Mendelian randomization studies on ischemic stroke: a field synopsis and systematic review

Abstract

Background

Ischemic stroke (IS) is a major cause of disability and death worldwide. However, previous observational studies failed to establish the causality between various exposures and IS occurrence. Currently, an increasing number of Mendelian randomization (MR) studies have been performed to investigate the causal effects of exposure factors on IS risk, but the results remain inconsistent and inconclusive. Thus, our systematic review summarized all published MR studies focusing on IS and its subtypes, and further identified causal relationships with robust evidence.

Methods

We retrieved PubMed and Embase databases for MR analyses exploring the correlation of genetically determined exposures with the risk of IS and its subtypes. For publications selected in the main analysis, we summed up these MR results and classified the strength of evidence into Grade 1–4 based on the significance and concordance of the findings from the primary and complementary MR methods, as well as the robustness of sensitivity analyses.

Results

After conducting a comprehensive search, a total of 207 published studies with 1199 MR associations were included. Among these causal correlations, 8 (Grade 1) and 81 (Grade 2) were graded as robust evidence, while 226 (Grade 3) and 884 (Grade 4) causal relationships were considered to have insufficient reliability. Specifically, with regard to IS susceptibility, genetically predicted diastolic blood pressure, type 2 diabetes, Parkinson’s disease, and primary aldosteronism were evaluated as the prominent robust causal determinants. As for IS subtypes, genetically determined higher vitamin K1 levels and primary aldosteronism were related to greater occurrence of large artery stroke and small vessel stroke, respectively, whereas increased adult height was associated with a lower risk of small vessel stroke but a higher risk of cardioembolic stroke.

Conclusions

Our field synopsis systematically summarized and evaluated the evidence strength for causality between various exposures and the occurrence of IS and its subtypes, and ultimately identified 89 causal relationships as convincing findings. We hope that our results can provide constructive and impressive insights for developing effective IS prevention strategies.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12967-025-06992-4.

Introduction

Stroke is one of the main causes of long-term disability and mortality worldwide. In accordance with the global burden of disease (GBD) study 2019, the number of patients with stroke reached 12.22 million, and deaths caused by stroke rose to 6.55 million [1]. In addition, stroke also brings a huge economic burden to patients, families, and society, with global costs of stroke reaching $2059 billion in 2019 [2]. As the most common subtype of stroke, ischemic stroke (IS) accounts for approximately 80% of stroke cases [3]. Thus, it is necessary to have an in-depth insight into various risk factors to decrease the morbidity and mortality of IS. Previous observational studies have assessed the relationships of numerous environmental factors on the risk of IS and its subtypes, including large artery stroke (LAS), small vessel stroke (SVS), and cardioembolic stroke (CES), and have identified dozens of potential risk factors for IS susceptibility [4–6]. However, the direction and magnitude of these associations are sometimes inconsistent and controversial, which reflects the limitations of observational studies, including confounding, reverse causality and small sample size [7, 8]. Therefore, effective analysis methods for evaluating causal associations are urgently needed.

Mendelian randomization (MR) analysis utilizes genetic variants, usually single nucleotide polymorphisms (SNPs), as instrumental variables (IVs) for exposure to assess the causal relationship between exposure and clinical outcome of interest. Since genetic variants are randomly assigned at conception and not disturbed by environmental factors, the results of MR analyses are less susceptible to confounding factors [9, 10]. Additionally, MR studies can effectively avoid the possibility of reverse causality, as genetic variants are not affected by disease development [11]. Furthermore, two-sample MR, the most commonly used study design in MR analysis, typically identifies genetic variants associated with exposures from large-scale genome-wide association studies (GWAS) and then applies these genetic variants to independent outcome datasets, instead of obtaining exposure and outcome data from each patient, overcoming the problem of small sample size in observational studies [12]. Although MR studies have leveraged inherent methodological strengths to investigate the potential causal effects of various risk factors on IS and its subtypes, the findings have not always been consistent. For instance, Deng et al. performed an MR analysis by using summary-level genetic data from the UK Biobank and the MEGASTROKE consortium and found no causal association between tea consumption and IS occurrence [13]. However, a more recent MR study by Gao et al., despite utilizing the same GWAS statistics, reported a protective causal role of tea consumption in IS [14]. This discrepancy could be attributed to specific differences in the employed methods, further underscoring the necessity of evaluating the robustness of MR analyses based on standardized design. Consequently, in this field synopsis, we systematically reviewed all published MR studies exploring the risk of IS and its subtypes and assessed the evidence strength from these MR analyses to gain novel insights into IS prevention strategies.

Methods

Due to the fact that the data were extracted from previously published MR studies, this systematic review was exempt from institutional review board approval. Moreover, the authors declare that all supporting data are available within the article and its online additional files.

Search strategy

To comprehensively review MR studies of IS and its subtypes, we performed a systematic literature search of Embase and PubMed databases as of April 1, 2025, using the following search strategies: “mendelian randomization/mendelian randomisation” and “stroke/cerebrovascular disease/IS/cerebral infarction/brain infarction”. In addition, we screened the references of the retrieved articles for any potentially eligible articles. The entire search strategy for PubMed and Embase databases is displayed in Additional file 1.

Study selection

For articles identified using the above search strategy, we carefully examined their titles, abstracts and full texts. Studies met the following criteria: (1) with an MR design; (2) relevant to the risk of IS or its subtypes; (3) associated with various exposures, including anthropometric traits, socio-economic and lifestyle factors, dietary factors, various diseases and associated traits, as well as drug targets; and (4) published in English. In contrast, the following articles were excluded: (1) case reports, reviews, letters, editorials, commentaries, communications, or conference abstracts; (2) studies without full text available for further analysis; (3) duplicated studies. The screening process of eligible MR studies is depicted in Fig. 1.

Fig. 1.

Flowchart of MR studies selection process

In addition, when more than one study was available to assess the causal effect of the same exposure factor on IS risk, the main analysis would be selected according to the following three steps: (1) we first prioritized analyses with rigorous study designs that used SNP selection criteria including a genome-wide significance threshold of P < 5 × 10⁻⁸, a stringent linkage disequilibrium cutoff (r² < 0.001), and F-statistic threshold above 10 to avoid weak instruments; (2) MR studies with the largest sample sizes were favored; (3) lastly, we tended to choose studies with exposure and outcome populations from the same ethnicity. The above screening process was completed independently by two authors, and any disagreements were resolved by discussion with a third author.

Data extraction

Two authors independently extracted the following data from eligible studies: first author’s name, year of publication, exposure and outcome of interest, sample size and ethnicity of exposure and outcome populations, number of genetic variants, P-value threshold, linkage disequilibrium (r²), instrument strength (F-statistic), variance explained (R²), effect estimates with 95% confidence intervals, P-value of effect estimates, study design (one-sample or two-sample MR), primary MR method and supplementary analysis methods. Specifically, when the exposure in MR studies was analyzed using multiple instrumental variables, we extracted the causal estimates from the inverse-variance weighted (IVW) method as the primary findings. Meanwhile, results from complementary methods, such as the weighted median, weighted mode, MR-Egger, and Mendelian Randomization robust adjusted profile score (MR-RAPS), were retrieved to assess consistency in the direction of effect and statistical significance. In contrast, when only a single genetic instrument was available for an exposure, causal estimates were sourced exclusively from the Wald ratio method, without the support of supplementary MR approaches. Lastly, due to the presence of heterogeneity and pleiotropy in MR studies, we extracted information on sensitivity analyses including leave-one-out analysis, MR Pleiotropy Residual Sum and Outlier (MR-PRESSO), MR Egger intercept, Cochran’s Q statistic, and the MR Steiger test to evaluate the credibility of the reported causal relationships.

Evaluation of robustness of associations identified by MR studies

The flowchart of the robustness assessment of the MR studies included in the main analysis is displayed in Fig. 2. The detailed grading criteria are presented as follows. Grade 1: All MR methods (IVW and supplementary) showed significant causal effects with consistent directions, and sensitivity analyses confirmed robustness. Grade 2: Primary IVW was significant, supplementary methods showed consistent directions, and sensitivity analyses confirmed robustness. Grade 3: At least one MR method was significant (with potential directionality inconsistencies). Grade 4: All MR methods were non-significant. Notably, statistical significance was determined using corrected P-values when multiple testing was reported; otherwise, we applied the nominal significance threshold of P < 0.05. Lastly, based on the abovementioned grading framework, Grade 1 and Grade 2 results were considered to provide robust evidence for the causal associations between various exposure factors and IS as well as its subtypes in our systematic review.

Fig. 2.

Flowchart of robustness assessment of MR studies included in main analysis

Results

Study identification

As depicted in Fig. 1, 2347 articles were identified from the PubMed and Embase databases. After removing 847 duplicated publications, we further excluded 773 irrelevant studies based on titles and abstracts. Next, through full-text screening, 382 studies were deleted and the specific reasons for the exclusion of full-text articles are displayed in Additional file 1. Ultimately, 345 articles were included in the present systematic review, and the detailed characteristics of all studies enrolled are presented in Additional file 2.

Characteristics of studies included in main analysis

Out of the 345 eligible articles, 207 MR studies were included in the main analysis (Fig. 1), and details of these MR analyses are provided in Additional file 3 [13–219]. Overall, we identified 1199 causal estimates from these 207 MR studies, which investigated the causal effects of 433 different exposures on IS and its subtypes. Of these 207 articles, the median number of SNPs used as instrumental variants was 13, ranging from 1 to 2,270. In addition, the median number of exposure sample size in MR studies was 166,161 (range 338 to 3,037,499). Regarding the number of outcome sample size, the median was 440,328 (range 11,312 to 1,614,080). Most of the data of SNPs for exposure were obtained from populations of European ancestry (n = 506; 98.4%), followed by Asian (n = 4; 0.8%) and multi-ancestry (n = 4; 0.8%), and the populations of outcome were derived from European (n = 490; 95.3%), multi-ancestry (n = 22; 4.4%) and Asian (n = 2; 0.3%). Moreover, 1,189 (99.9%) MR correlations used a two-sample MR design and 1 (0.1%) utilized a one-sample design.

Robustness of the evidence

Anthropometric traits

Following rigorous evaluation of 11 MR studies included in the main analysis, we ultimately identified 15 robust causal associations of anthropometric traits with the risk of IS and its subtypes (Fig. 3). Specifically, as comprehensive indicators reflecting body fat distribution and overall body size, robust MR results demonstrated that genetically predicted higher waist-to-hip ratio (WHR) and body mass index (BMI) were causally associated with an increased risk of IS and LAS [60, 92]. In addition, basic anthropometric traits such as adult height, waist circumference, and weight exhibited robust causal effects on specific stroke subtypes in MR analyses [110, 131]. For example, genetically predicted adult height, waist circumference, and weight causally increased the risk of CES [110, 131], however, per standard deviation (SD) increment in adult height was reported to significantly reduce the incidence of SVS by 15% in the MR analysis conducted by Linden et al. [110].

Fig. 3.

Robust evidence for the causal effects of anthropometric traits on ischemic stroke and its subtypes

Socio-economic and lifestyle factors

Socio-economic and lifestyle factors represented an important area in MR studies of IS, particularly given the limitations or infeasibility of conducting randomized controlled trials (RCTs). Within this category of exposures, 3 causal associations were assessed as grade 2 at the robust level (Fig. 4). Among these, the findings of the MR study indicated that genetically predicted educational attainment was associated with a significantly reduced risk of SVS [173]. As for sleep-related traits, Titova et al. found that genetic predisposition to short sleep duration was specifically associated with LAS, whereas their MR analyses yielded no causal relationship for other stroke subtypes [148]. Additionally, with respect to smoking, convincing MR evidence supported the idea that genetically determined higher lifetime smoking index causally elevated the risk of IS by 34% [92].

Fig. 4.

Robust evidence for the causal effects of socio-economic and lifestyle factors on ischemic stroke and its subtypes

Dietary factors

Among the dietary factors included in our systematic review, only 5 exposures were found to exert robust causal influence on IS and its subtypes, based on stringent selection criteria and comprehensive evaluation (Fig. 5). Of these associations, only the positive causal relationship between genetically predicted vitamin K1 concentrations and LAS was assessed as grade 1 [90], while the remaining dietary factors were assessed as grade 2. Specifically, MR studies focusing on various micronutrients clarified that genetic predisposition to per SD increase in circulating levels of iron elevated a 21% risk of CES [19]. In addition, MR analyses on the daily intake of common foods and beverages revealed robust causal links indicating that genetically determined higher consumption of dried fruit and tea protected against both IS and SVS [13, 14, 189].

Fig. 5.

Robust evidence for the causal effects of dietary factors on ischemic stroke and its subtypes

Cardiometabolic diseases and associated indices

As shown in Fig. 6, a total of 19 robust causal correlations were obtained from 29 MR analyses included in the main analysis, which investigated the associations of cardiometabolic diseases and related indices with the susceptibility to IS and its subtypes. Regarding blood pressure, the MR findings by Le et al. provided solid evidence that genetic liability to higher diastolic blood pressure (DBP) was causally related to the increased incidence of IS (Grade 1) [92]. However, although significant causal associations were observed between DBP and specific stroke subtypes, these findings failed to pass sensitivity analyses of pleiotropy or directional tests, thereby limiting the evidence strength of causal inference for specific IS subtypes [93, 139]. Similarly, in terms of blood lipid traits, although multiple MR studies reported significant causal relationships between low-density lipoprotein cholesterol (LDL-C) and the risk of IS as well as its subtypes, only the causal influence on overall IS development was graded as robust evidence (Grade 1) [92]. Moreover, genetically predicted higher levels of lipoprotein(a) elevated LAS occurrence, implying that lipoprotein(a) exerted its pathogenic influence primarily via atherosclerotic mechanisms, rather than uniformly affecting all stroke subtypes [71]. In addition, recently published MR studies also provided compelling insights that genetic predisposition to type 1 and type 2 diabetes (T2D), including various novel subtypes of type 2 diabetes, such as mild age-related diabetes and severe insulin-deficient diabetes, were causally linked to SVS incidence [134, 196].

Fig. 6.

Robust evidence for the causal effects of cardiometabolic diseases and associated indices on ischemic stroke and its subtypes

Non-cardiometabolic diseases and related indices

Beyond traditional vascular risk factors, a series of MR studies focusing on non-cardiovascular diseases identified 20 causal relationships classified as robust evidence (Fig. 7). Specifically, in the domain of neurological disorders, genetically predicted Parkinson's disease (PD) was found to be causally linked with the elevated occurrence of IS (Grade 1) [97]. Moreover, credible MR findings indicated that a greater genetic predisposition to primary aldosteronism (PA) had a detrimental causal influence on the risk of IS and SVS (Grade 1) [136]. Regarding autoimmune and inflammatory diseases, ankylosing spondylitis [122], systemic lupus erythematosus [49], and psoriasis [50] were identified to causally increase the risk of IS and/or its various subtypes, underscoring the potential critical role of chronic systemic inflammation in IS pathogenesis. As for the non-cardiometabolic biomarkers, two-sample MR studies evidently indicated that genetically determined forced vital capacity (FVC) was inversely related to SVS incidence [16], whereas higher levels of gamma-glutamyl transferase and estimated glomerular filtration rate were deemed causal determinants for CES and LAS development, respectively [81, 95].

Fig. 7.

Robust evidence for the causal effects of non-cardiometabolic diseases and associated indices on ischemic stroke and its subtypes

Drug targets

As displayed in Fig. 8, after systematically evaluating 33 MR analyses in main analysis, we identified 25 robust causal associations that shed light on potential pharmacological targets for the IS and its subtypes (Grade 2). As for the MR analyses investigating the causal effects of existing drug targets, a previous study leveraging genetic variants located within the genes encoding the lipid-lowering drugs unveiled that the reduced LDL-C concentration mediated by the 3-hydroxy-3-methylglutaryl-coenzyme A reductase (HMGCR) gene and Niemann-Pick C1-like 1 (NPC1L1) gene decreased the risk of SVS [170], which supported a possible protective role of HMGCR and NPC1L1 inhibitors against SVS. Similarly, Chen et al. observed strong evidence for the inverse causal relationship between the genetic prediction of sodium-glucose cotransporter protein 2 inhibitors, a novel class of orally administered antidiabetic agents, and the susceptibility to IS and CES [35]. In addition, robust MR findings were able to identify a wide range of undiscovered drug targets based on large-scale genome-wide datasets. For instance, in the MR analysis evaluating the causal impacts of 653 plasma proteins on the incidence of IS as well as its three major subtypes, several circulating biomarkers were recognized as robust candidate therapeutic targets. In detail, histo-blood group ABO system transferase (ABO) was found to be the potential target for LAS and CES, whereas scavenger receptor class A5 (SCARA5) and tumor necrosis factor-like weak inducer of apoptosis (TNFSF12) were suggested to be associated with lower occurrence of CES [36].

Fig. 8.

Robust evidence for the causal effects of drug targets on ischemic stroke and its subtypes

Discussion

Our field synopsis and systematic review summarized and assessed the robustness of the evidence for 1,199 causal estimates between a series of risk factors and the susceptibility to IS as well as its subtypes. Based on the rigorous evaluation framework, a total of 89 causal associations fulfilled the standard of robust evidence, including 8 classified as Grade 1 and 81 as Grade 2 (Additional file 4). In addition, 226 (Grade 3) and 884 (Grade 4) relationships were considered to have insufficient reliability (Additional file 5 and 6).

With respect to anthropometric traits, our field synopsis identified robust evidence that genetically determined adult height was causally associated with declined SVS risk, but elevated CES incidence [110]. Specifically, A large-scale observational study involving 33,197 participants found an inverse correlation between height and blood pressure [220]. Given that blood pressure is a recognized risk factor for SVS [221], it is plausible that the causality between adult height and reduced SVS risk may be partially mediated through blood pressure. On the contrary, the positive causal relationship between adult height and CES risk may be attributable to underlying structural cardiac changes. It was reported that taller people tended to have a larger left atrial diameter, which might increase the susceptibility to atrial fibrillation, thereby elevating the occurrence of cardioembolic stroke [222]. Additionally, robust evidence also supported the causal role of increased BMI and WHR in IS occurrence [92]. This finding was strongly supported by a prospective observational study involving 13,549 adults aged 45–64 years followed over a median of 16.9 years, which reported that higher BMI and WHR were linearly associated with increased risk of IS [223].

Regarding various lifestyle factors, the robust causal link between genetically determined short sleep duration and LAS occurrence may be explained by the following mechanism: chronic sleep loss could contribute to increased secretion of cortisol and specific pro-inflammatory factors such as TNF-α and interleukin-6, which in turn result in dyslipidemia, elevated arterial blood pressure and atherosclerosis, and ultimately lead to the development of IS [224–226]. As for smoking, credible evidence from MR analyses demonstrated a causal effect of higher lifetime smoking index on the risk of IS [92], which was aligned with the prospective observational studies conducted in substantial sample sizes [227, 228]. Furthermore, in mice undergoing transient middle cerebral artery occlusion, Li et al. found that cigarette smoke exposure could trigger the mobilization of peripheral monocytes and neutrophils, which may contribute to the exacerbation of neurological deficits and brain infarction [229].

For dietary factors, the recent MR analysis with robust evidence identified that higher tea consumption protected against SVS occurrence [13]. This association might be attributed to the antioxidative and anti-inflammatory properties of its components, such as caffeine, polyphenols, and flavonoids [230], which have been shown to mitigate ischemia/reperfusion injury and ameliorate endothelial dysfunction [231–234]. Moreover, robust MR findings suggested that genetically determined higher vitamin K1 levels could causally increase LAS risk [90], potentially due to the role of vitamin K1 in promoting atherogenesis, vascular calcification, and a prothrombotic state via activation of coagulation factors II, VII, IX, and X [235].

As for cardiometabolic diseases and associated indices, previous large-scale observational studies have proved that several well-known risk factors, including DBP [236, 237], type 1 and type 2 diabetes [238, 239], circulating levels of LDL-C [240, 241], and lipoprotein(a) [242] were closely related to the risk of IS and its subtypes. Moreover, a recent GWAS involving 42,393 Korean participants identified shared genetic susceptibility loci between diabetes and dyslipidemia and IS, suggesting a common genetic architecture underlying these diseases [243]. More importantly, compelling MR findings further supported the potential causal roles of these cardiometabolic disorders and related traits in the pathogenesis of IS [71, 92, 196]. Additionally, MR studies yielded valuable insights for evaluating the comparative effects of closely associated risk factors, for instance, multivariable MR analysis suggested that compared to other glycemic traits, 2 h glucose after an oral glucose challenge underscored the pivotal role in the development of IS and SVS [185]. Similarly, the MR framework proposed the independent causal impact of type 1 diabetes on SVS incidence, even after adjusting for various common cerebrovascular disease risk factors such as BMI, smoking, LDL, and triglycerides [196].

With regard to non-cardiometabolic diseases, two-sample MR analysis revealed a robust causal association between primary aldosteronism and the occurrence of IS and SVS [136], corroborated by a case–control study reporting a stroke incidence rate of 35 cases per 1000 person-years in patients with PA [244], nearly ten times higher than that observed in the general population [245]. Besides, in male mice, aldosterone was found to enhance superoxide production in cerebral arteries and to upregulate mRNA expression levels of pro-inflammatory cytokines in the brain, potentially impairing endothelial function and contributing to small vessel injury [246, 247]. Moreover, credible findings from MR studies supported the notion that PD was a causal risk factor for the development of IS [97], which may plausibly be attributed to the pathological accumulation of α-synuclein within the brain. In vivo experiments illustrated that transgenic mice with α-synuclein overexpression exhibited heightened susceptibility to ischemic brain injury compared to controls [248], while significant shrinkage in infarction volume was detected following knockdown of α-synuclein in rodents [249].

MR analyses were also applied to provide powerful support in the identification of drug targets. In our systematic review, compelling MR evidence from Xie et al. demonstrated that genetically proxied inhibition of NPC1L1 and HMGCR—the molecular targets of ezetimibe and statins, respectively—was causally associated with a lower risk of SVS [170]. Consistent with these genetic findings, a retrospective cohort study involving 11,009 participants showed that statin therapy targeting HMGCR could decrease IS risk by 24% [250]. Furthermore, a large-scale, double-blind, randomized trial found that combining ezetimibe with simvastatin resulted in a greater reduction in IS incidence compared with statin monotherapy [251, 252]. Moreover, robust MR results unveiled the protective role of SCARA5 for CES. A plausible explanation was the involvement of SCARA5 in iron homeostasis through ferritin-bound iron transport to organs such as the brain and heart [253], thus potentially reducing oxidative stress and attenuating atrial remodeling [254–256].

As the first field synopsis and systematic review focusing on MR associations with IS susceptibility, our study presented several major strengths. Firstly, we summarized the most updated and extensive evidence of MR studies so far to elucidate the causality of various exposures with IS as well as its subtypes. Additionally, for the assessment of evidence robustness, the present rigorous grading criteria we applied were on the basis of primary and complementary MR methods, as well as various sensitivity analyses, which enabled our results to be verified transparently and repetitively. However, the limitations of our study should also be acknowledged. First of all, although our systematic review covered a comprehensive set of risk and protective factors for IS, including anthropometric traits, socio-economic and lifestyle factors, dietary factors, various diseases and associated traits, as well as drug targets, other MR analyses investigating exposures beyond the predefined exposure range, such as brain imaging-derived phenotypes, were not included and evaluated. Secondly, details such as power calculations or the proportion of variance explained by selected genetic variants were often missing in most of the MR studies assessed in our review, making it challenging to distinguish whether non-significant findings truly reflected a lack of causal relationship or were simply due to insufficient statistical power. Thus, the introduction of MR reporting guidelines, such as the STROBE-MR checklist, is expected to improve the quality and transparency of future MR studies, thereby enhancing the robustness and interpretability of MR results. Lastly, since our genetic data were primarily based on individuals of European ancestry, the mentioned viewpoints revealing causal risk factors for IS and its subtypes should be applied to other ethnic groups with caution.

Conclusion

In summary, our systematic review comprehensively assessed the causal relationships of genetically predicted exposures, including anthropometric traits, socio-economic and lifestyle factors, dietary factors, various diseases and associated indices as well as drug targets, with the risk of IS and its major subtypes. After employing strict criteria, a total of 89 causal associations were identified robust evidence, among which the most compelling MR results included causal effects of genetically determined DBP, T2D, PD, and PA on overall IS risk, as well as correlations between vitamin K1 concentrations, PA, and adult height and occurrence of specific IS subtypes. We hope that the reliable MR findings obtained through robustness assessment in this systematic review could offer valuable insights into the etiology of ischemic stroke, and support clinicians in implementing more targeted prevention and management strategies for IS and its subtypes.

Abbreviations

GBD

Global burden of disease

IS

Ischemic stroke

LAS

Large artery stroke

SVS

Small vessel stroke

CES

Cardioembolic stroke

MR

Mendelian randomization

SNP

Single nucleotide polymorphism

IV

Instrumental variables

GWAS

Genome-wide association studies

IVW

Inverse-variance weighted

MR-RAPS

Mendelian randomization robust adjusted profile score

MR PRESSO

MR pleiotropy residual sum and outlier

WHR

Waist-to-hip ratio

BMI

Body mass index

SD

Standard deviation

RCTs

Randomized controlled trials

DBP

Diastolic blood pressure

LDL-C

Low-density lipoprotein cholesterol

T2D

Type 2 diabetes

PD

Parkinson's disease

PA

Primary aldosteronism

FVC

Forced vital capacity

HMGCR

3-Hydroxy-3-methylglutaryl-coenzyme A reductase

NPC1L1

GeneandNiemann-Pick C1-like 1

ABO

Histo-blood group ABO system transferase

SCARA5

Scavenger receptor class A5

TNFSF12

Tumor necrosis factor-like weak inducer of apoptosis

Author contributions

Conceptualization: Junyi Yang, Chen Han, Yue Zhang, Shutong Tan, and Xu Liu; Data curation: Qian Wu, Yumei Ma, Yuanjie Duan; Methodology: Yaxin Wang, Jinke Wang; Investigation: Binhui Liu, and Changqing Mu; Formal analysis: Junyi Yang, Chen Han, Yue Zhang and Shutong Tan; Project administration: Junyi Yang and Chen Han; Supervision: Jian Zhang and Xu Liu; Writing—original draft: Junyi Yang, Chen Han, Yue Zhang and Shutong Tan; Writing—review & editing: Xiaoqian Zhang, Jian Zhang and Xu Liu; Funding acquisition: Ruixia Zhu, Xiaoqian Zhang, Jian Zhang and Xu Liu. All authors read, critically revised, and approved the final manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (Grant No. 81400950), Natural Science Foundation of Liaoning Province (Grant No. 2019-MS-365, 2023-MSLH-406), Joint Program of the Science and Technology Plan of Liaoning Province (Grant No. 2023JH2/101700076).

Data availability

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

The current study is based on the summary results from the published studies. Ethical approval for each of the included studies can be found in the original publications.

Competing interests

The authors have declared no competing interests.