Central aortic pressure waveforms in hypertension

Abstract

The impact of hypertension on cardiovascular end organs is mediated by increases in steady-state and pulsatile components of blood pressure (BP). The pulsatile component of BP is determined by both a forward travelling pressure wave produced by ventricular ejection and a backward pressure wave generated by wave reflection off points of vascular tapering. Although the backward pressure wave contributes to events, the factors that influence the impact and the most appropriate method of measuring the adverse effects thereof are unclear. In the present thesis I explored several aspects of the determinants, impact and measurement of backward travelling pressure waves. An extended time of the forward pressure wave and increased speed of wave travel with an earlier return of the backward wave enhance overlap of these waves and hence augment pulse pressure (PP). The extent to which these factors influence the impact of backward wave pressures on end organ is nevertheless unknown. In 701 participants randomly selected from a community sample of African descent, I demonstrated that neither the time to the peak of the forward wave nor the speed of wave reflection were independently associated with echocardiographic left ventricular mass (LVM). However, there was a strong interaction between the time to the peak of the forward wave and central aortic PP (PPc) or peak backward wave pressure (Pb) to determine LVM and LV hypertrophy. Thus, the time to the peak of the forward pressure wave is a major determinant of the impact of backward wave pressures and hence PPc on LVMI. Augmentation of PPc (augmentation index [AIx]) is increased in women as compared to men. However, the contribution of the peak of the forward wave pressure (Pf) and Pb to the increases in AIx in women is uncertain. In some populations, Pf contributes more than Pb to the increases in AIx in women, but the role of Pf or Pb in determining sex differences in AIx in populations where Pb plays a major role in explaining variations in PPc, is unknown. In 1097 participants randomly selected from a community sample of African descent, I showed that as compared to men, women have an increased AIx. The most important change associated with sex differences in AIx was an increased reflected wave magnitude (RM=PPb/Pf x 100). Thus, in groups of African ancestry an increased wave reflection rather than forward wave pressures largely accounts for sex differences in AIx. The β-adrenergic receptor blocker, atenolol, is not as effective as alternative antihypertensive agents at preventing cardiovascular events, an action attributed to increases in AIx. However, whether this is a class effect of all β-adrenergic receptor blockers or an effect of atenolol on wave reflection is uncertain. In the present thesis, using simultaneous carotid pressure and aortic outlet tract velocity and diameter assessments and subsequently performing wave separation analysis, I showed that across a range of steady-state pressures (induced by phenylephrine administration), atenolol given to spontaneously hypertensive rats for 3 months resulted in increases in AIx associated with an increased time to peak Pf and an enhanced RM. Thus the deleterious effect of atenolol on AIx is attributed to both a class effect (heart rate-dependent increases in the time to peak Pf) as well as an effect specific to the vascular effects of atenolol on wave reflection. Although aortic backward wave function (RM and Pb) predicts events, the assessment requires wave separation analysis. Whilst AIx has been employed as an easy and reproducible approach to reflected wave function, it underestimates backward wave effects. In this present thesis, in 1367 participants from a community sample I assessed whether the second systolic shoulder of the peripheral pulse waveform [pP2]) more accurately indexes Pb effects than aortic pressure augmentation (Pa). Importantly, pP2 was more closely correlated with Pb than Pa. The largest difference was at the highest tertile of the reflected wave foot time, where the speed of wave reflection was lowest. Consequently, independent of brachial BP and additional confounders, Pb or pP2 but not Pa were independently associated with LVM. Thus, pP2 which is readily identified on a peripheral pulse wave, may be a useful surrogate of Pb effects. In conclusion, in the present thesis I show that an extended time to the peak of the forward wave pressure determines the impact of backward wave pressures on LVM; that atenolol-based antihypertensive therapy not only produced adverse effects on aortic function by extending the time to the peak of the forward wave through heart rate-related effects (class effect), but also increases wave reflection (specific to non-vasodilator β-adrenergic receptor blockers); that the adverse effects of female gender on aortic function in groups of African ancestry are through enhanced backward pressure waves and that backward wave pressure effects, although inadequately indexed by aortic augmented pressure, can be appropriately indexed by the second systolic shoulder of the peripheral pulse. These data provide further insights into the adverse effects and appropriate measures of central arterial wave reflection.