Ankle brachial index (ABI) measurement is a noninvasive method for detecting arterial occlusion of the lower extremities. The original method measures the blood pressure in both the brachia and ankles using the Doppler method. Since the 1990s, fully automated ABI measuring devices with oscillometric pressure cuffs have been developed in Japan, which simultaneously measure the blood pressures of both the brachia and ankles. Compared to the Doppler method, the oscillometric method has difficulty in correctly measuring weak blood pressure and has low accuracy in the evaluation of severe ischemic limbs [1]. However, on the other hand, such 4-cuff devices have enabled us to perform ABI tests easily, quickly, inexpensively, and without requiring special skills. The devices correctly measure normal and middle-low pressure; therefore, their widespread use has led to large-scale screening for patients at risk for peripheral artery disease (PAD), including those who are elderly, have a smoking history, diabetes, dyslipidemia, or hypertension [2].

Blood pressure decreases when arterial stenosis reaches more than 60-70% of the lumen. The sensitivity and specificity of ABI 0.90 or less for indicating significant stenosis of 50% or more in the lower extremity arteries are 90% and 95%, respectively [3]. In other words, the patients with milder arterial stenosis are undiagnosed. It is well known that pulse volume recording (PVR) sense proximal arterial stenosis earlier than ABI with its deformation characterized by an absence of a dicrotic notch, a prolonged systolic downslope, a rounded systolic peak, and a flattened wave with a delayed upstroke and run-off. As mentioned in the present manuscript, the automated oscillometric device records pulse volume waveforms of the limbs for 10 s by air plethysmography after measuring limb pressures [4, 5]. PVR does not require arterial compression and is less affected by arterial calcification. Therefore, PVR is a useful indicator for detecting PAD even in patients with diabetes mellitus or chronic renal failure, in whom ABI often shows a false high value [6]. In addition, the device calculates the upstroke time (UT) and percentage mean arterial pressure (%MAP) of each limb as the index of PVR. The calculation method of %MAP is expressed in the present manuscript [5]. That is, %MAP is the percentage of enclosed pulse volume area relative to the rectangle with the pulse’s base and amplitude [7]. As the height of the first sharp peak of the pulse wave decreases, the %MAP value increases (Fig. 1). UT and %MAP enable quantitative evaluation of the pulse waveform changes due to proximal arterial occlusive lesions with continuous variables.

Ankle pulse volume recordings (PVR) and calculation methods of upstroke time (UT) and percentage mean arterial pressure (%MAP). A broken line with two-headed arrows indicates the height at which the enclosed pulse wave area (pink area) is flattened in the rectangle (blue box) (See details of the formulas in the text and previous [7] and present manuscript [5]). Ankle PVRs of normal adults show a sharp and high first peak. In post-occlusive patterns, the height of the first peak decreases, the UT prolongs, and the %MAP value increases. In healthy young individuals, the first downslope is generally gradual, and the dicrotic notch often appears on a high position of the downslope with a dull shape, resulting in a high %MAP value

Actually, patients with PAD have been detected using ankle UT and %MAP before ABI decreases [8]. Patients with PAD generally have polyvascular disease and often have advanced atherosclerotic lesions in the coronary arteries and cerebral vessels. Therefore, abnormalities in ankle UT and %MAP may serve as markers for noninvasively detecting early or asymptomatic coronary artery disease (CAD) or cerebrovascular disease (CVD) among patients with normal ABI. However, there is no clear mention of the cut-off values for the detection of CAD or CVD.

For diagnosing significant PAD, The cut-off values of ankle UT ≥ 180 msec and %MAP ≥ 45% are commonly used [8]. The authors of the present study previously reported that an ankle UT ≥ 180 msec could be useful as a marker for CAD [9] and also, for CAD or CVD among patients with diabetes mellitus who often have incompressible calcified arteries [10]. The present study enrolled patients with a normal ABI (1.0 ≤ ABI < 1.4) in both legs and examined the optimal cut-off values of ankle UT and %MAP for detecting patients with CAD [5]. As the result, the cut-off values were set to UT ≥ 148 msec and %MAP ≥ 40.4%. Namely, it was optimal to identify the mild PAD in order to detect patients with significant CAD (coronary artery stenosis ≥50% or occlusion of at least one branch). The authors concluded that using UT ≥ 148 msec together with the traditional risk factor significantly improved the detection accuracy of CAD; however, using %MAP ≥ 40.4% did not.

An ankle UT ≥ 148 msec identified 152 out of 240 patients with CAD (sensitivity = 63.3%). In contrast, %MAP ≥ 40.4% identified only 92 patients (sensitivity = 38.3%). In addition, although 7.8% of the patients in Group 1 (with normal UT and %MAP) had CAD, only 4.0% of the patients in Group 2 (abnormal %MAP and normal UT) had CAD. Therefore, the diagnostic utility of %MAP was weak. These results corresponded with the results of Mitsutake et al. They investigated the relationship between UT or %MAP and coronary artery calcification (CAC) scores evaluated using multi-detector computed tomography. In the group with an intermediate CAC score, the average UT was 150 msec, and the average %MAP was 39% [11]. Moreover, Mitsutake et al. also concluded that %MAP did not remain as an independent factor for CAC score after a multivariate logistic regression analysis.

%MAP value increase may occur later than UT prolongation because flattering of pulse waveform occurs after the blood flow decreases significantly. Additionally, high ankle %MAP values do not always indicate atherosclerosis. For example, in healthy young individuals, the first downslope of the ankle pulse wave is generally gradual, and the dicrotic notch often appears on a high position of the downslope with a dull shape, resulting in a high %MAP value (Fig. 1) [7]. In a survey of healthy people around the age of 21 years, the average UT was as short as 144 msec, but the average %MAP was approximately 40% [7]. In the present study, although the difference was not significant at all, Group 2 had somewhat better backgrounds in terms of age, comorbidities, and smoking status than the other three groups. Whereas, not all participants had undergone CAG, so asymptomatic patients might not have been defined as having CAD.

Ankle UT is also affected by noises, including aortic stenosis. However, this may be rather advantageous for screening for atherosclerosis. Taken together with previous findings, UT may be a hopeful predictor of CAD. On the other hand, %MAP may be necessary to adjust the cut-off value according to the background, such as age, or to use it together with other findings, such as PVR raw waveform, to detect patients with CAD efficiently. The accuracy of the cut-off values should be validated by in various populations. Future research is expected to clarify the utilities of ankle UT and %MAP to identify patients with CAD and systemic atherosclerosis.