The January issue of the Journal of Hypertension opens with two meta-analyses on blood pressure (BP) regression to the mean, as calculated from randomized controlled trials. Salam et al. (pp. 16–23) confirm, on a total of 86 trials in about 350 000 patients, that, due to regression to the mean, office BP shows a reduction when initial values are higher and an increase when they are lower. They further show that this phenomenon is quantitatively so large as to make, in patients with a higher initial office BP, the BP-lowering effect greater than that due to antihypertensive treatment. As often mentioned in the past, regression to the mean makes the BP range of a given study group much smaller in the follow-up than in the screening or pretreatment phase, increasing the steepness of the relationship between BP and outcomes and thus upgrading the importance of BP as a risk factor for cardiovascular events. Significantly, in the other meta-analysis (Moore et al., pp. 24–29) regression to the mean is also reported for ambulatory BP whose day and night follow-up mean values were found to be lower and higher in patients with initially higher and lower values, respectively, similarly to office BP. To several investigators this will appear surprising because ambulatory mean BP values have often been reported to be similar with measurements performed at shorter or longer time intervals from the initial measurements, with in addition little or no BP-lowering effect also of prolonged periods under placebo. Indeed, this has been traditionally listed as an advantage of ambulatory over office BP measurements for the study of antihypertensive drugs. The meta-analysis of Moore et al. challenges this concept and calls for further investigation of this issue, hopefully by a database larger than that available in the present meta-analysis. The two papers are commented by the interesting Editorial of Messerli and Rexhaj (pp. 4–5) who provide information on the origin and the development of the concept as well as on its implications for the interpretation (and misinterpretation) of the results of clinical trials.
The following three articles provide new data on BP and haemodynamic measurements in hypertension. Stabouli et al. (pp. 30–36) tested the new oscillometric Sphygmocor XCEL device (AtCor Medical Pty Ltd, West Ryde, NSW, Australia) in individuals aged 6–20 years and found its ability to measure pulse wave velocity and central BP similar to that provided by the gold standard tonometric measurements. Central BP and pulse wave velocity are more and more frequently part of the clinical examination of adults and children with hypertension, making simplification of data collection by newer vs. classical methods, a relevant advantage. Palatini et al. (pp. 37–41) show that in very obese individuals cuffs that reflect the troncoconical shape of their arms provide lower BP values (−5/−3 mmHg) than cylindric cuffs, thereby reducing the well known overestimation of BP in overweight patients. This is not of marginal clinical relevance because overestimation of BP values in obese individuals may lead to unnecessary drug treatment. Finally, Seo et al. (pp. 42–49) show, in more than 1200 patients, that unobserved (or unattended) office and daytime mean BP measurements exhibited large discrepancies, the differences being particularly evident in patients with a high cardiovascular risk. This implies that unattended office BP measurements do not provide, as often claimed, BP values that are superimposable to the daily life ones, thereby sharing with them a superior prognostic significance. Unattended BP measurements my provide values that are closer to daily life BP because they are immune from the white-coat effect that characterizes classical attended office BP measurements. However, daily life BP is modulated by countless other factors that make discrepancies with office BP, no matter whether attended or unattended, by no means surprising. The issue is discussed by the editorial commentary of Kjeldsen and Mancia (pp. 6–8) also in relationship to the SPRINT trial, in which unattended office BP measurements were, at least as reported in the protocol, the BP measurement of choice.
The following series of articles focus on heart and vessels. Tadic et al. (pp. 50–56) show that hypertensive patients have lower left ventricular (LV) global longitudinal and circumferential strains than their normotensive counterparts, this being the case more in men than in women. Sex differences in the adaptation to hypertension are more and more frequently discovered, also with clinical and therapeutic implications. Milan et al. (pp. 57–64) describe the results of a study (ARGO-SIIA project) designed to determine the prevalence of ascending aorta dilatation in untreated hypertension. Compared with normotensive controls aortic dilatation was quite frequent (13%) in hypertensive individuals, its detection being frequently associated with an increase of LV mass. Aortic dilatation not rarely goes unreported by the echocardiographic examination of patients with an increased BP, whereas it deserves attention both because, as shown by Milan et al., is not rare and because the possibility that the combination of LV hypertrophy and aortic dilatation increases cardiovascular risk more than LV hypertrophy alone cannot be excluded. Pan et al. (pp. 65–72) describe the relationship between a large number of plasma biomarkers, including those related to inflammation, and expansion of extracellular volume or increased LV mass as assessed by cardiac magnetic resonance. Several biomarkers correlated with one or both abnormalities, scoring in favour of the possibility to predict in the future the development of hypertension-related abnormalities by plasma variables of simple collection. Izawa-Ishizawa et al. (pp. 73–83) describe a model of aortic dissection obtained in mice by inducing endothelial dysfunction via administration of L-NAME. Aortic dissection and aneurysmal rupture were much more frequent in mice administered L-NAME than in control mice, which supports a causal role of endothelial damage in the pathogenesis of this clinically dramatic condition. Significantly, administration of pitavastatin had protective effects, which suggests that statins may be a therapeutic coadjuvant, possibly because of their anti-inflammatory effect. Yang et al. (pp. 84–91) provide further data on the relationship between long-term (i.e. visit-to-visit) SBP variability and all-cause death, using a real life cohort from China. The interesting aspects of the study are that also long-term heart rate (HR) variability showed a relationship with fatal events; BP variability was prognostically more detrimental in patients with a higher HR variability, suggesting an interaction between these two phoenomena. Finally, Alves-Cabrato et al. (pp. 92–98) describe the results obtained by studying the prognostic value of ankle–brachial index (ABI) in an impressively large number of hypertensive patients with no previous cardiovascular events (n = 9126 out of a database of more than 44 000 patients). Compared with what is regarded as the normal value (<0.9) an increased ABI was associated, over a 6-year follow-up, with an increased cardiovascular risk, which was maximally evident for heart failure and reached 44% for values greater than 1.3. This strengthens the evidence in favour of including ABI among the measures of asymptomatic vascular damage in the screening of patients with high BP, as already recommended by several guidelines. The challenge is to conclusively prove that treatment-induced changes in ABI are predictive of patients’ protection, a goal still elusive for this as well as for other markers of vascular damage.
Several articles of the January issue deal with the pathogenetic, diagnostic and treatment aspects of renal damage or on the determinants of cardiovascular risk. Zhang et al. (pp. 99–108) examined the validity of five published equations that estimate 24-h urinary sodium excretion via time spot urine specimen, measuring spot and 24-h urinary sodium excretion in 99 healthy volunteers. No equation was particularly accurate and indeed with all equations an up to 60% misclassification of the true individual salt intake was observed. This confirms the limitations of spot urine as well as the need of more accurate measurements, keeping in mind, however, that complex measurements are incompatible with the size of epidemiological and intervention studies that need to be performed to give the long-term relationship among sodium intake and cardiovascular risk a reliable answer. Underwood et al. (pp. 109–115) provide evidence that in mice with Lewis polycystic kidney disease vasopressin as well as activation of the hypothalamic paraventricular nucleus (obtained by the inactivating effect of muscimol) contribute to the accompanying BP elevation. The obvious implication is that studying the effect of interventions against these mechanisms may have a therapeutic rationale. Khatir et al. (pp. 116–124) describe that, in patients with advanced chronic kidney disease, randomization to an additional vasodilating medical therapy (calcium channel and a renin–angiotensin system blocker) reduced BP and renal vascular resistance more effectively than randomization to a nonvasodilating therapy (a beta-blocker) which had a renal vasoconstrictor effect. Over a period of 18 months, however, this did not translate into a difference in cortical or medullary oxygenation values, the decline of glomerular filtration rate being also similar in the two groups. This adds to the evidence of trials and meta-analyses, that aggressive BP reductions may not carry any advantage to the kidney, even when, as suggested by the results of Khatir et al., drugs elicit a powerful renal vasodilation, thus in theory protecting organ perfusion in the face of the reduction in perfusion pressure. Lin et al. (pp. 125–134) report the results of a retrospective analysis of more than 7000 patients with acute kidney disease in which treatment made or did not make use of a mineralocorticoid receptor antagonist. Use of a mineralocorticoid receptor antagonist did not lower the rate of major cardiovascular events and death compared with nonmineralocorticoid-based treatments. However, although increasing the risk of hyperkalaemia, mineralocorticoid antagonist therapy was associated with a 17% significant reduction in the risk of dialysis, suggesting a favourable effect of this therapeutic approach in the transition from acute to chronic kidney disease. An insight into the pathogenetic role of aldosterone for chronic kidney disease as well as into the renal effects of mineralocorticoid receptor antagonists is provided by the editorial commentary of Maiolino and Calò (pp. 9–10).
The articles on determinants of cardiovascular risk focus on BP and other risk factors. Using an animal model of hypertension Cardoso et al. (pp. 135–143) report a favourable effect of chronic exercise (swimming) on hypertension-mediated memory dysfunction, possibly via modulation of the purinergic system in the brain. He et al. (pp. 144–153) show that euthyroid individuals with increased thyroid-stimulating hormone levels have a greater risk of developing hypertension. Kerr et al. (pp. 154–166) provide evidence that nicotine-containing electronic cigarettes did not increase BP but increased HR and reduced peak expiratory flow, indicating that some of the acute adverse effects of smoking are not completely avoided. Of note, both electronic cigarettes and tobacco smoking did not have a pressor effect, at variance from old studies in which smoking a cigarette was associated with a marked and prolonged pressor response. Kuwabara et al. (pp. 167–174) provide data from a Japanese nondiabetic population that higher baseline glucose levels are associated with an increased risk of incident hypertension, this being the case both in men and women. This adds to the already large body of evidence that blood glucose and BP are related by a sort of a positive interaction, the basis of which may be the stimulating effect of sympathetic activity on insulin resistance and secretion, and the sympathetic activation caused by the increased insulin levels. The epidemiological link of hyperglycaemia and BP is emphasized by the Editorial of Nilsson (pp. 11–12). Finally, Rosenbaum et al. (pp. 175–181) report on the central BP and vascular alterations of patients with congenital adrenal hyperplasia. Compared with well matched controls, 26 individuals with congenital adrenal hyperplasia exhibited an early increase of central BP. Somewhat surprisingly this was not accompanied by an increase of pulse wave velocity and carotid intima–media thickness, leading the authors to advance the hypothesis that in congenital adrenal hyperplasia vascular damage depends on and thus follows a BP elevation.
The following four articles deal with preeclampsia. In the article by Salazar et al. (pp. 182–186) evidence is presented that in high risk midpregnancy nocturnal hypertension is frequent and represents a strong predictor of preeclampsia or eclampsia. As discussed in the accompanying editorial of Webster (pp. 13–15), this provides further evidence on the importance of the information offered by ambulatory BP in pregnancy, strengthening the case for use of this approach on a routine basis. Further supporting data are shown in the second article in which Lv et al. (pp. 187–196) document a relationship of nocturnal hypertension and preeclampsia with an excessive coiling of the umbilical cord, probably via mechanical interference with fetoplacental vascular resistance. In this setting, a considerable interest has also the third article (Shao et al., pp. 197–205) which shows that early preeclampsia is associated with an imbalance of the levels of plasma testosterone and oestradiol, which correlates with the levels of procoagulation factors. Information on biomarkers that may predict preeclampsia is making considerable progress, and the hope of future clinical applications is by no means chimeric. Finally, Iwama et al. (pp. 206–215) provide a large descriptive evidence that in monochorionic and dichorionic twin pregnancies early, mid and late SBP or DBP values are somewhat greater than in the singleton pregnancy. The implications of these differences for gestational and perinatal problems remain to be studied.
Last but not least, Lind et al. (pp. 216–222) report the results of a genetic research that has a special element of interest because it links proteomic patterns with functional cardiovascular derangement, a genetic–mechanistic approach that is not yet common in genetic research. Lind et al. examined proteins known to be associated with cardiovascular risk for their relationship with endothelial dependent forearm vasodilation, elicited by intra-arterial acetylcholine infusion. Endothelial dependent vasodilation was selected because endothelial dysfunction is the first step in the cascade of events that leads to atherosclerosis. Out of the 84 proteins found to have a cardiovascular link in two related population studies, 15 showed an association with endothelial dysfunction, a replication being obtained in seven of them. Eventually, after adjustment for confounders, the plasma levels of cathepsin-D showed a significant inverse relationship with the endothelial-mediated vascular effect, a result that supports the validity of this research approach, which will hopefully become more and more common in the future. It is obvious that, once the genes are involved in hypertension and cardiovascular disease are identified, we need to know through which mechanisms they operate.
Conflicts of interest
There are no conflicts of interest.