Arterial hypertension

Arterial hypertension remains one of the main modifiable contributors to mortality and morbidity in the developed world. Its prevalence in developing nations is high and increasing [1]. However, arterial hypertension rarely causes morbidity per se, but rather by its deleterious effects on the heart and blood vessels [2]. The relationship between the three entities arterial hypertension, blood vessels and cardiac structure has been intensively investigated, but remains somewhat elusive. In this issue of the journal, Palmieri et al. [3] add to that literature. The authors retrospectively analysed the data from 667 selected participants in the losartan intervention for endpoint reduction in hypertension (LIFE) study, aiming to delineate relationships between arterial stiffness and left ventricular geometry in patients with arterial hypertension and target organ damage. A number of publications have originated from the same dataset demonstrating the value of post-hoc analyses of large clinical studies. Their study raises a number of issues of relevance to the exploration of the relationship between cardiac structure and arterial stiffness. Cardiac structure is a consequence of long-term myocardial remodelling and is not greatly influenced by acute changes in blood pressure and haemodynamics. Arterial hypertension increases left ventricular afterload chronically leading to physiological myocardial hypertrophy [4]. The degree of left ventricular hypertrophy reflects the average afterload to which the heart has been exposed during the course of hypertension [5,6]. In addition, left ventricular mass is influenced by body size and independently by gender, with mass in men approximately 1.5 times that in women [7]. Whilst isolated episodes of transient blood pressure elevation will not alter cardiac muscle mass, it is conceivable that they may alter model-based two-dimensional measurements, as used by Palmieri et al. [3], by changing estimated volumes [4,8]. Such an effect on a population-based study [3] is presumably small. By contrast, blood vessels are affected by arterial hypertension both acutely and chronically. Due to the non-linear pressure–volume relationship, which exists throughout the vasculature, arterial stiffness is positively related to distending pressure [9]. The higher the blood pressure at the time of the measurement, the higher the measured arterial stiffness without any change in the structure of the vessel. Such changes have been termed ‘functional’ as distinct from ‘structural’ modification of arterial stiffness. Attempts have been made to model non-linear arterial behaviour and thereby predict arterial stiffness at a different blood pressure. This has been fraught with difficulty and the potential for spurious results by under- or overcorrection [10–12]. In the long term, arterial hypertension leads to increases in conduit artery diameter [11,13], wall hypertrophy [11,14] and accelerated atherosclerosis [11,14, 15]. Whilst the early adaptive changes in arterial hypertension, as we now know, reduce isobaric arterial stiffness [11,16,17], this effect is later reversed by accelerated atherosclerosis increasing arterial stiffness above and beyond the initial decrease [18,19]. In general, the effect of the current level of blood pressure at the time of measurement far outweighs any long-term effects [11]. Other noteworthy determinants of arterial stiffness are vessel diameter (smaller vessels are stiffer) and age (older individuals have higher stiffness) [20]. There are many methods to determine regional or systemic arterial stiffness, as recently reviewed in depth [21]. The determination of systemic arterial compliance as the ratio of stroke volume over pulse pressure (SV/PP) assumes a model whereby all of stroke volume is ‘stored’ in the large systemic arteries during systole. This is obviously not the case [20]. Despite this, SV/PP is relatively easy to determine, conceptually simple, and relates well to other measures of central arterial stiffness. It has been investigated and published extensively for many decades [22,23]. Both compliance expressed as SV/PP and PP/SV have been used in the literature. In their study, Palmieri et al. [3] chose to further multiply PP/SV by body surface area yielding PP/SV index (PP/SVi). Whilst the latter is not described in detail in the methods or the reference they quote, the same authors have previously applied the Dubois formula calculating body surface area (m2) as 0.20247 × height (m)0.725 × weight (kg)0.425 [24] to be used in this index. Despite its previous use [22], the rationale for such a composite index remains elusive. Presumably this multiplication may correct for effects of body habitus on both stroke volume, arterial diameter, or both. In one of the very few published studies evaluating SV/PP and related indices as predictors of future cardiovascular morbidity and mortality, neither body height nor PP/SVi were predictors of future events [25]. It would be desirable to demonstrate in any studied cohort that SV/PP relates to body surface area before correcting for its effect. In the study by Palmieri et al. [3], PP/SVi was unrelated to either height or weight (compare their Tables 3 and 1, respectively). This could be the result of either a perfect correction or the absence of a relationship combined with small variability in height and weight in the group. The fact that cardiac structure is related to the long-term level of blood pressure, whereas measures of systemic compliance such as SV/PP are mostly instantaneous, will attenuate the significance of any relationship between cardiac and vascular parameters. The relationship in the present study appears to be much smaller than previously reported [26]. Both cardiac and arterial parameters are subject to multiple influences, only some of which have been accounted for in the study by Palmieri et al. [3]. Blood pressure differed significantly between the three tertiles of PP/SVi and, as discussed, this is relevant given that blood pressure ‘passively’ increases arterial stiffness. Consistent with this, mean pressure was a determinant of PP/SVi. One of the most important questions to be addressed in a study such as this is the predictive value of PP/SVi in relation to cardiac structure and function, independent of blood pressure. Thus, the data examining left ventricular structure and function (their Table 2), although corrected for age, gender and body surface area, should have been corrected for mean blood pressure. Similarly, the outcome of including mean pressure as an independent variable in subsequent analyses examining determinants of left ventricular function which also included PP/SVi as an independent variable (their Tables 4 and 5) would have been interesting. In particular, such analyses could have provided a comparison with previous data demonstrating that arterial stiffness measurements give no better predictive power than blood pressure alone in relation to left ventricular mass [27]. Differences between gender not only occur in left ventricular mass. Women, even at similar body height, also have vessels of smaller diameter than men. As expected, these vessels are stiffer in line with the general finding of smaller vessels being stiffer than larger vessels [20,28,29]. Possibly as an adaptation to the smaller size of the left ventricular outflow tract, ejection time is prolonged in women at identical heart rates compared to men [28,30]. Thus, despite the considerably smaller outflow tract in women, outflow velocity is similar to men, stroke volume is reduced only slightly and heart rate is increased slightly [20]. The net effect is an only slightly lower cardiac output in women. These gender differences in left ventricular and vessel anatomy and physiology make mixed gender data notoriously difficult to interpret, in particular where there is an imbalance in gender distribution. Gender was a determinant of many of the cardiac functional parameters in the study by Palmieri et al. [3] (their Table 4). This is evident in their Table 1 where a greater proportion of women is present in the highest tertile of PP/SVi compared to the lowest. Gender is thus likely to be an important determinant of PP/SVi and should have been included in the multivariate analysis presented [3]. In conclusion, it is not clear whether the reported relationships between arterial stiffness and concentric left ventricular geometry and diastolic function are truly independent of related variables including blood pressure and gender. Previous studies [27] suggest that elevated blood pressure is the common factor underlying both arterial and cardiac adaptation.