Epidemiology of Peripheral Artery Disease and Polyvascular Disease

Abstract

Atherosclerotic lower-extremity peripheral artery disease (PAD) is increasingly recognized as an important cause of cardiovascular morbidity and mortality that affects more than 230 million people worldwide. Traditional cardiovascular risk factors, including advanced age, smoking, and diabetes, are strongly linked to an increase risk of PAD. Although PAD has been historically underappreciated compared to coronary artery disease and stroke, greater attention on PAD in recent years has led to important new epidemiologic insights in the areas of thrombosis, inflammation, dyslipidemia, and microvascular disease. In addition, the concept of polyvascular disease, or clinically-evident atherosclerosis in multiple arterial beds, is increasingly identified as a particularly malignant cardiovascular disease worthy of special clinical attention and further study. It is noteworthy that PAD may increase the risk of adverse outcomes in similar or even greater magnitude than coronary disease or stroke. In this review, we highlight important new advances in the epidemiology of PAD with a particular focus on polyvascular disease, emerging biomarkers, and differential risk pathways for PAD compared to other atherosclerotic diseases.

Introduction

Lower extremity peripheral artery disease (PAD) is characterized by stenosis or occlusion anywhere from the aortoiliac segment to the pedal arteries.¹ Many different disease processes can cause obstruction in these vessels, but the vast majority of cases are due to atherosclerosis, which will be the focus of this review. Although PAD has long been underappreciated compared to coronary artery disease (CAD) and cerebrovascular disease, PAD is increasingly recognized as an important cause of cardiovascular morbidity and mortality. Although many risks are shared between PAD and other forms of atherosclerosis, epidemiologic data have increasingly made it clear that PAD warrants recognition as a unique entity. In this review, we provide a summary of key epidemiologic insights into PAD with important updates since last reviewed in this journal.² We also discuss PAD in the context of polyvascular disease, an important conceptual framework that has furthered our understanding of atherosclerotic risk.

Defining PAD and Polyvascular Disease

The scope of this review is limited to PAD due to atherosclerosis and does not include nonatherosclerotic causes, such as Buerger’s disease,³ vasculitis,⁴ and fibromuscular dysplasia.⁵ PAD can affect any artery perfusing the lower extremities. The classic symptom of PAD is intermittent claudication (IC), characterized by cramping, pain, or fatigue in the lower extremities brought on by exertion and relieves by rest, typically within 10 minutes.¹ Symptoms may occur in the buttocks, thigh, or calf and often correspond to the proximal level of arterial obstruction. Although IC can be detected as part of a standard medical history, there are numerous validated claudication questionnaires used in a variety of epidemiologic studies.⁶ Importantly, many patients with PAD do not experience typical IC and, instead, have atypical limb symptoms or no clear limb symptoms at all.⁷ The notion of “asymptomatic” PAD should be carefully recognized, as patients with PAD, regardless of leg symptoms, have poorer prognosis and reduced limb function compared to those without. In addition, some patients may reduce their daily activity to palliate limb symptoms.⁸

Chronic limb-threatening ischemia (CLTI), also referred to as critical limb ischemia, is a severe form of PAD defined by ischemic rest pain, tissue loss, or gangrene.^{1, 9} Symptoms should be present for a minimum of 2 weeks and be accompanied by objective limb hypoperfusion (see below). Rest pain typically involves the distal limb but may extend proximally in more severe cases, and it may worsen with elevation and improve with dependency. Acute limb ischemia (ALI) is another severe manifestation of PAD, defined by sudden, severe hypoperfusion of the limb over a period of <2 weeks.¹ Symptoms can include pain, pallor, pulselessness, poikilothermia, paresthesias, and paralysis with loss of sensation and motor function occurring in more severe cases. Although ALI may occur in the setting of PAD, it frequently occurs as a result of cardiac or artery-to-artery thromboembolism in the absence of significant peripheral atherosclerosis.

Since atherosclerosis is a systemic disease process, recent attention has focused on the coincidence of PAD with atherosclerosis in other arterial beds, most commonly CAD and cerebrovascular disease (occasionally including renal artery or mesenteric atherosclerosis). This framework, termed polyvascular disease, is defined by the presence of atherosclerosis in ≥2 arterial beds.¹⁰ Although definitions vary, qualifying atherosclerosis is typically either symptomatic or resulting in significant vessel stenosis (e.g. >50%).¹¹ Increasingly, polyvascular disease is being used to identify patients at heightened risk of cardiovascular and limb events who may benefit from intensive secondary preventive therapies. Indeed, the 2018 American Heart Association/American College of Cardiology (AHA/ACC) Cholesterol Guideline categorizes patients with two or more major atherosclerotic cardiovascular diseases as very high-risk.¹²

Diagnosis

Because the majority of patients with PAD lack typical leg symptoms, means of objectively diagnosing PAD are critical. In accordance with the 2016 AHA/ACC Guideline on the Management of Patients with Lower Extremity PAD, the resting ankle-brachial index (ABI) is considered the first-line test for diagnosing PAD.¹ In this test, a sphygmomanometer cuff and Doppler probe are used to calculate the systolic blood pressure in both brachial arteries as well as the bilateral dorsalis pedis and posterior tibial arteries with the patient in the supine position.¹³ The ABI is the ratio of the higher reading of the two ankle pressures (dorsalis pedis and posterior tibial) over the higher reading of right or left brachial pressure and should be calculated in each limb. An ABI ≤ 0.90 in either limb is diagnostic of PAD.¹ The sensitivity and specificity of the resting ABI are generally reported as >80%.^14–16 However, these measures are primarily derived from studies of symptomatic PAD, and the performance of the ABI in individuals with atypical or absent limb symptoms is less understood. Nonetheless, given the robust association of ABI with mortality and morbidity, the AHA/ACC 2016 PAD Guideline acknowledges that the measurement of ABI is reasonable in persons at increased risk of PAD even when leg symptoms are absent.¹ In the setting of non-compressible, calcific ankle vessels, a toe-brachial index (the ratio of toe systolic blood pressure to brachial systolic blood pressure) with a threshold of ≤ 0.70 can be used to diagnose PAD. In cases of suspected PAD but normal resting ABI, post-exercise ABI measures may increase the yield of testing.¹⁷

Additional imaging modalities, while not required to diagnose PAD, may be useful in assessing disease distribution and planning for revascularization.¹ Duplex ultrasonography, computerized tomography angiography, and magnetic resonance angiography are all widely available and capable of detecting arterial stenosis throughout the limb. Invasive angiography is the gold standard for diagnosing lower limb stenotic lesions, although it is typically reserved for procedural planning or in cases where prior testing is inconclusive. Additional methods to assess limb perfusion include toe arterial waveform analysis, transcutaneous oxygen pressure, and skin perfusion pressure. Although a full discussion of these methods is outside the scope of this review, they may be particularly helpful in guiding management of patients with suspected CLTI.¹⁸

Prevalence, Incidence, and Temporal Trends

• Overall PAD

A recent systematic review estimates that the global prevalence of PAD was 5.6% in 2015, indicating that approximately 236 million adults were living with PAD worldwide.¹⁹ The prevalence was higher in high-income countries than low- and middle-income countries (7.4% vs. 5.1%), but, reflecting the population size, most individuals with PAD (72.9%) were in low- and middle-income countries. Looking at this systematic review and the prior systematic review²⁰ from the same research group, the prevalence of PAD has increased from 2000 to 2015 by ~45% globally (~18% in high-income countries and ~58% in low- and middle-income countries) (Table 1).

Table 1.

Estimated global prevalence of peripheral artery disease in high-income and low- and middle-income countries

	Year of Estimation									Change over Time
	2000 (millions)			2010 (millions)			2015 (millions)			2000–2010 (millions)			2000–2015 (millions)
Age (years)	HICs	LMICs	Global	HICs	LMICs	Global	HICs	LMICs	Global	HICs	LMICs	Global	HICs	LMICs	Global
25–29	2.31	10.76	13.07	2.38	12.04	14.42	1.42	15.44	16.86	3.0	11.9	10.3	−38.6	43.5	29.0
30–34	2.80	11.47	14.27	2.76	12.34	15.10	1.86	15.56	17.43	−1.5	7.6	5.8	−33.6	35.7	22.1
35–39	3.49	11.25	14.73	3.34	13.78	17.12	2.38	15.75	18.13	−4.1	22.5	16.2	−31.7	40.0	23.1
40–44	4.07	11.14	15.21	3.94	14.71	18.65	3.09	17.25	20.34	−3.3	32.0	22.6	−24.1	54.9	33.7
45–49	4.53	11.41	15.94	4.85	14.35	19.21	3.89	18.20	22.09	7.1	25.8	20.5	−14.1	59.5	38.6
50–54	4.91	9.90	14.81	5.50	14.10	19.60	4.87	17.83	22.70	12.1	42.4	32.4	−0.8	80.1	53.3
55–59	4.53	9.11	13.64	5.95	14.17	20.12	5.72	16.65	22.37	31.3	55.5	47.5	26.3	82.7	64.0
60–64	5.34	9.07	14.42	6.24	11.79	18.03	6.46	16.18	22.63	16.8	29.9	25.1	20.9	78.3	57.0
65–69	5.29	8.42	13.70	5.55	10.12	15.67	7.22	12.61	19.83	4.9	20.3	14.3	36.6	49.8	44.7
70–74	5.59	6.95	12.55	6.04	9.02	15.06	6.82	9.89	16.70	8.0	29.7	20.1	21.9	42.2	33.1
75–79	4.81	4.96	9.77	5.37	7.01	12.38	6.81	8.12	14.93	11.7	41.4	26.8	41.6	63.7	52.8
80–84	3.11	3.02	6.12	4.72	4.40	9.12	6.05	5.22	11.27	52.0	45.8	48.9	94.7	73.1	84.1
85–89	2.25	1.41	3.66	3.03	2.09	5.12	4.52	2.68	7.19	34.8	47.9	39.8	101.2	89.9	96.6
≥90	1.17	0.54	1.72	1.61	0.86	2.47	2.99	1.16	4.15	37.2	58.8	44.1	154.7	113.2	141.7
Total	54.20	109.41	163.60	61.29	140.78	202.06	64.09	172.53	236.62	13.1	28.7	23.5	18.3	57.7	44.6

HICs refers to high-income countries; LMIC, low- and middle-income countries

Estimates derived from Song P, Fang Z, Wang H, Cai Y, Rahimi K, Zhu Y, Fowkes FGR, Fowkes FJI and Rudan I. Global and regional prevalence, burden, and risk factors for carotid atherosclerosis: a systematic review, meta-analysis, and modelling study. The Lancet Global health. 2020;8:e721-e729 and Fowkes FGR, Rudan D, Rudan I, Aboyans V, Denenberg JO, McDermott MM, Norman PE, Sampson UKA, Williams LJ, Mensah GA and Criqui MH. Comparison of global estimates of prevalence and risk factors for peripheral artery disease in 2000 and 2010: a systematic review and analysis. Lancet (London, England). 2013;382:1329–1340.

The prevalence of PAD in the US is estimated to be ~7%, affecting 8.5 million adults.²¹ However, we should note that this estimate is predominantly based on data collected in the 1990s. Notably, the National Health and Nutrition Examination Survey (NHANES) measured ABI only in its cycle of 1999–2004.²² Given the recent changes in the prevalence of atherosclerotic risk factors (e.g., epidemics of diabetes²³ and declining prevalence of conventional cigarette use²⁴), a more contemporary estimate of PAD prevalence is required. Another caveat is that data on the prevalence of PAD in individuals aged 80 years or older, the population most rapidly growing in the US,²⁵ are sparse.

• CLTI

Limited epidemiological data are available for CLTI. A study using data from MarketScan (database of millions of patients across the US from Medicare and Medicaid programs and large employers’ health plans) estimated that the prevalence of CLTI was 1.3% among individuals aged 40 years or older in a given year between 2003 and 2008.²⁶ In this specific dataset, this accounted for 11% of overall PAD. The prevalence of CLTI seemed largely constant in those years. Similarly, a study using data from the Nationwide Inpatient Sample (NIS) demonstrated that the number of admissions related to CLTI was generally stable between 2003 and 2011.²⁷

• Amputation

When we interpret statistics of non-traumatic lower extremity amputation in the context of PAD, we should keep in mind the two aspects of amputation: (1) an outcome of severe PAD and (2) an important therapeutic procedure to save a patient’s life or limb. The decline in the rate of non-traumatic lower extremity amputation in the US has been reported from the 1990s to the beginning of the 2000s.^{28, 29} However, a recent report using data from the NIS demonstrated a concerning increase in amputation rate after 2009 in persons with diabetes. Specifically, the age-adjusted rate of non-traumatic lower extremity amputation has increased by 50% from 2009 (3.1 per 1,000 adults) to 2015 (4.6 per 1,000 adults).³⁰ A similar increasing trend has been seen through 2018 among US veterans.³¹ This increase in amputation rate is mainly due to an increase in minor amputations below the ankle, although a full explanation for this negative trend remains unknown.^{29, 30} A recent review provides more details on the epidemiology of lower extremity amputations.³²

Racial Disparities of PAD

Prior analyses have generally demonstrated a higher prevalence of PAD among Black individuals compared to non-Hispanic White individuals.^33–35 For instance, an analysis from 7 different community-based cohorts showed that both male and female Black individuals had the greatest prevalence of PAD, defined by an ABI <0.90, among all race groups.³⁶ Among men, Hispanic, American Indian, and non-Hispanic White individuals had similar prevalence rates, while individuals of Asian descent had the lowest prevalence. These trends were similar for women except for American Indian individuals, who had a prevalence similar to Black women with advancing age. Lifetimes estimates of PAD risk, defined by an ABI <0.90, suggest that 30% of Black individuals and 20% of non-Hispanic White individuals will develop PAD.³⁷ Data linking race to incident PAD are limited. For a more detailed discussion, see the review entitled “Racial Disparities in PAD” in this compendium.

Traditional Cardiovascular Risk Factors and PAD/Polyvascular Disease

Many traditional cardiovascular risk factors are shared between PAD, CAD, and other atherosclerotic diseases. However, several important findings in recent years have emphasized a differential pattern between some risk factors and atherosclerosis in different beds (Table 2). Although these findings have advanced our epidemiologic understanding of atherosclerotic disease, the biology underlying this in many instances remains unclear.

Table 2.

Association between modifiable cardiovascular risk factors and atherosclerotic diseases

Risk Factor
Smoking	PAD > CAD/stroke
Diabetes	PAD > CAD/stroke
Low-Density Lipoprotein Cholesterol	PAD < CAD/stroke
Triglycerides	PAD > CAD/stroke
Hypertension	PAD = CAD/stroke*
Microvascular Disease	PAD > CAD/stroke

Table 2 displays the magnitude of risk associated with each risk factor for the components of polyvascular disease

PAD refers to peripheral artery disease; CAD, coronary artery disease

^*Data suggest the risk association may be strongest for stroke

• Age

As true in any atherosclerotic diseases, older age is one of the strongest risk factors of PAD. For example, in the NHANES, the prevalence of PAD was ~15% in individuals aged 70 years or older but ~1% in those aged 40–49 years.³⁸ In most racial and ethnic groups, the prevalence of PAD approximately doubles with each subsequent decade of life beginning with those aged 40–49 year.³⁶ This why the AHA/ACC 2016 PAD Guideline recommends screening of PAD using the ABI in adults aged 65 years or older (and younger adults according to other risk factors).¹

• Sex

Until the age of ~60–70 years, at a given age, male sex is known to have a higher risk of atherosclerotic diseases such as CAD and stroke compared to the female sex. However, several studies have shown that the prevalence of PAD is similar between men and women at any given age.^{21, 33, 38, 39} This similar prevalence between men and women may partially reflect the property of ankle blood pressure potentially being lower in individuals at shorter heights. Indeed, some experts have proposed a sex-specific threshold of ABI for detecting PAD.⁴⁰ However, to our knowledge, major clinical guidelines have not adopted sex-specific thresholds of ABI and consistently recommend ABI ≤0.9 to identify PAD regardless of sex.^{1, 41} Of note, a study including ~1,800 apparently healthy adults estimated that the potential impact of female sex on ABI is small (i.e., around −0.02) after accounting for other major demographic and clinical factors.⁴²

• Tobacco Use and Electronic Cigarettes

Smoking is one of the most important modifiable risk factors of atherosclerotic diseases. Its impact is particularly strong for PAD, and epidemiologic data have consistently shown current smoking is associated with a 2- to 3-fold increased risk of PAD.^{33, 34, 43–45} A recent report from the Atherosclerosis Risk in Communities (ARIC) study demonstrated that the relative risk (RR) according to pack-years ≥25 (vs. never smokers) was ~4 for PAD and 1.5–2 for CAD and stroke.⁴⁶ This study also showed that the impact of smoking persisted up to 30 years for PAD, whereas the elevated risk returned to the level of never smokers within 20 years for other atherosclerotic diseases (Figure 1). Importantly, a recent Mendelian randomization analysis demonstrated that smoking is indeed more likely to cause PAD than CAD and stroke.^47,48 Even among adolescents, cigarette smoking is associated with an increase in carotid intima-media thickness.⁴⁸ Therefore, we should continue our efforts for smoking prevention and cessation among all age groups. Also, any documents stating the impact of smoking on cardiovascular disease should take into account PAD; otherwise, the document would underestimate the harm of smoking on the cardiovascular system. Although long-term data are lacking, a body of evidence suggests potential harm of electronic cigarettes on the vascular system.^49–52

Figure 1. Risk of smoking and persistence of risk after cessation for major atherosclerotic diseases.

A – Multivariable adjusted hazard ratios for CAD, stroke, and PAD by pack-years of smoking in the Atherosclerosis Risk in Communities Study. B – Time-varying multivariable adjusted hazard ratios for CAD, stroke, and PAD by time since smoking cessation in the Atherosclerosis Risk in Communities Study.

CAD refers to coronary artery disease; PAD, peripheral artery disease; CI, confidence interval

Data derived from Ding N, Sang Y, Chen J, Ballew SH, Kalbaugh CA, Salameh MJ, Blaha MJ, Allison M, Heiss G, Selvin E, Coresh J and Matsushita K. Cigarette Smoking, Smoking Cessation, and Long-Term Risk of 3 Major Atherosclerotic Diseases. J Am Coll Cardiol. 2019;74:498–507.

• Hypertension

According to its high prevalence and strong pathophysiological contribution, a study using data from the Health Professionals Follow-up Study estimated that hypertension accounts for ~40% of PAD risk. A study using data from ARIC recently quantified the association of the new hypertension classification according to the AHA/ACC 2017 Hypertension Guideline.^{53, 54} This study demonstrates a dose-response relationship between systolic blood pressure and the risk of PAD, with a RR of ~2.6 if systolic blood pressure ≥140 mmHg and ~1.6 if systolic blood pressure 120–139 mmHg, compared to the reference <120 mmHg, among individuals not taking antihypertensive medications. On the other hand, the excess risk of PAD was only observed when diastolic blood pressure is ≥90 mmHg. This study with a long follow up over 25 years has uniquely shown that blood pressure (especially systolic) was robustly associated with incident CLTI.

• Diabetes

Together with smoking, diabetes is considered one of the strongest risk factors for PAD.⁵⁵ Indeed, a meta-analysis demonstrated the highest odds ratio (OR) of having PAD for current smoking (OR 2.1) followed by diabetes (OR 1.7).²⁰ Of importance, diabetes is particularly strongly associated with a severe form of PAD. For example, a US cohort study has recently reported that the hazard ratio (HR) of CLTI was 10.3 (95% confidence interval [CI] 4.8–22.5) in middle-aged adults with diagnosed diabetes and hemoglobin A1c ≥7% compared to those with no diagnosis of diabetes and hemoglobin A1c <5.7%. Moreover, diabetes accounts for ~70% of cases undergoing lower extremity amputation in the US.³⁰

• Lipoproteins

Despite strong data linking low-density lipoprotein cholesterol (LDL-C) to CAD, few epidemiologic studies have demonstrated a link between LDL-C and incident PAD.^{56, 57} In ARIC each standard deviation increment in LDL-C was associated with CAD (HR, 1.20 [95% CI, 1.12–1.28]) but not PAD (HR, 1.10 [95% CI, 0.97–1.25].⁵⁸ Data from the Women’s Health Study (WHS) similarly showed a strong risk association between LDL-C and incident CAD and cerebrovascular disease but not PAD.⁵⁹ In contrast, data from ARIC, WHS, and the Physicians’ Health Study (PHS) have demonstrated a strong link between high-density lipoprotein cholesterol (HDL-C) and total cholesterol:HDL-C ratio measures and incident PAD.^{56, 58, 59} Lipid subfraction analyses suggest that triglyceride-rich lipoprotein particles are the primary drivers of lipid-associated risk in PAD.^{56, 58–66}

Epidemiologic data linking lipoprotein(a) to incident PAD are mixed. Data from several small studies have suggested a link between lipoprotein(a) and incident,⁶¹ prevalent,⁶² and worsening PAD.⁶³ However, there was not a significant association between lipoprotein(a) and incident PAD in either WHS⁶⁴ or PHS.⁵⁶ In contrast, GWAS data have demonstrated an association between lipoprotein(a) and PAD.⁶⁵ In the ODYSSEY OUTCOMES trial (Evaluation of Cardiovascular Outcomes After an Acute Coronary Syndrome), the risk of PAD among participants in the placebo arm was associated with baseline lipoprotein(a) concentration rather than LDL-C.⁶⁶

Emerging Risk Pathways

• Inflammation

Inflammation has emerged as an important risk marker of systemic atherosclerosis, and numerous studies have linked inflammation biomarkers with incident PAD. In a prospective, nested, case-control study from the PHS, which included apparently healthy men free of cardiovascular disease, baseline levels of C-reactive protein (CRP) were independently associated with incident PAD (adjusted HR extreme quartiles 2.2; 95% CI, 1.1–4.8; P=0.04). The risk association was even greater for individuals requiring lower extremity revascularization. Similar associations for CRP were demonstrated in the Edinburgh Artery Study and WHS.^{64, 67} In ARIC, baseline high-sensitivity CRP concentration was associated with overall incident PAD (HR per standard deviation [log transformed] 1.34; 95% CI, 1.18–1.52) as well as incident CLTI (HR per standard deviation [log transformed] 1.34; 95% CI, 1.09–1.65).⁶⁸ Soluble intercellular adhesion molecule (sICAM)-1, a marker of leukocyte adhesion, was also associated with incident PAD in PHS (OR for highest vs. referent quartile 3.2; 95% CI, 1.4–7.4; P=0.008).⁶⁹

• Thrombosis

Thrombosis has emerged as an important therapeutic target in PAD, and there are numerous studies suggesting a link between coagulation cascade activation and PAD. D-dimer, a fibrin degradation product release in the blood during fibrinolysis, is a common biomarker of thrombosis. Among 370 individuals with PAD defined by an abnormal ABI, D-dimer was independently and inversely associated with several measures of functional capacity, including 6-minute walk distance and 4-meter walking velocity.⁷⁰ In the Biomarker Risk Assessment in Vulnerable Outpatients (BRAVO) Study, which followed 595 subjects with PAD for up to 3 years, D-dimer levels rose 2 months prior to adverse CAD events compared to individuals who did not suffer coronary events.⁷¹ D-dimer also independently associates with adverse limb events,⁷² severity of PAD,^{73, 74} and polyvascular disease.⁷⁵

Fibrinogen is an enzymatic target of thrombin also involved in clot formation. In a nested case-control study of previously healthy men in PHS, baseline fibrinogen concentration was associated with incident PAD (adjusted RR highest vs. referent quartile 2.2; 95% CI, 1.1–4.7; P=0.02).⁵⁶ Similarly, fibrinogen was associated with incident PAD in the Edinburgh Artery Study (RR 1.16; 95% CI, 1.05–1.17; P <0.01) and ARIC (RR highest vs. referent quartile 2.14; 95% [CI, 1.31–3.51; P=0.003) after adjusting for other cardiovascular risk factors.^{76, 77} Finally, markers of impaired fibrinolysis, including raised levels of plasminogen activator inhibitor-1 and tissue plasminogen activator antigen, are also linked to PAD.^{65, 78–80}

GWAS data have demonstrated a link between the Factor V Leiden variant and PAD.⁶⁵ Notably, this association has not been demonstrated for CAD. More recently, two randomized clinical trials have shown that the addition of low-dose rivaroxaban, a factor Xa inhibitor, to aspirin is effective at reducing limb events in patients with stable PAD⁷⁹ and in those following lower extremity revascularization.⁸⁰

• Platelet Activation

Enhanced platelet activation may also contribute to the development of PAD, although data supporting this primarily come from small case-control studies. Compared to healthy control subjects, patients with both IC and CLTI exhibit greater expression of platelet P-selectin,^{81, 82} a cell-surface adhesion molecule, as well as increased platelet aggregation,⁸¹ and this is proportional to disease severity.⁸¹ Individuals with PAD treated with the antiplatelet drug aspirin exhibit normal platelet aggregation, suggesting potential unresponsiveness to aspirin therapy in this population.⁸³ In the Edinburgh Artery Study, genetic variation in the platelet receptor GP IIIa locus was associated with a decreased incidence of IC over 5 years of follow-up.⁸⁴ More data are needed to determine whether there is a prospective link between platelet activation and incident PAD.

• Microvascular Disease

PAD has traditionally been recognized as a large artery disease. However, several recent studies have shown that microvascular disease may play an important role in the development and progression of PAD. For example, a US community-based cohort study with 9,371 adults has demonstrated that any retinopathy is associated with incident PAD and CLTI independently of other potential confounders such as diabetes and blood pressure.⁸⁵ Notably, this study has demonstrated that the HR according to any retinopathy was greater for CLTI and PAD (~3.4 and ~2.2, respectively) compared to stroke and CAD (<2). A similar pattern was shown for another measure of microvascular disease, albuminuria, in an international collaborative study including more than 800,000 individuals.⁸⁶ The robust association of retinopathy and nephropathy with lower extremity amputations was confirmed in US veterans.⁸⁷ Similarly, this study showed a stronger association between microvascular disease and amputation than major adverse cardiovascular events (MACE).

Biomarkers

Biomarkers have been widely used in cardiovascular medicine for disease diagnosis (e.g., high-sensitivity troponin for diagnosing acute myocardial infarction [MI]) and risk prediction (e.g., coronary artery calcium for guiding the intensity of lipid-lowering therapy). Although still exploratory, several potential biomarkers for PAD have been reported. For example, Wilson et al. conducted proteomic profiling in plasma samples from patients with PAD and persons without PAD and found that β2-microglobulin is uniquely elevated in the patients with PAD.⁸⁸ Another study also confirmed that β2-microglobulin is strongly and robustly associated with subsequent risk of PAD in the general population.⁸⁹ Also, biomarkers representing the pathophysiological pathways described above (e.g., hyperglycemia or inflammation) are likely to reflect the risk of PAD. Nonetheless, to our knowledge, to date, the most promising circulating biomarker of future PAD risk, especially its severe form, CLTI, is likely to be high-sensitivity cardiac troponin T (hs-cTnT). Specifically, a recent community-based study has found that individuals with hs-cTnT ≥14 ng/L had ~10-times (95% CI ~6 to ~15-times) higher risk of developing CLTI compared to those with hs-cTnT <3 ng/L after adjusting for various demographic and clinical covariates.⁹⁰ Corresponding relative risk was ~4 for hs-cTnT 9–13 ng/L, ~2.5 for 6–8 ng/L, and ~2 for 3–5 ng/L.

Complications

• Amputation

Although amputation is an important therapeutic option for severe cases of PAD, it remains a marker of poor cardiovascular outcomes. Among 186,338 Medicare recipients who underwent major lower extremity amputation from 2000–2008, mortality was 13.5% at 30 days, 48.3% at 1 year, and 70.9% at 3 years.⁹¹ Above-the-knee amputation was more strongly associated with death than below-the-knee amputation (HR 1.31; 95% CI, 1.25–1.36). There remains significant geographic variation in amputation rates for PAD across the U.S.,⁹² and both Black race and low socioeconomic status are associated with increased risk of amputation.⁹³

• Acute Limb Ischemia

Contemporary epidemiologic data on in the incidence of ALI are limited. Over a one year period (1980) in Sweden, there were 138 ALI cases out of a population of 1.5 million individuals.⁹⁴ Overall mortality was 19.5% among those with ALI. National healthcare data from Scotland in the 1970s and 1980s report the incidence of ALI was 1:27,000.⁹⁵ Mortality was stable at 30% during the period of study. Among individuals residing in Gloucestershire, England, in 1994, ALI incidence was 1:7,000; this figure rose to 1:6,000 when cases of acute bypass graft occlusion were included.⁹⁶ Of note, although these studies document acute thrombotic limb occlusion, they do not specify the presence or absence of PAD.

Importantly, recent clinical trials have begun reporting ALI outcomes. For instance, the EUCLID (Examining Use of Ticagrelor in PAD) trial of 13,885 individuals with symptomatic PAD reported 293 ALI events (0.8 per 100 patient-years).⁹⁷ 13% of those suffering ALI requiring major amputation. ALI was strongly associated with an increased risk of MACE (HR 1.4; 95% CI, 1.0–2.1; P=0.04), all-cause mortality (HR 3.3; 95% CI, 2.4–4.6; P<0.01), and major amputation (HR 34.2; 95% CI, 9.7–20.8; P<0.01). Similarly, among 3,787 subjects with symptomatic PAD in the Trial to Assess the Effects of Vorapaxar in Preventing Heart Attack and Stroke in Patients With Atherosclerosis–Thrombolysis in Myocardial Infarction 50 (TRA2°P-TIMI 50), ALI occurred in 108 individuals at a rate of 1.3%/year.⁹⁸

• Mortality

PAD is strongly linked with a heightened risk of death. According to the Centers for Disease Control and Prevention, in 2019 PAD was listed as the underlying cause of death in 11,753 individuals with 58,210 cases of any-mention deaths.^{99, 100} A meta-analysis of 48,294 subjects from 16 prospective cohorts found a strong association between ABI and mortality.¹⁰¹ The 10-year cardiovascular mortality for men with an ABI ≤0.90 compared to those with an ABI 1.11–1.40 was 18.7% vs. 4.4% (HR 2.9; 95% CI, 2.3–3.7) after adjusting for traditional atherosclerotic risk factors. In women, 10-year cardiovascular mortality was lower (12.6% vs. 4.1%) with a similar relative risk (HR 3.0; 95% CI, 2.0–4.4).¹⁰² Although older data suggested PAD-related mortality was primarily due to cardiovascular causes, recent data suggest many patients with PAD die from cancer-related causes.¹⁰²

Other CVDs

Since PAD is a subtype of atherosclerotic disease, it is not surprising that persons with PAD have a higher risk of other atherosclerotic diseases such as CAD and stroke.¹⁰³ Consequently, the AHA/ACC 2018 Cholesterol guideline acknowledges ABI ≤0.9 as a “risk enhancer” to be taken into account for guiding risk-based primary prevention beyond traditional atherosclerotic risk factors.¹²

This guideline takes into account symptomatic PAD (namely, a history of intermittent claudication with ABI <0.85 or lower extremity revascularization or amputation) as a type of major atherosclerotic CVD to classify patients with atherosclerotic CVD into very high-risk vs. high-risk for secondary prevention. Specifically, very high-risk is defined as a history of multiple major ASCVDs (namely, polyvascular disease) or one major ASCVD and multiple high-risk conditions such as older age or diabetes. The inclusion of symptomatic PAD in this classification is important since the Further Cardiovascular Outcomes Research with PCSK9 Inhibition in Subjects with Elevated Risk (FOURIER) trial¹⁰⁴ has demonstrated that the prognostic impact of symptomatic PAD is similar or even larger than that of MI or stroke. This observation highlights the importance of the concept of polyvascular disease.

Several studies have also demonstrated that individuals with PAD are at elevated risk of heart failure (HF).^105–107 For example, a report from ARIC, including 13,150 middle-aged US adults, showed that ABI ≤1.0 (vs. 1.01–1.40) conferred ~40% elevated risk of HF after accounting for potential confounders such as diabetes, smoking, and the presence of CAD or carotid atherosclerosis.¹⁰⁶ A recent report from the Multi-Ethnic Study of Atherosclerosis observed that the elevated risk related to low ABI was more evident for HF with reduced ejection than HF with preserved ejection fraction.¹⁰⁷ These observations will have important clinical implications given a body of evidence showing reduced risk of HF by SGLT2 inhibitors regardless of diabetic status.

• Physical function and quality of life

A number of studies have shown that persons with PAD, regardless of the presence/absence or types of leg symptoms, have reduced physical function.^108–112 For example, one of the largest studies using data from ARIC showed that persons with ABI ≤0.9 (PAD) and those with ABI 0.9–1.0 (borderline low ABI) had 2-times higher odds of having impaired physical function defined as a reduced score of a validated instrument, the Short Physical Performance Battery (Figure 2) after accounting for various potential confounders such as a history of stroke or HF.

Figure 2. Adjusted odds ratio of reduced physical function* according to categories of ABI.

* Defined as the Short Physical Performance Battery score ≤5 and adjusted for age, race, sex, education, smoking, alcohol, body mass index, antihypertensive medications, systolic blood pressure, lipids, lipid-lowering medications, history of coronary artery disease, stroke, or heart failure.

ABI refers to ankle-brachial index.

Several studies have also investigated the quality of life (QOL) related to PAD. In those studies, PAD is consistently associated with a reduction of the physical component of QOL (e.g., body pain or physical activity) regardless of age or geography of study populations.^113–116 In contrast, for the mental component of QOL, most studies showed no association or an association with a limited domain (e.g., vitality^{113, 115}).

• Cost

Several studies have reported that PAD considerably increases medical expenditure. For example, a study using data from the Agency for Healthcare Research and Quality Medical Expenditure Panel Surveys has estimated that the average annual expenditure per individual was $11,553 (95%CI $8,137-$14,968) for patients with PAD vs. $4,219 (95% CI, $4,064-$4,375) for those without.¹¹⁷ Increased prescription medication, inpatient care, and outpatient care all contributed to the higher medical expenditure in patients with PAD. The medical expenditure is particularly high in patients with severe PAD requiring major amputation. For example, a study using data from Medicare estimated that the annual average medical expenditure was $55,700 per patient after the procedure.¹¹⁸

Polyvascular Disease

Historically, patients with PAD were frequently recruited to clinical trials of atherosclerotic therapies because of their high risk for cardiovascular outcomes. However, outcomes were rarely reported in subgroups of individuals with PAD alone or polyvascular disease, and clinical guidelines recommended near uniform treatment of atherosclerotic risk factors regardless of the affected arterial bed. This began to change when early insights from clinical registries identified a greater risk for polyvascular disease.^{119, 120} Polyvascular disease is also an ideal opportunity for personalized medicine, as the potential tradeoffs of any given therapy, particularly novel antiplatet and antithrombotic regimens with higher bleeding risk, may be more acceptable in this patient population (Figure 3).

Figure 3. Cardiovascular Risks of Polyvascular Disease.

Atherosclerotic disease in a given vascular bed is directly linked to adverse outcomes in that same organ. Because polyvascular disease is indicative of systemic atherosclerosis, individuals with polyvascular disease are at heightened risk for cardiovascular events in all vascular territories. (Illustration credit: Ben Smith).

Few studies have examined risk factors associated with developing polyvascular disease. Using 2-sample Mendelian randomization in individuals with PAD as well as CAD, cerebrovascular disease, or abdominal aortic aneurysm, investigators sought to determine whether a causal link existed between atherogenic lipoproteins, blood pressure, glycated hemoglobin, and smoking.¹²¹ Genetically-predicted levels of small, dense LDL particles as well as systolic blood pressure were most strongly associated with polyvascular disease, although the analysis was limited by small sample size (n=2009).^{65, 122} Genomic analyses have demonstrated that variants in the chromosome 9p21 locus are associated with not only CAD and cerebrovascular disease but also a diagnosis of PAD⁶⁵ and low ABI measures.¹²²

Several clinical studies have documented an increased cardiovascular risk associated with polyvascular disease (Table 3). For instance, among 68,236 patients with a history of atherosclerotic disease or ≥3 atherosclerotic risk factors in the Reduction of Atherothrombosis for Continued Health (REACH) Registry, polyvascular disease was independently associated with an increased risk of MACE (HR 1.99; 95% CI, 1.78–2.24; P<0.001) and was a stronger risk marker than diabetes or a history of ischemic event in the prior year. In FOURIER, patients with polyvascular PAD were at highest risk of MACE among subjects in the placebo arm: MI or stroke only (Kaplan-Meier incidence rate at 30 months 7.6%), PAD only (10.3%), PAD + MI (14.9%).¹⁰⁴ Subsequent studies have shown that diabetes further amplifies risk of cardiovascular events among patients with polyvascular disease.^123–125

Table 3.

Cardiovascular Outcomes from Clinical Studies of Polyvascular Disease

Data Source	Study Type	Qualifying Vascular Territories	Patients with PAD/Total Sample (%)	Definition of MACEs	Impact on MACEs	Definition of MALEs	Impact on MALEs
CRUSADE131	Observational Registry	CAD – A or S PAD – S CVD – S	11,345/111,972 (10.1)	In-hospital mortality, MI, stroke, and HF	1 bed: OR 1.07 (95% CI, 1.02–1.12) 2 beds: OR 1.25 (95% CI, 1.18–1.33) 3 beds: OR 1.31 (95% CI, 1.15–1.46)	--	--
REACH120	Observational Registry	CAD – S PAD – S CVD – S	5,872/68,236 (8.6)	MI, stroke, or cardiovascular mortality	Polyvascular disease vs. atherosclerotic risk factors: HR 1.99 (95% CI, 1.78–2.24)	--	--
MarketScan + Medicare126	Observational Claims Data	CAD – A-S PAD – A-S CVD – A-S	156,129/943,232 (16.6)	MI, coronary revascularization, stroke, carotid revascularization, ALI, lower extremity revascularization, or amputation	2 vs. 1 bed: HR 1.45 (95% CI, 1.42–1.48) 3 vs. 1 bed: HR 2.08 (95% CI, 2.03–2.14)	ALI, lower extremity revascularization, or amputation	2 vs. 1 bed: HR 2.18 (95% CI, 2.02–2.36) 3 vs. 1 bed: HR 3.91 (95% CI, 3.58–4.28)
CAPRIE132	Randomized Controlled Trial	CAD – S PAD – S CVD – S	6,452/19,185 (33.6)	MI, stroke, or cardiovascular mortality	Polyvascular vs. single bed: RRR 12.4%; ARR 2.45%; NNT 41	--	--
TRA 2°P-TIMI 50133	Randomized Controlled Trial	CAD – S PAD – S CVD – S	3,787/26,449 (14.3)	MI, stroke, or cardiovascular mortality at 3 years	1 bed: 7.8% 2 beds: 14.7% 3 beds: 21.7%	Lower extremity revascularization	1 bed: 7.8% 2 beds: 14.7% 3 beds: 21.7%
FOURIER104	Randomized Controlled Trial	CAD – S PAD – S CVD – S	3,642/27,564 (13.2)	MI, stroke, or cardiovascular mortality	Kaplan-Meier incidence rate at 30 months PAD + MI or stroke: 14.9% PAD only: 10.3% MI or stroke only: 7.6%	--	--
PEGASUS-TIMI 54134	Randomized Controlled Trial	CAD – S PAD – A or S	1,143/21,162 (5.4)	MI, stroke, or cardiovascular mortality	PAD + CAD vs. PAD: HR 1.60 (95% CI, 1.20–2.13)	--	--
EUCLID135	Randomized Controlled Trial	CAD – S PAD – A or S CVD – S	13,885/13,885 (100.0)	MI, stroke, or cardiovascular mortality	PAD + cerebrovascular disease: HR 1.34 (95% CI, 1.15–1.57) PAD + CAD: HR 1.65 (95% CI, 1.43–1.91) PAD + CAD + cerebrovascular disease: HR 1.99 (95% CI, 1.69–2.34)	Lower extremity revascularization	PAD + CVD: HR 1.17 (95% CI, 1.02–1.35) PAD + CAD: HR 1.17 (95% CI, 1.03–1.34) PAD + CAD + CVD: HR 1.34 (95% CI, 1.15–1.57)
SAVOR-TIMI 53123	Randomized Controlled Trial	CAD – S PAD – S CVD – S	1,959/16,492 (11.9)	MI, stroke, or cardiovascular mortality	1 bed: HR 1.95 (95% CI, 1.63–2.35) 2 beds: HR 3.54 (95% CI, 2.84–4.43) 3 beds: HR 4.64 (95% CI, 2.87–7.15)	--	--
LEADER124	Randomized Controlled Trial	CAD – A,S PAD – S CVD – A or S	1,184/9,340 (12.7)	MI, stroke, or cardiovascular mortality	Polyvascular vs. single bed: HR 1.52 (95% CI, 1.33–1.73)	--	--
IMPROVE-IT125	Randomized Controlled Trial	CAD – S PAD – A or S CVD – S	1,005/18,144 (5.5)	MI, stroke, or cardiovascular mortality	Polyvascular vs. single bed: HR 1.55 (95% CI, 1.41–1.70)	Lower extremity revascularization	Polyvascular vs. single bed: HR 4.77 (95% CI, 3.75–6.06)
						Amputation	Polyvascular vs. single bed: HR 5.96 (95% CI, 2.74–12.94)

A refers to asymptomatic; A-S, symptoms not assessed in qualifying events; ALI, acute limb ischemia; ARR, absolute risk reduction; CAD, coronary artery disease; CAPRIE, Clopidogrel vs. Aspirin in Patients at Risk of Ischemic Events; CI, confidence interval; COMPASS, Cardiovascular Outcomes for People Using Anticoagulation Strategies; CRUSADE, Can Rapid Risk Stratification of Unstable Angina Patients Suppress Adverse Outcomes with Early Implementation of the ACC/AHA Guidelines; CVD, cerebrovascular disease; EUCLID, Effects of Ticagrelor and Clopidogrel in Patients with Peripheral Artery Disease; FOURIER, Further Cardiovascular Outcomes Research with PCSK9 Inhibition in Subjects with Elevated Risk; HF, heart failure; HR, hazard ratio; IMPROVE-IT, Improved Reduction of Outcomes: Vytorin Efficacy International Trial; LEADER, Liraglutide Effect and Action in Diabetes: Evaluation of Cardiovascular Results; MACE, major adverse cardiovascular events; MALE, major adverse limb events; MI, myocardial infarction; OR, odds ratio; PAD refers to peripheral artery disease; PEGASUS-TIMI 54, Patients With Prior Heart Attack Using Ticagrelor Compared to Placebo on a Background of Aspirin; REACH, Reduction of Atherothrombosis for Continued Health; RRR, relative risk reduction; S, symptomatic; SAVOR-TIMI 53, Saxagliptin Assessment of Vascular Outcomes Recorded in Patients with Diabetes Mellitus-TIMI 53; TRA2°P-TIMI 50=Trial to Assess the Effects of Vorapaxar in Preventing Heart Attack and Stroke in Patients With Atherosclerosis-TIMI 50

Recently, an analysis of 943,232 individuals from the MarketScan and Medicare databases demonstrated an increased incidence of atherosclerotic events (MI, coronary revascularization, stroke, carotid revascularization, ALI, peripheral artery revascularization, or major amputation) among patients with polyvascular disease: 1 bed (event rate per 1,000 person-years 40.8; 95% CI, 33.2–36.2), 2 bed (68.9; 95% CI, 67.9–70.0), 3 bed (119.5; 95% CI, 117.0–122.0).¹²⁶ PAD had an equivalent prognostic impact, if not greater, than CAD or cerebrovascular disease: PAD + CAD (event rate of atherosclerotic cardiovascular disease per 1,000 person-years 72.8; 95% CI, 71.0–74.7), PAD + cerebrovascular disease (63.9; 95% CI, 60.6–67.4), CAD + cerebrovascular disease (67.9; 95% CI, 66.4–69.3). Additionally, although atherosclerotic disease in a given bed was most strongly associated with its respective endpoint (e.g. CAD with MI, cerebrovascular disease with stroke, and PAD with limb events), all atherosclerotic endpoints were associated with polyvascular disease.

Future Directions

PAD has received greater attention in recent years, both in epidemiologic research and cardiovascular outcome trials, and we expect this will further our understanding of PAD development in the near future. Already, large cohorts such as the UK Biobank and the Million Veterans Program have yielded important insights into the genomic basis of PAD (detailed elsewhere in this issue).⁶⁵ Transcriptomic analysis of patients with PAD undergoing lower extremity revascularization can identify markers of future adverse events, and this technology potentially has the ability to identify risk markers of PAD subtypes, such as macrovascular versus microvascular disease as well as CLTI.¹²⁷ Finally, recent randomized, placebo-controlled trials have demonstrated a benefit of anti-inflammatory therapy in reducing cardiovascular events,^{128, 129} and preliminary data suggest such a strategy may also be effective in PAD.¹³⁰ We hope future trials will help advance our understanding of the role inflammation plays in PAD and polyvascular disease development.

Conclusions

PAD is associated with exertional limb symptoms, functional decline, and poor quality of life, and individuals with PAD at heightened risk of lower extremity ulceration, amputation, and death. The global prevalence of PAD continues to grow in parallel with the expanding burden of atherosclerotic risk factors and an aging population. Although many risk factors of PAD are shared with other cardiovascular diseases, some risk factors, such as smoking and dyslipidemia, have a differential impact on PAD risk compared to that of CAD. Increased focus on PAD, particularly in the setting of polyvascular disease, has furthered our understanding of atherosclerotic risk, and this paradigm is increasingly used in the design of cardiovascular therapeutic trials. We feel this focus is long overdue and is necessary to help us increase PAD awareness, detection, and management in the future.

Non-standard Abbreviations and Acronyms

ABI

ankle-brachial index

ALI

acute limb ischemia

ARIC

Atherosclerosis Risk in Communities Study

BMI

body mass index

CAD

coronary artery disease

CI

confidence interval

CLTI

chronic limb-threatening ischemia

EUCLID

Examining Use of Ticagrelor in Peripheral Artery Disease

FOURIER

Further Cardiovascular Outcomes Research with PCSK9 Inhibition in Subjects with Elevated Risk

HDL-C

high-density lipoprotein cholesterol

HF

heart failure

HR

hazard ratio

IC

intermittent claudication

LDL-C

low-density lipoprotein cholesterol

MACE

major adverse cardiovascular events

MI

myocardial infarction

NIS

Nationwide Inpatient Sample

OR

odds ratio

PAD

peripheral artery disease

PEGASUS

Prevention of Cardiovascular Events in Patients With Prior Heart Attack Using Ticagrelor Compared to Placebo on a Background of Aspirin

PHS

Physicians’ Health Study

REACH

Reduction of Atherothrombosis for Continued Health

RR

relative risk

WHS

Women’s Health Study