Heart and kidneys

Since the early twentieth century, when Bramwell and Hill [1] described the association of the velocity of the arterial pulse and the elasticity of arteries, there has been increasing interest in the use of pulse wave velocity as a surrogate measure of arterial stiffness, due principally to the intrinsic relationship of the speed of the pulse traveling along the artery wall and the mechanical properties of the wall determining the storage capacity of the vessel. In recent years, stiffness of large conduit arteries has been shown to have a powerful association with cardiovascular risk factors leading to increased mortality and morbidity, mainly due to consequent elevation of age-related arterial blood pressure (BP) and its sequelae [2]. The elucidation of arterial stiffness, specifically stiffness of the aortic trunk, as an independent risk factor has been made possible by the fact that measurement of pulse wave velocity, as a surrogate measure of arterial stiffness, can be readily performed noninvasively [3]. Vascular stiffness varies throughout the arterial tree, with a general increase from the central aorta toward the periphery; significant variability also occurring along the aortic trunk when measured along regional segments [4]. Age-related changes among different populations occur most prominently in the aorta compared with peripheral limbs, in which pulse wave velocity is readily measured [5]; changes have been related to physical and structural properties of the vessel wall [6]. The more pronounced age-related stiffening of the aorta compared with peripheral sites has been attributed to contribute to the development of isolated systolic hypertension in the elderly [7]. However, the differential change between aortic and peripheral arterial stiffness has been recognized as a potential marker of cardiovascular risk [8]. Indeed, the concept of stiffness gradient has been implicated in local changes affecting plaque rupture [9] to global effects related to risk prediction in renal disease [10,11]. The study by Bruno et al.[12] in this issue of the Journal of Hypertension addresses the concept of stiffness gradient in a novel way by comparing the stiffness of the aortic trunk as measured by carotid–femoral pulse wave velocity (cfPWV) and the carotid stiffness measured in a more direct manner by assuming the original concept of Bramwell and Hill [1] in relating changes in vessel cross-sectional area and pulse pressure (PP). Although the difference in carotid and aortic stiffness was not compared in the form of a stiffness gradient as in other studies, carotid stiffness and cfPWV were related to properties of dysfunction of separate target organs, namely the heart and kidneys, respectively. Notwithstanding the limitations stated by the authors, this study has a number of notable features that can pave the way for future significant investigations, principally on how to manage population normal values in specific disease groups. The study by Bruno et al.[12] was conducted in a group of 314 patients with essential hypertension and comparisons were made with a group of 110 age and sex-matched normal patients. cfPWV was measured with applanation tonometry of carotid and femoral arteries to register pulse transit times; carotid stiffness was measured using validated contour-tracking methodology applied to ultrasound images of carotid diameter [13]. Vessel distensibility was computed using changes in carotid diameter associated with PP. An important feature of this measure in this study is that carotid stiffness was expressed in units of pulse wave velocity (m/s), thus enabling direct comparisons with cfPWV. However, a significant caveat then arises that carotid stiffness is a local measure at a single location in the carotid artery (of approximately 1 cm), whereas cfPWV is a global average over the whole of the aortic trunk and iliac and femoral segments. In both control and hypertensive cohorts, carotid stiffness (6.19 and 6.91 m/s, respectively) was significantly lower than cfPWV (7.37 and 9.36 m/s, respectively). A significant observation here is that an estimated difference of mean BP between control (86.4 mmHg) and hypertensive (99.2 mmHg) cohorts (increase of 15%) was associated with only 11.6% change in carotid stiffness but a 27% change in cfPWV. This implies that the cfPWV measurement has more than double the pressure sensitivity than the carotid stiffness metric as reported in this study. The analysis procedure used in the study by Bruno et al.[12] is also informative in providing an insightful approach to the understanding of how to obtain useful information from normal population values of arterial stiffness. The two approaches involved using a similar cutoff threshold (90th percentile) obtained from the published reference values for the normal (nonhypertensive) population for carotid stiffness [14] and for cfPWV [15], and from the control cohort used for the study. When published reference values were used, there was no association of carotid stiffness or cfPWV with any target organ damage. However, when the local control cohort was used, there was a strong separation effect: increased carotid stiffness was independently associated with increased left ventricular mass (LVM) in the presence or absence of increase in cfPWV, and increased cfPWV was associated with reduction in estimated glomerular filtration rate (eGFR) also in the presence or absence of increased carotid stiffness. The effect was very pronounced for eGFR with large odds ratios (ORs), particularly in the presence of concomitant increase of cfPWV and carotid stiffness [OR 13.27 (3.86–45.58)], with a significant but relatively lower effect for LVM [OR 2.86 (1.15–7.09)]. Although the analyses in the study are robust, there is no clear indication of the possible reasons for the large discrepancy in the OR between the measures and particularly the rather large confidence intervals (CIs) for the association of cfPWV and eGFR (95% CI > 45). In addition to the intricate association of eGFR, hypertension and metabolic factors, one possible explanation might be the much reduced pressure sensitivity of carotid stiffness compared with cfPWV, as seen for the comparison between the hypertensive and control cohort in this study. Notwithstanding the smaller mean difference in carotid stiffness compared with cfPWV in the hypertensive and control cohorts, the strong association of only carotid stiffness and not cfPWV (a measure of overall aortic stiffness) with LVM would seem somewhat counterintuitive, as it is the impedance of the aorta that presents the majority of the load on the ejecting ventricle [16], with the carotid vasculature presenting only a relatively small part of the total load. However, as the authors explain [12], carotid stiffness may reflect a measure of the stiffness of the material of the proximal aorta distal to the aortic root, due to the close anatomical association, hence also reflecting the proximal aortic pulsatile ventricular load. A similar association of carotid stiffness with LVM has also been found in exercising individuals, in whom carotid stiffness was measured with similar ultrasound techniques [17]. The strong and independent association of cfPWV and eGFR suggests a potential hemodynamic mechanism involving increased pulsatility in the peripheral aortic trunk affecting kidney function through increased mechanical load on the kidney microvasculature as proposed in studies addressing effects of PP changes in arterial stiffness gradient in renal patients [8,10,11,18,19]. The study by Bruno et al. also describes a form of arterial stiffness gradient, but one in relation to pressure sensitivity. The stiffer aortic trunk and peripheral conduit arteries have a greater change with arterial pressure, and this is associated with organ damage targeting the kidney. A relatively less-stiff carotid region, with less pressure sensitivity is associated with organ damage targeting the heart. Although the study cannot establish causality, it offers an insight into the use of noninvasive techniques of assessing arterial function for enhanced investigations of effects of basic hemodynamics for better stratification of patient-specific cardiovascular risk, leading to potential improvement in treatment and management of cardiovascular disease. ACKNOWLEDGEMENTS Conflicts of interest There are no conflicts of interest.