Sodium, Arterial Stiffness, and Cardiovascular Mortality in Hypertensive Rats:

Abstract

Background:

We examined the effects of early high salt diet (HSD) and angiotensin II type 1 (AT1) receptor antagonist valsartan (Val) on mortality and carotid distensibility in surviving spontaneously hypertensive rats (SHRs).

Methods:

The HSD was initiated either early (week 4 after birth) or late (week 10), continued until 20 weeks of age, and compared to normal salt diet (NSD) groups. Valsartan was given from the fourth week after birth.

Results:

Eighty-six percent of the rats died in early HSD on placebo, 70% in early HSD on Val-3 mg, 35% in early HSD on Val-30 mg, and 13% in late HSD on placebo. Mean arterial pressure (MAP) was higher in the early HSD and late HSD groups on placebo compared with NSD. The Val-30 mg reduced MAP in all except early HSD groups. Distensibility at MAP (operational distensibility) was lower in late HSD on placebo than in NSD placebo groups. The Val-30 mg increased distensibility in NSD groups. There was no effect of Val in late HSD and early HSD groups. Operational distensibility was negatively correlated with MAP and salt and positively correlated with Val treatment. All animals receiving HSD showed a higher isobaric distensibility in early HSD than in late HSD groups and a smaller distensibility in rats treated with Val.

Conclusions:

Our results showed that administration of early HSD in SHR was associated together with a high mortality, a protective action of Val that increased longevity, and an increased level of isobaric distensibility. Survival in HSD groups suggest a direct role of angiotensin II in salt-induced cardiovascular mortality. This role is associated with MAP independent of changes in carotid stiffness. Am J Hypertens 2007;20: 319–325 © 2007 American Journal of Hypertension, Ltd.

Arterial stiffness, SHR model, salt, mortality, AT1 antagonist

The role of sodium (Na) in hypertensive subjects results from two principal clinical observations. First, a positive correlation has been widely observed between urinary Na excretion, an estimate of Na consumption, of various communities around the world, and the upward slope of systolic blood pressure (BP) with aging.^1–3 Second, dietary Na restriction lowers systolic BP and pulse pressure (PP), especially in older subjects with isolated or predominantly systolic hypertension, but has only a modest effect on diastolic BP.² In view of the role of arterial stiffness on the level of systolic BP and PP in older subjects, and on cardiovascular (CV) risk, it seems that the effects of Na may impact not only on small arteries but also on large arteries. In addition to the correlation between Na intake and pressure, it is relevant to determine, at different levels of Na intake, the relation between BP and distensibility.

Adequate methods have been developed to measure the BP distensibility of large arteries in rats.⁴ Because Na intake influences CV longevity, it is of interest to determine its role on the large artery wall for a given genetic background.^5–7 When increased Na diet is initiated in the spontaneously hypertensive rats (SHR) at different phases of development, various degrees of severity of hypertension and of CV longevity may be obtained for the same genetic background.^4,5,8

The effects of Na on the carotid arterial wall involve multiple alterations, including mechanical, hormonal, and growth factors.^5–7,9,10 Because trophic modifications, such as arterial wall hypertrophy, develop progressively with a high Na diet (HSD),^4–7,9,10 only long-term steady-state BP distensibility relationships may be evaluated. Because the renin-angiotensin system plays a major role in Na homeostasis, these curves should be established with and without long-term blockade of the angiotensin II type 1 (AT1) receptors.^11,12 We have previously reported that the AT1 receptor antagonist valsartan (Val) decreased mean arterial pressure (MAP) but was not able to reduce carotid artery stiffness in SHRs receiving a HSD from the 10th to the 20th week of age.¹² Here, our working hypothesis was that, in animals receiving a HSD from an early age, an increase in carotid arterial stiffness may be observed in association with more severe hypertensive vascular disease, which may in turn favor the occurrence of CV morbid events.

The first objective was to establish in SHRs the mortality rate and the BP distensibility curve of the carotid artery with HSD initiated either early or later, with and without long-term administration of the selective AT1 antagonist Val at low and high dosage.¹² The second objective was to evaluate whether, in addition to mechanical stress, BP-independent changes of carotid stiffness are involved in the severity of hypertensive vascular disease in SHR.

Methods

All procedures were carried out in accordance with institutional guidelines for animal experimentation. Male SHRs were obtained from Iffa Credo, Lyon, France. Animals were fed ad libidum and were housed at 25°C with a 12-h light/dark cycle. It is well established that in the SHR, animals given a high salt diet from an early age develop more severe hypertensive disease.^4–8 In a preliminary study we observed that early (≤5th week of age) administration of HSD was associated with a higher incidence of CV deaths than when HSD was started later. Thus, the present study involved two separate series of experiments: one on CV mortality according to the date of initiation of HSD, and another, on the study of carotid arterial hemodynamics in surviving animals. Weight and left ventricular hypertrophy (LVH) expressed as milligrams per gram of body weight were recorded.

Mortality Study

Mortality was studied in 152 SHRs. Of these, 45 were kept constantly on a normal Na diet (NSD, 0.4% NaCl) from birth to the 20th week of age. Between the 10th and the 20th week of age, SHRs on NSD were randomized into three groups receiving either AT1 blockade by Val at 30 mg/kg/d (n = 15), 3 mg/kg/d (n = 15), or vehicle alone (placebo) (n = 15).

In 62 SHRs, early HSD was initiated at 4 weeks of age and composed of 7% NaCl added to the normal diet until the 20th week of age. Rats were then randomized into three groups to receive either AT1 blockade by Val at 30 mg/kg/d (n = 20), 3 mg/kg/d (n = 20), or placebo (n = 22). In 45 SHRs, late HSD was initiated at 10 weeks of age and was composed of 7% NaCl added to the normal diet until the 20th week. These late HSD animals received either Val at 30 mg/kg/d (n = 15), 3 mg/kg/d (n = 15), or placebo (n = 15). In all groups, Val or placebo treatment was initiated at the 4th week of age up to the 20th week for survivors.

Previous studies have shown that Val administered orally at 30 mg/kg/d significantly decreases BP in 24 h in SHR on an NSD.¹² All treatments (NaCl and Val) were incorporated in the diet (prepared by Iffa Credo). The daily doses of Val were calculated on the basis of a mean consumption of 20 g of chow/d per rat. Intracarotid BP and distensibility were measured at the end of the treatment period in all surviving SHRs (n = 110). A histopathologic analysis of brain, heart, and lung tissues was performed in all rats that died.

Hemodynamic Study of Surviving Animals

We simultaneously recorded intra-arterial diameter (left carotid artery) and BP (right carotid artery), in pentobarbital-anaesthetized rats, and determined arterial distensibility as previously described.^12–14 Internal arterial diameter (D) was measured with an ultrasonic echo-tracking device (NIUS-01, Asulab SA, Neuchâtel, Switzerland). The relationship between the pressure (P) and the lumen cross-sectional area (LCSA) was fitted with the model of Langewouters et al¹⁵ using an arc tangent function. Distensibility, a derivative of this function, was used to assess the elastic behavior of the artery.

To compare distensibility–BP curves, we calculated first distensibility at MAP (operational distensibility), and second, distensibility at a BP level (isobaric distensibility) that was common to NSD animals (168 mm Hg) and also at a BP level common to both NSD (on placebo) and HSD groups (200 mm Hg), both early and late, on placebo and on Val. Operational distensibility represents an evaluation of the effective buffering carotid arterial function at MAP. Isobaric distensibility is an index of HSD and Val-induced changes in the arterial wall, which is independent of MAP level.

Statistical Analysis

Results were expressed as means ± SEM. Data were analyzed by use of two-way ANOVA, followed by a Fisher's test for multiple comparison. The survival analysis was based on the Cox model and calculation of the odds ratio, where CI represents the 95% confidence intervals. To identify independent predictors of operational distensibility, we used multivariable Cox regression analysis with stepwise selection. Variables included in multivariate models were MAP, salt, and Val treatment. A P value < .05 was considered as significant.

Results

Mortality Study

During the follow-up period, 86% of rats died in early HSD on placebo, 70% in early HSD on Val 3 mg, 35% in early HSD on Val 30 mg, and 13% in late HSD on placebo. No mortality was observed in the other groups. The probability of survival is shown in Fig. 1. Unadjusted Cox analysis shows that early HSD was significantly associated with a higher mortality rate. Administration of Val 30 mg reduced the mortality rate in early HSD, whereas Val 3 mg had no significant effect.

Cox analysis showing survival in spontaneously hypertensive rats as a function of sodium diet and angiotensin II type 1 receptor blockade (valsartan 3 or 30 mg). NSD = normal salt diet; HSD = high salt diet.

Figure 1.

In SHRs receiving early HSD, the odds ratio for death risk was 19.4 (95% CI 5.0–76.3) in response to placebo, whereas the odds ratio of Val 3 mg was 15.8 (95% CI 3.9–62.9) and that of Val 30 mg was 9.0 (95% CI 0.32–8.55).

At autopsy, a large number of SHRs receiving early HSD developed cardiac fibrosis, cerebral hemorrhage, or pulmonary edema. Administration of Val 30 mg reduced such alterations, particularly regarding pulmonary edema.

In Vivo Mechanical Properties of Carotid Artery at MAP in Surviving Animals

Table 1 shows the effects on body weight, BP, heart rate, and distensibility at MAP. Blood pressures and PP were significantly higher in early HSD and late HSD on placebo groups compared with NSD on placebo. Blood pressure levels were not different between early-HSD and late-HSD groups on placebo. There was no difference in heart rate between groups. In placebo groups, HSD produced a decrease in body weight compared to NSD animals.

Table 1

Mortality, hemodynamic parameters, and operational distensibility of the carotid artery in SHRs given HSD

	NSD			Early HSD			Late HSD
	Placebo	3 mg/kg/d	30 mg/kg/d	Placebo	3 mg/kg/d	30 mg/kg/d	Placebo	3 mg/kg/d	30 mg/kg/d
N rats (baseline number)	15	15	15	22	20	20	15	15	15
Survivors (number)	15	15	15	3	6	13	13	15	15
Survivors (%)	100	100	100	14	30	65	87	100	100
Weight (g)	408.0 ± 7.7	389.7 ± 5.0	386.7 ± 10.2	365.0 ± 12.6+	367.9 ± 9.0	376.2 ± 7.7	375.5 ± 8.6+	365.5 ± 11.5	370.5 ± 11.8
SBP (mm Hg)	222 ± 8	222 ± 7	172 ± 4*	268 ± 5+	265 ± 13	270 ± 6	260 ± 7+	243 ± 6	226 ± 7*
DBP (mm Hg)	162 ± 6	162 ± 6	130 ± 3*	188 ± 4+	194 ± 11	193 ± 5	191 ± 6+	172 ± 4*	157 ± 5*
MAP (mm Hg)	182 ± 6	182 ± 6	144 ± 3*	215 ± 4+	218 ± 12	218 ± 5	214 ± 6+	196 ± 5*	180 ± 5*
PP (mm Hg)	60 ± 3	60 ± 2	42 ± 3*	80 ± 4+	71 ± 3	77 ± 2	69 ± 2+	71 ± 2	69 ± 3
HR (beats/min)	349 ± 11	362 ± 11	337 ± 10	298 ± 3	341 ± 19	357 ± 10	347 ± 12	360 ± 9	363 ± 11
Distensibility at MAP (mm Hg-1.10–3)	1.96 ± 0.16	2.18 ± 0.21	4.43 ± 0.39*	1.53 ± 0.20	1.02 ± 0.13	1.28 ± 0.12	1.36 ± 0.11+	1.59 ± 0.13	1.60 ± 0.20

Values are mean ± SEM. See text for abbreviations.

+

P < .05 v NSD placebo animals.

*

P < .05 v placebo animals for the same group.

In NSD animals and late HSD, Val 30 mg produced a significant reduction in systolic BP, diastolic BP, and MAP, whereas Val 30 mg did not reduce BP in early HSD.

Pulse pressure was significantly decreased in NSD on Val 30 mg compared to the placebo group. There was no reduction of BP in response to Val 3 mg except in the late HSD group in which diastolic BP and MAP were significantly reduced. We have determined LVH in late HSD and NSD SHRs. Valsartan 30 mg significantly decreased LVH only in NSD animals (2.68 ± 0.06 v 2.21 ± 0.05 mg/g body weight; P < .05). Valsartan 3 mg has no effect in both groups.

Due to a higher mean BP, operational distensibility was significantly lower in late HSD on placebo than in NSD placebo groups. At mean BP, Val 30 mg, but not 3 mg, increased distensibility in NSD groups. There was no effect of Val in late HSD and early HSD groups. Multiple regression analysis in the totality of survivors show that operational distensibility was negatively correlated with MAP and salt, which represents 32% and 5% of total variance, respectively, and positively correlated with Val, which represents 5% of total variance.

Isobaric Distensibility in Surviving Animals

Fig. 2 shows the distensibility–BP curves in NSD, early HSD, and late HSD groups. Whereas the placebo and Val 3 mg curves never differ significantly, the Val 30 mg curve is shifted on the left and upward in NSD animals, remains unmodified by comparison with placebo in early HSD, and is shifted on the left in late HSD SHRs. Fig. 3 indicates isobaric distensibility (at 168 and 200 mm Hg) of the carotid artery. In NSD groups, the distensibility–BP curve of Val 30 mg was significantly shifted upward and leftward from that of placebo group. Distensibility at 168 mm Hg was significantly reduced in Val 30 mg than in placebo group indicating a pressure-independent increase of arterial stiffness.

Isobaric distensibilities of the carotid artery calculated within a common range of BP in NSD rats at 168 mm Hg and in NSD (on placebo) and HSD rats at 200 mm Hg (on placebo and valsartan). *P < .05 v placebo for the same sodium diet; +P < .05 v NSD placebo.

Figure 3.

Distensibility–BP curves in the carotid artery of NSD animals, early HSD, and late HSD animals treated by placebo or valsartan (3 and 30 mg/kg/d).

Figure 2.

In early HSD groups, the distensibility–BP curves and isobaric distensibility in Val-treated groups were not significantly different compared with the placebo group. In late HSD, the distensibility–BP curve in the Val 30 mg-treated group was shifted leftward from that of the placebo group indicating a lower isobaric distensibility at 200 mm Hg than in the placebo group. There was no effect of the small dose of Val (3 mg) in any group.

There was an increase in distensibility at 200 mm Hg in early and late HSD groups on placebo compared to the NSD group. There was no difference in isobaric distensibility between early HSD and late HSD groups on placebo. To investigate whether the age of HSD administration had an effect, a two-way analysis of variance in all animals receiving HSD was performed. This showed a greater isobaric distensibility in early HSD than in late HSD groups (P < .0005) and a lower distensibility in rats treated with Val (P < .002; interaction, not significant).

Thus, our results showed that administration of early HSD from 5 weeks of age in surviving SHR was associated with a high mortality rate, a protective action of Val, which increased longevity, and increased levels of isobaric distensibility.

Discussion

In this investigation we determined the chronic effects of an Na diet on CV mortality and carotid arterial stiffness in the SHR. We identified that, according to the early or late initiation of HSD, significantly different degrees of severity of hypertension, of growth retardation, and occurrence of CV mortality were observed. With early HSD, a high incidence of CV mortality was noted that was consistently (but incompletely) reversed by the association with angiotensin II AT1 blockade. Surviving animals still presented a high level of MAP and PP. With late HSD, practically no CV mortality occurred, except with placebo. The AT1 receptor blockade in late HSD animals was associated with reduced MAP, but no change in operational distensibility and PP. Only NSD animals on Val presented a marked increase in operational carotid distensibility and reduced PP. Finally, all Val-treated animals, including those with early HSD, exhibited a reduction in isobaric carotid distensibility, indicating a direct effect of AT1 blockade on arterial wall structure in survivors. These findings are consistent with those on operational distensibility, which showed that, independently of MAP, this parameter was modulated by sodium and angiotensin II type 1 receptor blockade.

Because the effects of early administration of salt on hemodynamics and distensibility were only studied in survivors, some of the results should be interpreted with caution. From the Cox analysis, it appears clearly that other factors than MAP might influence death, particularly the date of HSD initiation and the presence of AT1 blockade (Val 30 mg). From the study of operational distensibility, it appears also that the same observation can be made regarding factors influencing distensibility such as sodium or AT1 blockade. In early HSD animals, BP did not differ significantly in rats taking placebo or Val 30 mg. Taken together, these findings do not suggest that BP was the exclusive factor implicated in the mechanisms of death or carotid distensibility and that both parameters are consistently inter-related. This observation supports some observations already noted in older hypertensive humans.¹⁶

In early HSD animals, CV death occurred with a high incidence with placebo but was consistently reduced when a high dosage of Val was administered. Thus, the cause of death can probably be attributed to the deleterious consequences of early HSD on the renin-angiotensin system. This finding may indicate the contributive role of one or several pathophysiologic factors acting on the vessel wall such as growth factors (TGF-β),^10,16 changes in vascular smooth muscle (VSM) cells, or in extracellular matrix,^4,5,17,18 modifications in the equilibrium between vascular reactivity to angiotensin II and changes in the expression and density of carotid AT1 receptors,^12,19,20 or finally other unidentified factors. In early HSD animals, these mechanisms are certainly associated since: (1) CV mortality was largely but not completely reversed by AT1 receptor blockade at high, but not at low, dosage, (2) significant differences in isobaric distensibility were observed between early HSD and late HSD surviving animals, indicating consistent changes in VSM cells (Fig. 3), and (3) striking structural alterations of the CV system were observed at autopsy.

In late HSD animals, CV death was practically absent, with the exception of two SHRs taking placebo. Because survival in early HSD was increased by AT1 blockade and this treatment had a consistent effect on isobaric distensibility, it seems likely that prevention of CV death and the effect of Val on the large artery wall may be linked. Because Val decreased MAP in both NSD–HSD and late HSD animals, but decreased PP and LVH and increased operational carotid distensibility only in the former, it seems possible that salt would preclude any effect of Val on its consequent reduction of MAP on improvement of the elastic properties of the carotid artery of late HSD rats. In the recent years, we and other investigators^9,10,13,21 have shown that angiotensin II blockade may modulate BP and the stiffness of the carotid wall material through changes in endothelial function or mechanisms related to mechanotransduction.

Previous studies show that the BP response to AT1 receptor antagonist in SHRs on a normal salt diet reaches an average of 40 mm Hg.^12,22,23 It also appears that the degree of BP reduction in response to AT1 receptor antagonist depends to the baseline BP value. In our study, the BP response of 30 mg of Val (−44 mm Hg) is in accordance with these data. Concerning the molecules involved in arterial remodeling, it has been demonstrated in SHRs with late HSD that EIIIA fibronectin isoform markedly increased in the arterial wall and the increase of fibronectin was reversed in response to Val. However, reversibility is obtained only in animals with NSD, and not in late HSD, resulting in a significant interaction.¹² In the present study, a similar problem was observed when HSD animals are compared in terms of isobaric distensibility (Fig. 3). Whereas isobaric distensibility was increased in HSD groups on placebo compared to NSD animals, a significant reduction in distensibility was observed in both early HSD and late HSD animals on Val. Finally, our data of both isobaric and operational distensibility indicate that survival is associated to arterial structure and function independently of MAP. These findings fit with previous results obtained in experimental studies in salt-loaded stroke-prone SHRs,^24,25 and also in clinical studies of hypertensive subjects.^16,26,27,28

In conclusion, the present study showed consistent links in the SHR between CV longevity, sodium, and the renin-angiotensin system in relation with structure and function of large arteries. The high level of mortality in response to early HSD and the implication of AT1 receptors in survival, as well as in regulating arterial stiffness, may have important implications in the drug treatment of chronic hypertension. First, in hypertensive subjects taking various antihypertensive drug therapies, diastolic BP is often normalized (<90 mm Hg), whereas systolic BP remains elevated (>140 mm Hg), resulting in an increased PP.²⁶ In the SHR, increased PP may be totally normalized with antihypertensive treatment but only in the case of the association of NSD and AT1 blockade at a high dosage. Second, in severe hypertensive subjects with advanced renal failure or more than 50 years of age, increased PP is due to a MAP-independent increase in aortic stiffness and is reversed by blockade of the renin-angiotensin system, but only in the presence of salt and water depletion.^27–29 In this report we have observed quite similar findings in early HSD groups in which the percent change in surviving animals was associated with changes in isobaric carotid distensibility, independently of MAP. Taken together, the results suggest that both operational and isobaric carotid distensibility might be adequate predictors of cardiovascular survival, thus directing future research toward molecular mechanisms involved in the effects of early HSD on the renin-angiotensin system and components of the extracellular matrix determining vascular stiffness.

We thank Dr. Anne Safar for her great support and advice.