Arterial Stiffness and Ankle-Brachial Index – Cross-Sectional Study of 259 Primary Care Patients ≥50 Year-Old

Background

Peripheral artery disease (PAD) is a common clinical manifestation of atherosclerosis, acting more than 230 million adults globally [1]. Despite its significant health implications, PAD often remains underrecognized, even though it has a higher mortality rate compared to coronary heart disease [2]. Notably, in its early stages, PAD is frequently clinically asymptomatic, underscoring the need for accessible and reliable diagnostic methods [2].

Lower-extremity arterial disease (LEAD) is the most common form of PAD, and the ankle-brachial index (ABI) measurement with a Doppler probe is the primary non-invasive diagnostic approach [1]. The literature consistently defines an ABI of less than 0.9 as the accepted threshold for diagnosing LEAD [3]. However, the reliability and accuracy of this test hinge on various factors, including the experience of the examiner, patient positioning during the examination, and other procedural considerations [4].

Arterial stiffness, characterized by an impaired ability of arterial walls to accommodate volume changes, worsens with age and certain comorbidities [5,6]. Research has shown that arterial stiffness parameters are correlated with subclinical PAD and are predictors of disease progression [7–9].

Pulse wave velocity (PWV) measurement is the preferred method for assessing increased arterial stiffness, with both invasive and non-invasive techniques available [10,11]. Among the non-invasive options, oscillometric devices offer a practical and ambulatory solution for primary care settings [12,13].

The objective of our study was to evaluate the correlation between ABI and arterial stiffness parameters, specifically focusing on PWV. Additionally, we aimed to assess the correlation between PWV and established LEAD risk factors, determining whether PWV could serve as a valuable additional parameter for detecting LEAD.

Material and Methods

THE STUDY GROUP:

The study cohort was recruited by convenience sampling from primary care facility. Patients aged 50 years or older who voluntarily agreed to participate in the study were eligible. Exclusion criteria were a history of limb amputation, the presence of skin lesions or edema at the cuff placement site, and an inability to maintain the required examination position. We performed a sample size calculation before initiating the study using the formula designed for cross-sectional studies [14]. This calculation estimated the required number of participants to be 253.

METHODS OF EXAMINATION:

The initial phase of patient assessment involved obtaining a medical history, which included information on previously diagnosed cardiovascular conditions such as coronary artery disease, PAD, past cardiovascular events, diabetes mellitus, hypertension, atrial fibrillation, and chronic kidney disease. Additionally, we recorded the medications patients were currently taking and their smoking history. The Edinburgh Claudication Questionnaire was administered to all participants to assess the presence of claudication [5].

Physical examination entailed the palpation of lower-limb arteries, including the femoral, posterior tibial, and dorsalis pedis arteries, to assess the presence of palpable pulses. We also performed auscultation of the carotid and femoral arteries and abdominal aorta for detection of murmurs. Auscultatory blood pressure measurements were taken from both upper limbs, and the higher reading was recorded. Elevated blood pressure was defined as systolic blood pressure ≥140 mmHg or diastolic blood pressure ≥90 mmHg [3].

All participants underwent a body mass index (BMI) assessment using the Tanita MC780-MA device.

ARTERIAL STIFFNESS MEASUREMENT:

The primary method of arterial stiffness measurement involved PWV assessment using the oscillometric device Mobil-o-Graph Pulse Wave Analyzer (Mobil-o-Graph PWA, IEM, Germany). PWV measurements were taken at the brachial artery. This involved inflating a cuff placed on the arm above the elbow to diastolic pressure for 10 s, capturing the pulse wave [15].

Abnormal PWV levels were defined in 2 ways: (1) as PWV values exceeding universal fixed PWV threshold of 10 m/s (uPWVt) [15], and (2) based on individual patient characteristics, with Mobil-o-Graph PWA calculating PWV ranges according to sex, age, and blood pressure – individualized PWV threshold (iPWVt) as indicated by the manufacturer.

ANKLE-BRACHIAL INDEX MEASUREMENT:

Ankle-brachial index (ABI) measurements were conducted using a Doppler probe (Dopplex DMX Digital Doppler, Huntleigh Healthcare). An ABI below <0.9 was considered abnormal, and for subsequent analysis served as evidence of LEAD [3,16].

STATISTICAL ANALYSIS:

Statistical analysis was conducted using IBM SPSS Statistics 25 software (USA). Quantitative parameters are shown as the mean±standard deviation, and qualitative parameters are presented as numbers with percentages.

To explore the relationships between nominal data, we performed Spearman’s correlation test. We studied correlations between PWV values and Doppler ABI measurements. We also determined correlations between PWV values and various demographic and clinical factors, including age, sex, BMI, blood pressure measurements, previously diagnosed cardiovascular disease (CVD), smoking history, claudication presence, medication intake, absence of palpable pulse, hypertension, and diabetes. Associations between age groups and PWV were tested with the chi-square test.

A significance level of α=0.05 was established, with P values less than 0.05 considered statistically significant. Spearman’s rho was calculated to determine strength of the correlation.

Results

DEMOGRAPHIC DATA:

The demographic characteristics of the study participants are summarized in Table 1. The study comprised 259 participants with a mean age of 67.5±7.5 years, including 197 (76.1%) women and 62 (23.9%) men. Among the entire cohort, most were either current or former smokers, accounting for 186 (71.8%), while 190 (73.4%) had an abnormal body mass index of ≥25 kg/m2, and mean BMI 27.9±4.4. Hypertension was the most prevalent comorbidity, affecting 164 (63.3%) participants, followed by diabetes mellitus in 40 (15.4%). A history of coronary heart disease was noted in 23 (8.8%) individuals, and 10 (3.9%) reported past acute coronary syndrome, 5 (1.9%) experienced strokes, and 3 (1.2%) had a history of transient ischemic attacks (TIAs). Hypotensive medications were being taken by 155 (59.8%) patients, while 101 (38.9%) were on statin therapy. Furthermore, 50 (19.3%) participants were regularly using antiplatelet agents, and 44 (16.9%) were taking medications to lower blood glucose levels.

Regarding established evidence of PAD, 6/259 (2.3%) patients had previously received a diagnosis and treatment for LEAD. The Edinburgh Claudication Questionnaire was administered to the entire study group, with 25/259 cases (9.6%) yielding positive results.

MEASUREMENTS OF ABI AND PWV:

The measurements of both PWV and Doppler ABI were obtained in 259 patients, with a mean ABI result of 0.99±0.14. Results below <0.9 were observed in 44/259 (16.9%) of the reported measurements.

Oscillometric measurements of arterial stiffness parameters using the Mobil-o-Graph PWA were successfully conducted in 266 patients. The mean PWV for the examined group was 9.99±1.36 m/s. Overall, elevated PWV was recorded in 130/266 (48.8%) and 104/266 (39.1%) according to uPWVt and iPWVt, respectively.

Stratifying by age groups, 5 of 88 participants (5.7%) under 65 years had PWV >10 m/s (u PWVt), 115 of 168 (68.4%) between 65 and 80 years fell into this category, and all 10 participants over 80 years exhibited PWV >10 m/s (χ² test, P<0.001, V=0.62). When considering iPWVt, elevated PWV measurements were found in 25% of those aged 50–65, 45.8% of individuals aged 66–80, and 50% of those over 80 years (χ² test, P=0.004, V=0.20).

Table 2 presents mean values of PWV, ABI, and blood pressure of the study group.

CORRELATIONS BETWEEN PWV AND ABI:

To assess possible correlation between PWV values and ABI, we exclusively analyzed data from patients who underwent successful examinations for both variables (N=259). Among patients with Doppler ABI <0.9, 25/44 (56.8%) cases exhibited elevated PWV according to iPWVt versus 72/215 (33.5%) in those with ABI ≥0.9 (Table 3). Remarkably, among 44 patients with ABI <0.9, 19/44 (43.2%) had PWV in the normal range according to iPWVt.

Spearman’s correlation test showed a statistically significant correlation between abnormal PWV measurements for the individual threshold and abnormal ABI (P=0.03, r=0.18), while there was no significant correlation for the threshold of PWV measurements (P=0.057, r=0.12).

PULSE WAVE VELOCITY AND CARDIOVASCULAR RISK FACTORS:

The correlations between PWV values and LEAD risk factors, clinical symptoms, and signs are presented in Table 4. No noticeable correlation was found between PWV values and the sex of the participants (P=0.515 and P=0.775, for both uPWVt and iPWVt). A significant positive correlation was observed between increased PWV over 10 m/s and the presence of diabetes (P=0.04, r=0.18). Furthermore, a significant correlation was established between PWV (both cutoffs) and diagnosis of hypertension (P=0.009, r=0.16).

We found no correlation between positive Edinburgh Claudication Questionnaire results and elevated PWV (P=0.131 and P=0.336, respectively, for both PWV cutoffs). No correlation in arterial stiffness parameters and current or former smoking was found for uPWVt or iPWVt (P=0.134 and P=0.366).

We found correlations between medication intake and arterial stiffness parameters. Among those with PWV over 10 m/s, antiplatelet agents (0.027, r=0.014) and statins (P=0.009, r=0.16) were more frequently used.

There were no significant correlations between PWV and past TIA incidents (P=0.074, P=0.324), past strokes (P=0.159, P=0.332), past acute coronary syndrome (P=0.82, p=0.56), or a history of coronary heart disease (P=0.108; P=0.066).

Discussion

Our study validated a correlation between reduced ABI, a recognized diagnostic parameter for LEAD according to AHA/ACC guidelines [1,16], and heightened arterial stiffness, particularly in the form of PWV. Furthermore, our research affirmed a correlation between elevated PWV and certain well-established risk factors for PAD.

Some other authors have studied the correlation between arterial PWV and ankle-brachial index, as well as arterial stiffness parameters and established cardiovascular risk factors. A study by Piko et al [17] of patients with previously diagnosed ischemic heart disease concluded that lower ABI was correlated with higher PWV, and that PWV was positively correlated with age, BMI, and mean arterial pressure [17]. However, we are not aware of any research similar to ours, in which PWV and ABI relations were determined among unselected primary care patients.

Our study findings illuminate the high prevalence of arterial stiffness in primary care patients aged 50 and above, particularly when accompanied by other cardiovascular risk factors. Specifically, 48.9% and 37.1% of our study participants exhibited elevated arterial stiffness, defined by PWV values above uPWVt and iPWVt, respectively.

We present 2 interpretations of abnormal PWV values: a fixed universal threshold of PWV >10 m/s, and individualized thresholds computed by the measurement device considering patient-specific factors. Notably, the use of individualized thresholds, which take into account age, appears to be a better approach, given the significant influence of age on PWV. For instance, all patients aged over 80 years had PWV >10 m/s, while only 50% surpassed the individualized threshold. This suggests that individualized thresholds may better distinguish varying levels of stiffness and cardiovascular risk. An alternative approach, as suggested by Diaz et al [18], proposes separate normal ranges for each decade of life, emphasizing the lack of a universally applicable PWV threshold.

While an elevated PWV was associated with an abnormal ABI, it is worth noting that many patients with low ABI exhibited normal PWV according to iPWVt. In total, elevated arterial stiffness in our study population, as indicated by iPWVt, was more than twice more prevalent than decreased ABI 37.5% vs 16.9%, and much more prevalent than earlier recognizedLEAD (2.3%). Intriguingly, among patients with an ABI <0.9, a substantial proportion (43.2%) still demonstrated PWV within the normal range. This suggests that not all cases of atherosclerosis present with elevated PWV, or alternatively, that our method of assessing arterial stiffness may have limitations. Some authors have highlighted drawbacks in the PWV oscillometric method and identified valvular heart disease and blood flow impairment as potential sources of overestimation or underestimation of results [19]. Notably, we did not account for these factors in the past medical histories of our participants.

Pulse wave velocity may offer additional value beyond ABI index measurements, particularly in LEAD prevention, as stiffness parameters serve as early indicators of arterial wall alterations, and reducing arterial stiffness is a target in cardiovascular disease prevention [3,20].

We found a significant correlation between increased arterial stiffness parameters and high blood pressure. This has been confirmed by Diaz et al [18], who found that patients with hypertension exhibited increased arterial stiffness parameters across all age groups, with hypertension being a predominant cause of arterial stiffness in those aged over 50 years.

Additionally, patients with diabetes in our study also demonstrated a higher prevalence of elevated PWV (with statistical significance observed for the universal threshold only). Notably, among patients with diabetes, the combination of PWV and ABI measurements has been previously shown have superior in predictive ability for mortality in higher-risk individuals compared to isolated ABI measurements. Moreover, in patients with diabetes, ABI may not be a sensitive factor for the early diagnosis of peripheral artery disease [21,22].

In our attempt to explore the association between clinical symptoms, measured by the Edinburgh Claudication Questionnaire, we found no significant correlation between questionnaire results and elevated arterial stiffness. This aligns with previous findings that claudication is not a common symptom among PAD patients, despite its specificity as a symptom. Claudication symptoms are only present in 10–30% of PAD cases, making the Edinburgh Claudication Questionnaire a tool with good specificity but poor sensitivity [5].

The absence of peripheral pulses may not be highly sensitive as a sign of LEAD, particularly in the early stages of the disease [23]. However, our study did identify a correlation between the absence of the right posterior tibial artery pulse and abnormal PWV values in both interpretations, while on the left posterior tibial artery, only a correlation with PWV >10 m/s reached statistical significance. To sum up, our findings underscore the potential value of PWV measurements as an additional tool for assessing cardiovascular risk profiles, offering insights into vascular health.

It is crucial to acknowledge and address potential limitations of our study. Firstly, the size of our participant group may have been insufficient to capture all significant correlations, potentially limiting the generalizability of our findings. Additionally, our study group may not fully represent the broader population of primary care patients, as it consisted of individuals who responded to an office invitation, thus constituting a convenience sample.

Another limitation of our research was that ABI measurements, which we regarded as the diagnostic method of choice for non-invasive LEAD detection [24], have inherent limitations. ABI may not be sensitive enough for detecting early-stage disease or in patients with comorbidities such as diabetes mellitus [21].

An important consideration is the extensive use of medications among our study participants. Some of these medications have vasoactive properties and can influence arterial stiffness parameters. This introduces the possibility of medication-related biases in our results, which should be considered when interpreting our findings.

Finally, it should be noted that the Rho-values for Spearman’s correlations in our study were relatively low, reaching a maximum of 0.21. Those values represent a weak correlation according to the literature [25].