Q J Med 1999; 92: 595-600
© 1999 Association of Physicians
Similarities and differences between augmentation index and pulse wave velocity in the assessment of arterial stiffness
From the Department of Medicine, Clinical Pharmacology Unit, University of Cambridge, Addenbrooke's Hospital, Cambridge, UK
Received 15 April 1999 and in revised form 1 July 1999
Dr Yasmin, Clinical Pharmacology Unit, Addenbrooke's Hospital, Box 110, Hills Road, Cambridge CB2 2QQ
 |
Summary |
|---|
 Top Summary IntroductionMethodsResultsDiscussionReferences |
|---|
We investigated whether there was a correlation between the simultaneous assessments of augmentation index (AI) and pulse wave velocity (PWV), undertaken by the SphygmoCor system, and what were the principal factors responsible for differences in these two putative assessments of arterial stiffness, in 105 offspring (41 men, 64 women) aged 19–71 years, of patients with familial hypertension. Arterial stiffness was measured using the SphygmoCor pulse wave analysis system. AI and PWV correlated significantly and positively (r=0.29, p<0.005) and the strength of the correlation was greater when each gender was examined separately. This led us to observe several-fold higher AI in women (22.04±12) than in men (8.59±13) (p<0.001); the difference could be explained only in part by an inverse regression correlation between AI and height (r=-0.45; p<0.001), but not PWV. AI was also more influenced than PWV by heart rate and blood pressure. AI is strongly correlated with a previously validated estimate of arterial stiffness, PWV. It is probable that separate normal ranges should be established for men and women, while further studies determine what parameters other than height are responsible for the gender difference.
|
Introduction |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Diagnosis and prognosis in hypertension entails a web of paradoxes. Blood pressure is, literally, a property of liquid in a conduit, but outside of intensive care units is never measured directly. It is, indeed, rarely affected by the blood itself, being the product of cardiac output and the resistance to flow offered by small (50–200 µM) arteries; neither of these is readily amenable to measurement in patients. The heart is one of the target organs affected by high blood pressure, but the most common cause of death in hypertension is disease of the large artery walls. In most patients with hypertension, the primary tissue or organ at fault is unknown, but is unlikely to be either the heart or large arteries. For the measurement of blood pressure in patients, we resort to various imperfect measures of the brachial artery—a large artery whose structure is never overtly affected by hypertension, although subtle functional abnormalities have been demonstrated.1 Given the importance, therefore, to clinical practice of large arteries in hypertension, it is unsurprising that many investigators have sought to study the behaviour of their walls, and to investigate whether there are functional changes that account for some effects of hypertension. The main problems which have arisen from such studies are, firstly, the variety of properties that are studied—distensibility, compliance, incremental elastic modulus (Einc), pulse wave velocity2—with no easy way of comparing their validity and utility; secondly, the sophistication of the equipment, which limits wide application; and thirdly, the most accessible large arteries are not susceptible to the structural changes/diseases caused by hypertension.
Recently, the re-discovery by O'Rourke of the sphygmogram (which was used to study the pulse before the sphygmomanometer) has led to the development of techniques which offer a simple way of estimating changes in the central aorta that are likely to arise from the loss of arterial elasticity in the course of hypertension. The systolic pulse wave in the central aorta is augmented by the reflection of blood from arterial bifurcation points, principally that of the distal aorta itself; a perfectly elastic aorta absorbs all the pulse wave generated by ventricular contraction, whereas a completely rigid tube reflects a large proportion of the wave. The augmentation index is defined as the proportion of central pulse pressure due to the late systolic peak, which is in turn attributed to the reflected pulse wave. The SphygmoCor system uses an empirically-generated transfer function to calculate central pressure from the radial pulse waveform, which is measured by a hand-held tonometer. The same equipment can be used also at other superficial arterial sites, carotid and femoral, and by ECG-gating, the time for transmission of the arterial pulse wave between sites is calculated. Since this measurement, namely pulse wave velocity, is generally accepted as one of the valid estimates of arterial stiffness,3–4 we wished to determine whether augmentation index (AI) and pulse wave velocity (PWV) are closely correlated. Clearly, however, they are not identical, having (like many of the other measures mentioned above) different units of measurement. We therefore also investigated what parameters would separately influence AI and PWV. Other recent studies have sought to validate the measurement of AI, some using the SphygmoCor system, both by comparison with direct intra-arterial recordings, and by estimates of repeatability.5–8 It is clear from some of these that the general transfer function equation used to estimate central waveform has an error rate when applied to individual subjects.9 However, if AI is found to be yet another meaningful estimate of arterial stiffness, which is repeatable for each subject, then its simplicity will permit for the first time an empirical study of its value in predicting complications in hypertension, by its inclusion in large-scale outcome trials.
For our study, we undertook AI and PWV measurements in the offspring of hypertensive sibling pairs previously recruited by us for studies on the genetics of hypertension. The attraction of such subjects is the expectation of a wider spread of values than in an entirely unselected normotensive population, whilst blood pressure itself and drug treatment were not expected to be major confounders as they would be in a hypertensive population.
|
Methods |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Subjects
Invitation letters along with short questionnaires were mailed to the offspring of previously-identified affected sibling pairs in hypertension research clinics at the Clinical Pharmacology Unit, Addenbrooke's Hospital. Ethical approval was obtained from the Cambridge Health Authority, written consent was obtained from the participants. The response rate of participation was 95%. At the Clinical Pharmacology Unit clinic, all subjects filled in a detailed questionnaire regarding a history of their family and themselves, and their personal habits. A total of 105 untreated, healthy offspring aged 19–71 years were seen between March and June 1998. Pulse wave velocity was measured in 40 men and 59 women. Six subjects were on anti-hypertensive drugs. All subjects were included in the analyses and presented here since removal of these six subjects did not result in any substantive changes to the conclusions.
Measurements
Blood pressure was measured using the Datascope Accutorr 4 (Datascope Medical Company). Three supine blood pressure measurements were taken 1 min apart after 10 min rest, and the average of these readings was used in the analysis. Height was recorded to the nearest 0.1 cm using the Hand Held Digital Display Height Measure (CMS Weighing Equipment), and weight to the nearest 0.5 kg using the Stand-on Type Digital Weighing Scales (CMS Weighing Equipment).
Arterial applanation tonometry
The SphygmoCor system (pulse wave velocity system, PWV, and blood pressure analysis system, BPAS) was used to assess arterial stiffness (PWV Medical). The SphygmoCor is one of the recently-developed computerized, portable and simple-to-use devices used to assess pulse waveforms. Aortic pulse waveform, augmentation index, and central aortic pressure were derived at the radial artery by applanation tonometry. The radial and carotid artery sites were used to assess the PWV. The radial and carotid artery pressure waves and amplitude were recorded non-invasively with a pencil-type probe (at the base of neck for the common carotid artery and over the right radial artery). The probe incorporates a high-fidelity strain-gauge transducer at the tip, which has a small pressure-sensitive ceramic sensor area with a frequency response of >2 kHz that is coplanar with a longer area (7 mm diameter) of flat surface in contact with the skin overlying the arterial pulse (Millar Instruments). The probe's technology is based on the principle of applanation tonometry, as used in ocular tonometry for the assessment of intraocular pressure.10 The probe was held on the skin over the maximal arterial pulsation by hand and pressed down on the artery against the underlying bone. Recordings were taken when a reproducible signal was obtained with high amplitude excursion (usually two screens or 10 consecutive beats to cover a complete respiratory cycle are needed for subsequent analysis).
Arterial PWV was determined by the foot-to-foot flow wave velocity method.11–12 The foot of the flow wave was identified between the two recording sites as the beginning of the sharp systolic up-stroke. The time delay was measured between the feet of the flow waves recorded at these different points and designated as pulse transmit time. The distance traveled by the pulse wave was measured over the surface of the body with a tape measure. ECG gating permitted the time lapse between pulse waves at the carotid and radial sites to be calculated from sequential rather than simultaneous measurements. PWV was calculated as the distance : transit time ratio and is expressed as meters per second. AI (the difference between early and late pressure peaks divided by pulse pressure)13–14 was calculated by a computer algorithm derived from invasive pressure and flow data, and is expressed as a percentage (Figure 1
).
|
P/PP. Statistical analysis
Statistical analysis used SPSS for Windows (version 6.1.4). Students t-test was used to compare the group differences, and data were expressed as means and standard deviations. Multiple regression analyses were done to study the relationships of the arterial stiffness measures to other haemodynamic and clinical variables. The repeatability of AI and PWV measurements was established by a paired two-tailed t-test. There were no significant differences between the first and the second measurements either for AI (p=0.90) or PWV (p=0.16). For AI, the mean difference was -0.18±4.69% with 95% CI -3.30 to 2.97%, and for PWV the mean difference was -0.32±0.69 m/s with 95% CI -0.78 to 0.14 m/s. The coefficient of repeatability as defined by the British Standards Institution (RC2=
Di2/n, where Di2 is the absolute difference between measurements and n is the number of subjects) was 4.47% for AI and 0.7 m/s for PWV.15 |
Results |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
The characteristics of the study population are summarized in Table 1
. Because of the large and unexpected difference in AI between men and women, results are shown separately for the two sexes. Also, significant gender differences existed in mean blood pressure, PWV and other variables after adjusting for age (Table 1). A further multiple regression analysis (below) showed that the higher PWV in the men was accounted for by the slightly higher BP, whereas the several-fold higher AI in women was not pressure-related.
|
Correlations between the two measures of arterial stiffness and all other measured variables are shown in Table 2
. Importantly, the two measures of arterial stiffness (AI and PWV) were significantly correlated with each other (Figure 2). The coefficient value observed was 0.29 (p<0.005). A strong inverse correlation was observed between AI and body height (Figure 3), with an r value of -0.45 (p<0.001). No significant heterogeneity existed between the correlations for either AI and PWV with other variables, except for AI and PWV with body height (p<0.001).
|
|
|
The relationship between the two measures of stiffness was much stronger when assessed separately in men and women, being 0.42 in men (Figure 4
) and 0.56 in women (Figure 5). By contrast, the negative correlation between AI and height in the whole cohort lost significance when analysed separately in men (r=-0.12) and women (r=-0.15).
|
|
Table 3
shows the multivariate regression analysis of the influence on AI of age, sex, central aortic systolic blood pressure, heart rate, height and PWV. AI was independently influenced by all six of these clinical parameters, with an adjusted R2 of 0.61.
|
|
Discussion |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
The major findings of this study are that the two measures of arterial stiffness correlated significantly with each other, that AI is much higher in women, and that AI is negatively correlated with height. The first observation provides useful support for the use of AI to assess arterial stiffness. This is potentially important because applanation tonometry at the radial artery is now technically the easiest and quickest of available methods to assess arterial stiffness. Five minutes with a specialist nurse is sufficient to incorporate AI into routine clinic visits; this should encourage the evaluation of arterial stiffness in prediction of vascular complications by inclusion of AI measurements in future outcome trials. However, our other two observations, and the much higher correlation of AI and PWV within genders than overall, suggest that gender is a major confounder, or at least, that different normal ranges need to be established for men and women.
The finding of a correlation between the two measures is consistent with the other single published report.16 However, our results differ in three ways. Firstly, we observed this relationship in healthy, untreated subjects as compared to the previous report which investigated patients with end-stage renal disease and controls, which could possibly result in a positive finding with the two indices of stiffness and over-estimate the correlation, as isolated systolic hypertension and pulse pressure is high in such patients, leading to increased arterial stiffness.11 Secondly, pulse wave velocity was measured between carotid and femoral arterial sites with a doppler flow velocity record technique, which is again different from the sites and method of recording pressure waveforms in the present study. Finally, to our knowledge, the correlation between these two measures within each gender has not been explored to date with the SphygmoCor method. Hence we believe our findings to be unique with respect to the method and the sample studied.
The finding of increased AI in women is consistent with the published literature.17 Our multiple regression shows that the effect is only partly explained by lower height in women. Contrasting with the peripheral (brachial) SBP, the central (aortic) SBP was not different in men and women, a finding in accord with a previous study.17 The inverse correlation found between AI and body height in the whole study group has also been noted previously,16,18–21 and is probably due to earlier reflection of the aortic wave in short people. The strength of the association in the whole group was greater compared to other studies. The absence of a significant relation between AI and height within each gender is not surprising, in that the correlation values observed are similar to those found in another study which investigated the influence of gender on central arterial pressure waveform using much larger numbers.20 The result is probably due to the smaller numbers in subsets, although the weaker influences of height and gender in the multiple regression than in the simple correlation suggests that height may in part be a surrogate for other variables between the sexes.
Arterial stiffness (measured as AI and PWV) is influenced by a number of physiological and anatomical properties. However, with repeatability studies, checks using Bland and Altman plots22 and other measures of reliability, and with contemporary computer technology, it is now quite feasible to investigate AI and PWV accurately with some practice. The excellent correlation (r=0.98, p<0.001) among the repeated measurements for AI, and a strong correlation (r=0.71, p<0.025) for PWV observed in this study does suggest that pulse wave analysis (PWA) is a simple, reproducible non-invasive technique with which to measure AI and PWV. Linking indirect determination of pulse contour by sphygmography with sphygmomanometry, may provide an improvement in management of patients with various illnesses of the arterial system.
In conclusion, the study supports use of AI as a measure of arterial stiffness, but highlights a difference in how AI and PWV are influenced by gender and height. The ease of AI measurement should permit its larger-scale use within trials that can determine whether AI predicts long-term morbidity, thus justifying measurement of arterial stiffness in everyday practice.
|
Acknowledgments |
|---|
We would like to thank all the participants in this study. This work was supported by the British Heart Foundation Project Grant (No: PG95169), UK.
|
References |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
1. Laurent S, Lacolley P, Brunel P, Laloux B, Pannier B, Safar M. Flow-dependent vasodilation of brachial artery in essential hypertension. Am J Physiol 1990; 258:H1004–11.
2.
Bank AJ, Kaiser DR. Smooth muscle relaxation: effects on arterial compliance, distensibility, elastic modulus, and pulse wave velocity. Hypertension 1998; 32:356–9.
3. Nichols WW, O'Rourke MF. McDonalds blood flow in arteries, 3rd edn. Edward Arnold, 1990.
4.
Young T. On the function of the heart and arteries: the Croonian lecture. Phil Trans Roy Soc 1809; 99:1–31.
5. Wilkinson IB, Fuchs SA, Jansen IM, Spratt JC, Murray GD, Cockroft JR, Webb DJ. Reproducibility of pulse wave velocity and augmentation index measured by pulse wave analysis. J Hypert 1998; 16:2079–84.[Web of Science][Medline]
6.
Benetos A, Laurent S, Hocks AP, Boutouyrie PH, Safar M. Arterial alterations with aging and high blood pressure: a noninvasive study of carotid and femoral arteries. Arterioscler Thromb 1993; 13:90–7.
7. Benetos A, Asmar RG, Gautier S, Salvi P, Safar M. Heterogeneity of the arterial tree in essential hypertension: a noninvasive study of terminal aorta and common carotid artery. J Hum Hypertens 1994; 8:501–7.[Web of Science][Medline]
8.
Chen CH, Ting C-T, Nussbacher A, Nevo E, Kass DA, Pak P, Wang SP, Chang MS, Yin FCP. Validation of carotid artery tonometry as a measure of estimating augmentation index of ascending aortic pressure. Hypertension 1996; 27:168–75.
9.
Chen CH, Nevo E, Fetics B, Pak PH, Yin FCP, Maugham L, Kass DA. Estimation of central aortic pressure waveform by mathematical transformation of radial tonometry pressure: validation of generalised transfer function. Circulation 1997; 95:1827–36.
10. Kelly R, Hayward C, Ganis J, Daley J, Avolio A, O'Rourke M. Noninvasive registration of the arterial pressure pulse wave form using high-fidelity applanation tonometry. J Vasc Med Biol 1989; 1:142–9.
11. London GM, Marchais SJ, Safar ME, Genest AF, Guerin AP, Metivier F, Chedid K, London AM. Aortic and large artery compliance in end-stage renal failure. Kidney Int 1990; 37:137–42.[Web of Science][Medline]
12.
Avolio AO, Chen SG, Wang RP, Zhang CL, Li MF, O'Rorke MF. Effects of aging on changing arterial compliance and left ventricular load in Northern Chinese urban community. Circulation 1983; 68:50–8.
13.
Takazawa K, Nobuhiro T, Takeda K, Kurosu F, Ibukiyama C. Underestimation of vasodilator effects of nitroglycerin by upper limb blood pressure. Hypertension 1995; 26:520–3.
14.
Murgo JP, Westerhof N, Altobelli SA. Aortic impedance in normal man: relationship to pressure wave forms. Circulation 1980; 62:105–16.
15. British Standards Institution. Precision of test methods I: guide for the determination and reproducibility for a standard test method (BS5497, part 1). London, British Standards Institution, 1979.
16.
Marchais SJ, Guerin AP, Pannier BM, Levy BI, Safar ME, London GM. Wave reflections and cardiac hypertrophy in chronic uremia: influence of body size. Hypertension 1993; 22:876–83.
17.
London GM, Guerin AP, Pannier BM, Marchais SJ, Stimpel M. Influence of sex on arterial haemodynamics of blood pressure: role of body height. Hypertension 1995; 26:514–19.
18. London GM, Guerin AP, Pannier BM, Marchais SJ, Metivier F. Body height as a determinant of carotid pulse contour in humans. Hypertension 1992; 10 (suppl 6):S93–5.
19.
London GM, Guerin A, Pannier BM, Marchais SJ, Benetos A, Safar ME. Increased systolic pressure in chronic uremia: role of arterial wave reflections. Hypertension 1992; 20:10–19.
20. Hayward CS and Kelly RP. Gender-related differences in the central arterial pressure waveform. J Am Coll Cardiol 1997; 30:1863–71.[Abstract]
21.
Smulyan H, Marchais SJ, Pannier BM, Guerin AP, Safar ME, London GM. Influence of body height on pulsatile arterial haemodynamic data. J Am Coll Cardiol 1998; 31:1103–9.
22. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 88:307–10.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  K. Berend and M. Levi Is it time to celebrate a century of blood pressure management? Nephrol. Dial. Transplant., August 1, 2008; 23(8): 2558 - 2562. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S S M Chen, A E Donald, C Storry, J P Halcox, P Bonhoeffer, and J E Deanfield Impact of aortic stenting on peripheral vascular function and daytime systolic blood pressure in adult coarctation Heart, July 1, 2008; 94(7): 919 - 924. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. J. Sommerfield, I. B. Wilkinson, D. J. Webb, and B. M. Frier Vessel wall stiffness in type 1 diabetes and the central hemodynamic effects of acute hypoglycemia Am J Physiol Endocrinol Metab, November 1, 2007; 293(5): E1274 - E1279. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H.-M. Su, K.-T. Lee, C.-S. Chu, M.-Y. Lee, T.-H. Lin, W.-C. Voon, S.-H. Sheu, and W.-T. Lai Effects of Heart Rate on Brachial-Ankle Pulse Wave Velocity and Ankle-Brachial Pressure Index in Patients Without Significant Organic Heart Disease Angiology, February 1, 2007; 58(1): 67 - 74. [Abstract] [PDF]
|
|
|
|
 |
|
 M. F. O'Rourke, A. Avolio, A. Qasem, S. C. Millasseau, J. M. Ritter, and P. J. Chowienczyk Clinical Assessment of Wave Reflection * Response: Augmentation Index and the Radial-to-Aortic Transfer Function Hypertension, November 1, 2003; 42 (5): e15 - e16. [Full Text] [PDF]
|
|
|
|
|
|
G. E. McVeigh Pulse Waveform Analysis and Arterial Wall Properties Hypertension, May 1, 2003; 41(5): 1010 - 1011. [Full Text] [PDF]
|
|
|
|
 |
|
 S. A. Hope, D. B. Tay, I. T. Meredith, and J. D. Cameron Comparison of generalized and gender-specific transfer functions for the derivation of aortic waveforms Am J Physiol Heart Circ Physiol, September 1, 2002; 283(3): H1150 - H1156. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Lantelme, C. Mestre, M. Lievre, A. Gressard, and H. Milon Heart Rate: An Important Confounder of Pulse Wave Velocity Assessment Hypertension, June 1, 2002; 39(6): 1083 - 1087. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. L. Tycho Vuurmans, W. H. Boer, W.-J. W. Bos, P. J. Blankestijn, and H. A. Koomans Contribution of Volume Overload and Angiotensin II to the Increased Pulse Wave Velocity of Hemodialysis Patients J. Am. Soc. Nephrol., January 1, 2002; 13(1): 177 - 183. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Asmar Effect of telmisartan on arterial distensibility and central blood pressure in patients with mild to moderate hypertension and Type 2 diabetes mellitus Journal of Renin-Angiotensin-Aldosterone System, June 1, 2001; 2(2_suppl): S8 - S11. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||