Apolipoprotein and peripheral artery disease: Mendelian randomization analysis

Abstract

Background

The prevalence of peripheral arterial disease (PAD) is on the rise globally, leading to adverse clinical outcomes. Our aim was to investigate the causal relationship between apolipoprotein and PAD, as well as the potential mediating role of smoking, diabetes, hypertension, myocardial infarction, and ischemic stroke.

Methods

We employed two-sample Mendelian randomization (TSMR) to assess the causal effect of apoB/A1 on the risk of PAD and potential mediators (smoking, diabetes mellitus, hypertension, myocardial infarction, and ischemic stroke), as well as the causal effect of those mediators on PAD. The use of multivariate MR (MVMR) allowed us to explore and quantify the mediating role of these factors in the causal association between apoB/A1 and the risk of PAD.

Results

Our MR analysis showed that each standard deviation increase in apoB/A1 increased the risk of PAD by 46% (OR = 1.460, 95% CI: 1.255–1.697, P = 8.74E-07). Hypertension, myocardial infarction, and ischemic stroke were ultimately recognized as the mediators of the causal relationship between apoB/A1 and PAD, explaining 7.5%, 19.7%, and 62.5% of the causal effect, respectively, and the proportion combined of the three together was 81%.

Conclusions

These studies demonstrated that elevated apoB/A1 increases the risk of developing PAD and that this association may be mediated by hypertension, myocardial infarction, and ischemic stroke.

Keywords

Peripheral arterial disease apolipoprotein mendelian randomization mediation

Introduction

Peripheral artery disease (PAD), defined as atherosclerotic disease affecting the arteries that supply the lower extremities, impacts over 200 million individuals globally and is linked to a significantly heightened risk of adverse cardiovascular events and mortality.¹ PAD and coronary artery disease (CAD) represent distinct manifestations of atherosclerosis, sharing similar risk factors.² Dyslipidemia is closely associated with the development and advancement of atherosclerosis.³ Specifically, elevated serum total cholesterol (TC) and low-density lipoprotein cholesterol (LDL-C) have been identified as significant factors in the onset and progression of PAD and CAD.⁴ As the complement to traditional lipid markers,⁵ apolipoproteins levels are recognized as crucial predictors of cardiovascular disease risk.⁶

Apolipoprotein B (apoB) is a constituent of low-density lipoprotein (LDL), intermediate-density lipoprotein (IDL), and very low-density lipoprotein (VLDL), while apolipoprotein A1 (apoA1) is the primary apolipoprotein associated with high-density lipoprotein (HDL). The apoB/A1 ratio has been identified as a reliable predictor of coronary artery disease (CAD) in obese patients.⁷ Higher apoB/A1 ratio is not only robustly associated with incident CAD⁸ but also with hypoechoic and unstable plaques.⁹ Similar to CAD, apoB has been established as an independent risk factor for PAD¹⁰ and play a significant role in the pathogenesis and risk stratification.¹¹ In contrast, apoA1 is negatively correlated with PAD.¹² A meta-analysis revealed that individuals with PAD had elevated apoB levels and decreased apoA1 levels, along with increased apoB/A1 ratios, compared to non-PAD controls.⁶

The adverse impact of smoking, diabetes, hypertension, and hyperlipidemia on PAD has been extensively documented.¹³ Among these factors, smoking and diabetes are identified as the primary risk factors for PAD, while hypertension and dyslipidemia have a relatively lesser impact compared to CAD.^15,16 Smoking is known to double the risk of PAD, with the risk escalating proportionally with the number of cigarettes smoked and age at which smoking began, with the highest risk associated with smoking initiation before the age of 16.¹⁷ Diabetes, on the other hand, is strongly linked to the severity of PAD.¹⁸ In the United States, approximately 70% of non-traumatic lower limb amputations occur in diabetic patients, despite diabetes having an overall prevalence of 12%.¹⁹

The significant association between hyperlipidemia and PAD has been extensively validated.¹⁴ In addition, a Mendelian randomization study by Hoek AG et al. demonstrated a causal relationship between genetically predicted type 2 diabetes mellitus, initiation of smoking, heavy smoking, and an increased risk of PAD.⁶ While the genetic link between apoB/A1 levels and PAD remains inconclusive, an elevated apoB/A1 ratio has been found associated with the risk of hypertension,¹² diabetes,¹⁵ myocardial infarction,¹⁶ and ischemic stroke,¹⁷ all of which are also risk factors for PAD.

By utilizing Mendelian randomization, a commonly utilized method that overcomes the limitations of traditional observational research, we can explore causal relationships between exposures and outcomes, thereby addressing gaps in the evidence. Therefore, we aimed to estimate the effect of apoB/A1 ratio on the risk of PAD by two-sample MR (TSMR) and quantify the roles of smoking, diabetes, hypertension, ischemic stroke, and myocardial infarction as potential mediators by multivariable Mendelian randomization (MVMR).

Method

MR study design

The research design depicted in Figure 1 outlines the methodology. Independent single nucleotide polymorphisms (SNPs)¹⁸ of genome-wide association study (GWAS)¹⁹ served as instrumental variables (IVs) to represent exposure (ratio of apolipoprotein B to apolipoprotein A1 levels), mediators, and outcome (peripheral artery disease), without necessitating ethical approval as publicly available data was utilized. Initially, we used two-sample Mendelian randomization (TSMR)²⁰ method to investigate the causal link between Ratio and PAD. Subsequently, we referred to existing literature to identify potential mediators influencing the anticipated causal connection between Ratio and PAD.²¹ Finally, we assessed the mediating roles of specific factors utilizing two-step MVMR (multivariable Mendelian randomization)²² and figured out the corresponding proportion of the mediating effect.

Figure 1.

The overall MR study design. (a) The total effect of Ratio (ratio of apolipoprotein B to apolipoprotein A1 levels) on PAD (peripheral artery disease), denoted as c, was derived from the result of TSMR (two-sample MR). (b) The total effect was decomposed into: (i) indirect effect using a two-step approach (where a represents the total impact of Ratio on the mediator, and b represents the effect of the mediator on PAD while adjusting for Ratio) and the product method (a × b) and (ii) direct effect (c′ = c – a × b). As for the mediation by 2 or 3 mediators combined, the indirect effect was calculated by the difference method (c – c′). Proportion mediated was the indirect effect divided by the total effect.

Data source

We gathered the summarized data from the genome-wide association studies (GWAS) for each trait, which is presented in detail in Table 1. The ratio of apolipoprotein B to apolipoprotein A1 levels has been suggested to predict vascular disease risk due to the strong link between apolipoprotein and serum lipid levels.²³ One meta-analysis indicated that a higher Ratio of apolipoprotein B to apolipoprotein A1 levels was found in PAD patients.²⁴ The genetic instruments for apolipoprotein were selected from a large-scale molecular profiling and genotyping in 115,082 European ancestry individuals²¹ (ID: ebi-a-GCST90092810).

Table 1.

Data sources used in Mendelian randomization analyses.

	Trait	Sample size	Population	Reference	GWAS-ID	SNPs	Year
Exposure	Ratio of apolipoprotein B to apolipoprotein A1 levels	115,082	European	Richardson TG et al.	ebi-a-GCST90092810	11,590,399	2022
Outcome	Peripheral artery disease	213,639	European	NA	finn-b-I9_PAD	16,380,453	2021
Mediators	Body Mass Index	461,460	European	NA	ukb-b-19953	9,851,867	2018
	Pack years of smoking	142,387	European	Ben Elsworth et al.	ukb-b-10831	9,851,867	2018
	Type 2 diabetes	490,089	European	Sakaue S et al.	ebi-a-GCST90018926	24,167,560	2021
	Hypertension	484,598	European	D<U+00F6>nerta<U+015F> HM	ebi-a-GCST90038604	9,587,836	2021
	Ischemic stroke	440,328	European	Malik R et al.	ebi-a-GCST005843	7,537,579	2018
	Myocardial infarction	200,641	European	NA	finn-b-I9_MI	16,380,433	2021

The outcome of our study was peripheral artery disease (PAD), which is the third most common cause of atherosclerotic vascular morbidity, following coronary heart disease and stroke.²³ We extract SNPs related to PAD from GWAS databases (ID: finn-b-I9_PAD), which comprises 7098 cases and 206,541 controls of European ancestry.²⁵

The mediators in the relationship between Ratio and PAD were chosen based on the following inclusion criteria²⁶: (1) The exposure and mediating factors have a causal relationship; (2) There is a causality between the mediating factors and the outcome, regardless of whether the exposure factors are adjusted for or not. On the basis of our literature review, we ultimately identified 6 candidate mediators in line with the criteria to evaluate the role of mediation between Ratio and PAD, including BMI (ukb-b-19953), pack years of smoking (ukb-b-10831),²⁷ type 2 diabetes (ebi-a-GCST90018926),²⁸ hypertension (ebi-a-GCST90038604),²⁹ ischemic stroke (ebi-a-GCST005843),³⁰ and myocardial infarction (finn-b-I9_MI).³¹

Selection of instrumental variables

The fundamental requirements for genetic variation to fulfill the instrumental variable assumptions in this study are as follows: (1) The genetic variant must be associated with the exposure; (2) the genetic variant must not be associated with any confounders of the exposure-outcome association; and (3) the genetic variant should not affect the outcome, except possibly through its association with the exposure.³² The authors guarantee that all the data supporting the findings are fully accessible without restrictions through consortium websites or can be obtained from consortia upon request.

To ensure the reliability of our study, IVs were carefully selected based on stringent criteria. Specifically, SNPs were considered as valid IVs if they showed a significant genome-wide association with the exposure (at P < 5E-08). Besides, SNPs with a minor allele frequency (MAF) in the outcome greater than 0.01 and a linkage disequilibrium (LD) r² of less than 0.001 within a 10,000 kb distance were chosen as IVs.

To evaluate whether these IVs were strong enough to explain the causal relationships, we computed F-statistic to evaluate the strength of the IVs. IVs with an F-statistic <10 were regarded as weak instruments,³³ which may lead to biased results. F-statistic was computed with the formula: F = R² × (N−2)/(1−R²), where R² (the coefficient of determination) represents the variance of the phenotype explained by each genetic variant in exposure. R² was calculated with the formula: R² = 2 × β² × EAF × (1−EAF)/[2 × β² × EAF × (1−EAF) + 2 × (SE)² × N × EAF × (1−EAF)],³⁴ where β represents the effect size of the SNP, EAF signifies the effect allele frequency, SE denotes the standard error of the effect size, and N denotes the total number of individuals in the GWAS study.

Statistical analysis

This study utilized a two-step Mendelian randomization (MR) approach to specifically explore the mediating effects of current smoking status, type 2 diabetes, hypertension, ischemic stroke, and myocardial infarction on the relationship between Ratio (Ratio of apolipoprotein B to apolipoprotein A1 levels) and peripheral artery disease (PAD). In the first step, we used TSMR to evaluate the causal link among exposure (Ratio), outcome (PAD), and mediators by 5 different MR methods. The inverse-variance weighting (IVW) method with random effects was our primary MR analysis, while MR Egger, weighted median, simple mode, and weighted mode methods were used as supplementary analyses.³⁵ In the second step, the MVMR method was utilized to investigate and quantify the mediating effects of mediators.³⁶ The indirect effects were estimated through the coefficient-product method and the proportion of mediation (PM) was calculated by dividing the indirect effect (multiplying the Ratio-mediators effect with the Ratio-adjusted effect of the mediator on PAD) by the total effect (Ratio-PAD effect). In addition, we utilized the difference method to calculate the direct effect of Ratio on PAD and determine the combined proportion of mediation, by simultaneously adjusting the genetic effect for multiple mediators (such as hypertension + ischemic stroke + myocardial infarction). Then, the combined indirect effect of considered mediators was derived as the residual of the total effect. We systematically investigated all possible meaningful combinations of mediators to identify the combination with the most substantial proportion mediated.

We deemed associations to be statistically significant (i.e., the causal effect of exposure on the outcome is significant) if the P-values in the IVW fell below 0.05. Effect estimates were presented as odds ratios (ORs) with corresponding 95% confidence intervals (CIs).

To ensure the accuracy and reliability of the our study, various sensitivity analyses were conducted. We evaluated the heterogeneity using the Cochran’s Q statistic, computed through the Mr Egger and IVW methods.^37,38 A P-value exceeding 0.05 was indicative of no heterogeneity. In cases where heterogeneity was present, a random-effects IVW method was utilized to determine the causal link. If not, a fixed-effects model was applied. The potential pleiotropy of IVs was checked using the intercept p-value from the MR Egger regression, with P > .05 signifying no potential pleiotropy.³⁹ The R software (version 4.4.0) along with the R packages “TwoSampleMR” and “MVMR” were utilized to perform all statistical analyses in the MR analysis.

Result

Genetic instruments

After rigorous exclusion of SNPs failing to meet stringent quality control criteria (P < 5E-08, R² < 0.001, F >10), along with the removal of palindromic and incompatible SNPs, we utilized the remaining SNPs for MR analysis. Comprehensive details regarding SNPs and their relationships with Ratio, mediators, and PAD are available in the Supplementary Material.

Overall causal effect of Ratio on PAD and mediators

According to the primary IVW method utilized, each SD increase in Ratio was associated with a 46% higher risk of PAD (OR = 1.460, 95% CI = 1.255–1.697, P = 8.74E-07; Figure 2). Other models also demonstrated the same significant association with the IVW method (Table 2).

Figure 2.

The results of two-sample MR analysis and MVMR. Forest plot displays the results of univariable and multivariable MR analyses. MR, Mendelian randomization; MVMR, multivariable MR analyses; CI, confidence interval; IVW, inverse-variance weighted; Ratio, the ratio of apolipoprotein B to apolipoprotein A1 levels.

Table 2.

Two-sample MR analysis of effect of Ratio of apolipoprotein B to apolipoprotein A1 levels on PAD.

Exposure	Outcome	Methods	N of SNPs	OR (95%CI)	P
Ratio	PAD	MR Egger	66	1.519 (1.169, 1.973)	2.63E-03
		Weighted median	66	1.593 (1.339, 1.895)	1.46E-07
		IVW	66	1.460 (1.255, 1.697)	8.74E-07
		Simple mode	66	1.799 (1.233, 2.623)	3.31E-03
		Weighted mode	66	1.677 (1.398, 2.013)	5.44E-07

Ratio, ratio of apolipoprotein B to apolipoprotein A1 levels; PAD, peripheral artery disease; MR, Mendelian randomization; SNPs, single nucleotide polymorphisms; IVW, inverse-variance weighted; OR, odds ratio; CI, confidence interval.

Boldness indicates statistical significance.

The effect of Ratio on each mediator is shown in Table S1. A causal link was established between Ratio and elevated risks of hypertension (OR = 1.020, 95% CI = 1.010–1.031, P = 2.32E-04), ischemic stroke (OR = 1.138, 95% CI = 1.058–1.223, P = .000), and myocardial infarction (OR = 1.873, 95% CI = 1.618–2.168, P = 4.66E-17), although we failed to find the relationship between Ratio on BMI (P = .112), pack years of smoking (P = .656), and type 2 diabetes (P = .855).

Overall causal effects of mediators on PAD

Utilizing the IVW method, it was found that BMI (OR = 1.611, 95% CI = 1.420–1.829, P = 1.40E-13, as is shown in Table S2), pack years of smoking (OR = 2.879, 95% CI = 1.882–4.405, P = 1.09E-06), type 2 diabetes (OR = 1.321, 95% CI = 1.251–1.396, P = 2.72E-23), hypertension (OR = 3.272, 95% CI = 2.223–4.818, P = 1.87E-09), ischemic stroke (OR = 1.744, 95% CI = 1.305–2.332, P = 1.74E-04), and myocardial infarction (OR = 1.448, 95% CI = 1.229–1.705, P = 9.28E-06) were all causally associated with an increased risk of PAD.

Effects of mediators on PAD controlling for Ratio

TSMR analysis mentioned above identified that Ratio had a causal effect on hypertension, ischemic stroke, and myocardial infarction, and these mediators also demonstrated causal effects on PAD. Consequently, we ultimately chose these 3 factors to serve as mediators in the connection between Ratio and PAD, excluding current smoking status and type 2 diabetes. After adjusting for Ratio, MVMR revealed that hypertension was associated with almost 4 times the risk of PAD (OR = 3.989, 95% CI = 2.648–6.011, P = 3.71E-11, Figure 2) and the proportion mediated (PM) of hypertension on the causal link between Ratio and PAD was estimated at 7.5% (as is shown in Figure 3). Moreover, after adjusting for Ratio, ischemic stroke was linked with 78% increase in PAD risk (OR = 1.780, 95% CI = 1.405–2.255, P = 1.82E-06, Figure 2), suggesting ischemic stroke mediated 19.7% of the causal effect between Ratio and PAD (as is shown in Figure 3). Furthermore, after adjusting for Ratio, myocardial infarction was associated with 45.8% increase in PAD risk (OR = 1.458, 95% CI = 1.298–1.638, P = 2.08E-10, Figure 2), revealing myocardial infarction mediated surprisingly 62.5% of the causal effect between Ratio and PAD (as is shown in Figure 3).

Figure 3.

Estimate of the effect of Ratio on PAD explained by mediators. Forest plot displays the proportion mediated by each mediator or mediators combined. PM, proportion mediated; CI, confidence interval; All, hypertension + ischemic stroke + myocardial infarction; Ratio, the ratio of apolipoprotein B to apolipoprotein A1 levels; PAD, peripheral artery disease.

Combined proportion mediated

We figured out the proportion mediated by various combinations of mediating variables to identify the combination that best explained the mediation in the Ratio–PAD association (as is shown in Figure 3). Among the two-mediator combinations, the combination of hypertension + ischemic stroke mediated 22.5% of the total effect (as is shown in Figure 3), while the hypertension + myocardial infarction combination showed the highest proportion mediated (82.3% as is shown in Figure 3). It’s worth noting that when adjusting for both ischemic stroke and myocardial infarction, the effect of Ratio on PAD was not significant (P = .221, as is shown in Figure 2). When all the mediators are considered, we didn’t observe a significant increase in the proportion mediated compared to the combination of hypertension + myocardial infarction.

Sensitivity analysis

We observed the presence of potential heterogeneity from genetic instruments among the included SNPs through the utilization of Cochran’s Q statistic within the IVW and MR Egger models (as is shown in Table 3), indicating potential pleiotropy of SNP effects, although it was acceptable when using the random-effects IVW method. The result of MR Egger regression analysis indicated that no apparent evidence of overall directional pleiotropy in our results, further confirming the reliability of the causal links (as is shown in Table 4).

Table 3.

MR heterogeneity test.

Exposure	Outcome	N of SNPs	Methods	Q statistic	Q_df	Q P-value
Ratio	PAD	66	MR Egger	138.0629	64	2.32E-07
Ratio	PAD	66	IVW	138.3507	65	3.20E-07
Ratio	Pack years of smoking	67	MR Egger	111.8882	65	2.71E-04
Ratio	Pack years of smoking	67	IVW	119.0192	66	6.94E-05
Pack years of smoking	PAD	10	MR Egger	11.2759	8	0.1865
Pack years of smoking	PAD	10	IVW	11.2762	9	0.2573
Ratio	Type 2 diabetes	70	MR Egger	388.1427	68	2.29E-46
Ratio	Type 2 diabetes	70	IVW	404.3071	69	6.63E-49
Type 2 diabetes	PAD	173	MR Egger	241.0639	171	3.33E-04
Type 2 diabetes	PAD	173	IVW	242.1255	172	3.40E-04
Ratio	Hypertension	70	MR Egger	527.2880	68	3.26E-72
Ratio	Hypertension	70	IVW	528.3507	69	5.74E-72
Hypertension	PAD	250	MR Egger	317.2200	248	1.94E-03
Hypertension	PAD	250	IVW	318.5728	249	1.89E-03
Ratio	Ischemic stroke	58	MR Egger	100.3868	56	2.50E-04
Ratio	Ischemic stroke	58	IVW	103.5148	57	1.62E-04
Ischemic stroke	PAD	18	MR Egger	59.9410	16	5.35E-07
Ischemic stroke	PAD	18	IVW	60.3646	17	9.15E-07
Ratio	Myocardial infarction	66	MR Egger	171.6720	64	8.73E-12
Ratio	Myocardial infarction	66	IVW	181.7105	65	5.56E-13
Myocardial infarction	PAD	13	MR Egger	49.0550	11	9.25E-07
Myocardial infarction	PAD	13	IVW	49.2138	12	1.92E-06

Boldness indicates statistical significance.

Table 4.

MR directional pleiotropy test.

Exposure	Outcome	Egger_intercept	Se	P
Ratio	PAD	−0.0028	0.0078	0.7161
Ratio	Pack years of smoking	0.0027	0.0013	0.4590
Pack years of smoking	PAD	−0.0003	0.0189	0.9895
Ratio	Type 2 diabetes	0.0073	0.0043	0.0970
Type 2 diabetes	PAD	−0.0036	0.0041	0.3867
Ratio	Hypertension	0.0002	0.0005	0.7124
Hypertension	PAD	0.0043	0.0042	0.3048
Ratio	Ischemic stroke	0.0049	0.0037	0.1919
Ischemic stroke	PAD	0.0171	0.0510	0.7411
Ratio	Myocardial infarction	0.0142	0.0073	0.0575
Myocardial infarction	PAD	−0.0083	0.0440	0.8537

Ratio, ratio of apolipoprotein B to apolipoprotein A1 levels; PAD, peripheral artery disease; MR, Mendelian randomization; Se, standard error.

Discussion

Despite the increasing prevalence of PAD globally and the adverse clinical outcomes it can lead to, PAD remains understudied and under-recognized compared to other atherosclerotic diseases such as myocardial infarction and stroke.¹⁵ The classic symptom of peripheral arterial disease is intermittent claudication (IC), which is characterized by cramping, pain, or fatigue in the lower extremities on exertion and usually resolves within 10 minutes of rest.⁴⁰ It can lead to serious complications, including limb ischemia and amputation as well as future stroke and myocardial infarction, which can severely impair function and quality of life.² However, patients with PAD are clinically asymptomatic, so identifying patients at high risk for PAD in the asymptomatic phase is critical to improving prognosis.⁴⁰

The overall prevalence of peripheral arterial disease (PAD) is similar between men and women.^41–43 Smoking and diabetes, the most critical risk factors for atherosclerotic disease, have a particularly significant impact on PAD. Notably, Mendelian randomization analyses indicate that smoking and diabetes are more likely to contribute to PAD than coronary artery disease and stroke,⁴⁴ which aligns with our findings. In our two-sample MR analysis, we found that both pack years of smoking and type 2 diabetes were genetically associated with PAD. Specifically, each standard deviation increase in pack years of smoking nearly doubled the risk of PAD (OR = 2.879, P = 1.09E-06), and the prevalence of PAD in diabetic patients was 32% higher (OR = 1.321, P = 2.72E-23, as shown in Table S2). Obesity (BMI ≥28 kg/m²) is also one of the major risk factors for atherosclerosis, increasing its incidence and severity by 2-4 times.⁴⁵ Our MR results revealed that elevated BMI significantly increased the risk of developing PAD (OR = 1.611, P = 1.40E-13).

In the United States, approximately 50% of patients with PAD also have coronary artery disease or cerebrovascular disease.⁴⁶ A Mendelian randomization study demonstrated that elevated levels of lipids, vLDL, LDL, and IDL, as well as high blood pressure, are associated with multivessel involvement in PAD patients.⁴⁷ Furthermore, patients with multivessel disease have a higher risk compared to those without multivessel disease.⁴⁸ Motivated by these findings, we validated the genetic causality of hypertension, ischemic stroke, and myocardial infarction with PAD using two-sample MR analysis (as shown in Table S2). Therefore, we believe that reducing the incidence of adverse cardiovascular events is crucial for preventing PAD.⁴⁹

Since each atherogenic particle contains one apoB molecule, the concentration of apoB is considered a direct measure of the total amount of atherogenic lipoproteins in circulation.⁵⁰ On the other hand, among patients with established cardiovascular disease, low levels of apoA1 are associated with higher rates of cardiovascular events, mortality,⁵¹ and severe limb ischemia.⁵² The apoB/A1 ratio is thought to reflect the balance between arteriolar internalization of cholesterol and its retrograde transport back to the liver,⁵³ and variations in this ratio are strongly correlated with atherosclerosis.²⁴ In this paper, we chose apolB/A1 as an exposure factor because we believe that rather than choosing one of these indicators alone, their ratio can be a more dynamic response to apolipoprotein levels. As shown in Table 2, our two-sample MR analysis confirmed a genetic causal relationship between the apoB/A1 Ratio and PAD. Specifically, for each 1 SD increase in the Ratio, there was a 46% increase in the risk of PAD (OR = 1.460, P = 8.74E-07).

Inspired by the aforementioned epidemiological evidence and existing findings, we hypothesized that the correlation between apoB/A1 and PAD might be mediated through other factors. Therefore, we initially selected BMI, smoking, type 2 diabetes mellitus, hypertension, ischemic stroke, and myocardial infarction as potential mediators and employed the two-step MVMR method to investigate their mediating effects in the relationship between apoB/A1 and PAD (as shown in Figure 2). Unfortunately, we failed to find positive genetic associations between apoB/A1 and BMI, smoking, or type 2 diabetes mellitus, although negative associations existed (as shown in Table S1). Ultimately, we selected only hypertension, ischemic stroke, and myocardial infarction as mediators for our MVMR analysis. After adjusting for the apoB/A1 ratio, the genetic associations of these mediators with PAD remained significant (as shown in Table S3). The mediating effects were 7.5% for hypertension, 19.7% for ischemic stroke, and 62.5% for myocardial infarction. The combined mediation effects of hypertension and ischemic stroke, and hypertension and myocardial infarction, were 22.5% and 82.3%, respectively (as shown in Figure 3). The total combined effect of all three mediators was 81.0%, likely due to some sample overlap (as shown in Figure 3). Our findings indicate that preventing hypertension, ischemic stroke, and myocardial infarction—adverse cardiovascular events—could be an effective strategy to curb the prevalence and progression of PAD, especially in individuals with hyperlipidemia.

Through MR analysis, we revealed a causal relationship between apolipoproteins and cardiovascular risk factors and PAD, demonstrating that an elevated ratio of apolipoprotein B to apolipoprotein A1 levels increases the risk of PAD. Notably, 81% of this genetic correlation was mediated by three factors: hypertension, ischemic stroke, and myocardial infarction. Given that such a significant portion of the association is mediated by these three conditions, timely and effective clinical interventions could potentially reduce the risk of PAD, particularly in populations with hyperlipidemia. However, it is important to note that this relationship may vary according to individual differences, and caution should be exercised when interpreting our results. Future research on PAD should aim to elucidate the causative genes, analyze structural variants, and validate these findings across different populations, pedigrees, and genders to be a better reference for the future prevention and treatment of clinical PAD.

Our study has several strengths. First, unlike traditional observational studies, Mendelian Randomization (MR) mitigates the effects of confounding and reverse causality on outcomes. Second, we employed multiple methods for MR analysis and further estimated the combined effects of multiple mediators on outcomes using MVMR based on TSMR, thereby enhancing the reliability of our findings. Third, all datasets included in this study were derived from large genome-wide studies of European origin, and the F-statistics of all instrumental variables were greater than 10, avoiding the effects of population differences and weak instrumental variables on the reliability of the results. Finally, our MR Egger regression test did not reveal any horizontal pleiotropy, further demonstrating the robustness of our findings.

Despite these intriguing results, we have to recognize some limitations of our study. First, the GWAS data were not stratified for confounders such as sex and age. Although MR analyses are less susceptible to confounding factors, we cannot completely rule out their potential impact on the results. Second, the instrumental variables for our MR study were primarily from European populations, so we cannot ignore the potential impact of racial differences on the results. Third, despite the robustness of our MR study, it is challenging to completely rule out the presence of pleiotropy in any MR analysis. Fourth, although we used MVMR to examine the mediating effects of apoB/A1 on the occurrence of PAD and their respective mediator proportions, these mediator proportions may interact with each other. The combined proportions accounting for such interactions have not been quantified. Finally, MR results reflect lifetime genetic exposure, and inferences drawn from them are applicable to the pathogenesis of PAD, but MR analyses have limited utility for extrapolating to the progression of the disease, prognosis, and the outcome of clinical interventions.

Conclusion

In conclusion, our study not only paves the way for future investigations but also provides a valuable reference for clinicians. It highlights the importance of raising awareness about PAD and informs preventive measures and pharmacological treatments for this increasingly prevalent disease.

Supplemental Material

Supplemental Material - Apolipoprotein and peripheral artery disease: Mendelian randomization analysis

Supplemental Material for Apolipoprotein and peripheral artery disease: Mendelian randomization analysis by Chen-Xin Wan, Yu-Shu Gong, and Tao Xu in Vascular.

Footnotes

Author contributions

W-CX: conception and design, data mining, writing the article, analysis, and interpretation. G-YS: analysis and figure drawing. XT: interpretation and critical revision of the article. All authors contributed to the article and approved the submitted version.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Ethical statement

ORCID iD

Tao Xu

Supplemental material

Supplemental material for this article is available online.

References

Fowkes

Aboyans

Fowkes

, et al. Peripheral artery disease: epidemiology and global perspectives. Nat Rev Cardiol 2017; 14(3): 156–170.

Criqui

Aboyans

. Epidemiology of peripheral artery disease. Circ Res 2015; 116(9): 1509–1526.

Khan

Lyytikainen

, et al. Serum apolipoprotein A-I concentration differs in coronary and peripheral artery disease. Scand J Clin Lab Invest 2020; 80(5): 370–374.

Aboyans

Ricco

Bartelink

MLEL

, et al. Editor’s choice - 2017 ESC guidelines on the diagnosis and treatment of peripheral arterial diseases, in collaboration with the European society for vascular surgery (ESVS). Eur J Vasc Endovasc Surg 2018; 55(3): 305–368.

Tunstall-Pedoe

Peters

SAE

Woodward

, et al. Twenty-year predictors of peripheral arterial disease compared with coronary heart disease in the scottish heart health extended cohort (SHHEC). J Am Heart Assoc 2017; 6(9).

Hoek

Van Oort

Elders

PJM

, et al. Causal association of cardiovascular risk factors and lifestyle behaviors with peripheral artery disease: a mendelian randomization approach. J Am Heart Assoc 2022; 11(16): e025644.

Oksala

Seppälä

Hernesniemi

, et al. Complementary prediction of cardiovascular events by estimated apo- and lipoprotein concentrations in the working age population. The Health 2000 Study. Ann Med 2013; 45(2): 141–148.

Tikkanen

Jägerroos

Holmes

, et al. Metabolic biomarker discovery for risk of peripheral artery disease compared with coronary artery disease: lipoprotein and metabolite profiling of 31 657 individuals from 5 prospective cohorts. J Am Heart Assoc 2021; 10(23): e021995.

Panayiotou

Griffin

Georgiou

, et al. ApoB/ApoA1 ratio and subclinical atherosclerosis. Int Angiol 2008; 27(1): 74–80.

10.

Levin

Zuber

Walker

, et al. Prioritizing the role of major lipoproteins and subfractions as risk factors for peripheral artery disease. Circulation 2021; 144(5): 353–364.

11.

Sniderman

Robinson

. ApoB in clinical care: pro and con. Atherosclerosis 2019; 282: 169–175.

12.

Aday

Lawler

Cook

, et al. Lipoprotein particle profiles, standard lipids, and peripheral artery disease incidence. Circulation 2018; 138(21): 2330–2341.

13.

Conte

Bradbury

Kolh

, et al. Global vascular guidelines on the management of chronic limb-threatening ischemia. J Vasc Surg 2019; 69(6s): 3S–125S.

14.

Aboyans

Ricco

Bartelink

MLEL

, et al. 2017 ESC guidelines on the diagnosis and treatment of peripheral arterial diseases, in collaboration with the European society for vascular surgery (ESVS): document covering atherosclerotic disease of extracranial carotid and vertebral, mesenteric, renal, upper and lower extremity arteries endorsed by: the European stroke organization (ESO)the task force for the diagnosis and treatment of peripheral arterial diseases of the European society of cardiology (ESC) and of the European society for vascular surgery (ESVS). Eur Heart J 2018; 39(9): 763–816.

15.

Criqui

Matsushita

Aboyans

, et al. Lower extremity peripheral artery disease: contemporary epidemiology, management gaps, and future directions: a scientific statement from the American heart association. Circulation 2021; 144(9): e171–e191.

16.

Walldius

Jungner

. The apoB/apoA-I ratio: a strong, new risk factor for cardiovascular disease and a target for lipid-lowering therapy--a review of the evidence. J Intern Med 2006; 259(5): 493–519.

17.

Koren-Morag

Goldbourt

Graff

, et al. Apolipoproteins B and AI and the risk of ischemic cerebrovascular events in patients with pre-existing atherothrombotic disease. J Neurol Sci 2008; 270(1–2): 82–87.

18.

Azizzadeh-Roodpish

Garzon

Mainali

. Classifying single nucleotide polymorphisms in humans. Mol Genet Genom 2021; 296(5): 1161–1173.

19.

Uffelmann

Huang

Munung

, et al. Genome-wide association studies. Nat Rev Methods Primers 2021; 1(1): 59.

20.

Hemani

Tilling

Davey Smith

. Orienting the causal relationship between imprecisely measured traits using GWAS summary data. PLoS Genet 2017; 13(11): e1007081.

21.

Richmond

Hemani

Tilling

, et al. Challenges and novel approaches for investigating molecular mediation. Hum Mol Genet 2016; 25(R2): R149–R156.

22.

Burgess

Thompson

. Multivariable Mendelian randomization: the use of pleiotropic genetic variants to estimate causal effects. Am J Epidemiol 2015; 181(4): 251–260.

23.

Dong

Chen

Wang

, et al. Apolipoprotein A1, B levels, and their ratio and the risk of a first stroke: a meta-analysis and case-control study. Metab Brain Dis 2015; 30(6): 1319–1330.

24.

Forte

Calcaterra

Lupoli

, et al. Association of apolipoprotein levels with peripheral arterial disease: a meta-analysis of literature studies. Eur J Prev Cardiol 2022; 28(18): 1980–1990.

25.

Huang

X-W

Pang

S-W

Zhang

, et al. Assessment of causal associations between obesity and peripheral artery disease: a bidirectional Mendelian randomization study. Front Cardiovasc Med 2024; 11: 1332530.

26.

Lan

Xie

Lan

, et al. Association and mediation between educational attainment and respiratory diseases: a Mendelian randomization study. Respir Res 2024; 25(1): 115.

27.

Lyon

Andrews

Elsworth

, et al. The variant call format provides efficient and robust storage of GWAS summary statistics. Genome Biol 2021; 22(1): 32.

28.

Sakaue

Kanai

Tanigawa

, et al. A cross-population atlas of genetic associations for 220 human phenotypes. Nat Genet 2021; 53(10): 1415–1424.

29.

Dönertaş

Fabian

Valenzuela

, et al. Common genetic associations between age-related diseases. Nat Aging 2021; 1(4): 400–412.

30.

Malik

Chauhan

Traylor

, et al. Multiancestry genome-wide association study of 520,000 subjects identifies 32 loci associated with stroke and stroke subtypes. Nat Genet 2018; 50(4): 524–537.

31.

Cho

Kwon

J-M

Kim

K-H

, et al. Artificial intelligence algorithm for detecting myocardial infarction using six-lead electrocardiography. Sci Rep 2020; 10(1): 20495.

32.

Emdin

Khera

Natarajan

, et al. Genetic association of waist-to-hip ratio with cardiometabolic traits, type 2 diabetes, and coronary heart disease. JAMA 2017; 317(6): 626–634.

33.

Burgess

Small

Thompson

. A review of instrumental variable estimators for Mendelian randomization. Stat Methods Med Res 2017; 26(5): 2333–2355.

34.

Shim

Chasman

Smith

, et al. A multivariate genome-wide association analysis of 10 LDL subfractions, and their response to statin treatment, in 1868 Caucasians. Plos One 2015; 10(4): e0120758.

35.

Burgess

Bowden

Fall

, et al. Sensitivity analyses for robust causal inference from mendelian randomization analyses with multiple genetic variants. Epidemiology 2017; 28(1): 30–42.

36.

Carter

Sanderson

Hammerton

, et al. Mendelian randomisation for mediation analysis: current methods and challenges for implementation. Eur J Epidemiol 2021; 36(5): 465–478.

37.

Bowden

Davey Smith

Burgess

. Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression. Int J Epidemiol 2015; 44(2): 512–525.

38.

Burgess

Thompson

. Interpreting findings from Mendelian randomization using the MR-Egger method. Eur J Epidemiol 2017; 32(5): 377–389.

39.

Verbanck

Chen

Neale

, et al. Detection of widespread horizontal pleiotropy in causal relationships inferred from Mendelian randomization between complex traits and diseases. Nat Genet 2018; 50(5): 693–698.

40.

Gerhard-Herman

Gornik

Barrett

, et al. 2016 AHA/ACC guideline on the management of patients with lower extremity peripheral artery disease: a report of the American college of cardiology/American heart association task force on clinical practice guidelines. Circulation 2017; 135(12): e726–e779.

41.

Selvin

Erlinger

. Prevalence of and risk factors for peripheral arterial disease in the United States: results from the National Health and Nutrition Examination Survey, 1999-2000. Circulation 2004; 110(6): 738–743.

42.

Mcdermott

Liu

Criqui

, et al. Ankle-brachial index and subclinical cardiac and carotid disease: the multi-ethnic study of atherosclerosis. Am J Epidemiol 2005; 162(1): 33–41.

43.

Allison

Cushman

Solomon

, et al. Ethnicity and risk factors for change in the ankle-brachial index: the Multi-Ethnic Study of Atherosclerosis. J Vasc Surg 2009; 50(5): 1049–1056.

44.

Levin

Klarin

Assimes

, et al. Genetics of smoking and risk of atherosclerotic cardiovascular diseases: a mendelian randomization study. JAMA Netw Open 2021; 4(1): e2034461.

45.

Liu

Shi

, et al. Adipokines, adiposity, and atherosclerosis. Cell Mol Life Sci 2022; 79(5): 272.

46.

Huxley

Yatsuya

Lutsey

, et al. Impact of age at smoking initiation, dosage, and time since quitting on cardiovascular disease in african americans and whites: the atherosclerosis risk in communities study. Am J Epidemiol 2012; 175(8): 816–826.

47.

Dikilitas

Satterfield

Kullo

. Risk factors for polyvascular involvement in patients with peripheral artery disease: a mendelian randomization study. J Am Heart Assoc 2020; 9(24): e017740.

48.

Tran

Spertus

Mena-Hurtado

, et al. Association of disease-specific health status with long-term survival in peripheral artery disease. J Am Heart Assoc 2022; 11(4): e022232.

49.

Jang

Cho

, et al. Comparison of cardiovascular risk factors for peripheral artery disease and coronary artery disease in the Korean population. Korean Circ J 2013; 43(5): 316–328.

50.

Langlois

Chapman

Cobbaert

, et al. Quantifying atherogenic lipoproteins: current and future challenges in the era of personalized medicine and very low concentrations of LDL cholesterol. A consensus statement from EAS and EFLM. Clin Chem 2018; 64(7): 1006–1033.

51.

Pol

Held

Westerbergh

, et al. Dyslipidemia and risk of cardiovascular events in patients with atrial fibrillation treated with oral anticoagulation therapy: insights from the ARISTOTLE (apixaban for reduction in stroke and other thromboembolic events in atrial fibrillation) trial. J Am Heart Assoc 2018; 7(3).

52.

Ahnström

Gottsäter

Lindblad

, et al. Plasma concentrations of apolipoproteins A-I, B, and M in patients with critical limb ischemia. Clin Biochem 2010; 43(6): 599–603.

53.

Walldius

Jungner

. Apolipoprotein B and apolipoprotein A-I: risk indicators of coronary heart disease and targets for lipid-modifying therapy. J Intern Med 2004; 255(2): 188–205.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.29 MB