Abstract
The authors consecutively assessed various arterial pulse‐wave velocity (PWV) indices and ankle‐brachial index (ABI) by an automatic device (VP2000, OMRON Health Care Co. Ltd., Kyota, Japan) in outpatients with ≥ 1 cardiovascular risk. PAD was defined as ABI ≤ 0.9. Among 2309 outpatients (mean age 62.4 years), worse renal function was associated with higher brachial‐ankle PWV, heart‐carotid PWV, heart‐femoral PWV (hf‐PWV), and lower ABI (all P < .001). Multivariate regression models showed independent associations between lower eGFR, lower ABI (Coef: 0.42 & 0.41 for right and left), higher hf‐PWV (Coef: −11.4 [95% CI: −15.4, −7.3]) and greater PAD risk (adjusted OR: 0.83 [95% CI: 0.76, 0.91], all P < .05). eGFR set at 77 mL/min/1.73m2 was observed to be useful clinical cutoff (c‐statistics: 0.67) for identifying PAD (P for ΔAUROC: .009; likelihood X 2: 93.82 to 137.43, P < .001) when superimposed on clinical risks. This study suggested early renal insufficiency is tightly linked to region‐specific vascular stiffness and PAD.
Keywords: ankle‐brachial index, peripheral artery disease, pulse‐wave velocity, renal insufficiency, vascular stiffness
1. INTRODUCTION
Chronic kidney disease (CKD) is a risk factor for the development of cardiovascular disease (CVD) and is further associated with adverse outcomes in individuals with CVD.1, 2, 3, 4, 5 Peripheral artery disease (PAD) and arterial stiffness are both well‐known predictors of CVD events and mortality.6 It has been proposed that they derive from endothelial dysfunction or inflammation tightly linked to the pathophysiology of atherosclerosis.7, 8 Although patients with PAD or increased arterial stiffness share common risk factors with CAD, those with known PAD further displayed a higher risk for all relevant cardiovascular events or all‐cause mortality compared to their counterparts.9, 10
More recently, CKD has been shown to elicit inflammatory responses and present additional risk components for the development of CVD. To date, noninvasive pulse‐wave velocity (PWV), including segmental arterial PWV and derived ankle‐brachial index (ABI), is a convenient tool for documenting the presence of arterial stiffness or PAD in wide clinical settings.11, 12 Nevertheless, the association between advanced stage CKD, multisegment arterial stiffness, or abnormal ankle‐brachial index (ABI) in the range favoring the diagnosis of PAD from outpatient clinics remains largely unexplored. Therefore, we sought to investigate this issue in our current study.
2. METHODS
2.1. Study participants and design
All participants were recruited from cardiovascular outpatient clinics at MacKay Memorial Hospital, a tertiary center in Northern Taiwan, from August 2009 to July 2014. The study design was cross‐sectional and retrospective. Our current work was approved by the local ethics committee in accordance with the Declaration of Helsinki. In brief, 2310 consecutive patients were enrolled at the outpatient clinics with risks of PAD. The risk factors included age (male > 45 years, female > 55 years), family history of PAD, high‐density lipoprotein cholesterol (male < 40 mg/dL, female < 50 mg/dL), hypertension, hypercholesterolemia, diabetes, and personal history of smoking. All patients received complete biochemical analysis, various PWV measures including right and left side brachial‐ankle PWV (ba‐PWV), heart‐carotid PWV (hc‐PWV), heart‐femoral PWV (hf‐PWV), and ABI assessment on extremities from both sides. Demographic information, clinical symptoms/signs, and medical histories were all obtained by senior cardiologists through face‐to‐face interviews during outpatient clinic visits. We further categorized eGFR into 3 groups: < 60, 60‐89, and ≧ 90 mL/min/1.73 m2 based on Chronic Kidney Disease Epidemiology Collaboration (CKD‐EPI) formula and related eGFR information to various vascular stiffness measures (including hc‐PWV, hf‐PWV, ba‐PWV, and ABI). In our current analysis, patients with known history of cardiovascular surgery, rheumatic heart disease, atrial fibrillation, or any arrhythmia that rendered stable measures of arterial waves less likely were excluded. Additionally, patients with previous pacemaker implantation, overt renal insufficiency (eGFR < 15 mL/min/1.73 m2), or known dialysis history were also excluded from the study.
2.2. Anthropometric measures and baseline risk factors
Anthropometric parameters including body height, body weight, and waist circumference were measured by experienced research nurses. Hypertension was defined by current usage of antihypertensive agents or documented presence of abnormally high blood pressure, including systolic blood pressure (SBP) ≥ 140 mm Hg and/or diastolic blood pressure (DBP) ≥ 90 mm Hg, from two different measures. Hyperlipidemia treatment was defined by usage of any lipid‐lowering medicine (statin or fibrate use) from medication lists. Diabetes was defined by fasting blood glucose > 126 mg/dL or known history or any current usage of diabetes mellitus (DM) medications. Smoking was defined as current, active tobacco usage. Coronary artery disease was defined by known myocardial infarction history or any significant coronary stenosis with angina symptoms requiring percutaneous coronary intervention or history of prior stent deployment. Stroke was defined as either hemorrhagic or ischemic cerebrovascular events from medical history. In our current work, cardiovascular disease was defined as the existence of either coronary artery disease or stroke.
2.3. Measures of region‐specific segmental artery pulse‐wave velocity and ankle‐brachial index
Arterial stiffness was diagnosed using pulse‐wave velocity. After resting for 5 minutes in a supine position, bilateral brachial‐ankle pulse‐wave velocity (ba‐PWV), systolic blood pressure, and diastolic blood pressure from 4 limbs were measured simultaneously with an automated machine (VP‐2000; Collin Corp., Komaki, Japan). Right and left side ankle‐brachial indices were calculated by the highest pressure on the dorsal and posterior tibial artery from the right and left sides, respectively, with the same method used for the highest brachial artery pressure assessment. Of the two ABI measurements for each patient, we selected the lowest ABI for study use. The carotid artery and femoral artery pulse waveforms were also measured and recorded by a single experienced technician. Region‐specific central aortic heart‐femoral pulse‐wave velocity (hf‐PWV) and heart‐carotid pulse‐wave velocity (hc‐PWV) were also obtained from the time delay between the rapid upstroke of the heart‐femoral artery and the heart‐carotid artery by an automatic device (VP‐2000; Collin Corp., Komaki, Japan). PWV was defined and calculated as the distance between the 2 arterial sites divided by the time delay between the 2 arterial point sites and presented as centimeters per second.13 According to the recommendation of the American College of Cardiology and the American Heart Association (ACC/AHA) guidelines,14 ABI results should be uniformly reported with uncompressed values categorized as high when the value is above 1.40, normal when the value is between 1.00 and 1.40, borderline when between 0.91 and 0.99, and clinically abnormal when ABI is 0.90 or less. Our current study adopted a stricter definition of PAD for those with ABI ≤ 0.90 only in any given side of extremity measures.
2.4. Laboratory measurements
Overnight fasting blood serum and plasma samples were collected for glucose, lipid profiles (total cholesterol, triglyceride, low‐density lipoprotein cholesterol, high‐density lipoprotein cholesterol), and other biochemical measurements including renal function. Serum samples were collected using standard sampling tubes or tubes containing separating gel. After ensuring individualized patient samples, calibrators and controls were set at an ambient temperature (20‐25°C) and the measurement was taken within 2 hours. High‐sensitivity C‐reactive protein (hs‐CRP) level was determined by a highly sensitive, latex particle enhanced immunoassay using Elecsys 2010 (Hitachi Corp. Hitachinaka Ibaraki, Japan). Serum B‐type natriuretic peptide (BNP) concentrations were measured using fluorescence‐based microtiter plate assay with a coefficient of variation (CV) of 10.4% (Alere Biosite Triage, San Diego, CA, USA). Renal function was determined as estimated glomerular filtration rate (eGFR) using either the Modification of Diet in Renal Disease (MDRD) formula: eGFR (mL/min/1.73 m2) = 186.3 × (serum creatinine−1.154) × (age−0.203) × 0.742 (if female),15 or CKD‐EPI method as previously described. 16
2.5. Statistical analysis
Continuous data are displayed as mean and standard deviation (SD), specifically indicated with categorical variables and presented as proportions or percentages of distributions. Cuzick's nonparametric trend test, with additional application of the arterial stiffness assessment, derived measures (including PWV and ABI), eGFR (≥ 90, 60‐90, < 60 mL/min/1.73 m2), and CKD‐EPI formula (Table 1, Figure 1), was used to examine the trends in demographic information. After accounting for clinical confounders including age, sex, body mass index (BMI), systolic blood pressure, fasting blood glucose, total cholesterol, high‐density lipoprotein, diabetes, hypertension, cardiovascular disease, hyperlipidemia treatment, and active smoking (Table 2), uni‐ and multivariate linear regression models (Table 2) were used to explore relationships between eGFR, PWV, and ABI indices and stepwise logistic regression models (with odds ratio [ORs]) were used to establish the risk of worsening renal function and PAD with ABI cutoffs (Table 3). Model discrimination and incremental diagnostic values were estimated by using a likelihood ratio test. The evaluated model X 2 differences identified PAD by additionally incorporating eGFR after it was superimposed on clinical risk factors such as smoking behavior (Figure 2). Receiver operating characteristic curves (ROC) were used to choose the most optimal clinical cutoff of eGFR (either MDRD or CKD‐EPI methods, respectively) for identifying PAD. Maximal Youden's index was chosen for optimal clinical cutoff of MDRD or CKD‐EPI with highest summation of sensitivity and specificity. Considering renal function was tightly linked to several key baseline measures (eg, age, gender, BMI, or blood pressure) or vascular stiffness measures, we further conducted interaction analysis between eGFR and PWV to predict PAD (Figure 3).
Table 1.
Basic information for study participants across EPI‐CKD eGFR tertiles groups
| CKD‐EPI eGFR categories | All | eGFR group 1 | eGFR group 2 | eGFR group 3 | Trend P/X 2 |
|---|---|---|---|---|---|
| (n = 2310) | < 60 (n = 402) | 60‐90 (n = 858) | ≥ 90 (n = 1050) | ||
| Baseline characteristics | |||||
| Age, y | 61.7 (12.4) | 72.4 (10.7) | 65.6 (11.4)* | 55.0 (10.0)*# | <.001 |
| Women, % | 1206 (52.2%) | 202 (50.37%) | 381 (44.41%) | 622 (59.24%) | <.001 |
| BMI, kg/m2 | 25.9 (6.9) | 25.79 (4.24) | 26.07 (9.90) | 25.88 (4.30) | .82 |
| SBP, mm Hg | 136.5 (21.7) | 143.43 (24.20) | 136.25 (20.99)* | 134.11 (20.59)* | <.001 |
| DBP, mm Hg | 78.89 (13.17) | 78.62 (14.12) | 78.92 (12.79) | 78.98 (13.11) | .45 |
| HR, beats/min | 70.7 (13.3) | 71.4 (15.8) | 69.9 (12.4) | 71.0 (12.8) | .38 |
| Biochemical data | |||||
| Fasting glucose, mg/dL | 115.68 (37.63) | 126.19 (51.56) | 114.31 (32.73)* | 112.85 (34.36)* | <.001 |
| Total cholesterol, mg/dL | 196.14 (41.98) | 186.46 (45.02) | 195.63 (39.90)*# | 200.26 (41.83)*# | <.001 |
| Triglyceride, mg/dL | 139.23 (100.1) | 156.13 (116.25) | 134.87 (84.00)* | 136.43 (104.88)* | <.001 |
| LDL, mg/dL | 118.09 (35.4) | 108.73 (34.95) | 119.18(34.92)* | 120.75 (35.42)* | <.001 |
| HDL, mg/dL | 48.23 (14.89) | 45.52 (14.40) | 48.19 (15.10)* | 49.3 (14.78)* | <.001 |
| eGFR (MDRD), mL/min/1.73 m2 | 81.56 (27.44) | 40.7 (15.0) | 73.9 (10.4)* | 103.0 (19.6)*# | <.001 |
| eGFR (CKD‐EPI), mL/min/1.73 m2 | 82.04 (23.51) | 40.2 (14.7) | 77.0 (8.3)* | 101.5 (8.3)*# | <.001 |
| Hs‐CRP, mg/dL | 0.24 (0.74) | 0.62 (1.27) | 0.32 (0.68)* | 0.25 (0.45)* | <.001 |
| BNP, pg/mL | 96.89 (335.83) | 302.45 (680.73) | 64.11 (161.65)* | 33.57 (87.60)* | <.001 |
| Medical history | |||||
| HTN, n (%) | 1778 (77%) | 331 (82.75%) | 686 (79.95%) | 760 (72.38%) | <.001 |
| DM, n (%) | 672 (29.1%) | 178 (44.50%) | 241 (28.09%) | 254 (24.19%) | <.001 |
| CVD, n (%) | 467 (20.2%) | 133 (33.17%) | 170 (19.81%) | 164 (15.62%) | <.001 |
| Hyperlipidemia, n (%) | 1283 (55.6%) | 230 (57.5%) | 499 (58.16%) | 554 (52.86%) | .049 |
| Active smoker, n (%) | 567 (24.6%) | 92 (23.00%) | 210 (24.48%) | 265 (25.24%) | .67 |
ABI, ankle‐brachial index; BMI, body mass index; CVD, cardiovascular disease; DBP, diastolic blood pressure; DM, diabetes mellitus; eGFR, estimated glomerular filtration rate; HDL, high‐density lipoprotein; hsCRP, high‐sensitivity CRP; HR, heart rate; HTN, hypertension; LDL, low‐density lipoprotein; PAD, peripheral artery disease; PWV, pulse wave velocity; SBP, systolic blood pressure.
Abnormal ABI, either right or left ABI ≤ 0.9 or ABI > 1.4.
*P < .05 compared to Group 1, # P < .05 compared to Group 2.
Figure 1.
Worse eGFR categories (< 60, 60‐90, ≥90 mL/min/1.73 m2) were associated with higher vascular stiffness measures (including hc‐PWV [A], hf‐PWV [B], ba‐PWV (mean) [C] and lower ABI (mean) [D] using MDRD and CKD‐EPI formula (all trend P < .001). *P < .05 between groups in same eGFR category
Table 2.
Univariate and multivariate linear regression evaluating the associations between eGFR (CKD‐EPI) and various peripheral arterial parameters
| Per 10 units reduction of eGFR | Univariate model | Univariate model | Multivariate models | (Adjusted R 2) | Multivariate models | (Adjusted R 2) |
|---|---|---|---|---|---|---|
| CKD‐EPI | MDRD | CKD‐EPI | CKD‐EPI | MDRD | MDRD | |
| Coef. (95%CI) | Coef. (95%CI) | Coef. (95%CI) | Coef. (95%CI) | |||
| Right brachial‐ankle PWV | −49.8 (−55.4, −44.2)a | −31.3 (−36.2, −26.3)a | −5.7 (−12.0, 0.54) | 0.37 | −1.9 (−6.7, 2.9) | 0.37 |
| Left brachial‐ankle PWV | −49.8 (−55.6, −44.1)a | −31.2 (−35.3, −25.1)a | −3.1 (−9.6, 3.4) | 0.37 | 0.76 (−4.2, 5.7) | 0.37 |
| Heart‐carotid PWV | −29.1 (−34.5, −23.7)a | −20.1 (−24.9, −15.4)a | −2.9 (−9.5, 3.65) | 0.20 | −2.6 (−7.7, 2.5) | 0.20 |
| Heart‐femoral PWV | −50.2 (−55.1, −45.2)a | −35.1 (−39.5, −30.7)a | −16.6 (−21.9, −11.4)a | 0.48 | −11.4 (−15.4, −7.3)a | 0.48 |
| Right ABI, % change | 0.89 (0.67, 1.11)a | 0.68 (0.49, 0.86)a | 0.57 (0.29, 0.86)a | 0.08 | 0.42 (0.21, 0.64)a | 0.08 |
| Left ABI, % change | 0.93 (0.71, 1.14)a | 0.68 (0.50, 0.87)a | 0.60 (0.32, 0.88)a | 0.07 | 0.41 (0.20, 0.63)a | 0.07 |
ABI, ankle‐brachial index; CI, confidence interval; PWV, pulse wave velocity.
Model 1, Univariate analysis.
Model 2, adjusted for age, sex, body mass index.
Model 3: adjusted for age, sex, body mass index, systolic blood pressure, fasting blood glucose, total cholesterol, high‐density lipoprotein, diabetes, hypertension, cardiovascular disease, hyperlipidemia treatment, and active smoking.
P < .05 with confidence interval 95%.
Table 3.
The comparisons and associations of PAD existence with baseline characters and vascular stiffness indices
| Non‐PAD | PAD | Univariate model for PAD | Multivariate model for PAD | |||
|---|---|---|---|---|---|---|
| (n = 2064) | (n = 243) | Odds Ratio (OR) | 95% CI | Odds Ratio (OR) | 95% CI | |
| Baseline characteristics | ||||||
| Age, y | 61.05 (12.12) | 67.03 (13.82)2 | 1.04 | (1.02‐1.05) | 1.03 | (1.02‐1.04) |
| Women, % | 1096 (53.02%) | 109 (45.04%)2 | 0.73 | (0.56‐0.95) | ─ | ─ |
| BMI, kg/m2 | 25.97 (7.14) | 25.62 (4.47) | 0.99 | (0.96‐1.02) | ─ | ─ |
| SBP, mm Hg | 135.17 (21.36) | 147.09 (24.53)2 | 1.02 | (1.02‐1.03) | 1.03 | (1.02‐1.04) |
| DBP, mm Hg | 78.69 (13.20) | 79.98 (14.74) | 1.01 | (1.0‐1.02) | ─ | ─ |
| HR, beats/min | 70.3 (13.0) | 74.1 (15.5)2 | 1.02 | (1.01‐1.03) | ─ | ─ |
| Biochemical data | ||||||
| Fasting glucose, mg/dL | 114.62 (35.99) | 124.83 (48.62)2 | 1.01 | (1.003‐1.01) | ─ | ─ |
| Total cholesterol, mg/dL | 196.43 (41.49) | 193.8 (45.91) | 1 | (1.0‐1.002) | ─ | ─ |
| Triglyceride, mg/dL | 137.28 (98.59) | 156.23 (111.08)2 | 1.001 | (1.00‐1.003) | ─ | ─ |
| LDL, mg/dL | 118.51 (35.09) | 114.44 (37.9) | 1 | (0.99‐1.00) | ─ | ─ |
| HDL, mg/dL | 48.65 (15.09) | 44.65 (12.51)2 | 0.98 | (0.97‐0.99) | 0.98 | (0.97‐0.99) |
| eGFR (MDRD), mL/min/1.73 m2 | 83.36 (26.55) | 66.19 (30.01)2 | 0.78b | (0.97‐0.98) | ─ | ─ |
| eGFR (CKD‐EPI), mL/min/1.73 m2 | 83.81 (22.13) | 66.77 (28.89)2 | 0.76b | (0.72‐0.80) | 0.83b | (0.76‐0.91) |
| < 60 | 304 (14.70%) | 98 (40.50%)2 | ─ | ─ | ─ | ─ |
| 60‐90 | 777 (37.57%) | 81 (33.47%)2 | 0.32 | (0.23‐0.45) | ─ | ─ |
| ≥ 90 | 987 (47.73%) | 63 (26.03%)2 | 0.2 | (0.14‐0.28) | ─ | ─ |
| Medical history | ||||||
| HTN, n (%) | 1594 (77.12%) | 183 (75.93%) | 0.92 | (0.67‐1.25) | ─ | ─ |
| DM, n (%) | 574 (27.77%) | 99 (41.08%)2 | 2.12 | (1.62‐2.77) | 1.48 | (1.01‐2.15) |
| CAD, n (%) | 365 (17.66%) | 64 (26.56%)2 | 1.8 | (1.34‐2.43) | 1.41 | (1.002, 1.98) |
| Hyperlipidemia drugs, n (%) | 1164 (56.34%) | 119 (49.58%)2 | 0.77 | (0.59‐1.004) | ─ | ─ |
| Active smoker, n (%) | 489 (23.66%) | 78 (32.37%)2 | 1.55 | (1.16‐2.07) | 1.71 | (1.18‐2.49) |
ABI, ankle‐brachial index; BMI, body mass index; CVD, cardiovascular disease; DBP, diastolic blood pressure; DM, diabetes mellitus; eGFR, estimated glomerular filtration rate; HDL, high‐density lipoprotein; hsCRP, high‐sensitivity CRP; HR, heart rate; HTN, hypertension; LDL, low‐density lipoprotein; PAD, peripheral artery disease; PWV, pulse wave velocity; SBP, systolic blood pressure.
Figure 2.
Increasing likelihood ratio (LR) was observed when Age (Model 1, X 2: 43.6), Age + medical history (Hx) including hypertension, diabetes, hyperlipidemia, cardiovascular disorders (Model 2, X 2: 86.87), and active smoking (Model 3, X 2: 98.46) was added in models. Significantly incremental prediction value was observed in predicting higher peripheral artery disease risk (defined as ≤ 0.9) when eGFR was added on age, medical histories (Hx), and active smoker (LR X 2: 98.46 expanded to 137.43, P < .001)
Figure 3.
Worse renal functions in terms of lower eGFR set at 77 mL/min/1.73 m2 were associated with higher risk of peripheral artery disease, which showed no effect modifications by age (< 65, ≥ 65 years), gender (male vs Female), BMI (< 25, ≥ 25 kg/m2), SBP (< 134, ≥ 134 mm Hg), and various vascular stiffness measures (including ba‐PWV, hc‐PWV and hf‐PWV)
All statistical tests were 2‐tailed, and P < .05 was considered statistically significant. Analyses were performed with Stata version 11 (Stata Corp., College Station, TX, USA) and SAS 9.2 version (SAS Institute, Cary,NC, USA).
3. RESULTS
3.1. Study participants
Table 1, using the CKD‐EPI formula (< 60, 60‐90, ≥ 90 mL/min/1.73 m2), summarizes the baseline clinical characteristics of the 2310 (mean age 62.3 ± 12.6 years, 52.0% women) study participants based on eGFR categories. The mean eGFR values determined by CKD‐EPI and MDRD formulas were 82.0 ± 23.5 and 81.6 ± 27.4 mL/min/1.73 m2, respectively, with more than 80% participants having eGFR ≧ 60 mL/min/1.73 m2 for both formulas. The mean difference was estimated to be −0.461 (CI −0.930 to 0.007) and limits of agreement (reference range for difference) were −23.416 to 22.49 (by Bland‐Altman analysis). Patients with eGFR < 60 mL/min/1.73 m2 were more likely to be male, were significantly older, had higher systolic blood pressure, had more elevated fasting glucose, and had worse lipid profiles (all P < .05). Significantly higher proportions of hypertension, diabetes, CAD prevalence, and hyperlipidemia medication use were also observed in lower eGFR categories (all P < .05). Participants with worse renal function results were associated with substantially higher hs‐CRP and BNP levels compared to those with better renal eGFR (Table 1, all P < .001).
3.2. The associations among glomerular filtration rate, region‐specific segmental vascular stiffness measures, and ABI
Lower eGFR was associated with higher ba‐PWV (r = −.35 & −.34 for right and left side), higher hc‐PWV (r = −.25), hf‐PWV (r = −.47), and lower ABI of both sides (0.16 and 0.17 for right and left side, all P < .001) (Table S1). Compared to MDRD formula, CKD‐EPI, in general, seemed to show slightly higher coefficients in magnitude with vascular stiffness measures and ABI assessment (Table S1). Further, worse renal function, in terms of lower eGFR, MDRD, and CKD‐EPI tertiles, was associated with graded increase of brachial‐ankle PWV measures on both sides, which paralleled a trend of higher segmental heart‐carotid PWV and higher heart‐femoral PWV measures (Table S2, all P for trend: < .001). A similar trend toward lower ABI measures was also observed across three eGFR categories (Table S2, all P for trend < .001). The univariate model demonstrates consistent associations among eGFR, PWV, and ABI measures, and the multivariate‐adjusted models showed that lower eGFR was independently associated with higher hf‐PWV (Coef: −16.6 [95% CI: −21.9, −11.4]) and lower ABI on both sides (Coef: 0.57 [95% CI: 0.29, 0.86], 0.60 [95% CI: 0.32, 0.88] for right and left sides per 10 units eGFR reduction, respectively, all P < .05) (Table 2).
3.3. The associations among glomerular filtration rate, ABI, and PAD
Compared to highest eGFR (≥ 90) category, PAD showed 4 times higher prevalence in the lowest eGFR category (6.0% vs 24.4%, Table S2). Table 3 displays the baseline characteristics of study participants with and without PAD. Participants with PAD were more likely to be male, were more likely to be active smokers, had more advanced age, and had higher clinical comorbidities. Additionally, they had higher blood pressure, fasting glucose levels, and triglyceride levels; lower HDL levels; and more elevated hs‐CRP and BNP levels and showed significantly worse renal function (all P < .05). In the stepwise multivariate linear regression model, with smoking included, higher eGFR levels (per 10 units reduction) were independently associated with lower opportunity of PAD (nonadjusted OR: 0.76 [95% CI: 0.72, 0.80]; adjusted OR: 0.83 [95% CI: 0.76, 0.91], both P < .001).
3.4. Incremental value of renal function in predicting PAD
Although medical histories superimposed on age substantially expanded the prediction model (from Model 1 [Age]: 43.6 to Model 2 [Age + Medical Histories]: 86.87, P < 0.001) in identifying PAD, eGFR based on CKD‐EPI (as Model 4) superimposed on age, medical histories and active smoking (as Model 3) further expanded the prediction model 3 substantially (Model 3 [Age + Medical Histories + Active Smoking]: 93.82 to Model 4: 137.43, P < .001) (Figure 2). A similar trend was also observed by incorporating eGFR using MDRD formula (data not shown). By using ROC, eGFR set at 77 and 70.6 mL/min/1.73 m2 alone (by CKD‐EPI and MDRD formula, respectively) showed most optimal cutoff (sensitivity: 67.6%, specificity: 58.7% and sensitivity: 69.3%, specificity: 56.3%; both concordance statistics [c‐statistics]: 0.67, 95% CI: 0.63‐0.71) in identifying PAD in our current work (Figure S1). A modestly increased incremental prediction model was observed when eGFR was added to age, medical histories, and active smoking status, which further expanded c‐statistics from 0.67 [95% CI: 0.63 to 0.71] to 0.70 [95% CI: 0.66 to 0.74], P for ΔAUROC: .009). Age, gender, BMI, systolic blood pressure (by using the median value of 134 mm Hg as a categorical variable), and various vascular stiffness measures (using median values as a categorical variable) did not significantly influence the associations between worse eGFR (set at 77 mL/min/1.73 m2) and PAD (Figure 3).
4. DISCUSSION
Our current findings can be 2‐fold. First, worse renal function was associated with graded alterations in vascular reactivity and increased region‐specific vascular stiffness, mainly manifested as higher ba‐PWV, hc‐PWV, and hf‐PWV and increased ABI. The association between worse renal function and region‐specific arterial stiffness, in particular, hf‐PWV and ABI, remained unchanged after adjusting for several clinical confounders. Further, we also observed an independent association between renal insufficiency and higher PAD risk, which may start to happen at a relatively early stage of renal insufficiency at eGFR cutoffs of 77 and 70.6 mL/min/1.73 m2 for CKD‐EPI and MDRD formula, respectively.
The main clinical feature of PAD is defined as the blockage or reduction of blood flow to lower extremities, which shares similar pathophysiology with atherosclerosis leading to limb ischemia and clinical symptoms of disability.17, 18 To date, it has been proposed that endothelial dysfunction may play a central role in the initiation, progression, and overt clinical onset of atherosclerosis and vascular stiffness,19, 20, 21 which not only correlates with the severity of the circulatory failure of the affected limb but also triggers circulating plasma markers of inflammation.22, 23 Endothelial dysfunction may be complicated by increased arterial stiffness causing microvasculature disorder24 and has also shown connections to the formation of peripheral artery disease 25, 26 and CKD.27 In addition, CKD and atherosclerotic diseases 5, 28, 29, 30 are highly interrelated. Estimates from the Third National Health and Nutrition Examination Survey (NHANES III) showed that the prevalence of PAD in patients with CKD is 24%.31 The previous study also showed a more advanced stage of CKD significantly correlated with the presence of PAD.31, 32, 33, 34, 35 Consistent with prior reports, our study revealed a high PAD prevalence of 24.4% in patients with eGFR < 60 mL/min/1.73 m2 using CKD‐EPI formula, which doubled the risk of PAD (15.4%) in participants with eGFR higher than 60 mL/min/1.73 m2.
To date, noninvasive measures in assessing pulse‐wave velocity is a clinically feasible and convenient method for objectively documenting the presence of arterial stiffness and may also serve as a strong clinical marker and prognosticator of cardiovascular events and mortality.36, 37 Our findings indicated that region‐specific, segmental vascular stiffness measures of hf‐PWV were highly associated with renal functional decline in a relatively early stage, which remained independent after considering key clinical confounders. This finding could be perfectly explained by the anatomical coverage of descending aorta and renal arteries encompassed by PWV measurement (such as hf‐PWV rather than ba‐ or hc‐PWV).38, 39 On the other hand, assessment of hf‐PWV may also overlap with ABI estimates of ankle‐brachial peak blood measures. In our literature review, concomitant endothelial dysfunction and arterial stiffness have been shown to enhance microvascular complications in diabetes, which mimics renal glomerular ultrastructural pathophysiology.24 The failure in the correlation between eGFR and ba‐PWV may be better explained by the inconsistency of blood flow path or a mixture of vascular biology character (both central and peripheral) as mentioned previously.40
In our study, renal function with eGFR set at 77 and 70.6 mL/min/1.73 m2 by using CKD‐EPI and MDRD formulas seemed to provide the most optimal cutoff in identifying participants at risk for PAD. CKD‐EPI as an estimate of renal function has been more widely used as a potentially better41 estimate to more accurately reflect true renal functional decline in participants not reaching end‐stage renal function, especially in participants with GFR > 60 mL/min per 1.73 m2 16 Meanwhile, we observed that renal function, when added to traditional CVD risks, further provided incremental values and expanded the prediction model significantly in identifying PAD.42, 43 Of note, our current data challenged traditional eGFR cutoffs set at 60 mL/min/1.73 m2 as starting point for CVD and may provide a useful clinical reference for a more aggressive screening cut point for possible existence of PAD in patients. Our findings, therefore, call for an increased awareness of PAD existence in an early stage CKD population and possibly call for a demand for more aggressive PAD screening in early‐stage CKD when more than 1 CVD risk was met. This finding, in aggregate, could potentially avert adverse limb ischemia as well as CVD events if modification of risk factors were intensified and early interventions, or treatment in any form, were given. This study also supports a PAD screening strategy for patients who may have potential CKD risk in future larger‐scale studies.
4.1. Limitations
Our current study has several limitations. First, the generalizability of our current work was mainly derived from outpatients, and whether our findings are applicable to community‐based population remains unexplored well. Further, vascular stiffness measures used in current work is noninvasive, and data were restricted to one single center. Even though, indirect measures of arterial stiffness by using PWV and ABI as screening tool for peripheral artery disease has been widely utilized in clinical settings. Third, our findings were mainly cross‐sectional in study design, and thus longitudinal studies for the validation of outcome‐driven thresholds are necessary in future studies.
5. CONCLUSIONS
Worse renal function in terms of lower eGFR was associated with higher PWV from various arterial segments; lower ABI was tightly linked to higher risk PAD participants. Although increasing age and several conventional cardiovascular risk factors may help to identify particpants at risk for PAD, participants with worse renal function further expanded the prediction model successfully, which indicated a need to screen for the clinical presence of PAD in an individual with chronic kidney disease.
CONFLICT OF INTEREST
The authors have nothing to disclose.
Supporting information
ACKNOWLEDGEMENTS
This research was supported by Ministry of Science and Technology (Taiwan) (NSC‐101‐2314‐B‐195−020, NSC103‐ 2314‐B‐010‐005‐MY3, 103‐2314‐B‐195‐001‐MY3, 101‐2314‐B‐ 195‐020‐MY1, MOST 103‐2314‐B‐195‐006‐MY3, MOST 105‐2632‐B‐715‐001, MOST 106‐2314‐B‐195‐008‐MY2, NSC102‐ 2314‐B‐002‐046‐MY3), National Taiwan University (NTU‐ CDP‐103R7879, 1104R7879 and 105R7879) and Taipei Medical University (TMU103‐AE1‐B14), and funds from MacKay Memorial Hospital (10271, 10248, 10220, 10253, 10375, 10358, E‐ 102003), the Taiwan Foundation for Geriatric Emergency and Critical Care (2014, 2015), and Hsinchu MacKay Memorial Hospital (E‐101‐11).
Lin Y‐H, Sung K‐T, Tsai C‐T, et al. The relationship of renal function to segmental vascular stiffness, ankle‐brachial index, and peripheral artery disease. J Clin Hypertens. 2018;20:1027–1035. 10.1111/jch.13297
[Correction updated on June 14, 2018, after initial online publication: Affiliation of author Ta‐Chuan Hung has been updated.]
Footnotes
Denotes per 10 units changes.
Denotes P < 0.05 vs Non‐PAD group.
Contributor Information
Ming‐Cheng Peng, Email: mcpeng@ms2.mmh.org.tw.
Ta‐Chuan Hung, Email: hung0787@ms67.hinet.net.
REFERENCES
- 1. Muntner P, He J, Hamm L, Loria C, Whelton PK. Renal insufficiency and subsequent death resulting from cardiovascular disease in the United States. J Am Soc Nephrol. 2002;13:745‐753. [DOI] [PubMed] [Google Scholar]
- 2. Wright RS, Reeder GS, Herzog CA, et al. Acute myocardial infarction and renal dysfunction: a high‐risk combination. Ann Intern Med. 2002;137:563‐570. [DOI] [PubMed] [Google Scholar]
- 3. Ohtake T, Kobayashi S, Moriya H, et al. High prevalence of occult coronary artery stenosis in patients with chronic kidney disease at the initiation of renal replacement therapy: an angiographic examination. J Am Soc Nephrol. 2005;16:1141‐1148. [DOI] [PubMed] [Google Scholar]
- 4. Ix JH, Shlipak MG, Liu HH, Schiller NB, Whooley MA. Association between renal insufficiency and inducible ischemia in patients with coronary artery disease: the heart and soul study. J Am Soc Nephrol. 2003;14:3233‐3238. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Sarnak MJ, Levey AS, Schoolwerth AC, et al. Kidney disease as a risk factor for the development of cardiovascular disease: a statement from the American Heart Association Councils on Kidney in Cardiovascular Disease, High Blood Pressure Research, Clinical Cardiology, and Epidemiology and Prevention. Circulation. 2003;108:2154‐2169. [DOI] [PubMed] [Google Scholar]
- 6. Gillum RF. Peripheral arterial occlusive disease of the extremities in the United States: hospitalization and mortality. Am Heart J. 1990;120(6 Pt 1):1414‐1418. [DOI] [PubMed] [Google Scholar]
- 7. Brevetti G, Silvestro A, Schiano V, Chiariello M. Endothelial dysfunction and cardiovascular risk prediction in peripheral arterial disease: an additive value of flow‐mediated dilation to ankle‐brachial pressure index. Circulation. 2003;108:2093‐2098. [DOI] [PubMed] [Google Scholar]
- 8. Hadi HA, Carr CS, Al SJ. Endothelial dysfunction: cardiovascular risk factors, therapy, and outcome. Vasc Health Risk Manag. 2005;1:183‐198. [PMC free article] [PubMed] [Google Scholar]
- 9. Criqui MH, Langer RD, Fronek A, et al. Mortality over a period of 10 years in patients with the peripheral arterial disease. N Engl J Med. 1992;326:381‐386. [DOI] [PubMed] [Google Scholar]
- 10. Rossi E, Biasucci LM, Citterio F, et al. Risk of myocardial infarction and angina in patients with severe peripheral vascular disease: predictive role of C‐reactive protein. Circulation. 2002;105:800‐803. [DOI] [PubMed] [Google Scholar]
- 11. Fowkes FG, Murray GD, Butcher I, et al. Development and validation of an ankle‐brachial index risk model for the prediction of cardiovascular events. Eur J Prev Cardiol. 2014;21:310‐320. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Lange SF, Trampisch HJ, Pittrow D, et al. Profound influence of different methods for determination of the ankle‐brachial index on the prevalence estimate of peripheral arterial disease. BMC Public Health. 2007;7:147. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Mahmud A, Feely J. Arterial stiffness is related to systemic inflammation in essential hypertension. Hypertension. 2005;46:1118‐1122. [DOI] [PubMed] [Google Scholar]
- 14. 2011 Writing Group Members ; 2005 Writing Committee Members ; ACCF/AHA Task Force Members . 2011 ACCF/AHA focused update of the guideline for the management of patients with peripheral artery disease (updating the 2005 Guideline): a report of the American College of Cardiology Foundation/American Heart Association Task Force on practice guidelines. Circulation. 2011;124:2020‐2045. [DOI] [PubMed] [Google Scholar]
- 15. Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, Roth D. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of Diet in Renal Disease Study Group. Ann Intern Med. 1999;130:461‐470. [DOI] [PubMed] [Google Scholar]
- 16. White SL, Polkinghorne KR, Atkins RC, Chadban SJ. Comparison of the prevalence and mortality risk of CKD Australia using CKD Epidemiology Collaboration (CKD‐EPI) and Modification of Diet in Renal Disease (MDRD) Study GFR estimating equations: the AusDiab (Australian Diabetes Obesity and Lifestyle) Study. Am J Kidney Dis. 2010;55:660‐670. [DOI] [PubMed] [Google Scholar]
- 17. Grenon SM, Chong K, Alley H, et al. Walking disability in patients with peripheral artery disease is associated with the arterial endothelial function. J Vasc Surg. 2014;59:1025‐1034. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Widener JM. Peripheral arterial disease and disability from NHANES 2001‐2004 data. J Vasc Nurs. 2011;29:104‐112. [DOI] [PubMed] [Google Scholar]
- 19. Davignon J, Ganz P. Role of endothelial dysfunction in atherosclerosis. Circulation. 2004; 109(23 Suppl 1): III27‐III32. [DOI] [PubMed] [Google Scholar]
- 20. Mudau M, Genis A, Lochner A, Strijdom H. Endothelial dysfunction: the early predictor of atherosclerosis. Cardiovasc J Afr. 2012;23:222‐231. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Anderson TJ. Arterial stiffness or endothelial dysfunction as a surrogate marker of vascular risk. Can J Cardiol. 2006; 22(Suppl B):72B‐80B. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Brevetti G, Martone VD, de Cristofaro T, et al. High levels of adhesion molecules are associated with impaired endothelium‐dependent vasodilation in patients with the peripheral arterial disease. Thromb Haemost. 2001;85:63‐66. [PubMed] [Google Scholar]
- 23. Brevetti G, Silvestro A, Di Giacomo S, et al. Endothelial dysfunction in peripheral arterial disease is related to increase in plasma markers of inflammation and severity of peripheral circulatory impairment but not to classic risk factors and atherosclerotic burden. J Vasc Surg. 2003;38:374‐379. [DOI] [PubMed] [Google Scholar]
- 24. Siasos G, Gouliopoulos N, Moschos MM, et al. Role of endothelial dysfunction and arterial stiffness in the development of diabetic retinopathy. Diabetes Care. 2015. Jan;38:e9‐e10. [DOI] [PubMed] [Google Scholar]
- 25. Heinen Y, Stegemann E, Sansone R, et al. A local association between endothelial dysfunction and intimal hyperplasia: relevance in peripheral artery disease. J Am Heart Assoc. 2015;4:e001472. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Marti CN, Gheorghiade M, Kalogeropoulos AP, Georgiopoulou VV, Quyyumi AA, Butler J. Endothelial dysfunction, arterial stiffness, and heart failure. J Am Coll Cardiol. 2012;60:1455‐1469. [DOI] [PubMed] [Google Scholar]
- 27. Ioannou K, Stel VS, Dounousi E, et al. Inflammation, endothelial dysfunction and increased left ventricular mass in chronic kidney disease (CKD) patients: a longitudinal study. PLoS One. 2013;10:e0138461. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Fried LF, Shlipak MG, Crump C, et al. Renal insufficiency as a predictor of cardiovascular outcomes and mortality in elderly individuals. J Am Coll Cardiol. 2003;41:1364‐1372. [DOI] [PubMed] [Google Scholar]
- 29. Furuhashi T, Moroi M, Joki N, et al. Prediction of cardiovascular events in pre‐dialysis chronic kidney disease patients with normal SPECT myocardial perfusion imaging. J Cardiol. 2014;63:154‐158. [DOI] [PubMed] [Google Scholar]
- 30. Zhang QL, Brenner H, Koenig W, Rothenbacher D. Prognostic value of chronic kidney disease in patients with coronary heart disease: role of estimating equations. Atherosclerosis. 2010;211:342‐347. [DOI] [PubMed] [Google Scholar]
- 31. O'Hare AM, Glidden DV, Fox CS, Hsu CY. High prevalence of peripheral arterial disease in persons with renal insufficiency: results from the National Health and Nutrition Examination Survey 1999‐2000. Circulation. 2004;109:320‐323. [DOI] [PubMed] [Google Scholar]
- 32. Baber U, Mann D, Shimbo D, Woodward M, Olin JW, Muntner P. Combined role of reduced estimated glomerular filtration rate and microalbuminuria on the prevalence of peripheral arterial disease. Am J Cardiol. 2009;104:1446‐1451. [DOI] [PubMed] [Google Scholar]
- 33. Mostaza JM, Suarez C, Manzano L, et al. Relationship between ankle‐brachial index and chronic kidney disease in hypertensive patients with no known cardiovascular disease. J Am Soc Nephrol. 2006;17(12 Suppl 3):S201‐S205. [DOI] [PubMed] [Google Scholar]
- 34. Yamasaki S, Izawa A, Koshikawa M, et al. Association between estimated glomerular filtration rate and peripheral arterial disease. J Cardiol. 2015;66:430‐434. [DOI] [PubMed] [Google Scholar]
- 35. Wattanakit K. Kidney function and risk of peripheral arterial disease: results from the Atherosclerosis Risk in Communities (ARIC) Study. J Am Soc Nephrol. 2007;18:629‐636. [DOI] [PubMed] [Google Scholar]
- 36. Fowkes FG, Price JF, Stewart MC, et al. Aspirin for prevention of cardiovascular events in a general population screened for a low ankle brachial index: a randomized controlled trial. JAMA. 2010;303:841‐848. [DOI] [PubMed] [Google Scholar]
- 37. Collaboration ABI, Fowkes FG, Murray GD, et al. Ankle brachial index combined with Framingham Risk Score to predict cardiovascular events and mortality: a meta‐analysis. JAMA. 2008;300:197‐208. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Salvi P, Parati G. Chronic kidney disease: arterial stiffness and renal function–a complex relationship. Nat Rev Nephrol. 2015;11:11‐13. [DOI] [PubMed] [Google Scholar]
- 39. Georgianos PI, Sarafidis PA, Liakopoulos V. Arterial stiffness: a novel risk factor for kidney injury progression? Am J Hypertens. 2015;28:958‐965. 10.1093/ajh/hpv004. Epub 2015 Feb 15. [DOI] [PubMed] [Google Scholar]
- 40. Choo J, Shin C, Barinas‐Mitchell E, et al. Regional pulse wave velocities and their cardiovascular risk factors among healthy middle‐aged men: a cross‐sectional population‐based study. BMC Cardiovasc Disord. 2014;13:5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Matsushita K1, Mahmoodi BK, Woodward M, et al. Comparison of risk prediction using the CKD‐EPI equation and the MDRD study equation for estimated glomerular filtration rate. JAMA. 2012;307:1941‐1951. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Ooi QL, Tow FK, Deva R, et al. The microvasculature in chronic kidney disease. Clin J Am Soc Nephrol. 2011;6:1872‐1878. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43. Maeda H, Sugiyama S, Jinnouchi H, et al. Advanced peripheral microvascular endothelial dysfunction and polyvascular disease in patients with high cardiovascular risk. J Cardiol. 2016;67:455‐462. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials