High sodium intake and arterial stiffness

The occurrence of a close relationship between high blood pressure (BP) and sodium intake with diet is widely acknowledged and supported by numerous studies. Among them, the results of Intersalt ^[1], an international study on the link between electrolyte excretion and BP, analysing data from more than 10 000 patients, have clearly shown the role of a low-sodium diet in reducing SBP, especially in the elderly, with the possibility of modulating the increase in SBP values with age, and hence of reducing cardiovascular mortality. When considering the mechanisms underlying the association between high sodium intake and BP increase, a role for a concomitant increase in arterial stiffness has been suggested. Indeed, the relationship between high sodium intake with diet, high BP values and increased arterial stiffness is not certainly a recent discovery, given that 3700 years ago the Chinese physician Huang Ti Nei Ching Su Wein already wrote: ‘…therefore if large amounts of salt are taken, the pulse will stiffen and harden’.

For a long time, the association between high sodium intake and increased BP levels was attributed exclusively to an increase in peripheral systemic resistance and to an action of sodium on resistance small arteries ^[2], although up to now the actual mechanisms underlying this relationship have not yet been fully clarified. More recently it has been shown that changes in sodium plasma levels may not only exert their effects on small resistance arteries, but may also affect the structure and function of large elastic arteries ^[3].

An alteration in the extracellular matrix of arterial wall may be considered one of the most important factors related to arterial stiffness induced by excessive sodium intake with diet. The extracellular matrix includes multiple structural proteins, comprising collagen and elastin, and is regulated by matrix metalloproteinases (MMPs) ^[4]. MMP2 and MMP9 are activated by high-sodium diet, leading to increased transforming growth factor-beta1 (TGFß-1) ^[4,5]. Another pressure-independent mechanism through which vascular damage can be caused by high sodium intake is related to modulation of endothelial cell function ^[6,7] and oxidative stress ^[8]. The endothelium sodium channel seems to be a major regulator of cellular mechanisms generating this vascular dysfunction ^[8,9]. In fact, high-salt diet, even without altering BP, affects arterial endothelial cell function and may stimulate aortic endothelial cell production of active TGFß-1 ^[7]. The overexpression of this cytokine may promote a fibrogenic effect on the arterial wall, thus altering its mechanical properties ^[10]. An excessive arterial fibrosis and extracellular matrix deposition may be responsible for an amplification of ageing-related vascular damage and arterial stiffening. Moreover, high-sodium diet has been shown to be associated with stimulation of aortic AT₁ receptors ^[11], and, on such a background, it is important to acknowledge that the vascular damage induced by high sodium intake with diet can be modulated by genetic factors, in particular by the polymorphism of AT₁-receptor genes ^[12] and aldosterone synthase genes ^[13]. These genetic polymorphisms are of particular importance especially in the elderly and in hypertensive patients ^[10,12]. A number of experimental research studies on rodents have shown that a high-sodium diet is associated with increased mortality, and that mortality was significantly reduced when the high-sodium diet was associated with the intake of selective angiotensin II blockers ^[14,15].

The first large studies on humans focusing on the relationship between arterial stiffness and sodium intake date back to the work done in the 1980s by Avolio et al. These authors, in a first study ^[16] showed that aortic distensibility, assessed by carotid–femoral pulse wave velocity (PWV), was consistently lower in a Chinese rural community than in an urban population, when comparing individuals with the same arterial pressure level and at the same age. Prevalence of hypertension, in general, was also significantly higher in the urban group. These between group differences in vascular haemodynamics were independent of concomitant differences in arterial pressure and were related to a different dietary salt consumption in the rural as compared with the urban group. In fact, salt-intake, as determined by urinary sodium excretion, averaged 13.3-g NaCl/day in the urban and 7.3 g/day in the rural population. In a second study, involving a normotensive urban Australian population ^[17], lower PWV values were found in volunteers who were following a low-salt diet compared with a reference group matched for age and arterial pressure under free diet without restriction of salt intake. Also in this case, the between group difference in PWV was independent of BP levels. However, despite these important clinical findings and the solid pathophysiologic assumptions behind them, the results of subsequent randomized controlled trials conducted to highlight the relationship between arterial stiffness and sodium intake with diet led to controversial results.

The meta-analysis published by D’Elia et al.^[18] in this issue of the Journal of Hypertension was specifically aimed at addressing these controversies and at better clarifying the relationship between sodium intake with diet and arterial stiffness. The authors considered 11 randomized controlled trials, including 14 cohorts, in which aortic distensibility was assessed by carotid–femoral PWV, which is considered as the gold standard method in the field of noninvasive measurement of arterial stiffness. The characteristics of the population studied were quite different from one another. In fact, the different studies were performed either on healthy volunteers ^[19], obese patients ^[20,21], hypertensive patients ^[22–28], women with a history of preeclampsia ^[29], patients with chronic kidney disease ^[24] or patients with type-2 diabetes ^[26]. In 12 out of the 14 cohorts, SBP values were significantly lower in the low-salt than in the high-salt group. Only in four cohorts ^{[23,24,27,28]}, all of them including hypertensive patients, PWV values were found to be significantly lower in the low-salt group compared with the high-salt group. In almost all studies, however, only a relatively small number of patients were enrolled, and this is probably one of the main factors that might have contributed to the failure in reaching a statistically significant difference in PWV between high and low-salt groups. Another factor that might have limited the power of the studies included in this meta-analysis could have been the relatively short duration of a given level of sodium intake with the diet assigned to each group. In fact, both the high and low-sodium diets were scheduled to last only a few weeks, ranging from 1 to 6 in the different studies. Although this study design may allow on one side to highlight the short-term haemodynamic changes induced by a high or low-sodium diet, respectively, on the other hand it does not allow for an evaluation of permanent structural alterations induced by any given level of prolonged sodium intake.

Another important confounding factor was the inclusion of treated hypertensive patients in four out of 11 studies ^[24–27]. In particular, two classes of drugs (diuretics and drugs active on the renin–angiotensin–aldosterone system, RAAS) can affect in a relevant manner the link between sodium intake, BP levels and arterial stiffness. The possible role of diuretics in affecting such an interaction is obvious, as they are crucial in the excretion of sodium and in the Na⁺/K⁺ balance, but also RAAS blockers can play a crucial role. It is worth noting that, as mentioned before, in the arterial system, the Angiotensin II AT₁ receptors are upregulated by a high-salt diet and are particularly sensitive to RAAS blockers ^[30]. Thus, treatment with inhibitors of the RAAS should be able to mitigate the adverse effects of high-salt diet on the endothelium and the arterial wall stiffening ^[31]. In the studies on treated hypertensive patients included in this meta-analysis, the role played by these drugs was not specifically investigated, and it is likely that this is an important confounding factor able to affect the results of these studies and consequently of their meta-analyses.

Augmentation index was positively related with salt intake only in one of the cohorts ^[29]. This is not surprising as many other factors, in addition to the viscoelastic properties of the aortic wall (heart rate, sex, patient size, systemic vascular resistance, drugs, left ventricular function, Windkessel effect etc. …) may contribute to modify this parameter. This is one of the reasons for which augmentation index is not currently considered to be a reliable or a specific marker of arterial stiffness.

Furthermore, the results of this interesting meta-analysis indicate the occurrence of a direct association between restriction of dietary sodium intake and decrease in carotid–femoral PWV in the randomized controlled trials included in this article. The study by D’Elia et al.^[18] does indeed show that an average reduction in salt intake of 5 g/day is associated with a 2.8% reduction in carotid–femoral PWV. The authors also showed how this reduction in PWV was independent of the reduction in BP values. However, the relationship between decrease in PWV and reduced sodium intake might be even more complex than how it was described in this article. The close interdependence between BP and PWV is a well known phenomenon in cardiovascular physiology and has been repeatedly shown in studies on vascular haemodynamics ^[32,33], which highlighted the invariable occurrence of an association between BP and PWV. Indeed, a reduction in aortic distensibility causes an increase in SBP values due to impaired aorta ability to dampen the left ventricular ejection (Windkessel effect). The early return of reflection waves in stiffer arteries, due to increased forward and backward pulse wave transmission speed, may cause a further rise in SBP values in ascending aorta due to the overlap of forward and backward waves during the systolic phase of the cardiac cycle. On the other hand, almost all studies considered in this review (12 out of 14 cohorts) showed significantly higher SBP and DBP values associated with high-salt intake. As it is well known that mean arterial pressure (MAP) may affect arterial distensibility ^[34], at least a weak increase in PWV values due to functional changes is to be expected in the presence of increased SBP and DBP values. Only when high BP values persist for a long time, structural alterations in the arterial wall of the elastic large arteries may occur, leading to an increase in arterial stiffness which is at that time independent from BP levels. A long-lasting BP increase is well known to induce organ damage in the cardiovascular system. In particular, overproduction of collagen fibres, with consequent decrease in the elastin/collagen fibres ratio, may well result in arterial stiffening. In this context, the pathophysiologically expected interrelationships between BP and PWV may be associated with the above-mentioned BP-independent alterations in arterial stiffness, induced by high-sodium diet on the arterial wall over prolonged time intervals. The framework of these interactions is therefore extremely complex, and it appears very difficult, even from a statistical point of view, to reliably distinguish between BP-dependent and BP-independent changes in viscoelastic properties of the large arteries in relation to the effects of sodium intake (Fig. 1).

FIGURE 1: The complex mechanisms underlying the relationship between high salt intake with diet, blood pressure and arterial stiffness. BP, blood pressure; EnNaC, endothelium sodium channel; MAP, mean arterial pressure; MMP, matrix metalloproteinases; NO, nitric oxide; ROS, reactive oxygen species; TGF, transforming growth factor.

Last, but not least, we should consider that sodium intake may modify the neural autonomic modulation of the cardiovascular system. In particular, it may modify sympathetic activity, also in this case in quite a complex manner. In a previous study by our group ^[35], we found that salt-sensitive hypertensive patients, as defined by conventional categorical classification, did exhibit alterations of autonomic cardiovascular control which were different at different levels of sodium loading. Salt sensitivity of 34 essential hypertensive patients was assessed on a continuous basis by the salt sensitivity index (SSI) quantified as the ratio of the change in brachial oscillometric MAP (ΔMAP), expressed in mmHg, between the high-sodium and the low-sodium diet periods, with the corresponding change in urinary sodium excretion rate (ΔUNaV) expressed in mmol/l/day, multiplied by a factor of 1000 to facilitate readability of results, namely: SSI = (ΔMAP/ΔUNaV)/1000 [mmHg/(mol/l/day)]. Beat-by-beat finger BP was also recorded after each diet period, and autonomic cardiovascular control was evaluated by spectral analysis of BP and pulse interval variability and by the related assessment of spontaneous baroreflex sensitivity (BRS) (sequence technique) ^[36–38]. Salt sensitivity and BRS were inversely related to each other during both low and high sodium intake, starting from low values of the SSI. All spectral indexes of pulse interval variability, except the ratio between low-frequency and high-frequency powers, were inversely related to SSI after high sodium intake. In patients with lower salt sensitivity, BRS and pulse interval power in the high-frequency band were higher after high sodium intake than after low sodium intake. Conversely, patients with a higher SSI showed lower values of BRS and pulse interval power in the high-frequency band, which were not influenced by salt intake. These data indicate an impairment of parasympathetic cardiac modulation (quantified by BRS and heart rate variability indices in the high-frequency band) associated with increasing values of the SSI. Given the reciprocal changes that usually characterize modifications in sympathetic and parasympathetic cardiac modulation, these findings are in line with the evidence of sympathetic activation in salt-sensitive patients ^[39]. In fact, in patients with a lower salt sensitivity and a preserved autonomic cardiovascular modulation, a high-sodium diet was accompanied by higher BRS and heart rate variability values than a low-sodium diet. This may indicate that, when the reflex cardiovascular regulation is preserved, high sodium intake may activate cardiopulmonary receptors through an increase in plasma volume, leading to a reflex reduction in sympathetic efferent activity ^[40], whereas the opposite may occur under low sodium intake ^[41]. Such a modulation was not found in patients displaying the highest degree of salt sensitivity, in whom no changes in their impaired autonomic cardiovascular control were observed with manipulation of their sodium intake. These observations suggest that the salt-dependent increase in BP observed in hypertensive patients with the higher degree of SSI might at least in part depend on their impaired arterial baroreflex function or on their inability to increase BRS and reduce sympathetic activity in response to the increase in plasma volume determined by sodium loading ^[39–46].

If we then consider the relevance of all these findings to arterial stiffness changes, we should not disregard the possibility that not only the level of sodium intake, but also the individual degree of sodium sensitivity, which is likely to have a genetic background, might be involved in explaining the complex relation between salt intake and arterial stiffness.

The interesting work by D’Elia et al.^[18] offers elements for a deeper insight. However, additional pathophysiological, epidemiological and intervention studies appear still to be needed, to further disentangle the complex interactions that are responsible for the association between different levels of sodium intake with daily diet and different individual degrees of arterial stiffness, at different levels of BP.

ACKNOWLEDGEMENTS

Conflicts of interest

There are no conflicts of interest.