Aortic stenosis intricacies beyond the valve: A bidirectional Mendelian randomization study

Abstract

Aims

Numerous observational studies have reported an association between calcific aortic valve stenosis (CAVS) and an increased risk of other cardiovascular diseases (CVDs). However, the potential causal association is uncertain.

Methods and results

A bidirectional Mendelian randomization (MR) analysis was performed to confirm the causal association between CAVS and the risk of seven common CVDs. The meta‐analysis was used to estimate the combined causal effect with multiple MR results. The MR analysis indicated that genetically predicted CAVS is associated with the increased risk of heart failure (odds ratio [OR] = 1.064; 95% confidence interval [CI] [1.008, 1.123]; P = 0.025), large‐artery stroke (OR = 1.256; 95% CI [1.112, 1.418]; P < 0.001), cardioembolic stroke (OR = 1.228; 95% CI [1.117, 1.349]; P < 0.001) and atrial fibrillation (OR = 1.100; 95% CI [1.061, 1.140]; P < 0.001) without heterogeneity and pleiotropy. Genetically predicted hypertension (FinnGen: OR = 1.170; 95% CI [1.055, 1.297]; P = 0.003 and UK Biobank: OR = 2.456; 95% CI: 1.571, 3.840; P = 8.16 × 10⁻⁵) increased the risk of CAVS.

Conclusions

This comprehensive MR study revealed the causal relationship between genetically predicted CAVS and the increased risk of various CVDs and the causal relationship between hypertension and CAVS risk. Given the harmful causal impact of CAVS on adverse cardiovascular events, it is crucial to prioritize the prevention and management of CVD in individuals with CAVS. Blood pressure management is an effective preventive intervention.

Introduction

Calcific aortic valve stenosis (CAVS) is the most common and severe valvular heart disease in high‐income countries and a major public health burden. ¹ The prevalence of CAVS ranges from 10% to 15% in individuals over 75 years with projections indicating a likelihood of more than doubling by the year 2040. ² Many individuals diagnosed with CAVS not only experience dysfunction of the aortic valve but also frequently present with concurrent cardiovascular diseases (CVDs). ³ , ⁴ Meanwhile, CAVS shares numerous common risk factors with CVD, such as older age, male sex, dyslipidaemia, adiposity, cigarette smoking and high blood pressure. ⁵ Noteworthy, the mortality rate in even asymptomatic patients with mild and moderate CAVS was higher than that observed in a control population. ⁶ , ⁷ Cardiovascular mortality rate is positively correlated with the CAVS severity, ⁸ and approximately half of the patients succumb to CVD. ⁶ , ⁸ However, these findings are derived from traditional observational studies, which are susceptible to confounding and reverse causation. The presence of shared cardiovascular risk factors may contribute to uncertainty regarding whether CAVS directly causes adverse cardiovascular events.

Mendelian randomization (MR) analysis has been widely used to decipher the underlying causality of a trait on an outcome. ⁹ MR analysis employs randomly allocated genetic alleles as instrumental variables (IVs), which is similar to the random allocation of intervention in randomized clinical trials and can thus avoid confounding and reverse causality bias in nonrandomized observational studies, thereby strengthening the causal inference. ¹⁰

Currently, there is still uncertainty regarding the independent causal effect of CAVS on CVD. The present study aimed to examine the causal relationships between CAVS on CVD using a two‐sample MR. Knowledge in this field can enhance our comprehension of the relationship between CAVS and CVD and provide valuable insights for developing effective intervention strategies to prevent cardiovascular adverse events.

Methods

Study design

Figure 1 presents an overview of the study design. By two‐sample MR analyses, we investigate the causal relationship between genetically determined CAVS on the risk of common CVDs.

Figure 1.

Overview of the study design. CAVS, calcific aortic valve stenosis; CVD, cardiovascular disease; SNP, single‐nucleotide polymorphism.

Firstly, we assessed the causality between CAVS and various CVDs with exposure using a CAVS genome‐wide association study (GWAS) meta‐analysis of 10 cohorts. Afterward, we performed replication MR analyses with exposure using the CAVS GWAS from the FinnGen study. Finally, we combined the results from the above MR analysis. Moreover, we conducted a reverse‐directional MR analysis to explore the causal relationship of CVDs on CAVS.

Data sources

We utilized summary‐level data from publicly available GWASs to obtain data sources for exposures and outcomes. To minimize bias arising from population stratification, the study exclusively employed GWAS datasets comprising individuals of European ancestry. Due to the data in the study from the publicly available GWASs, the necessity for obtaining approval from the ethical committee was exempted.

In discovery stage, we sourced CAVS summary statistics from a GWAS meta‐analysis of 10 cohorts involving 13 765 European ancestry cases and 640 102 controls, including Vanderbilt University Biobank, CAVS‐France1, CAVS‐France2, CAVS‐France3, deCODE, Genetic Epidemiology Research on Adult Health and Aging (GERA), Malmö Diet and Cancer Study, Penn Medicine Biobank, UK Biobank and Umeå University. ¹¹ In the replication stage, CAVS summary statistics are derived from the FinnGen GWAS database (https://r9.finngen.fi/).

Summary statistics data for various CVDs contained coronary heart disease (CHD), myocardial infarction (MI), hypertension, heart failure (HF), conduction disorders, atrial fibrillation (AF), total stroke, large artery stroke (LAS), small‐vessel stroke (SVS) and cardioembolic stroke (CES), which derived from UK Biobank, FinnGen database and other GWAS database. Table 1 lists a detailed description of all the GWAS datasets associated with all exposure and outcomes.

Table 1.

Detailed information on data sources

Phenotype	Consortium	Case/Control	Ancestry	Year	PubMed ID
Exposure
Calcific aortic valve stenosis	10 cohorts	13 765/640 102	European	2023	37 038 246
	FinnGen	9153/368 124	European	2023	NA
Outcome
Coronary heart disease	CARDIoGRAMplusC4D	60 801/123 504	77% European	2015	26 343 387
	FinnGen	43 518/333 759	European	2023	NA
	CARDIoGRAMplusC4D + UK Biobank	122 733/424 528	European	2017	29 212 778
Myocardial infarction	CARDIoGRAMplusC4D	43 676/128 199	77% European	2015	26 343 387
	FinnGen	24 185/313 400	European	2023	NA
	CARDIoGRAMplusC4D + UK Biobank	14 825/44 000	European	2021	33 532 862
Hypertension	FinnGen	111 581/265 626	European	2023	NA
	UK Biobank	119 731/343 202	European	2018	MRC‐IEU
Atrial fibrillation	6 cohorts	60 620/970 216	European	2018	30 061 737
	FinnGen	45 766/191 924	European	2023	NA
Conduction disorders	FinnGen	9949/286 109	European	2023	NA
	UK Biobank	1055/360 139	European	2018	Neale lab
Heart failure	HERMES	47 309/930 014	European	2020	31 919 418
	FinnGen	26 872/349 361	European	2023	NA
	UK Biobank	1088/360 106	European	2018	Neale lab
Total stroke	MEGASTROKE	40 585/406 111	European	2018	29 531 354
	FinnGen	39 818/271 817	European	2023	NA
	UK Biobank	7055/454 825	European	2018	MRC‐IEU
Large artery stroke	MEGASTROKE	4373/406 111	European	2018	29 531 354
Small‐vessel stroke	MEGASTROKE	5386/406 111	European	2018	29 531 354
Cardioembolic stroke	MEGASTROKE	7193/406 111	European	2018	29 531 354

CARDIoGRAMplusC4D, Coronary Artery Disease Genome wide Replication and Meta‐analysis (CARDIoGRAM) plus the Coronary Artery Disease (C4D) Genetics Consortium; HERMES, Heart Failure Molecular Epidemiology for Therapeutic Targets.

Instrumental variable selection

To obtain sufficient single‐nucleotide polymorphisms (SNPs), SNPs were considered with a threshold of P < 5 × 10⁻⁸. The presence of linkage disequilibrium was assessed using the linkage disequilibrium analyses (r ² < 0.001; distance threshold, 10 000 kb). PhenoScanner ¹² was utilized to find used SNPs and assess their association with potential confounders, including adiposity traits, lipids, glucose metabolism‐related traits, blood pressure, psychiatric traits, smoking and drinking habits, immune‐mediated inflammatory diseases and sleep traits. The SNPs strongly associated with the potential confounders (P < 5 × 10⁻⁸) were excluded. To address the issue of weak instrument bias, we used the F‐statistic, and an F‐statistic exceeding 10 was deemed sufficient to establish a robust relationship between the IVs and the exposure. ¹³

Statistical analyses

The MR study employed R software (version 4.2.0) and various R packages including TwoSampleMR, MR‐PRESSO and meta for statistical analysis. Multiple MR approaches were conducted to investigate causality. All methods are predicated on distinct assumptions regarding the efficacy of IVs. The inverse‐variance weighted (IVW) approach was used as the main method in MR analyses, which consistently produced reliable results when all IVs were found to be valid. ¹⁴ We also used MR‐pleiotropy residual sum and outlier (MR‐PRESSO) to observe and correct for outliers. ¹⁵ MR‐Egger ¹⁶ weighted median ¹⁷ and MR‐PRESSO approaches were additionally employed as secondary analysis methods. Finally, we combined the IVW results from at least two MR analyses utilizing meta‐analysis.

To assess the reliability of the findings, we conducted a sequence of sensitivity analyses. The assessment of heterogeneity of the utilized SNPs was conducted through the application of Cochrane's Q test, P values < 0.05 indicated that significant heterogeneity was present. Additionally, the MR‐Egger regression method was utilized to estimate the existence of directional pleiotropy among the IVs by examining its intercept.

Bonferroni correction (corrected P = 0.05/25 = 0.002) was used to address the issue of multiple testing. P values ranging from 0.002 to 0.05 were deemed to indicate potential causality between the exposures and the outcomes, while P values < 0.002 were interpreted as signifying a significant causal relationship.

Results

Characteristics of the IVs

The detailed characteristics of the selected SNPs for predicting CAVS are provided in Tables S1 and S2. Each SNP met the predetermined screening criteria outlined in the Methods section. Moreover, all employed SNPs exhibited an F‐statistic >10, which indicates the lack of any bias associated with weak instruments in the assessment of IV strength.

Causal estimates of CAVS on CVDs

The results of the genetically determined associations between CAVS and common CVDs are shown in Table 2 and Figure 2 .

Table 2.

Mendelian randomization estimates of the causal associations between CAVS and cardiovascular diseases

Exposure	Outcome	OR (95% CI)	P IVW	P heterogeneity	P pleiotropy
CAVS from 10 cohorts (discovery stage)
	Heart failure (HERMES)	1.109 (1.039, 1.184)	0.002	0.109	0.500
	Heart failure (FinnGen)	1.115 (1.029, 1.209)	0.008	0.533	0.400
	Hypertension (FinnGen)	1.023 (0.971, 1.078)	0.385	0.597	0.094
	Coronary heart disease (CARDIoGRAMplusC4D)	1.062 (0.982, 1.149)	0.131	0.484	0.282
	Coronary heart disease (FinnGen)	1.146 (1.070, 1.227)	<0.001	0.079	0.688
	Pooled	1.109 (1.053, 1.168)	<0.001
	Myocardial infarction (CARDIoGRAMplusC4D)	0.994 (0.911, 1.085)	0.894	0.342	0.091
	Myocardial infarction (FinnGen)	1.161 (1.062, 1.270)	0.001	0.115	0.637
	Pooled	1.074 (0.922, 1.250)	0.358
	Total stroke (MEGASTROKE)	1.073 (0.993, 1.159)	0.075	0.181	0.764
	Total stroke (FinnGen)	1.055 (0.982, 1.133)	0.142	0.301	0.623
	Large artery stroke (MEGASTROKE)	1.346 (1.094, 1.657)	0.005	0.095	0.844
	Small vessel stroke (MEGASTROKE)	1.142 (0.940, 1.387)	0.182	0.058	0.767
	Cardioembolic stroke (MEGASTROKE)	1.310 (1.114, 1.541)	0.001	0.504	0.553
	Atrial fibrillation (FinnGen)	1.109 (1.020, 1.205)	0.015	0.066	0.076
	Conduction disorders (FinnGen)	1.151 (1.015, 1.306)	0.028	0.715	0.659
CAVS from FinnGen (replication stage)
	Heart failure (HERMES)	1.071 (1.022, 1.123)	0.004	0.416	0.790
	Heart failure (UK Biobank)	1.001 (1.000, 1.001)	0.198	0.398	0.579
	Pooled	1.031 (0.965, 1.101)	0.349
	Hypertension (UK Biobank)	1.003 (0.994, 1.012)	0.535	0.001	0.466
	Coronary heart disease (CARDIoGRAMplusC4D + UK Biobank)	1.007 (0.965, 1.050)	0.762	0.057	0.927
	Myocardial infarction (CARDIoGRAMplusC4D + UK Biobank)	1.009 (0.964, 1.056)	0.713	0.117	0.670
	Total stroke (MEGASTROKE)	1.080 (1.021, 1.142)	0.007	0.301	0.441
	Total stroke (UK Biobank)	1.003 (1.001, 1.005)	<0.001	0.409	0.120
	Large artery stroke (MEGASTROKE)	1.211 (1.042, 1.407)	0.012	0.625	0.762
	Small vessel stroke (MEGASTROKE)	1.002 (0.871, 1.153)	0.976	0.878	0.977
	Cardioembolic stroke (MEGASTROKE)	1.187 (1.057, 1.334)	0.004	0.418	0.757
	Atrial fibrillation (6 cohorts)	1.098 (1.055, 1.143)	<0.001	0.294	0.827
	Conduction disorders (UK Biobank)	1.001 (1.000, 1.001)	0.100	0.713	0.990

CARDIoGRAMplusC4D, Coronary Artery Disease Genome wide Replication and Meta‐analysis (CARDIoGRAM) plus the Coronary Artery Disease (C4D) Genetics consortium; CAVS; calcific aortic valve stenosis; CI, confidence interval; HERMES, Heart Failure Molecular Epidemiology for Therapeutic Targets; IVW, inverse variance weighted; OR, odds ratio.

Figure 2.

Mendelian randomization estimates of the causal associations between CAVS and cardiovascular diseases using the inverse‐variance weighted method. CAVS, calcific aortic valve stenosis; CI, confidence interval; OR, odds ratio; SNP, single‐nucleotide polymorphisms.

In the discovery stage, the results of IVW analyses indicated that genetically determined CAVS (10 cohorts) was associated with an increased risk of HF (HERMES: OR = 1.109; 95% CI [1.039, 1.184]; P = 0.002 and FinnGen: OR = 1.115; 95% CI [1.029, 1.209]; P = 0.008), CHD (meta: OR = 1.109; 95% CI [1.053, 1.168]; P < 0.001), LAS (OR = 1.346; 95% CI [1.094, 1.657]; P = 0.005), CES (OR = 1.310; 95% CI [1.114, 1.541]; P = 0.001), AF (OR = 1.109; 95% CI [1.020, 1.205]; P = 0.015), conduction disorders (OR = 1.151; 95% CI [1.015, 1.306]; P = 0.028), while little evidence suggested that genetically determined CAVS (10 cohorts) was associated with hypertension, MI, total stroke and SVS (Table 2 and Table S3).

In the replication stage, the CAVS GWAS from the FinnGen database was used for the replication study. The replication results of IVW analysis indicated that genetically determined CAVS (FinnGen) was associated with an increased risk of total stroke (MEGASTROKE: OR = 1.080; 95% CI [1.021, 1.142]; P = 0.007 and UK Biobank: OR = 1.003; 95% CI [1.001, 1.005]; P < 0.001), LAS (OR = 1.211; 95% CI [1.042, 1.407]; P = 0.012), CES (OR = 1.187; 95% CI [1.057, 1.334]; P = 0.004), AF (OR = 1.098; 95% CI [1.055, 1.143]; P < 0.001), while there is insufficient evidence suggesting that genetically determined CAVS (FinnGen) was associated with HF, hypertension, CHD, MI, total stroke, SVS and conduction disorders (Table 2 and Table S4).

Combined all results of CVD risk from the meta‐analysis

After combining multiple MR results from different data sources by meta‐analysis, we found a suggestive causality of CAVS on HF (OR = 1.064; 95% CI [1.008, 1.123]; P = 0.025), and significant causality of CAVS on LAS (OR = 1.256; 95% CI [1.112, 1.418]; P < 0.001), CES (OR = 1.228; 95% CI [1.117, 1.349]; P < 0.001), and AF (OR = 1.100; 95% CI [1.061, 1.140]; P < 0.001), while little evidence suggested that genetically determined CAVS was associated with other CVDs after meta‐analysis (Figure  2 ).

Causal estimates of CVDs on CAVS

Afterward, we further assessed the causality of several CVDs on CAVS by reverse‐direction MR analysis. We observed a causal relationship between genetically determined hypertension (FinnGen: OR = 1.170; 95% CI [1.055, 1.297]; P = 0.003 and UK Biobank: OR = 2.456; 95% CI: 1.571, 3.840; P = 8.16 × 10⁻⁵) and an increased risk of CAVS (Table  3 ), while there is insufficient evidence suggesting that genetically determined HF, CHD, total stroke and AF was associated with CAVS.

Table 3.

Mendelian randomization estimates of the causal associations between cardiovascular diseases and CAVS

Exposure	Outcome	OR (95% CI)	P IVW	P heterogeneity	P pleiotropy
Heart failure (HERMES)	CAVS (10 cohorts)	0.722 (0.394, 1.322)	0.291	0.033	0.630
Heart failure (HERMES)	CAVS (FinnGen)	0.863 (0.535, 1.391)	0.545	0.005	0.067
Hypertension (FinnGen)	CAVS (10 cohorts)	1.170 (1.055, 1.297)	0.003	6.31 × 10 −10	0.275
Hypertension (UK Biobank)	CAVS (FinnGen)	2.456 (1.571, 3.840)	8.16 × 10 −5	1.96 × 10 −17	0.296
Coronary heart disease (CARDIoGRAMplusC4D)	CAVS (10 cohorts)	1.065 (0.866, 1.309)	0.550	1.57 × 10 −04	0.175
Coronary heart disease (CARDIoGRAMplusC4D)	CAVS (FinnGen)	0.976 (0.818, 1.166)	0.790	9.30 × 10 −05	0.142
Total stroke (MEGASTROKE)	CAVS (10 cohorts)	0.894 (0.679, 1.176)	0.422	0.849	0.672
Total stroke (MEGASTROKE)	CAVS (FinnGen)	0.943 (0.762, 1.168)	0.592	0.259	0.215
Atrial fibrillation (6 cohorts)	CAVS (10 cohorts)	0.999 (0.924, 1.079)	0.974	3.64 × 10 −09	0.581
Atrial fibrillation (6 cohorts)	CAVS (FinnGen)	0.991 (0.927, 1.059)	0.780	6.20 × 10 −09	0.549

CAVS, calcific aortic valve stenosis; CI, confidence interval; OR, odds ratio; SNP, single‐nucleotide polymorphisms.

Sensitivity analyses

In the whole MR analysis, used IVs revealed no pleiotropy (Table 2 and Table S5). part MR analyses revealed heterogeneity, and we conducted the random‐effect IVW method in these MR analyses, while other analyses used the fixed‐effects IVW method (Table 2 and Table S5). The majority of statistical models exhibited consistent directional estimates with the IVW analysis (Tables S3–S5).

Discussion

Numerous observational studies have reported an association between CAVS and adverse cardiovascular events. In this study, by comprehensive and systematical MR analysis, we confirmed the causality of CAVS on the increased risk of HF, LAS, CES and AF after meta‐analysis. In addition, we revealed causal correlation between hypertension and the CAVS risk by inverse‐direction MR analysis.

Among symptomatic aortic valve stenosis (AVS) patients, three‐quarters will experience HF, undergo valve replacement, or succumb within 5 years. ¹⁸ In a recent systematic review involving 12 143 moderate AVS patients with a median follow‐up duration of nearly 4 years, AVS was found to elevate the risk of HF. ¹⁹ Even asymptomatic patients with mild and moderate AVS exhibited a higher cardiovascular mortality rate, primarily associated with HF, compared with the control group. ⁶ , ⁷ , ⁸ Our results of MR analysis revealed the causal relationship of CAVS on HF. Traditionally, HF, in individuals with severe CAVS, serves as an indicator of disease progression. In the initial stages, the heightened afterload resulting from aortic valve constriction triggers adaptive left ventricular hypertrophy, serving to uphold wall stress and maintain cardiac output. Ultimately, this process undergoes decompensation, leading patients to transition from hypertrophy to HF, accompanied by the onset of symptoms and adverse cardiovascular events. ²⁰

Among individuals with severe CAVS necessitating intervention, approximately 40% to 75% exhibit coexisting CAD. ³ In addition, aortic sclerosis (ASc) is typically considered the early stage of AVS, ²¹ affecting 25% of individuals over the age of 65. Approximately 2% of individuals with ASc progress to clinical AVS each year. A meta‐analysis comprising 22 studies indicated that ASc is associated with a 68% increased risk of coronary events. ²² Nevertheless, our MR analysis did not find the causality of CAVS on CAD and MI after meta‐analysis. CAVS is highly prevalent in the aging population and shares many common risk factors with CAD. ⁵ Previous findings of observational studies may be influenced by the shared cardiovascular risk factors.

The Framingham Heart Study has indicated that severe AVS is significantly associated with AF risk. ²³ In addition, a prospective study reported that approximately one‐third of patients without AF undergoing transcatheter aortic valve implantation developed new‐onset AF. ²⁴ Our finding further confirmed the causality of CAVS on AF. On the one hand, AVS can elevate the left end‐diastolic filling pressure, leading to left atrial dilatation, a factor that has been associated with AF risk. ²⁵ , ²⁶ On the other hand, alleles associated with an increased risk of AVS are connected to heightened expression of IL6 and IL6‐AS1 in fibroblasts. ²⁷ By promoting fibrosis, fibroblasts can accelerate the development of AF. Moreover, by producing IL‐6 and proinflammatory cytokines, macrophages regulate fibrosis, thereby promoting the development of AF. ²⁸ Notably, among patients undergoing transcatheter aortic valve replacement, the emergence of new‐onset AF is linked to a higher risk of stroke and HF when compared with pre‐existing AF or the absence of AF. ²⁹ Based on our findings, the AF caused by CAVS may play a partial effect in the development of HF and stroke caused by CAVS.

Aortic valve calcification (AVC) is the main hallmark of AVS. Quantifying AVC in the severe AVS range has proven predictive of a poorer prognosis in patients. ³⁰ , ³¹ Numerous studies detected a statistically significant correlation between AVC and stroke and brain infarcts. ³² , ³³ Moreover, a meta‐analysis containing 22 studies reported ASc increased 27% risk of stroke, ²² and an association between aortic valve disease and overt stroke has been demonstrated specifically in the presence of AVS, ³⁴ which is consistent with our findings of causal analysis. Two potential mechanistic pathways may cause the causal association. On the one hand, AF was the major predictor of stroke in patients with AVS. ³⁵ It is suggested that the stroke outcomes associated with CAVS may be partially attributed to AF. Our study revealed the causal relationship of CAVS on CES and AF, further confirming the viewpoint. On the other hand, a study suggested that, on autopsy, systemic embolism, including involvement of the right middle cerebral artery, was discovered in one‐third of cases with CAVS. ³⁶ A potential physiological mechanism could implicate that turbulent blood flow, induced by AVS, leads to the fragmentation of red cells and the release of adenosine diphosphate and thromboplastin. This process gives rise to non‐calcific embolism and microthrombus formation. ³⁷

Large‐scale cohort studies have suggested that hypertension has independent associations with incident AVS. ³⁸ , ³⁹ Recently, a one‐sample MR study reported that elevated systolic blood pressure is associated with an increased risk of AVS. ⁴⁰ However, the overlap of the population due to using a one‐sample could introduce a bias in estimating MR towards estimates based on observational associations. ⁴¹ We used a two‐sample MR analysis to further confirm the causality of hypertension and CASV after a meta‐analysis of two reliable results. The potential mechanisms may include an increase in the diastolic transvalvular pressure gradient due to hypertension. The heightened pressure gradient is detected by the valve leaflets, leading to remodelling characterized by increased collagen deposition, inflammation and endothelial dysfunction. ⁴² , ⁴³ , ⁴⁴ To date, no drugs have shown compelling efficacy in reversing or halting the progression of the disease. ¹⁸ Blood pressure‐lowering treatment may be an effective strategy in the prevention of CAVS.

Inevitably, our study has some limitations that must be acknowledged. Firstly, all study participants belonged to the European ethnic group and originated from high‐income economy countries. Further studies of diverse ethnic populations are warranted. Secondly, the existence of heterogeneity in the utilized SNPs could introduce potential biases and affect the robustness of our MR findings. Thirdly, our study focused solely on examining the causal relationship between common CVDs with two or more GWAS and CAVS. Further investigation is warranted to explore the associations of CAVS with other diseases.

Conclusions

This MR study provided support for the causal relationship between genetically predicted CAVS and the increased risk of HF, LAS, CES and AF, and revealed positive causal correlations between genetically predicted hypertension and the CAVS risk. Considering the detrimental impact of CAVS on adverse cardiovascular events, it is imperative to prioritize the prevention and management of CVD in individuals afflicted with CAVS. Blood pressure management may be an effective preventive intervention strategy.

Conflict of interest

The authors declare no competing interests.

Funding

This work was supported by the Natural Science Foundation of Ningbo Municipality (2022J254) and Medical and Health Research Project of Zhejiang Province (WKJ‐ZJ‐2446).

Supporting information

Table S1. Characteristics of SNPs used as instrumental variables for CAVS (10 cohorts).

Table S2. Characteristics of SNPs used as instrumental variables for CAVS (FinnGen).

Table S3. MR estimating the associations between CAVS (10 cohorts) and CVDs in discovery stage.

Table S4. MR estimating the associations between CAVS (FinnGen) and CVDs in replication stage.

Table S5. MR estimating the associations between CVDs and CAVS.

Cai, D. , and Liu, J. (2025) Aortic stenosis intricacies beyond the valve: A bidirectional Mendelian randomization study. ESC Heart Failure, 12: 3276–3284. 10.1002/ehf2.15348.

Data availability statement

All the data used in the present study had been publicly available. The original contributions presented in the study are included in the article or the supporting information; further inquiries can be directed to the corresponding author.