North risk of incident stroke and and risk of disease in Asian compared with European ancestry A in in an population [published correction appears in PubMed Scopus Google Scholar In the in and study that Asian with T2DM of stroke and and of and disease compared with European and A J Woodward M and associated with of a PubMed Scopus Google Scholar study was to East and is that Asia there is cardiovascular disease D. Epidemiological features of cardiovascular disease in Asia.JACC Asia. 2021; 1: 1-13Crossref PubMed Scopus (84) Google Scholar are was for the outcome in Europeans, for which there were only and the were because of the outcome variables were extracted from is that there were in the of as a the outcome between East Asian and European analyses were as the analyses were by the of large for was for East Asians as whereas the European included and this for the of IHD from disease to the of each exposure on risk of IHD was between East Asians and Europeans. HF and IS are diseases. for with different etiologies that between ancestry and this to the risk between ancestry there is between of the exposure and outcome the of by only using associated with the exposure at p <5 × with associated is to the potential risk of there was evidence of in of the determined by of the the between the and sensitivity analyses that bias is to have the this study used data of East Asian individuals to evaluate the between traditional cardiovascular risk factors and diseases in a population that is relative to the proportion of the population. This demonstrated a similar risk for IHD and AF between East Asians and Europeans but that elevated the risk for IS and HF in East This study an research that further cardiovascular disease prevention and treatment in East Asian populations. The have to with with

Topics & Concepts