CAVI (Cardio-Ankle Vascular Index) is an index of global arterial stiffness, including properties of both large elastic and more peripheral arteries, which has been proposed to measure the degree of arteriosclerosis non-invasively. CAVI was reported to be independent of blood pressure measured at the time of its evaluation, at variance from the classic assessment of arterial wall properties through carotid-femoral pulse wave velocity. There is however a need to explore more in detail the relationship between CAVI and blood pressure, by focusing not only on office blood pressure but also on 24 h blood pressure profiles, which can be obtained by ambulatory blood pressure monitoring. This technique is able to assess 24 h, daytime and night-time blood pressure mean values and blood pressure variability. How the 24 h blood pressure profiles, which carry relevant prognostic information, relate with CAVI, which is an established marker of vascular organ damage, is an important issue for present and future research.
Assessment of arterial wall properties, and in particular of arterial stiffness, is becoming increasingly important in clinical medicine, since arterial stiffness is acknowledged not only as a determinant of blood pressure (BP) regulation, but also as an early marker of subclinical target organ damage in hypertension. Many parameters and methodological approaches have been so far proposed to explore arterial wall properties and to quantify arterial stiffness, and each of them has been shown to carry both advantages and disadvantages in relation to everyday clinical applications. Among these methods, carotid-to-femoral pulse wave velocity (PWV) is the one most frequently applied both in clinical and in research settings, and is currently recommended as the most clinically relevant measure for aortic stiffness.1 Its use has been recommended in the last European Society of Hypertension/European Society of Cardiology Hypertension Management Guidelines, with a supporting statement of Class IIa level B evidence, as a clinically useful approach aimed at detecting large artery stiffness as a type of hypertensive target organ damage.2
However, PWV through large elastic arteries is also strictly dependent on BP levels at the time of measurement. This obviously affects its clinical application, especially in the presence of elevated BP levels, and may have substantial implications both in risk assessment and in monitoring the effects of treatment on arterial wall properties in individual patients. Moreover, the routine use of PWV for risk stratification is made difficult by the relative complexity of its assessment, which needs well-trained operators and standardized procedures.3
Some years ago, a new approach to the evaluation of arterial stiffness was proposed through calculation of Cardio-Ankle Vascular Index (CAVI),4 computed by combining the stiffness parameter β and the Bramwell-Hill formula.5,6 The PWV included in the equation is the heart-ankle PWV, making CAVI suitable to evaluate global vascular wall stiffness across the aorta, including its segments from the proximal ascending aorta down to the femoral and tibial arteries.7 The inclusion of the β parameter in the calculation makes CAVI independent from BP values at the time of its measurement.8 Overall, CAVI, with respect to PWV, offers some significant advantages: besides being non-invasive, it is easier to obtain, BP independent, represents a global index of both central and peripheral stiffness, and thus it appears to be a suitable candidate for the routine evaluation of vascular organ damage and for the estimation of cardiovascular risk, especially in hypertensive subjects.9 CAVI is now emerging as a predictor of future cardiovascular events in longitudinal studies, independently of traditional cardiovascular risk factors.10,11
In the clinical management of hypertensive patients, a frequently used methodology for an accurate assessment of BP levels in daily life is represented by 24 h ambulatory BP monitoring (ABPM), which provides plenty of information on systolic and diastolic day and night BP profiles, on ambulatory average BP levels, as well as on indices of BP variability over the same time periods.12 The latter indices are particularly relevant in relation to arterial wall properties, because BP variability has been shown to be related, among other factors, also to the degree of arterial stiffness.13 This may influence the aortic systolic BP achieved in correspondence to each cardiac contraction, and therefore cardiac afterload. The degree of systolic BP fluctuations with increased aortic stiffness is also influenced by a possible impairment of the carotid and aortic baroreceptors activity, due to their reduced stimulation determined by the altered distensibility of stiffer arteries.
In spite of the importance of the information provided by ABPM, it has not been clearly elucidated yet whether and how much the variables derived from 24 h ABPM relate to CAVI. Indeed, the relationship of CAVI with variables derived from analysis of 24 h ambulatory BP profiles, and how all these parameters may interact in determining the development of vascular organ damage and vascular aging in hypertensive individuals, remain an important target for future research.
CAVI and 24 h ambulatory blood pressure monitoring
The relationship between arterial stiffness and BP levels, and the association of high BP levels with the loss of arterial elastic properties have been widely investigated, although most of the studies have been based only on office BP measurement. On the contrary, the relationship between 24 h BP profiles and arterial mechanical properties still needs to be considered more in detail. In this context, most information has been obtained through assessment of the association of BP mean levels and variability with carotid-femoral PWV, while little is known of the relationship between such 24 h BP parameters and CAVI. CAVI was indeed shown to be a pressure independent parameter for the evaluation of arterial mechanical properties, when focusing on office or clinic BP values. Since BP is subjected to continuous variations during the 24 h, however, it would be important to explore the possible relationship between CAVI and these dynamic features of ambulatory BP. This would also allow us to overcome possible biases due to transient variations in BP at the time of office measurement and to the inherent inaccuracy of isolated office BP readings.12,14
In the general population as well as in hypertensive patients,15,16 arterial stiffness assessed by carotid-femoral PWV is significantly related not only to conventional office BP measurements but also to 24 h ambulatory BP levels, and has been shown to carry prognostic information. Even after adjustment for mean 24 h BP, arterial stiffness measured as carotid-femoral PWV, seems to maintain its prognostic significance,15 although the nature of this predicting value is not clear, given the intrinsic BP dependency of PWV.17
In our centre, we are currently exploring whether a parameter such as CAVI, reported to be independent from office BP values, would correlate with ABPM derived mean BP values and with cardiovascular organ damage in essential hypertensive patients. We are assessing such a relationship in a population of essential hypertensive patients across an age range from 14 to 90 years. In this study, we are performing a thorough clinical evaluation in these patients, accompanied by measurement of CAVI with the VaSera device (Fukuda Denshi, Tokyo, Japan), and by 24 h ABPM (TM-2430, A&D, Tokyo, Japan). The study population will include patients with an equal distribution between the two genders. Considering office BP values, an interim analysis of our ongoing study seems to suggest a tendency, at univariate analysis, for a statistically significant relation not only between CAVI and age, but also between CAVI and office BP values. The relationship between CAVI and office BP, however, appears to be much weaker in a multiple regression analysis, while the relationship of CAVI with age seems to remain significant. Given that 24 h ABPM is known to be a better predictor of cardiovascular events than office BP measurements in hypertensive patients, we are also exploring the relationship between CAVI and ABPM values. During the same preliminary analysis, in a univariate model a relationship also seems to be present between CAVI and some ambulatory BP values although such relationship, also in this case, seems to become weaker when analysed within the framework of a multivariate model including age. These preliminary observations need however to be confirmed in the final analysis when our study is completed. Our data collection is continuing, and the final aim of this ongoing study is indeed to clarify whether, in a hypertensive population, a BP independent measure of arterial stiffness such as CAVI relates with ambulatory BP values, an issue which has both clinical and pathophysiological relevance.
The aim of our study will also be to explore the relationship between CAVI, not only with average ambulatory BP levels over 24 h but also with changes in the 24 h ambulatory BP profile, in the same hypertensive patients. It should be emphasized that there might be substantial differences between the BP measured in the doctor's office and the average BP values recorded during the day and night, with the identification of specific BP phenotypes. When office BP is significantly higher than average ambulatory BP values this may be due to a white coat effect raising BP during a doctor’s visit. As a consequence, when office BP is repeatedly above normal reference levels while 24 h ambulatory BP remains within a normal range, this condition is defined as ‘white coat’ or ‘isolated office’ hypertension.12,14 Evidence is available that such a condition carries a lower risk of organ damage or events than sustained hypertension, i.e. a condition where both office and ambulatory BP are elevated. However, some studies suggest that white coat hypertension might be the sign of an alteration in the arterial tree properties that might not be entirely innocent. Indeed, a strong association was found between a surrogate measure of the white-coat effect, based on the difference between office and ambulatory BP values, and arterial stiffness, with the white-coat effect being suggested to be a clinical manifestation of increased arterial stiffness.18 How CAVI relates with this commonly found BP phenotype is an important issue that needs to be investigated in future research.
An association with an increased risk of organ damage and cardiovascular events has been reported even more clearly for an opposite condition, characterized by the occurrence of normal office BP values associated with elevated ambulatory BP levels, known as ‘masked hypertension’.12,14 This BP phenotype, which is highly prevalent in patients with conditions like obstructive sleep apnea and type-2 diabetes mellitus, has been shown to carry an independent association with measures of arterial stiffness in such patients.19,20 Many common pathophysiological mechanisms and risk factors, like obesity and hypertension, are shared between masked hypertension and arterial stiffness, and can interact to increase the probability of vascular damage. As a consequence of these findings, it has been suggested that an early diagnosis of the vascular wall alterations by PWV or by CAVI assessment might be helpful for a preventive management of later cardiovascular complications in masked hypertension.
It has been suggested that circadian variations in blood pressure may affect the risk of cardiovascular events, such as acute coronary and aortic syndromes.21–23 This is the case for nocturnal BP reduction (‘dipping’) and for morning BP surge. In particular, the degree of morning BP surge, which can be evaluated using ABPM, has been shown to independently predict cardiovascular outcomes.24 Evidence is also available that, in hypertensive subjects, morning BP surge is associated with arterial stiffness evaluated as carotid-femoral PWV.25 Similarly to BP, which typically shows circadian BP changes over 24 h following the wake-sleep cycle, also arterial stiffness, which is modulated by vascular tone and by BP levels, may display circadian variations. However, available studies have shown inconsistent results regarding the possible occurrence of diurnal variations in PWV,26–28 although these inconsistencies could partly reflect lack of standardization in the methodology of data collection in these papers. Interestingly, in one study, CAVI was found to be higher when measured in the morning than at other times over the 24 h, even after adjustment for mean BP levels.29 Current guidelines for assessing arterial stiffness1 already recommend to standardize the time of arterial stiffness measurements in research studies, a suggestion which would apply also to CAVI measurement. Nevertheless, the available evidence on the occurrence of significant circadian variations in CAVI seems to be yet insufficient to recommend any specific time of the day for CAVI assessment in the everyday clinical practice, an issue which thus require additional investigations to be better clarified.
Data are also available to support the concept that the BP change from day- to night-time is an important determinant of arterial stiffness, as demonstrated by measuring aortic PWV values.16 However, the relationship between arterial stiffness and dipper-status appears to be of some complexity, being characterized as having a J-shaped nature, with extreme-dippers and reverse-dippers having higher carotid-femoral PWV values than dippers. This relationship remains significant even after multiple adjustment for covariates, with reverse-dippers showing the highest PWV values.30 Increased sympathetic tone at night has been suggested as a potential mechanism for this altered BP and vascular wall properties profile, being in the end responsible for the reported increase in target organ damage and for a worse cardiovascular prognosis in these patients. The relationship between dipper/non-dipper pattern and elastic properties of the arterial system, was evaluated using CAVI in a study considering a large population of hypertensive diabetic patients.31 In this study, a linear correlation between the percentage of BP nocturnal decline and CAVI was found, with considerably elevated values of CAVI in patients with a reverse-dipping pattern (Figure 1). Therefore, in the hypertensive diabetic patients, which are already characterized by an increased cardiovascular risk, the association with reverse dipping and the alteration in vascular properties determined by CAVI seems to be a frequent association, which needs to be regarded specifically by the physician.
The relationship of CAVI with these specific 24 h BP pattern alterations needs to be more widely investigated in future research, and in our study we will try to at least in part fill in this gap of knowledge.
CAVI and ABPM-derived indices of short term blood pressure variability
Epidemiological and observational clinical studies have repeatedly shown that cardiovascular risk is not only related to an increase in mean arterial pressure values or to alterations in the pattern of BP changes from wake to sleep and from sleep to wake, but also to the degree of short term 24 h BP variability (BPV), independently from the contribution of the increased average values of systolic or diastolic BP.32 Actually the degree of short term BPV has been repeatedly shown to represent an additional correlate of, and a causal factor for hypertension-related cardiovascular complications.32
BPV consists of different components which can be assessed over different time intervals and with different methods: (1) very short term BPV, assessed by means of continuous beat-to-beat BP recordings, (2) short term BPV, assessed within a 24 h period, by 24 h ABPM monitoring, (3) mid-term BPV, assessed day-to-day through home BP monitoring, or (4) long term BPV, assessed visit-to-visit or across different seasons by means of repeated office BP measures or through repeated performance of home or ambulatory BP monitoring. Each of these BPV components, when showing an increase, has been related to the development or the progression of hypertensive target organ damage and to an increased risk of cardiovascular events and mortality.32
BPV in itself is a physiological expression of spontaneous cardiovascular regulation, and reflects the degree of sympathetic activation,33 the effectiveness of cardiovascular baroreflex modulation34 as well as the effects of other intrinsic control mechanisms, both at rest and in response to a number of environmental factors35,36 including physical exercise and emotional stress. The baroreflex modulation of cardiovascular homeostasis may itself affect changes in sympathetic activity,37 causing reflex changes of arterial tone and modifications of cardiac output (including vagally mediated changes in heart rate), that may affect BPV. Indeed, a reduced arterial baroreflex sensitivity has been reported to be associated with an increased 24 h BPV.36,38
The arterial baroreflex is in fact one of the most important mechanisms for short-term control of BP. This reflex system consists of stretch receptors located in the carotid arteries, at the carotid sinus level in proximity to the origin of the internal carotid artery, as well as in the wall of the aortic arch. Actually, the aortic arch and carotid arteries are prevalently elastic vessels, particularly liable to arteriosclerotic degeneration with advancing age and in case of endothelial alterations due to atherosclerotic processes. It is important to underline that arterial baroreceptors respond to the extent of recurrent stretching/relaxation of carotid/aortic arterial walls in response to changes in pulsatile BP, rather than directly to changes in BP values themselves. Thus, in subjects with increased arterial stiffness, whose vessels are less distensible than individuals with more elastic arteries in response to BP changes, baroreflex responsiveness will be reduced, and such a baroreflex dysfunction might affect the degree of short term BPV.
This pathophysiological hypothesis is supported by studies showing that, in animals as well as in humans, increased local carotid stiffness is associated with reduced cardiovagal baroreflex sensitivity.39,40 Okada et al.41 clearly demonstrated a significant inverse correlation between the sympathetic baroreflex sensitivity and the degree of carotid stiffness as well as between sympathetic baroreflex sensitivity and aortic stiffness, assessed by carotid-femoral PWV.
Apart from the baroreflex control, BPV may be influenced directly from large artery mechanics. The passive effects of arterial mechanical properties clearly influence BP pulsatile behaviour, so that in a stiff arterial system an increased stroke volume variability and an increase in sympathetic activity in response to environmental challenges may produce wider BP fluctuations, compared to what occurs in an elastic arterial system. This phenomenon can be understood by considering the ‘Windkessel’ properties of the large arteries. In a stiffer model, stroke volume cannot be properly cushioned, and an increase in systolic BP is accompanied by a reduction in diastolic BP, resulting in a wider pulse pressure. Considering the physiological variations in stroke volume over 24 h, short-term systolic BPV will therefore be amplified in the presence of increased large artery stiffness.
Relevant determinants of short-term BPV are therefore both the active mechanisms involved in the reflex closed-loop control of BP (i.e. the arterial baroreflex), and the passive mechanical elastic properties of large conduit arteries. These mechanisms are inextricably intertwined with each other and with other mechanisms influencing BP variations in vivo, such as mechanical (e.g. ventilation), neurohumoral and psychological factors, and may contribute to modify BPV in a complex manner.
However, determining the relative contribution to BPV by large artery stiffness and fluctuations in stroke volume is made difficult by the fact that in vivo active and passive mechanisms of BP control are not only influenced by the above mentioned intrinsic mechanical, neurohumoral or psychological factors, but also by external, environmental factors. This has led some authors to develop models of simulated systemic circulation under controlled conditions in a laboratory setting in order to better explore the specific effect of large artery mechanics on BPV. In a recent study, Avolio et al.42 developed a lumped parameter model of systemic circulation in order to quantify the effects of large artery stiffness on BPV due to changes in vascular (total peripheral resistance) or cardiac (stroke volume and heart rate) properties without the interference of active mechanisms involved in closed loop control (i.e. autonomic reflex modulation). This study showed that BPV is related to the degree of pressure dependency of arterial compliance. Since the stiffness of large arteries is known to increase with age and with increases in BP (i.e. distending pressure), this study also evaluated the relative contribution of these factors to BPV. It was shown that the hemodynamic effect of age-related increases in large artery stiffness is an increase in BPV, which is clearly evident in the elderly as compared to the young, when exposed to the same BP values. In other words, the effect of age-related changes in large artery stiffness seems to increase BPV in the elderly at similar pressures as in young individuals. Interestingly, the study showed that the optimal mean BP value, where BP variability is minimal, is also age-dependent due to the known age-related changes of pressure dependency on aortic stiffness.43 These findings are relevant since this phenomenon might explain the increased risk of vascular events (i.e. small vessel disease in the brain) in subjects with mean BP controlled with antihypertensive treatment, but with a persisting increase in BPV, as reported in several studies.44 CAVI would ideally provide an easy way to evaluate the pressure-dependency of arterial stiffness, as it is derived from the β parameter, which is a measure of the exponent of the vascular BP-diameter relationship. However, the possibility to estimate the blood-pressure dependency of the arterial system by CAVI, and to relate it to short-term BPV needs to be evaluated in physiological and clinical studies.
Evidence is available that arterial stiffness and short term BPV are closely interrelated phenomena (Figure 2). A reduced baroreflex sensitivity, which is a characteristic feature of the altered autonomic cardiac modulation in hypertension, might thus be one of the factors, together with the accompanying increase in arterial stiffness, responsible not only for the increased BPV typical of hypertensive subjects,45 but also for the higher speed of changes in beat-to-beat systolic BP fluctuations reported to occur in hypertensive patients as compared with normotensive individuals.46 The effects of changes in baroreflex sensitivity on BPV should be considered in the context of the complex interactions among different mechanisms involved in cardiovascular regulation, including among them chemoreflex influences which may play a major role in conditions of hypoxia exposure.
In two large cohorts of treated and untreated patients, we recently demonstrated that short-term 24 h BPV and large artery stiffness, the latter assessed as carotid-femoral PWV, are significantly related to each other in human hypertension.13 The relationship between short-term BPV and arterial stiffness was confirmed after excluding multiple confounding factors, and survived after accounting for differences in average BP levels, thus demonstrating the independence of this association from office and 24-h average BP. As it is possible to quantify BPV from ABPM in different ways, various measures of BPV were explored. The stronger association with carotid-femoral PWV was found with measures of BPV focusing on short-term changes of systolic BP, such as those assessed as average real variability (ARV, the average difference between successive BP measurements over 24 h, Figure 3) and weighted-24 h standard deviation (i.e. the average of daytime and night-time standard deviation separately considered and weighted for the different duration of daytime and night-time, a calculation which therefore excludes the influence exerted on 24 h standard deviation by the day-night BP changes). Indeed, both these measures of short term BPV are free from any influence by the degree of day-night BP changes. No measure of diastolic BPV was associated with arterial stiffness in this study, which is not surprising if we consider that, with the stiffening of large arteries, diastolic BP is progressively lower and less modulated.
Also mid and long term BP variability may be associated with the degree of BPV. An association of home day-to-day BPV with PWV was reported, independent of other known risk factors, in Japanese patients with type 2 diabetes.47 Within-visit systolic BPV, estimated during a single clinic visit, was also found to have a closer correlation than mean SBP levels with cardiovascular risk factors and with CAVI.41 CAVI is of particular interest in this context because, as previously discussed, it is affected by changes in wall properties of both elastic and muscular arteries, and because it is independent from mean BP levels.
An association between short term systolic BPV parameters, obtained from ambulatory BP monitoring and PWV, has been reported also in young healthy volunteers.48 Conversely, visit-to-visit systolic BPV showed no correlation with arterial stiffness, measured with CAVI, nor with mean systolic BP, in a small cohort of treated Japanese hypertensive patient. On the other hand, it showed a correlation with left ventricular diastolic dysfunction, therefore suggesting a possible role of arterial stiffness in explaining such an alteration.49
The increasing evidence of a significant association between short term BPV and arterial stiffness, emphasizes the importance of arterial wall property changes in relation to the effectiveness of cardiovascular control mechanisms and to the degree of hemodynamic fluctuations. These findings may suggest that attention to mean BP levels in the management of patients with arterial hypertension may not be enough to achieve cardiovascular protection. The additional assessment of the degree of BPV by ABPM and, at the same time, of the degree of arterial stiffness, appear therefore both to be clinically relevant, and the possibility of their direct association deserves to be further assessed in future studies. These studies should also include assessment of vascular wall properties through CAVI measurement, i.e. through the assessment of a parameter shown to be independent from BP levels.
CAVI has been proposed as a tool for the assessment of atherosclerotic cardiovascular alterations and its ability as independent predictor of cardiovascular events is now emerging.7,10,11 Although the ability of CAVI to evaluate arterial wall properties independently from BP values at the time of measurement makes this index promising in the perspective of an accurate evaluation of how arterial stiffness translates into an alteration of the 24 h blood pressure profile, data from large cohorts about the relationship of this parameter with 24 h ambulatory BP levels, with indices of short-term BPV and with the occurrence of specific alterations in 24 h BP profiles are currently lacking. The importance of these issues strongly underlines the need for further investigations aimed at filling this gap of knowledge, while at the same time further supporting the suggested predictive ability of CAVI in hypertensive patients.
Conflict of interest: G.P. has received honoraria for lectures by Fukuda Denshi. The other authors have no conflicts of interest to declare.