Estimation of Central Systolic Blood Pressure from Peripheral Pressure Waves using a Novel Second Systolic Pressure-Based Method in Normal and Heritable Hypercholesterolemic Rabbits

Abstract

Aim: Central systolic blood pressure (cSBP) was closely related to hypertension-related organ damage rather than peripheral systolic blood pressure (pSBP). We aimed to estimate cSBP from pSBP without generalized transfer function in normal and Kurosawa and Kusanagi-hypercholesterolemic (KHC) rabbits aged 12 months.

Methods: Two catheter-tip transducers were advanced into the ascending aorta (AA) and distal end of the right brachial artery (Br) through the right common carotid and right radial arteries, respectively, under pentobarbital anesthesia. Pressure waves in response to the intravenous administration of angiotensin II and sodium nitroprusside were simultaneously recorded in AA and Br under regular cardiac pacing.

Results: The first (pSBP) and second peaks (pSBP₂) of the brachial blood pressure and their average (pSBP_m) were significantly correlated with cSBP, despite Murgo’s wave pattern of central pressure waves in both rabbit groups. In Bland–Altman plot and its modification as a function of the peripheral augmentation index (pAI) analyses, the differences between pSBP and cSBP decreased, and those between pSBP₂ and cSBP increased significantly in their average- or pAI-dependent manner, with undeniable mean biases in both rabbit groups. When the same analyses for SBP_m were performed instead, the mean bias was around zero, with reduced variance in the two rabbit groups. The observed pressure or pAI-dependent systematic biases for pSBP and pSBP₂ disappeared, representing the precise feature of pSBP_m as a cSBP estimate.

Conclusions: We conclude that pSBP_m could be more precise than pSBP₂ as a cSBP estimate, irrespective of blood pressure levels, pAI, or the presence of atherosclerosis.

See editorial vol. 30: 1108-1110

Introduction

Augmentation of pressure waves in the aortic root due to reflected waves is important to assess arterial stiffness because of aging, atherosclerosis, hypertension, and so forth^{1, 2)}. In the Anglo-Scandinavian Cardiac Outcomes Trial–Conduit Artery Functional Evaluation (ASCOT–CAFE) study, the amlodipine-based treatment of patients with hypertension was shown to achieve a significant central systolic blood pressure (cSBP) lowering effect, despite the effect on the peripheral systolic blood pressure (pSBP) equivalent to atenolol-based treatment, which would be associated with the reduction in cardiovascular events³⁾. Recent studies have been revealed that cSBP was more closely related to the left ventricular mass index, carotid intima–media thickness, and pulse-wave velocity than pSBP^{4, 5)}. cSBP was also more predictive of cardiovascular events^4-6), vascular damage^4-6), and organ failure^{4, 5, 7)}.

The peripheral second systolic blood pressure (pSBP₂) as well as cSBP mainly reflects the central augmentation peak due to reflection waves from the entire arteries. Central pressure waves have often been derived from tonometric radial pressure waves recorded noninvasively using generalized transfer function (GTF)^{2, 8-10)}. This method has been widely accepted, although the accuracy or reproducibility of GTF is still controversial^{11, 12)}. Pauca et al.¹³⁾ demonstrated, without using GTF, that the second peak of the radial artery pressure waves measured invasively was almost in good agreement with cSBP in elderly patients with hypertension. However, Hickson et al.¹⁴⁾ observed considerable difference between radial pSBP₂ and cSBP in younger normotensive individuals. We also showed that pSBP₂ was lower than cSBP in the Kurosawa and Kusanagi-hypercholesterolemic (KHC) as well as normal rabbit groups at lower blood pressure levels because of the infusion of sodium nitroprusside (NTP). However, pSBP₂ approximated cSBP before and after the infusion of angiotensin II (Ang II) in the KHC rabbit group¹⁵⁾. Thus, cSBP estimation bias of the pSBP₂-based method remains even in the presence of atherosclerosis.

Some attempts have been made to estimate cSBP without GTF from peripheral pressure or flow waves^16-18). Miyashita et al.¹⁹⁾ reported that during cardiac catheterization in 20 patients, noninvasive radial pSBP₂ underestimated invasive micromanometric cSBP in conditions of lower peripheral AI or greater pulse pressure amplification, which was broadly corrected by replacing pSBP₂ with the simple arithmetic mean of pSBP and pSBP₂ (pSBP_m).

In this study, we attempted to evaluate the pSBP_m correction of pSBP₂ as a new non-GTF-based estimate of cSBP at different blood pressure levels as well as in conditions with or without atherosclerosis in rabbits, that is, in young adult normal and KHC rabbits, an animal model for spontaneous hypercholesterolemia and atherosclerosis^{20, 21)}.

Methods

1. Animals

Thirteen male Japanese White and 12 male KHC rabbits, aged 12 months and weighing 3.8±0.4 and 3.5±0.3 kg [mean±standard deviation (SD)], respectively, were purchased from Japan Laboratory Animals Inc. (Nerima, Tokyo, Japan). They were bred in an air-conditioned room (25±3℃ room temperature, 50±10% relative humidity, and a 12L/12D light and dark cycle) and given a cholesterol-free commercial rabbit chow (RC–4, Oriental Yeast, Co., Ltd., Tokyo, Japan), at 100 g a day. This study was approved by the Experimental Animal Committee of Fukushima Medical University and was performed in accordance with the Guidelines for the Care and Use of Laboratory Animals published by the US National Institutes of Health.

2. Surgical Procedure

The surgical procedure was similar to that described previously¹⁵⁾. The rabbits were anesthetized by the intravenous administration of pentobarbital sodium (Nembutal, Abbott Laboratories, Inc., North Chicago, IL, USA) at a dose of 30 mg/kg, intubated through tracheotomy, and fixed supine. Butorphanol tartrate was administered intramuscularly at a dose of 0.2 mg/kg to reduce pain. Procaine chloride was locally applied to the incision area for pain relief. If needed, pentobarbital sodium was added at 5 mg/kg about every 20 min. Two catheter-tip micromanometers (2Fr, Millar Instruments Inc., Dallas, TX, USA) were introduced into the ascending aorta (AA) and distal end of the right brachial artery (Br) through the right common carotid and right radial arteries, respectively. Two polyethylene catheters for vasoactive drug infusion were placed at the superior vena cava through the right maxillary vein. The chest was carefully opened through median sternotomy under voluntary breathing. Electrodes for cardiac pacing were applied to the right atrium near the sinoatrial node.

3. Pressure Wave Recording

Ang II or NTP was infused using a syringe pump at doses of 30–40 ng/kg/min and 20–30 µg/kg/min, respectively, until the mean arterial pressure (MAP) level reached about 140 or 80 mmHg, respectively. Pressures at the AA and Br were simultaneously recorded in a personal computer (PowerBook G4 M9691J/A, Apple Inc., Cupertino, CA, USA) through an analog-to-digital converter (PowerLab System/16SP, AD Instruments Inc., Sydney, Australia) every 0.1 ms (10 kHz) under regular cardiac pacing. The MAP was derived from the original pressure by using a high-cut filter with a time constant of 2.5 s.

4. Analyses of Pressure Waves

Twenty successive pressure waves were analyzed before and after the infusion of Ang II and NTP at MAP levels of about 140 and 80 mmHg, respectively. The time at the second zero crossing of the fourth derivative of the raw pressure waves from above to below was taken as a peak of early systolic waves (first peak/shoulder) in the case of the type A pattern classified by Murgo et al.²²⁾, whereas the time at the third zero crossing of the fourth derivative from below to above was taken as a peak of late systolic waves (second peak/shoulder) in the case of the type C pattern. When the first peak/shoulder of the raw pressure waves was almost equal to the late systolic peak (within ±2.0 mmHg), we defined the wave as the type B pattern. Central (cAI; %) and peripheral augmentation indices (pAI; %) were determined as (P₂–P₁)/P₂ and P₂/P₁, where P₁ and P₂ were the heights of the early and late systolic waves, respectively. We analyzed a difference between the peripheral and central systolic pressures using the Bland–Altman plot, a statistical method to compare two measurements techniques by plotting the differences between the two techniques against the averages of the two techniques.

5. Statistical Analyses

The data were analyzed using a two-way analysis of variance (ANOVA). The individual cardiovascular parameter was compared between the normal and KHC rabbit groups, and before and after the vasoactive drug administration using Scheffe’s multiple-comparison test, if a significant difference was observed in ANOVA. Pearson’s simple correlation coefficient and regression equation were calculated between two variables. We calculated the statistical t-value using the correlation coefficient (r) and sample size (n) to test the significance of the correlation coefficient. The p-value was calculated using a t-distribution with n−2 degrees of freedom. A p-value ＜0.05 was considered significant.

Results

1. Cardiovascular Parameters

Table 1 summarizes the cardiovascular parameters before and after the infusion of vasoactive drugs in the normal and KHC rabbit groups. The values of cSBP (p＜0.001), pSBP (p＜0.01), pSBP₂ (p＜0.001), central diastolic blood pressure (cDBP; p＜0.01), and peripheral diastolic blood pressure (pDBP; p＜0.01); and cAI (p＜0.001) and pAI (p＜0.01) before the drug infusion were significantly greater in the KHC rabbit group than in the normal rabbit group. The cDBP values were almost the same as the pDBP values in each rabbit group. These values increased and decreased significantly because of the infusion of Ang II and NTP, respectively.

Table 1. Cardiovascular parameters in the normal and KHC rabbit groups before and after the infusion of vasoactive drugs

	Normal			KHC
	NTP	Control	Ang II	NTP	Control	Ang II
cSBP (mmHg)	95.7±5.3	122.5±5.5＊＊＊b, c	154.4±7.9	94.1±5.3	136.8±5.5＊＊＊a, b, c	157.0±7.3
pSBP (mmHg)	103.1±5.6	129.6±6.6＊＊＊b, c	156.9±7.7	100.9±7.1	141.7±6.2＊＊a,＊＊＊b, c	156.5±6.9
pSBP2 (mmHg)	87.0±5.2	117.9±5.9＊＊＊b, c	152.3±9.5	86.7±5.3	134.8±5.8＊＊＊a, b, c	156.5±6.8
cDBP (mmHg)	73.2±3.1	99.1±5.6＊＊＊b, c	126.9±5.8	73.2±3.4	110.1±4.4＊＊a,＊＊＊b, c	129.7±4.8
pDBP (mmHg)	73.8±3.4	98.9±5.3＊＊＊b, c	125.9±5.8	72.9±3.4	110.2±4.8＊＊a,＊＊＊b, c	129.0±5.3
cAI (%)	12.9±7.5	25.5±8.5＊＊b, ＊＊＊c	43.8±7.9	19.2±5.7	38.7±6.4＊＊＊a, b, ＊c	47.8±5.4
pAI (%)	44.9±11.5	62.3±7.0＊＊b,＊＊＊, c	85.1±11.7	49.0±6.5	78.3±5.1＊＊a,＊＊＊, b, c	99.5±9.7＊a

Values are mean±SD. ^＊; p＜0.05, ^＊＊; p＜0.01, ^＊＊＊; p＜0.001, a; KHC rabbit group vs normal rabbit group, b; control vs NTP, c; control vs Ang II

2. Change in Pressure Waveform in Response to Vasoactive Drugs

Fig.1 shows the changes in pressure waveform in response to the intravenous infusion of Ang II and NTP in normal (A) and KHC (B) rabbits. Pressure waves increased and decreased in amplitude in response to Ang II and NTP in both rabbits. The second systolic peak of brachial pressure waves was almost equal to the aortic systolic pressure before the infusion of vasoactive drugs (Control) in the KHC rabbits and after the infusion of Ang II in both rabbit groups; whereas it reduced apart from the aortic systolic peak because of the infusion of NTP.

Fig.1. Examples of pressure waveform at the AA and right brachial artery in the normal (A) and KHC (B) rabbits at different MAP levels because of the intravenous infusion of vasoactive drugs

Ang II, angiotensin II; NTP, sodium nitroprusside; Control, before the administration of Ang II or NTP. Arrows represent the first peak/shoulder and second peak of the pressure waves at the asending aorta (AA) and brachial artery, respectively. Broken lines represent pSBP₂. Pressure wave at the AA in the normal rabbit during the infusion of NTP shows the type B pattern.

3. Relationship between Central and Peripheral Blood Pressures

The type C pattern was detected only in one normal rabbit when the MAP level decreased to around 80 mmHg during the administration of NTP. The type B pattern was observed in one normal rabbit before NTP administration and in one KHC and six normal rabbits during the administration of NTP. Other waves were the type A pattern in the normal and KHC rabbit groups, irrespective of the administration of vasoactive drugs. We analyzed correlations between cSBP and pSBP, between cSBP and pSBP₂, and between cSBP and pSBP_m separately in type A and type B patterns in the normal and KHC rabbit groups. The values of pSBP (r=0.978, p=0.000 for the normal rabbit group and r=0.989, p=0.000 for the KHC rabbit group; Fig.2A), pSBP₂ (r=0.967, p=0.000 for the normal rabbit group and r=0.996, p=0.000 for the KHC rabbit group; Fig.2B), and the average between pSBP and pSBP₂ (pSBP_m; r=0.974, p=0.000 for the normal rabbit group and r=0.995, p=0.000 for the KHC rabbit group; Fig.2C) were strongly correlated with cSBP in the type A pattern in the two strains. The values of pSBP (r=0.972, p=0.000; Fig.2D), pSBP₂ (r=0.873, p=0.000; Fig.2E), and pSBP_m (r=0.981, p=0.000; Fig.2F) also significantly correlated with cSBP in the type B pattern in the normal strain.

Fig.2.

Correlation diagrams between cSBP and pSBP in the type A pattern (A) and type B pattern (D); between cSBP and pSBP₂ in the type A pattern (B) and type B pattern (E); and between cSBP and pSBP_m in the type A pattern (C) and type B pattern (F) in the normal and KHC rabbit groups

The values of pPP (r=0.716, p=0.000 for the normal rabbit group and r=0.666, p=0.000 for the KHC rabbit group), pPP₂ (r=0.748, p=0.000 for the normal rabbit group and r=0.872, p=0.000 for the KHC rabbit group), and pPP_m (r=0.841, p=0.000 for the normal rabbit group and r=0.894, p=0.000 for the KHC rabbit group) showed strong positive correlations with cPP in the two strains (Fig.3A, B and C). The differences in pPP (ΔpPP) before and after the infusion of vasoactive drugs showed significant correlations with ΔcPP in the normal rabbit group (r=0.520, p=0.001) but not in the KHC rabbit group (r=0.258, p=0.129; Fig.4A). The differences in pPP₂ (ΔpPP₂; r=0.763, p=0.000 for the normal rabbit group and r=0.835, p=0.000 for the KHC rabbit group) and pPP_m (ΔpPP_m; r=0.772, p=0.000 for the normal rabbit group and r=0.888, p=0.000 for the KHC rabbit group) correlated significantly with the differences in cPP (ΔcPP; Fig.4B and C).

Fig.3. Correlation diagrams between cPP and pPP (A), between cPP and pPP₂ (B), and between cPP and pPP_m (C) in the normal and KHC rabbit groups

cPP correlated strongly with pPP, pPP₂, and pPP_m.

Fig.4. Correlation diagrams between ΔcPP and ΔpPP (A), between ΔcPP and ΔpPP₂ (B), and between ΔcPP and ΔpPP_m (C) in the normal and KHC rabbit groups

ΔcPP, ΔpPP, ΔpPP₂, and ΔpPP_m were the differences in cPP, pPP, pPP₂, and pPP_m before and after the administration of vasoactive drugs.

ΔpPP, ΔpPP₂, and ΔpPP_m correlated significantly with ΔcPP in the two rabbit strains, except for ΔpPP in the KHC rabbit group.

4. Bland–Altman Plot

When the difference between pSBP and cSBP was plotted against their average in the Bland–Altman plot, the difference scattered in a wide range and decreased significantly with increasing their average in the normal (r=−0.485, p=0.013) and KHC (r=−0.564, p=0.001) rabbit groups. Limits of agreement for each comparison were set within the average difference±1.96 SD of the difference. The mean bias±1.96 SD between pSBP and cSBP was 5.7±7.5 and 3.8±8.9 mmHg in the normal and KHC rabbit groups, respectively (Fig.5A). The difference between pSBP₂ and cSBP also varied considerably and increased significantly with increasing their average in the normal (r=0.719, p=0.001) and KHC (r=0.758, p=0.000) rabbit groups. The mean bias±1.96 SD between pSBP₂ and cSBP was −5.1±7.8 and −3.3±7.5 mmHg in the normal and KHC rabbit groups, respectively (Fig.5B). Employing pSBP_m instead of pSBP or pSBP₂, the mean bias (±1.96 SD) between pSBP_m and cSBP was almost zero in the normal (0.3±4.7 mmHg) and KHC (0.2±5.3 mmHg) rabbit groups (Fig.5C). There was no significant correlation between the difference and the average in the normal (r=0.173, p=0.291) and KHC (r=0.046, p=0.790) rabbit groups (Fig.5C). Most bias data were distributed within the limits of agreement (Fig.5A, B and C).

Fig.5. Bland–Altman plots of the differences between pSBP and cSBP (A), between pSBP₂ and cSBP (B), and between pSBP_m and cSBP (C) against each average in the normal and KHC rabbit groups

Horizontal solid and dot-and-dash lines represent the mean bias and 1.96 SD between pSBP and cSBP (A), between pSBP₂ and cSBP (B), and between pSBP_m and cSBP (C) in the normal and KHC rabbit groups, respectively. A bold black line shows zero bias. The mean bias between pSBP and cSBP (±1.96 SD) was 5.7±7.5 and 3.8±8.9 for the normal and KHC rabbit groups, respectively. The mean bias between pSBP₂ and cSBP (±SD) was –5.1±7.8 and −3.3±7.5 for the normal and KHC rabbit groups, respectively. By applying the average between pSBP and pSBP₂ (pSBP_m) instead of pSBP or pSBP₂, the differences between pSBP and cSBP, and between pSBP₂ and cSBP canceled each other. Consistent bias was minimized and distributed over a relatively narrow range around zero (0.3±4.7 and 0.2±5.3 for the normal and KHC rabbit groups, respectively).

Then, the Bland–Altman plot was modified by substituting pAI, which represents pulse pressure amplification, for the average pressure value as the horizontal axis. The pSBP showed systematic bias from cSBP, exhibiting a significant negative correlation with pAI in the normal (r=−0.643, p=0.000) and KHC (r=−0.724, p=0.000) rabbit groups (Fig.6A). The pSBP₂ also showed systematic bias from cSBP, which was a significant positive correlation with pAI in the normal (r=0.816, p=0.000) and KHC (r=0.756, p=0.000) rabbit groups (Fig.6B). When pSBP or pSBP₂ was replaced with pSBP_m, the difference between pSBP_m and cSBP was distributed with a smaller variance and mean bias of around zero, showing no significant correlation with pAI in the normal (r=0.156, p=0.344) and KHC (r=0.072, p=0.677) rabbit groups (Fig.6C). Most bias data were distributed within the limits of agreement (Fig.6A, B and C).

Fig.6. Modified Bland–Altman plots depicting the difference between pSBP and cSBP (A), between pSBP₂ and cSBP (B), and between pSBP_m and cSBP (C) as the function of the peripheral augmentation index in the normal and KHC rabbit groups

P₁ and P₂ are the height of the early and late systolic waves, respectively. Horizontal solid and dot-and-dash lines represent the mean bias and 1.96 SD between pSBP and cSBP (A), between pSBP₂ and cSBP (B), and between pSBP_m and cSBP (C) in the normal and KHC rabbit groups, respectively. A bold black line shows zero bias. The mean bias between pSBP and cSBP (±1.96 SD) was 5.7±7.5 and 3.8±8.9 for the normal and KHC rabbit groups, respectively. The mean bias between pSBP₂ and cSBP (±SD) was −5.1±7.8 and −3.3±7.5 for the normal and KHC rabbit groups, respectively. By applying the average between pSBP and pSBP₂ (pSBP_m) instead of pSBP or pSBP₂, the negative and positive correlations canceled each other. The consistent bias (±1.96 SD) between pSBP_m and cSBP was minimized and scattered over a relatively narrow range around zero (0.3±4.7 and 0.2±5.3 for the normal and KHC rabbit groups, respectively).

Discussion

Rabbits are suitable for the study of lipoprotein metabolism and atherosclerosis²³⁾. These rabbit models enable us to acquire precise invasive simultaneous multisite pressure wave recordings in experimentally altered pathophysiological or hemodynamic conditions, which are difficult in humans. The pressure waveform of rabbits was also reported to closely resemble those of humans²⁴⁾. These factors convinced us to expect the extrapolation of rabbit data to humans. This experimental study, using precise simultaneous multisite pressure wave measurements over the wide range perturbation of blood pressure, reconfirmed that the average between pSBP and pSBP₂ (pSBP_m) is a more precise estimate for cSBP than pSBP₂, regardless of blood pressure levels, peripheral vascular tone, pulse pressure amplification, and atherosclerotic conditions.

Central pulse waves, consisting of forward and backward waves, propagate throughout the body, including the upper limbs. How the central pulse waves are distorted in the brachial or radial artery depends mainly on transmission characteristics of the artery in the forelimb. Therefore, central pulse waves can be estimated from peripheral pulse waves recorded by applanation tonometry using GTF. This method is widely employed in most clinical and epidemiological investigations, although the usefulness or reproducibility of GTF is controversial^{11, 12)}.

On the other hand, some investigators attempted to estimate cSBP from radial pressure waves without using GTF. Takazawa et al.²⁵⁾ first reported that late systolic shoulder (pSBP₂) of the digital photoplethysmogram well reflected the changes in central pressure waves measured invasively in response to Ang II and nitroglycerin. Pauca et al.¹³⁾ demonstrated, using simultaneous invasive measurements of central aortic and peripheral radial pressure waves, that pSBP₂ was almost equivalent to cSBP in elderly patients with hypertension. Munir et al.²⁶⁾ also reported the good agreement between cSBP and pSBP₂ after the decreasing blood pressure level due to the administration of glyceryl trinitrate and during cardiac pacing. Recently, Chen et al.²⁷⁾ proposed new methods to estimate aortic SBP from peripheral pressure waves measured noninvasively using an oscillometric device. They estimated aortic SBP using a multivariate prediction model based on radial pressure waves measured invasively and showed that the differences between brachial SBP₂ and aortic SBP, and between brachial SBP₂ and estimated aortic SBP were almost zero and distributed within ±2 SD over a wide pressure range. However, Hickson et al.¹⁴⁾ observed considerable discrepancy between radial SBP₂ and cSBP, especially in younger individuals with lower blood pressure levels, although overall, pSBP₂ was a good estimate for cSBP with a smaller difference between them in subjects of a wide age range. They suggest that the dissociation may be related to essential problems as to the contour of pressure waves, because they failed to reduce the difference between radial SBP₂ and cSBP by correcting radial SBP₂ for AI and by mathematical or statistical methods.

In this study, cAI and pAI were significantly greater in the KHC rabbit group than in the normal rabbit group before the administration of vasoactive drugs. We observed previously that young adult KHC rabbits exhibited atherosclerotic lesions that covered 48.6%±19.8% (mean±SD) of the aortic intimal surface, mild hypertension, increases in aortic pulse-wave velocity, arterial input impedance, and reflection coefficient compared with those in the age-matched control rabbits^28-30). There was significant positive correlation between cAI and pAI in the normal (r=0.842, p=0.000) and KHC (r=0.864, p=0.000) rabbit groups. These increases in magnitude and accelerated return of reflected waves would contribute to the increased cAI and pAI. The bias between pSBP₂ and cSBP was minimum during the pressor response to Ang II and maximum during the depressor response to NTP in the Bland–Altman plot, which had a significant positive correlation with their average (Fig.5) or pAI (modified Bland–Altman plot; Fig.6). The difference between pSBP and cSBP showed a negative correlation with their average or pAI. By applying pSBP_m, which is the average between pSBP and pSBP₂, instead of pSBP or pSBP₂, the consistent bias between pSBP_m and cSBP was minimized (almost zero), with less variance in the original and modified Bland–Altman plot analyses in both the rabbit groups.

The observed pressure and pAI-dependent systematic biases were counterbalanced or canceled each other by applying pSBP_m, which would contribute to markedly improved accuracy. Wilkinson et al.³¹⁾ demonstrated that hypercholesterolemia intensified aortic reflection pressure waves and AI. This study showed that pSBP_m approximated cSBP well, even in rabbits with the atherosclerotic aorta.

Miyashita¹⁹⁾ analyzed invasive central aortic and radial tonometry pressure waves in 20 patients during cardiac catheterization using discrete Fourier transformation. The 2^nd systolic (augmentation) peak of the peripheral pressure waves mostly consists of the 1^st and 2^nd harmonic components of the central pressure waves, which changed a little because of propagation from the central aorta. On the other hand, the 1^st systolic peak of the peripheral pressure waves includes the 3^rd and higher harmonic components that are amplified because of transmission from the central to peripheral arteries. These features of pressure wave transmission along the forelimb arteries could account for pSBP₂ as a good estimate of cSBP. However, this study and data reported elsewhere³¹⁾ have suggested its imperfectness by AI (augmentation/amplification)-dependent systematic bias of pSBP₂, as shown in Fig.6B. This feature of systematic bias may represent the transmission property of the forelimb arteries. The transmission property also alters pSBP inversely depending on AI, as shown in Fig.6A, possibly because of characteristic differences in their harmonic components.

This phenomenon could be related to shifts or the distortion of harmonic component distribution that is theoretically thought to be also AI dependent, that is, the lower the AI, the greater the higher-frequency harmonic components. This feature of pressure waveform was recently reported as harmonic distortion in mice³²⁾.

Based on the above mechanisms, pSBPm could be a better estimate of cSBP than pSBP₂, by canceling AI dependent systematic biases. The pSBP_m could be a useful, simple, and precise estimate of cSBP in clinical practice, which may compensate for the shortcomings and limitations of the pSBP₂-based cSBP estimation. Further investigations are necessary to elucidate the detailed mechanism by which the pSBP_m approximates cSBP in animal and human studies.

In conclusion, pSBP_m (the average between pSBP and pSBP₂) is a more precise estimate for cSBP than pSBP₂, regardless of blood pressure levels, peripheral vascular tone, pulse pressure amplification, and atherosclerotic conditions. It could compensate for or overcome the known shortcomings of SBP₂-based cSBP estimation.

Acknowledgement

The present study was supported by Japan Society for the Promotion of Science Grants-in-Aid for Scientific Research (Grant Number 23500522). The authors are also grateful to Prof. Hiroshi Miyashita for instructive cooperation, to Dr. Masahiko Kusanagi for offering KHC rabbits and to Mr. Haruyuki Wago for technical assistance in the present study.

Conflict of Interest

The authors have no conflict of interest concerning the present study.

References

• 1)  O’Rourke MF, Hashimoto J: Mechanical factors in arterial aging. A clinical perspective. J Am Coll Cardiol, 2007; 50: 1-13
• 2)  O’Rourke MF, Adji A: An updated clinical primer on large artery mechanics: implications of pulse waveform analysis and arterial tonometry. Curr Opin Cardiol, 2005; 20: 275-281
• 3)  Williams B, Lacy PS, Thom SM, Cruickshank K, Stanton A, Collier D, Hughes AD, Thurston H, O’Rourke M: Differential impact of blood pressure-lowering drugs on central aortic pressure and clinical outcomes: principal results of the Conduit Artery Function Evaluation (CAFE) study. Circulation, 2006; 113: 1213-1225
• 4)  Kollias A, Lagou S, Zeniodi ME, Boubouchairopoulou N, Stergiou GS: Association of central versus brachial blood pressure with target-organ damage: systematic review and meta-analysis. Hypertension, 2016; 67: 183-190
• 5)  Cheng HM, ChuangSY, Wang TD, Kario K, Buranakitjaroen P, Chia YC, Divinagracia R, Hoshide S, Minh HV, Nailes J, Park S, Shin J, Siddique S, Sison J, Soenarta AA, Sogunuru GP, Sukonthasarn A, Tay JC, Teo BW, Turana Y, Verma N, Zhang Y, Wang JG, Chen CH: Central blood pressure for the management of hypertension: Is it a practical clinical tool in current practice? J Clin Hypertens, 2020; 22: 391-406
• 6)  Roman MJ, Devereux RB, Kizer JR, Lee ET, Galloway JM, Ali T, Umans JG, Howard BV: Central pressure more strongly relates to vascular disease and outcome than does brachial pressure: the Strong Heart Study. Hypertension, 2007; 50: 197-203
• 7)  Wang KL, Cheng HM, Chuang SY, Spurgeon HA, Ting CT, Lakatta EG, Yin FCP, Chou P, Chen CH: Central or peripheral systolic or pulse pressure: which best relates to target organs and future mortality? J Hypertens, 2009; 27: 461-467
• 8)  Karamanoglu M, O’Rourke MF, Avolio AP, Kelly RP: An analysis of the relationship between central aortic and peripheral upper limb pressure waves in man. Eur Heart J, 1993; 14: 160-167
• 9)  Chen CH, Nevo E, Fetics B, Pak PH, Yin FCP, Maughan WL, Kass DA: Estimation of central aortic pressure waveform by mathematical transformation of radial tonometry pressure: validation of generalized transfer function. Circulation, 1997; 95: 1827-1836
• 10)  Melenovsky V, Borlaug BA, Fetics B, Kessler K, Shively L, Kass DA: Estimation of central pressure augmentation using automated radial artery tonometry. J Hypertens, 2007; 25: 1403-1409
• 11)  Hope SA, Tay DB, Meredith IT, Cameron JD: Use of arterial transfer functions for the derivation of aortic waveform characteristics. J Hypertens, 2003; 21: 1299-1305
• 12)  Millasseau SC, Patel SJ, Redwood SR, Ritter JM, Chowienczyk PJ: Pulse wave reflection assessed from the peripheral pulse: is a transfer function necessary? Hypertension, 2003; 41: 1016-1020
• 13)  Pauca AL, Kon ND, O’Rourke MF: The second peak of the radial artery pressure wave represents aortic systolic pressure in hypertensive and elderly patients. Br J Anaesth, 2004; 92: 651-657
• 14)  Hickson SS, Butlin M, Mir FA, Graggaber J, Cheriyan J, Khan F, Grace AA, Cockcroft JR, Wilkinson IB, McEniery CM: on behalf of the Anglo-Cardiff Collaboration Trial Investigators. The accuracy of central SBP determined from the second systolic peak of the peripheral pressure waveform. J Hypertens, 2009; 27: 1784-1788
• 15)  Katsuda S, Miyashita H, Shimada K, Miyawaki Y, Kojima I, Shiogai Y, Hazama A: Subservient relationship of the peripheral second systolic pressure peak to the central hemodynamic parameters is preserved, irrespective of atherosclerosis progression in hypercholesterolemic rabbits. Hypertens Res, 2014; 37: 19-25
• 16)  Li P, T.-M. Laleg-Kirati: Central blood pressure estimation from distal PPG measurement using semiclassical signal analysis features. IEEE Access, 2021; 9: 44963-44973
• 17)  Chemla D, Millasseau S, Hamzaoui O, Teboul JL, Monnet X, Michard F, Jozwiak M: New method to estimate central systolic blood pressure from peripheral pressure: A proof of concept and validation study. Front Cardiovasc Med, 2021; Dec: 1-10, doi: 10.3389/fcvm.2021.772613
• 18)  Miyashita H: Clinical assessment of central blood pressure. Curr Hypertens Rev, 2012; 8: 80-90
• 19)  Miyashita H, Katsuda S: Basis of monitoring central blood pressure and hemodynamic parameters by peripheral arterial pulse waveform analyses. Conf Proc IEEE Eng Med Biol Soc (10.1109/EMBC.2013.6609477), 2013; 221-224
• 20)  Kurosawa T, Kusanagi M, Yamasaki Y, Senga Y, Yamamoto T: New mutant rabbit strain with hypercholesterolemia and atherosclerotic lesions produced by serial inbreeding. Lab Anim Sci, 1995; 45: 385-392
• 21)  Katsuda S, Miyake M, Kobayashi D, Hazama A, Kusanagi M, Takazawa K: Does the augmentation index of pulse waves truly increase with progression of atherosclerosis? An experimental study with hypercholesterolemic rabbits. Am J Hypertens, 2013; 26: 311-317
• 22)  Murgo JP, Westerhof N, Giolma JP, Altobelli SA: Aortic input impedance in normal man: relationship to pressure wave forms. Circulation, 1980; 62: 105-116
• 23)  Fan J, Kitajima S, Watanabe T, Xu J, Zhang J, Liu E, Chen YE: Rabbit models for the study of human atherosclerosis: from pathophysiological mechanisms to translational medicine. Pharmacol Ther, 2015; 146: 104-119
• 24)  Nichols WW, O’Rourke MF, Vlachopoulos C: Central arterial pressure. In: McDonald’s Blood Flow in Arteries. 6th ed. London: Hodders Arnold, 2011; pp. 226-227
• 25)  Takazawa K, Tanaka N, Takeda K, Jurosu F, Ibukiyama C: Underestimation of vasodilator effects of nitroglycerin by upper limb blood pressure. Hypertension, 1995; 26: 520-523
• 26)  Munir S, Guilcher A, Kamalesh T, Clapp B, Redwood S, Marber M, Chowienczyk P: Peripheral augmentation index defines the relationship between central and peripheral pulse pressure. Hypertension, 2008; 51: 112-118
• 27)  Cheng HM, Wang KL, Chen YH, Lin SJ, Chen LC, Sung SH, Ding PY, Yu WC, Chen JW, Chen CH: Estimation of central systolic blood pressure using an oscillometric blood pressure monitor. Hypertens Res, 2010; 33: 592-599
• 28)  Katsuda S, Miyashita H, Hasegawa M, Machida N, Kusanagi M, Yamasaki M, Waki H, Hazama A: Characteristic change in local pulse wave velocity in different segments of the atherosclerotic aorta in KHC rabbits. Am J Hypertension, 2004; 17: 181-187
• 29)  Miyashita H, Katsuda S, Shimizu T, Hazama A, Shimada K, Yada T: Potential of hemodynamic waveform analysis to detect early-stage atherosclerotic lesion of the aorta. Circ Cont, 2003; 24: 360-370 (in Japanese with abstract in English)
• 30)  Katsuda S, Miyashita H, Takazawa K, Machida N, Kusanagi M, Miyake M, Hazama A: Mild hypertension in young Kurosawa and Kusanagi-Hypercholesterolemic (KHC) rabbits. Physiol Meas, 2006; 27: 1361-1371
• 31)  Wilkinson IB, Prasad K, Hall IR, Thomas A, MacCallum H, Webb DJ, Frenneaux MP, Cockcroft JR: Increased central pulse pressure and augmentation index in subjects with hypercholesterolemia. J Am Coll Cardiol, 2002; 39: 1005-1011
• 32)  Milkovich N, Gkousioudi A, Seta F, Suki B, Zhang Y: Harmonic Distortion of Blood Pressure Waveform as a Measure of Arterial Stiffness. Front Bioeng Biotechnol (10.3389/fbioe.2022.842754) 2022; 10: 842754