Peripheral artery disease and risk of kidney outcomes: The Atherosclerosis Risk in Communities (ARIC) study

Abstract

Background and aims:

The potential impact of peripheral artery disease (PAD) on kidney outcomes is not well understood. The aim of this study was to explore the association between PAD and end-stage kidney disease (ESKD) and chronic kidney disease (CKD).

Methods:

Among 14,051 participants (mean age 54 [SD 6 years]) from the Atherosclerosis Risk in Communities study, we categorized PAD status as symptomatic PAD (intermittent claudication or leg revascularization), asymptomatic PAD (ankle-brachial index [ABI] ≤0.90 without clinical history of symptoms), and ABI 0.91–1.00, 1.01–1.10, 1.11–1.20 (reference), 1.21–1.30, and >1.30. We evaluated their associations with two kidney outcomes: ESKD (the need of renal replacement therapy or death due to kidney disease) and CKD (ESKD cases or an estimated glomerular filtration rate (eGFR) <60 mL/min/1.73m² with a ≥25% decline from the baseline) using multivariable Cox proportional hazards models.

Results:

Over ~30 years of follow-up, there were 598 cases of incident ESKD and 4,686 cases of incident CKD. After adjusting for potential confounders, both symptomatic PAD and asymptomatic PAD conferred a significantly elevated risk of ESKD (hazard ratio 2.28 [95% confidence interval 1.23–4.22] and 1.75 [1.19–2.57], respectively). Corresponding estimates for CKD were 1.54 (1.14–2.09) and 1.63 (1.38–1.93). Borderline low ABI 0.91–1.00 also showed elevated risk of adverse kidney outcomes after adjustment for demographic variables. Largely consistent results were observed across demographic and clinical subgroups.

Conclusions:

Symptomatic PAD and asymptomatic PAD were independently associated with an elevated risk of ESKD and CKD. These results highlight the importance of monitoring kidney function in persons with PAD, even when symptoms are absent.

1. INTRODUCTION

Peripheral artery disease (PAD) is a prevalent clinical condition that affects over 230 million individuals worldwide.^{1, 2} PAD is associated with a multitude of adverse outcomes, including cardiovascular morbidity, leg amputation, and infection.^{1, 3–5} According to the Global Burden of Disease, from 2007 to 2017, the number of deaths as a result of PAD has increased by 55.7%.⁶ Despite the severity of these outcomes, over 50% of those with PAD are asymptomatic, coinciding with the high rates of underdiagnosis and potentially missed opportunities to receive any relevant preventive therapies.^{7, 8} Thus, clinical guidelines recommend screening high-risk individuals for PAD using ankle-brachial index (ABI), the primary non-invasive diagnostic test of PAD.⁴ This recommendation is important since the literature has shown that patients with coronary heart disease but without prior diagnosis of PAD frequently experience intermittent claudication.⁹

The bi-directional relationship between cardiovascular disease and kidney outcomes is widely recognized as cardiorenal syndrome.¹⁰ However, studies that explore PAD and subsequent risk of kidney outcomes are actually limited. For example, a clinical study of PAD patients found that lower ABI is an independent risk factor for the development of end-stage kidney disease (ESKD).¹¹ A few community-based studies have also found that lower ABI is associated with a higher risk of surrogate kidney endpoints such as estimated glomerular filtration rate (eGFR) decline and an increase in serum creatinine levels.¹²⁻¹⁴ To our knowledge, no studies have quantified the association between PAD and incident ESKD in the community. ESKD is an important outcome considering both the severity of the disease and its impact on medical expenditure.¹⁵

Therefore, the aim of this study was to assess the association of PAD with incident ESKD as our primary outcome and incident CKD as a secondary outcome in a community-based cohort, the Atherosclerosis Risk in Communities (ARIC) study. Since clinical guidelines such as the American Heart Association/American College of Cardiology (AHA/ACC) lipid guidelines have different recommendations regarding asymptomatic and symptomatic PAD, we were interested in seeing if there is a difference in kidney outcomes between these two types of PAD.¹⁶ We were additionally interested in examining the full range of ABI values among those without PAD regarding kidney prognosis.

2. PATIENTS AND METHODS

2.1. Study participants

The ARIC study is comprised of 15,792 participants aged 45–64 years from four communities in the United States: Washington County, Maryland; Forsyth County, North Carolina; suburbs of Minneapolis, Minnesota, and Jackson, Mississippi.¹⁷ Of 15,792 participants, in the present study, we excluded those who self-identified as non-White or non-Black and the Black participants from the Minnesota and Washington County cohorts due to small numbers (n = 103) and those who had missing data on ABI (n = 570) and covariates (n = 1048). For our primary outcome, incident ESKD, we also excluded participants with an eGFR <15 ml/min/1.73m² or had missing data on the outcome variable (n = 20). We additionally excluded those with eGFR <60 ml/min/1.73m² (n = 164) for our secondary outcome of incident CKD. The final sample consisted of 14,051 participants for the ESKD analysis and 13,887 for the CKD analysis (Supplemental Figure 1). The institutional review boards at each field location approved the ARIC Study protocol and each participant provided informed consent.

2.2. ABI measurement and PAD definitions

ABI measurements were taken in a single leg using the standardized ARIC protocol, which has been described elsewhere.¹⁸ In brief, the Dinamap Model 1846 SX, an automated oscillometric device was used for the blood pressure measurements (Criticon, Tampa, FL). The ankle that the measurement was taken in was randomly selected.¹⁸ The blood pressure cuff was wrapped around the participant’s selected ankle while lying in the prone position. After calibration, two measurements of the ankle blood pressure were taken with 5–8 minutes in between. Brachial blood pressure was assessed three times in the supine position with a 5-minute rest between measurements. An average of the last two measurements was used to compute the final value. To calculate ABI, the average of the two ankle blood pressure measurements was divided by the average of the brachial blood pressure measurements. The reliability of ABI based on this method was found to be 0.7.^{19, 20}

The definition of symptomatic PAD included history of leg revascularization or intermittent claudication. For leg revascularization, participants self-reported whether they had a balloon angioplasty of the lower extremity arteries. Intermittent claudication was defined by leg pain while walking that was relieved within 10 minutes of resting according to the Rose questionnaire.^{21, 22} Asymptomatic PAD was considered an ABI ≤0.90 without symptomatic PAD. All remaining participants were considered not having PAD and classified by ABI values.

2.3. Kidney outcomes

Incident ESKD was determined based on cases requiring renal replacement therapy (i.e., dialysis or kidney transplantation) through the linkage to the United States Renal Data System (USRDS) national registry,²³ or death with kidney disease listed as the first position on the death certificate (International Classification of Diseases, Ninth Revision [ICD-9] and corresponding Tenth Revision [ICD-10] codes listed in Supplemental Table 1). USRDS linkage was current until July 2017. Incident CKD was defined as any ESKD events described above or a newly identified eGFR <60 mL/min/1.73m² at any subsequent visits with a 25% or greater decline from the baseline visit. eGFR was assessed based on the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) creatinine equation.²⁴

2.4. Covariates

Age, race, sex, and education level were collected based on self-report.²⁵ Smoking and drinking status were reported by the participant via questionnaires. Body mass index (BMI) was computed by dividing the participant’s weight (kg) by the square of their height (m). Blood pressure was measured by certified technicians three times after 5-minute rest while the participant was in the seated position. The final two measurements were averaged to obtain the reported results.²⁶ Participants were asked to bring in their medications, which were recorded by trained staff.²⁷ Total and high-density lipoprotein (HDL) cholesterol levels were assessed using enzymatic methods in accordance with the National Cholesterol Education Program guidelines.²⁸ eGFR was assessed via the CKD-EPI creatinine equation. Diabetes mellitus was assessed based on a prior self-reported physician diagnosis, fasting glucose of ≥126 mg/dL, nonfasting glucose ≥200 mg/dL, or medication use for diabetes.²⁵ Prevalent stroke and coronary heart disease were based on self-report.²⁹ Prevalent heart failure was determined based on use of medication and the Gothenburg criteria.²⁹

2.5. Statistical analysis

We divided our cohort into PAD/ABI categories: symptomatic PAD, asymptomatic PAD, and other ABI categories of 0.91–1.00, 1.01–1.10, 1.11–1.20, 1.21–1.30, and >1.30.^{30, 31} We presented summary statistics of the participants as mean (SD) or as count (percentage) for continuous and categorical variables, respectively. To assess the cumulative incidence of ESKD and CKD by PAD/ABI categories, we used the Kaplan-Meier method.

We used Cox proportional hazards models to evaluate the association of PAD/ABI categories with subsequent risk of ESKD and CKD after accounting for potential confounders. Model 1 adjusted for demographic variables, including age, sex, race, and study site. Model 2 additionally accounted for education level, total cholesterol, HDL cholesterol, eGFR, systolic blood pressure, antihypertensive drugs, BMI, drinking status, smoking status, diabetes, prevalent heart failure, prevalent stroke, and prevalent coronary heart disease. The ABI 1.11–1.20 category was used as the reference because it was the most prevalent in our study and was used as the reference in prior studies.³⁰ To visualize the association of ABI with ESKD and CKD on a continuous scale, we modeled ABI using a restricted cubic spline, which contained knots at the 5, 35, 65, and 95 percentiles of ABI,³² after excluding participants with symptomatic PAD. The reference ABI in these continuous analyses was 1.15, the midpoint of the 1.11–1.20 category.

In sensitivity analyses, we divided the study population into subgroups based on age, sex, race, smoking status, diabetes, hypertension, and prevalent coronary heart disease. Using Wald tests for interaction, we determined whether there were significant differences in the association of PAD status with incident ESKD or CKD across each subgroup. To obtain reliable estimates in each subgroup, we modeled PAD status as a dichotomous variable of participants with symptomatic or asymptomatic PAD vs. all remaining participants. The threshold for statistical significance was a two-sided p-value of less than 0.05. R version 3.6.1 was used to conduct all analyses (R Foundation for Statistical Computing, Vienna, Austria).

3. RESULTS

3.1. Baseline characteristics

The 14,051 participants had a mean age of 54.1 (SD 5.8) years and were 74% White (Table 1). There were 118 participants (0.8%) with symptomatic PAD, 397 (2.8%) with asymptomatic PAD, and 1420 (10.1%) with an ABI >1.30. Both those with symptomatic and asymptomatic PAD were more likely to have diabetes, hypertension, prevalent coronary heart disease, and prevalent heart failure compared to the reference ABI of 1.11–1.20. Those with ABI >1.30 were more likely to be White and male.

Table 1:

Participant characteristics by symptomatic PAD/ABI categories

		PAD/ABI Categories
	Overall	Symptomatic PAD	Asymptomatic PAD	ABI 0.91–1.00	ABI 1.01–1.10	ABI 1.11–1.20	ABI 1.21–1.30	ABI >1.30
n	14051	118	397	1212	3255	4478	3171	1420
Age, years	54.1 (5.8)	56.6 (5.6)	55.8 (5.7)	53.7 (5.8)	53.7 (5.7)	53.9 (5.7)	54.3 (5.8)	55.2 (5.7)
White (%)	10397 (74.0)	94 (79.7)	271 (68.3)	892 (73.6)	2362 (72.6)	3290 (73.5)	2381 (75.1)	1107 (78.0)
Male (%)	6296 (44.8)	68 (57.6)	137 (34.5)	293 (24.2)	1113 (34.2)	2057 (45.9)	1757 (55.4)	871 (61.3)
Education Level (%)
Completed Less Than High School	3227 (23.0)	37 (31.4)	144 (36.3)	334 (27.6)	747 (22.9)	1024 (22.9)	649 (20.5)	292 (20.6)
Completed High School or Equivalent	5772 (41.1)	54 (45.8)	153 (38.5)	516 (42.6)	1392 (42.8)	1845 (41.2)	1229 (38.8)	583 (41.1)
At Least Some College	5052 (36.0)	27 (22.9)	100 (25.2)	362 (29.9)	1116 (34.3)	1609 (35.9)	1293 (40.8)	545 (38.4)
BMI, kg/m2	27.6 (5.3)	28.1 (5.2)	27.9 (6.3)	28.3 (6.2)	27.6 (5.6)	27.4 (5.0)	27.4 (4.9)	28.1 (5.4)
Systolic Blood Pressure, mmHg	121.0 (18.7)	124.8 (22.5)	125.9 (22.7)	122.8 (20.7)	122.1 (20.0)	120.6 (18.4)	119.8 (17.1)	119.7 (16.3)
Current Smoker (%)	3648 (26.0)	53 (44.9)	192 (48.4)	386 (31.8)	892 (27.4)	1119 (25.0)	718 (22.6)	288 (20.3)
Current Drinker (%)	7930 (56.4)	73 (61.9)	184 (46.3)	660 (54.5)	1826 (56.1)	2534 (56.6)	1851 (58.4)	802 (56.5)
HDL, mg/dL	51.8 (17.2)	45.4 (15.2)	49.9 (17.1)	53.1 (17.1)	53.5 (17.6)	51.9 (17.4)	50.6 (16.7)	49.7 (16.4)
Total Cholesterol, mg/dL	214.7 (41.8)	227.3 (48.7)	225.6 (48.6)	218.0 (42.5)	216.4 (42.3)	213.8 (41.4)	212.0 (40.2)	212.2 (41.0)
eGFR	102.5 (15.3)	97.1 (18.7)	101.1 (19.4)	103.2 (16.0)	103.3 (15.1)	102.6 (15.5)	102.4 (14.5)	100.9 (14.8)
Diabetes (%)	1629 (11.6)	23 (19.5)	90 (22.7)	156 (12.9)	383 (11.8)	497 (11.1)	317 (10.0)	163 (11.5)
Hypertension (%)	4062 (28.9)	51 (43.2)	163 (41.1)	393 (32.4)	1003 (30.8)	1275 (28.5)	800 (25.2)	377 (26.5)
Antihypertensive Medication (%)	3510 (25.0)	47 (39.8)	156 (39.3)	340 (28.1)	876 (26.9)	1093 (24.4)	666 (21.0)	332 (23.4)
Prevalent Coronary Heart Disease (%)	676 (4.8)	27 (22.9)	44 (11.1)	70 (5.8)	135 (4.1)	179 (4.0)	143 (4.5)	78 (5.5)
Prevalent Heart Failure (%)	642 (4.6)	13 (11.0)	36 (9.1)	79 (6.5)	175 (5.4)	178 (4.0)	114 (3.6)	47 (3.3)
Stroke (%)	253 (1.8)	<11 (<11)	<11 (<11)	22 (1.8)	62 (1.9)	90 (2.0)	42 (1.3)	23 (1.6)

^aABI = ankle-brachial index, PAD = peripheral artery disease, BMI = body mass index, HDL = high-density lipoprotein, eGFR = estimated glomerular filtration rate

3.2. Risk of ESKD and CKD

There were 598 events of ESKD and 4686 events of CKD over a median follow-up time of 27.6 and 24.0 years (max 32.1 years for both), respectively. For both outcomes, the highest cumulative incidence was among participants with PAD. At 25 years, the cumulative incidence for ESKD was 10.2% for symptomatic PAD and 11.7% for asymptomatic PAD whereas the corresponding cumulative incidence for the other groups ranged from 3.9% to 5.3% (Figure 1A). The cumulative incidences of CKD were 53.8% for symptomatic PAD, 53.6% for asymptomatic PAD, and ~32% in the other groups (Figure 1B).

Figure 1:

Cumulative incidence of (A) ESKD and (B) CKD.

Figure illustrates the cumulative incidence of ESKD and CKD by symptomatic PAD status and ABI category over 30 years. Participants with PAD, regardless of symptoms, had the highest cumulative incidence over time.

Abbreviations: ABI = ankle-brachial index, CKD = chronic kidney disease, ESKD = end-stage kidney disease, PAD = peripheral artery disease

After adjusting for demographic covariates, the hazard ratio (HR) of ESKD was 3.66 (95%CI 1.99–6.74) in symptomatic PAD and 3.00 (2.07–4.37) in asymptomatic PAD compared to the reference ABI of 1.11–1.20 (Model 1 in upper half of Table 2). Borderline low ABI also had a significant HR of 1.49 (95%CI 1.11–2.01). The results remained statistically significant in the symptomatic and asymptomatic PAD groups after adjusting for additional clinical covariates, with HRs of 2.28 (1.23–4.22) and 1.75 (1.19–2.57), respectively (Model 2 in upper half of Table 2). There was no association between high ABI (>1.30) and ESKD outcomes in either model. The results were largely consistent when we investigated incident CKD as an outcome (lower half of Table 2), although the strength of the associations was slightly weaker than the ESRD analysis. We confirmed the overall dose-response relationship between lower ABI and both kidney disease outcomes when ABI was modeled on a continuous scale in participants without symptomatic PAD (Figure 2).

Table 2:

Incidence rate and adjusted hazard ratios of CKD and ESKD according to PAD/ABI categories

End-Stage Kidney Disease
	Cases/N	Model 1	Model 2
Symptomatic PAD	11/118	3.66 (1.99–6.74)	2.28 (1.23–4.22)
Asymptomatic PAD	33/397	3.00 (2.07–4.37)	1.75 (1.19–2.57)
ABI 0.91–1.00	59/1212	1.49 (1.11–2.01)	1.10 (0.82–1.48)
ABI 1.01–1.10	132/3255	1.11 (0.89–1.40)	1.03 (0.82–1.30)
ABI 1.11–1.20	179/4478	Ref.	Ref.
ABI 1.21–1.30	121/3171	0.90 (0.71–1.13)	1.02 (0.81–1.29)
ABI >1.30	63/1420	0.99 (0.74–1.32)	1.10 (0.82–1.47)
Chronic Kidney Disease
	Cases/N	Model 1	Model 2
Symptomatic PAD	44/112	2.06 (1.53–2.79)	1.54 (1.14–2.09)
Asymptomatic PAD	156/380	2.09 (1.77–2.47)	1.63 (1.38–1.93)
ABI 0.91–1.00	398/1198	1.21 (1.08–1.35)	1.05 (0.94–1.17)
ABI 1.01–1.10	1071/3222	1.06 (0.98–1.15)	1.02 (0.94–1.11)
ABI 1.11–1.20	1502/4426	Ref.	Ref.
ABI 1.21–1.30	1042/3143	0.92 (0.85–1.00)	0.96 (0.89–1.04)
ABI >1.30	473/1406	0.89 (0.80–0.98)	0.90 (0.81–1.00)

Model 1 – Adjusted for sex, age, race, and study site

Model 2 – Adjusted for sex, age, race, study site, education level, body mass index, total cholesterol, high-density lipoprotein cholesterol, estimated glomerular filtration rate, systolic blood pressure, antihypertensive medication, drinking status, smoking status, diabetes, stroke, prevalent coronary heart disease, and prevalent heart failure

^aBolded values represent significant hazard ratios

PAD = peripheral artery disease, ABI = ankle-brachial index

Figure 2:

Adjusted hazard ratio of (A) ESKD and (B) CKD.

Using cubic splines, figure shows the association between ABI and both kidney outcomes in participants without symptomatic PAD. There was a dose-response relationship.

Hazard ratios in reference to ABI of 1.15 and adjusted for sex, age, race, study site, education level, body mass index, total cholesterol, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, diastolic blood pressure, systolic blood pressure, antihypertensive medication, drinking status, smoking status, diabetes, stroke, prevalent coronary heart disease, and prevalent heart failure. Restricted cubic spline created with knots at the 5, 35, 65, and 95 percentiles of ABI.

Includes only participants without symptomatic peripheral artery disease

Abbreviations same as Figure 1.

3.3. Sensitivity analyses

Among age, sex, race, smoking status, diabetes, hypertension, and prevalent coronary heart disease subgroups, the association between PAD status (symptomatic and asymptomatic PAD combined vs. no PAD) and incident ESKD was largely consistent, with the only significant interaction by smoking status (p-for-interaction 0.03) (Figure 3A). The results were generally similar for incident CKD (Figure 3B), although there was a significant interaction by sex, with a stronger association among males compared to females (p-for-interaction 0.01).

Figure 3:

Hazard ratios of (A) ESKD and (B) CKD.

Subgroup analyses highlight that the association between PAD status and both ESKD and CKD were largely consistent. However, for CKD there was a stronger association among males compared to females.

Adjusted for sex, age, race, study site, education level, body mass index, total cholesterol, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, diastolic blood pressure, systolic blood pressure, antihypertensive medication, drinking status, smoking status, diabetes, stroke, prevalent coronary heart disease, and prevalent heart failure, PAD

ABI = ankle-brachial index, PAD = peripheral artery disease, CI = confidence interval

PAD defined as symptomatic PAD or asymptomatic PAD (ABI ≤ 0.90), No PAD defined as all other ABI values

We observed largely similar results for ESKD and CKD when we used ABI 1.21–1.30 as the referent (Supplemental Table 2).

4. DISCUSSION

In the present study of community-dwelling adults, we found that both symptomatic and asymptomatic PAD was associated with a significantly elevated risk of both ESKD and CKD over ~30 years of follow-up, independent of potential confounders such as diabetes, hypertension, and other cardiovascular diseases. Adjusted HRs of incident ESKD and CKD were 1.5–2 for both symptomatic PAD and asymptomatic PAD (Graphical Abstract). 25-year cumulative incidence for ESKD was ~10% in symptomatic PAD and asymptomatic PAD whereas the corresponding cumulative incidence for those without PAD was 4–5%. These results were generally consistent in major subgroups by demographics (e.g., sex and age) and clinical conditions (e.g., diabetes status). In some analyses, borderline low ABI 0.91–1.00, but not high ABI >1.30, demonstrated an elevated risk of ESKD and CKD.

Our study is largely consistent with a limited number of previous studies,^{11, 13, 14} but is unique in various aspects. Specifically, to the best of our knowledge, this is the first prospective community-based cohort study examining the association between ABI and future risk of ESKD, a hard kidney endpoint with significant implications on patients and medical expenditure.³³ We also uniquely investigated both symptomatic PAD and asymptomatic PAD. Furthermore, we demonstrated the long-term prognostic value of ABI for evaluating the risk of kidney outcomes.

PAD and kidney disease both share similar risk factors, including diabetes and hypertension. However, the association of PAD with incident kidney outcomes was independent of these clinical conditions, suggesting pathophysiological mechanisms linking PAD to kidney disease. For example, as recognized in polyvascular disease, atherosclerosis can affect multiple arterial beds, including the renal artery.³⁴ Indeed, renal artery stenosis can limit blood flow to the kidneys³⁵ and induce ischemic nephropathy.³⁶ Inflammation may also play a role. More specifically, inflammation is known to be a key mediator in the progression of atherosclerosis³⁷ and similarly is shown to cause fibrotic changes in the kidneys.³⁸ Nonetheless, future studies are still required to explore exact mechanisms behind elevated risk of kidney outcomes in persons with PAD. The identification of those mechanisms may guide therapeutic options for preventing CKD progression in individuals with atherosclerotic diseases, including PAD.

Our findings suggest that ESKD and CKD are underrecognized relevant complications for individuals with PAD. Therefore, it is important to monitor kidney function in people with PAD, and the use of renoprotective medications can be important in this clinical population. Examples of these medications can include antihypertensives, such as angiotensin-converting enzyme inhibitors (ACEIs) or angiotensin receptor blockers (ARBs).³⁹ Yet among patients with PAD, less than one-third receive ACEIs or ARBs, even when these medications are clinically indicated.⁴⁰ Moreover, the AHA/ACC Hypertension Guideline provides some specific recommendation for patients with hypertension and coronary heart disease or stroke but not for patients with PAD.⁴¹ However, our study suggests the need of special attention to those with PAD who are at elevated risk for important complications of hypertension, adverse kidney outcomes. Additional medications, such as sodium-glucose transport protein 2 (SGLT2) inhibitors, have been shown to have renoprotective effects as well.^{42, 43} Thus, it would be worth testing these medications in eligible patients with PAD.

In addition, our findings on asymptomatic PAD build upon the prognostic value of PAD screening, although this concept is still controversial among expert organizations.^{4, 44, 45} The US Preventive Services Task Force recommends against PAD screening using ABI.⁴⁴ However, their position is mainly from the perspective of whether ABI can improve cardiovascular risk prediction beyond traditional risk factors. In this context, PAD has been shown to be associated with a broad range of adverse outcomes,^{46, 47} and our study adds kidney outcomes to this list. Thus, there seems to be a need of comprehensive discussion regarding the value of identifying persons with PAD.

There are a few study limitations to consider. Probably the most important one is due to the method of ABI measurement. Specifically, an oscillometric device was used to take the ABI measurements instead of the gold standard using a Doppler probe. Indeed, some studies have shown misclassification of ABI using oscillometric devices compared to Doppler devices.⁴⁸ In addition, ABI measurements were taken in a single leg,¹⁹ which is likely to further contribute to misclassification of ABI. Taken together, our study is likely to underestimate the association of PAD with incident ESKD and CKD. Also, the study participants were White or Black, and as a result, our findings may not be simply generalizable to other races or ethnicities.

To conclude, in a middle-aged cohort of White and Black participants, both symptomatic PAD and asymptomatic PAD were independently associated with an elevated risk of ESKD and CKD. Our results highlight the importance of monitoring kidney function in those with PAD, even when symptoms are absent.

Figure 4.

Asymptomatic and symptomatic PAD increase the risk of ESKD and CKD. Using data from over 14,000 patients, we found that both asymptomatic and symptomatic peripheral artery disease increased the risk of end-stage kidney disease and chronic kidney disease by approximately 2x and 1.5x, respectively, after adjusting for potentially confounding variables.

Highlights.

• In ~14,000 community-dwelling middle-aged US adults, we have explored the association between peripheral artery disease (PAD) and kidney outcomes.

• Participants with PAD had a 1.5–2x higher risk of incident end-stage kidney disease (ESKD) and chronic kidney disease.

• These findings were significant regardless of whether the participant had symptomatic or asymptomatic PAD.

• The 25-year cumulative incidence for ESKD was ~10% in symptomatic PAD and asymptomatic PAD vs. 4–5% in no PAD.