A salient feature of the pressure and flow velocity waveforms recorded in the ascending aorta is the marked difference in wave morphology. They have similar contours only for the first part of systole up to the time of peak flow, after which the wave morphology diverges, where pressure can continue to increase at the same time that flow decreases. After closure of the aortic valve, flow during diastole is essentially zero, while pressure can show secondary oscillations superimposed on a decaying profile until the beginning of the next cardiac systole. The difference in the time-dependent profile is explained in terms of the haemodynamic impedance to cardiac ejection presented by the aorta and the arterial vasculature . The frequency domain representation of the impedance shows an inverse frequency dependency of impedance modulus and a negative phase angle at low-frequency components, indicating that flow leads pressure. This is essentially an expression of the capacitive load due principally to the windkessel properties of the aorta, and the whole impedance spectrum quantifies the steady and pulsatile arterial load . When considering the aortic trunk as a distributed system, the transmission line concept is used to describe wave propagation along a series of distributed segments that can also be described in terms of uniform transmission lines with equivalent windkessel elements . This invokes wave reflection at points of impedance mismatch. Indeed, whether one uses a lumped parameter model or a distributed wave propagation model, the difference between pressure and flow wave shape is attributed to the elastic nature of the arterial load and expressed as a combination of storage and wave reflection phenomena .
The early work on analysis of the impedance spectrum suggested the simple concept of a single tube due to the ‘resonance’ spectral features of a frequency of minimum modulus and zero phase [4,5]. This suggested the possibility of a virtual reflecting site, where waves returning to the origin would appear at certain time points in the cardiac cycle and would depend on the aortic pulse wave velocity (PWV) [4–6]. Hence, waves would arrive early in systole in a stiff system but later (possibly in diastole) in a more compliant system. Although the vascular impedance representation of the arterial load has been recently recast in terms of reservoir capacitive properties with reduced effect of secondary waves as analysed in the time domain , there are examples that indicate quite a prominent role of wave reflection in determining the pressure wave shape. In the aortic pressure wave of the kangaroo, secondary waves are of such high magnitude that peak pressure can occur in diastole, after closure of the aortic valve, or the incisura can be seen on the ascending phase in late systole , rather than the conventional descending phase.
Using the notion that secondary oscillations can augment the stress on the contracting ventricle, waveform features have been used to quantify the augmented pressure by analysing the pressure waveform features by computing an augmentation index (AIx) [9,10]. This has been more widely used with the recent developments of methods to determine the aortic pressure noninvasively from the peripheral pulse  and expressed principally as a systolic AIx (sAIx), that is, the amount of augmented pressure after peak flow. When secondary inflections are found in late systole, it is described as a negative AIx. However, sAIx while showing relationships with age and variable association with arterial stiffness , is also highly dependent on heart rate .
In the current article in the Journal, Heim et al.  describe an enhanced method of pulse wave analysis by including an extension of sAIx by computing an equivalent diastolic AIx (dAIx). This is done by determining the difference in the value of the secondary diastolic wave from the theoretical exponential decay during diastole. The rationale for the study was based on the basic principles outlined above regarding wave reflection. Due to changes in aortic stiffness affecting PWV and consequently times for arrival of reflected waves, any increase in augmented pressure in diastole should be associated with a reduced effect in diastole. Hence the authors hypothesized that there should be an inverse relationship between sAIx and dAIx.
The study was conducted in a group of healthy individuals comprising 48 men and 45 women (19–70 years). Central aortic pressure was reconstructed from the radial pulse using the SphygmoCor device employing a validated radial transfer function [11,15]. Measurements were made in both sitting and supine positions and carotid-femoral PWV (cfPWV) in the supine position. Comparisons of sAIx and dAIX were made between men and women, positions and association with cfPWV.
The main finding of the study was indeed a reciprocal relationship between sAIx and dAIx, as well as a positive association of sAIx with cfPWV and a negative association for dAIx . This suggests that increased stiffness resulting in higher cfPWV will be associated with relatively higher systolic augmentation and reduced diastolic augmentation. This finding by Heim et al.  goes some way to quantifying the qualitative patterns seen by Kelly et al.  in radial, femoral and carotid pulse waveforms recorded by applanation tonometry in over 1000 individuals aged between 2 and 91 years. With advancing age the secondary oscillations seen in diastole diminish, whereas the late systolic augmentation increases. An associated and important aspect of the findings of the study by Heim et al.  is that the aortic waveform was not actually measured, but reconstructed from tonometric recordings of the radial pulse using a transfer function [11,15]. Indeed, this lends support to this technique insofar as the derived central aortic waveform provides similar information in the waveform features as those from measured waveforms in the carotid artery, a surrogate site for the central aorta.
Whereas the study provides evidence that sAIx and dAIx are inversely related, there is a discontinuity in this association in relation to heart rate. sAIx has a strong dependency on heart rate , but no such dependency was found for dAIx. Indeed, the authors advocate the use of dAIx because of this heart rate independency. However, this has to be qualified, since dAIx is relatively smaller in magnitude than the corresponding sAIx and so would also have a much lower discriminating power. Why two linearly related variables should show a divergent relationship with heart rate is not readily apparent. It may be related to the shortened duration of ejection with higher heart rates resulting in the reflected wave arriving with the same delay, but at a later relative time in systole corresponding to a point of increased flow deceleration. Since pressure undergoes a decay after flow cessation due to closure of the aortic valve, it might be thought that the diastolic component of augmentation is less affected by the changes in systolic and diastolic times with changes in heart rate.
The explanation proposed by Heim et al.  for the origin of the components that produce sAIx and dAIx has potential conceptual problems. They suggest that there are two separate reflections, one arriving in systole around 150 ms and the other arriving in diastole around 400 ms. Whereas the reflected wave causing sAIx would seem to be reasonable as a reflection within the dimensions of the body for physiological values of PWV, the same values of PWV would place the site of reflection for the diastolic wave causing dAIx outside the body dimensions. The authors do provide a qualification for this apparent conceptual anomaly in arguing that the concept of single tubes with single terminations at a certain distance may not be applicable to distributed and dispersive systems . One notion that was not considered in terms of the interpretation of sAIx and dAIx is that it may, in fact, be a single reflective phenomenon with increased dispersion, affected by the summation to the incident wave interrupted by the closure of the aortic valve. In humans there seems to be a fortuitous allometric scaling so that the closure of the valve generally coincides with the beginning of the secondary wave. However, the kangaroo aortic pressure illustrated that this is not a general occurrence, and that closure of the aortic valve can coincide with the peak and not the trough of the secondary wave .
The study by Heim et al.  suggests the use of a novel parameter, dAIx as an addition to the conventionally used sAIx. The authors point out that it may be a more robust parameter as there is no dependence upon heart rate. They also confirm the increased pressure augmentation seen in women compared to men, as well as changes in measured AIx due to body posture. However, they do not fully address the complex relationship of postural effects on aortic pressure distribution, some of it related to gravitational effects [18–20]. There is also some question as to the interpretation of the phenomenon of the reciprocal relation between sAIx and dAIx. However, the findings are of interest, not only insofar as they extend the possibilities of quantification of arterial function using pulse wave analysis, but also in demonstrating that this can be achieved using aortic pressure waveforms which were not registered directly by intravascular or external sensors, but derived noninvasively from the peripheral radial pulse.
Conflicts of interest
There are no conflicts of interest.
1. Nichols WW, O’Rourke MF, Vlachopoulos C. McDonald's blood flow in arteries
, 6th ed. London: Hodder Arnold; 2011.
2. Avolio A. Input impedance of distributed arterial structures as used in investigations of underlying concepts in arterial haemodynamics. Med Biol Eng Comput
3. Avolio AP. Multibranched model of the human arterial system. Med Biol Eng Comput
4. O’Rourke MF, Taylor MG. Input impedance of the systemic circulation. Circ Res
5. Taylor MG. Wave transmission through an assembly of randomly branching elastic tubes. Biophys J
6. Avolio AP, O’Rourke MF, Mang K, Bason PT, Gow BS. A comparative study of pulsatile arterial hemodynamics in rabbits and guinea pigs. Am J Physiol
7. Tyberg JV, Davies JE, Wang Z, Whitelaw WA, Flewitt JA, Shrive NG, et al. Wave intensity analysis and the development of the reservoir-wave approach. Med Biol Eng Comput
8. Nichols WW, Avolio AP, O’Rourke MF. Ascending aortic impedance patterns in the kangaroo: their explanation and relation to pressure waveforms. Circ Res
9. Yaginuma T, Noda T, Tsuchiya M, Takazawa K, Tanaka H, Kotoda K, Hosoda S. Interaction of left ventricular contraction and aortic input impedance in experimental and clinical studies. Jpn Circ J
10. Murgo JP, Westerhof N, Giolma JP, Altobelli SA. Manipulation of ascending aortic pressure and flow wave reflections with the Valsalva maneuver: relationship to input impedance. Circulation
11. 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
12. Sakurai M, Yamakado T, Kurachi H, Kato T, Kuroda K, Ishisu R, et al. The relationship between aortic augmentation index and pulse wave velocity: an invasive study. J Hypertens
13. Wilkinson IB, MacCallum H, Flint L, Cockcroft JR, Newby DE, Webb DJ. The influence of heart rate on augmentation index and central arterial pressure in humans. J Physiol (Lond)
14. Heim A, Liaudet L, Waeber B, Feihl F. Pulse wave analysis of aortic pressure: diastole should also be considered. J Hypertens
15. Chen CH, Nevo E, Fetics B, Pak PH, Yin FC, Maughan WL, Kass DA. Estimation of central aortic pressure waveform by mathematical transformation of radial tonometry pressure. Validation of generalized transfer function. Circulation
16. Kelly R, Hayward C, Avolio A, O’Rourke M. Noninvasive determination of age-related changes in the human arterial pulse. Circulation
17. Westerhof BE, van den Wijngaard JP, Murgo JP, Westerhof N. Location of a reflection site is elusive: consequences for the calculation of aortic pulse wave velocity. Hypertension
18. Huijben, A MT, Mattace-Raso FUS, Deinum J, Lenders J, van den Meiracker AH. Aortic augmentation index and pulse wave velocity in response to head-up tilting: effect of autonomic failure. J Hypertens
19. Avolio A, Butlin M, Xu K. Reflections on determinants of augmentation index. J Hypertens
20. Nurnberger J, Michalski R, Turk TR, Opazo Saez A, Witzke O, Kribben A. Can arterial stiffness parameters be measured in the sitting position? Hypertens Res