Persistent Vascular Dysfunction Following an Acute Non-Pharmacological Reduction in Blood Pressure in Hypertensive Patients

Abstract

Vascular dysfunction, an independent risk factor for cardiovascular disease, often persists in patients with hypertension, despite improvements in blood pressure control induced by antihypertensive medications. As some of these medications may directly affect vascular function, this study sought to comprehensively examine the impact of reducing blood pressure, by a non-pharmacological approach (5 days of sodium restriction), on vascular function in 22 hypertensive individuals (14 males/8 females, 50±10 y). Following a 2-week withdrawal of antihypertensive medications, two 5-day dietary phases, liberal sodium (LS, 200 mmol/d) followed by restricted sodium (RS, 10mmol/day), were completed. Resting blood pressure was assessed and vascular function, at both the conduit and microvascular levels, was evaluated by brachial artery flow mediated dilation (FMD), reactive hyperemia, progressive handgrip exercise (HG), and passive leg movement (PLM). Despite a sodium restriction-induced fall in blood pressure (LS: 141±14/85±9; RS 124±12/79±9 mmHg, p<0.01 for both systolic and diastolic blood pressure), FMD (LS: 4.6±1.8%; RS: 5.1±2.1%, p=0.27), and reactive hyperemia (LS: 548±201; RS: 615±206 ml, p=0.08) were not altered. Similarly, brachial artery vasodilation during HG exercise was not different between conditions (LS: Δ 0.36±0.19 mm; RS: Δ 0.42±0.18 mm, p=0.16). Lastly, PLM-induced changes in peak blood flow (LS: 5.3±2.5; RS: 5.8±3.6ml/min/mmHg, p=0.30) and the total vasodilatory response (LS: 2.0[0.9–2.5] vs. RS: 1.7[1.1–2.6]ml/min/mmHg; p = 0.5) were also not different between conditions. Thus, vascular dysfunction, at both the conduit and microvascular levels, persists in patients with hypertension even when blood pressure is acutely reduced by a non-pharmacological approach.

Introduction:

Vascular dysfunction is a hallmark characteristic of hypertension and an independent risk factor for cardiovascular disease [1–4]. Indeed, a negative relationship between blood pressure and vascular function, as measured by flow mediated dilation (FMD), has been clearly established in untreated hypertensive patients [5]. However, this relationship may not hold true when blood pressure is lowered by antihypertensive medications, as these medications may directly affect vascular function [5–9]. Recent evidence indicates that vascular dysfunction persists despite a reduction in blood pressure as a result of antihypertensive medications and this may contribute to the augmented cardiovascular risk in patients that have been treated for high blood pressure compared to untreated patients, when compared at a given blood pressure [10–13]. A better understanding of this persistent vascular dysfunction is critical, especially in light of the revised blood pressure control guidelines, aimed at intensively lowering blood pressure to minimize cardiovascular risks [14], and may help to explain why patients with hypertension remain at an elevated risk of cardiovascular disease despite improved blood pressure facilitated by medication.

Isolating the direct impact of blood pressure on vascular function in patients with hypertension who are taking antihypertensive medications requires non-pharmacological strategies to alter blood pressure in order to circumvent the potential direct effects of these medications on vascular function. One such approach examined the impact of increased blood pressure caused by acute pressure raising physical activity (weightlifting) on vascular function and reported remarkable reductions in vascular function, highlighting the rapid adaptability of the vascular endothelium to brief hypertensive exposure [15,16]. Alternatively, increasing or restricting dietary sodium intake is a highly effective approach to illicit sustained alterations in blood pressure and examine the effects on vascular function [17–19]. In salt sensitive individuals, acutely increasing sodium intake (7 days) increased blood pressure and diminished vascular function [20]. Interestingly, chronic sodium restriction (4 weeks) in patients with hypertension lowered resting blood pressure and improved vascular function, however, the improvements in vascular function were associated with increased nitric oxide (NO) bioavailability and were independent of changes in blood pressure [21]. Thus, it is difficult to ascertain from chronic studies whether lowering blood pressure directly impacts vascular function. Acute sodium restriction in patients with hypertension resulted in improvements in central arterial (carotid) compliance which were strongly related to reductions in resting blood pressure, however, the impact on peripheral vascular function was not assessed [22]. Thus, the effect of acutely lowering blood pressure, non-pharmacologically, on vascular function in patients with hypertension remains to be determined and would be useful to better elucidate the potential “cause and effect” between high blood pressure and vascular function in the hypertensive population.

This study sought to determine if acutely lowering blood pressure, through a sodium restricted diet, would alter vascular function in patients with hypertension. Specifically, utilizing a comprehensive battery of vascular function tests, providing assessments of conduit artery and microvascular function, this study tested the hypothesis that an acute reduction in blood pressure would not yield a concomitant improvement in vascular function. If confirmed, this would highlight the importance of developing and implementing therapies in patients with hypertension aimed at directly targeting vascular dysfunction, to further minimize cardiovascular risk.

Methods:

Patients

Twenty-two patients diagnosed with essential hypertension and prescribed antihypertensive medication (14 males/8 females, 51 ± 10 y, height 173 ± 11 cm, weight 99 ± 23 kg) volunteered to participate in this investigation. Patients refrained from medication use for at least two weeks, with physician oversight, prior to experimentation. Written informed consent was obtained from all subjects prior to inclusion in the study. The Institutional Review Boards of the University of Utah and the Salt Lake City Veterans Affairs Medical Center approved all protocols.

Protocols

All patients were studied in a thermoneutral environment and reported to the laboratory fasted and not having performed vigorous exercise within 24 h of the study. Prior to the experimental phase, patients reported to the laboratory on a preliminary day to complete a health history questionnaire, physical examination, and a venous blood draw. During the experimental phase of the protocol, participants completed 5 days of liberal sodium (LS, 200mmol/day) followed by 5 days of restricted sodium (RS, 10 mmol/day). Each diet also contained 100 mmol/day potassium and 20 mmol/day calcium, as previously described [23,24]. In accordance with previous studies using these diets [23,24], participants consumed their normal diet during the liberal salt phase and were given high sodium foods (V8 or chicken broth) to increase sodium intake to 200mmol/day. A 3-day dietary recall was performed and analyzed by a bionutritionist to estimate salt intake prior to the liberal salt diet. This information was then used to determine how much additional sodium intake was required from the V8 or chicken broth to increase sodium intake to 200 mmol/day. This information was also used to ensure adequate caloric and macronutrient intake during the reduced salt diet. Sodium balance was confirmed by 24 hour urine collection. Testing of the two experimental conditions was separated by 7 days and occurred on the fifth day of either LS or RS. Meals for the RS diet were prepared by the University of Utah’s Center for Clinical and Translation Science bio-nutritionist using the Harris-Benedict equation to calculate daily energy requirements with adjustments for activity level [25].

Resting Blood Pressure

Resting blood pressure was assessed at the beginning of each experimental visit. After instrumentation, participants sat, relaxed, upright in a chair, legs uncrossed, and the back and arms supported. A blood pressure cuff (Tango M2, SunTech Medical, Morrisville, NC) was placed on the upper arm, level with the right atrium. Care was taken to ensure that true resting conditions were achieved before blood pressure was assessed. Lights were dimmed and the subject sat undisturbed for 5 minutes with the researchers out of view. Three automated blood pressure measurements separated by 1 min were recorded and averaged to determine resting blood pressure [14,26].

Overview of Vascular Function Assessments

Data acquired during the vascular function assessments (FMD, HG, and PLM in the right arm and leg, respectively) included arterial blood velocity and diameter (brachial and common femoral arteries) using a Doppler ultrasound system (Logiq e9, GE Medical Systems, Milwaukee, WI) operating in duplex mode according to published guidelines [27]. Blood velocity was collected with the Doppler frequency optimized (3–5 MHz) in high-pulse repetition frequency mode (2–25 kHz) and vessel images were collected with the imaging frequency optimized (9–14 MHz). Sample volume was also optimized in relation to vessel diameter and centered within the vessel. During the FMD, reactive hyperemia, and HG tests ECG R-wave gated images were collected via video output from the Loqiq e9 for offline analysis of brachial artery diameter by automated edge-detection software (Medical Imaging Applications, Coralville, IA). For all tests, the shear rate was calculated as: shear rate (s⁻¹) = 8Vmean/diameter, blood flow (BF) was calculated with the formula: BF (ml·min⁻¹) = (Vmean × π (vessel diameter/2)² × 60), and vascular conductance (VC) was calculated as: VC (ml/min/mmHg) = BF / mean arterial blood pressure.

Conduit artery function was measured in the brachial artery approximately midway between the antecubital and axillary regions utilizing the flow mediated dilation (FMD). Forearm microvascular function was assessed simultaneously with FMD using the post-occlusive reactive hyperemia technique, which provides a measure of dilation in the downstream resistance vessels [27]. FMD was performed in accordance with current recommendations [28]. Briefly, patients rested in the supine posture for ~20 min prior to the start of data collection and remained in this position throughout the FMD assessment. FMD was performed with a blood pressure cuff placed on the right arm, just distal to the elbow. The FMD protocol consisted of 30 s of baseline data acquisition prior to inflation of the blood pressure cuff (250 mmHg for 5 min) and subsequent data acquisition for 2 min after cuff deflation. This two minute data acquisition, post cuff deflation, allowed the measurement of both FMD and reactive hyperemia. FMD was quantified as the maximal percentage change in brachial artery diameter after cuff release. Reactive hyperemia was calculated as the summed blood flow response over time after normalizing for baseline blood flow.

Static-intermittent progressive handgrip (HG) exercise was performed following an adequate recovery from the FMD cuff occlusion (at least 15 min) to assess shear-induced conduit vessel vasodilation [29]. Brachial artery diameter and blood velocity was measured via doppler ultrasound approximately midway between the antecubital and axillary regions. HG was performed in the supine position with the right arm extended and abducted to 90°. Maximal voluntary contraction force (MVC) was calculated as the greatest of, at least, three maximal contractions of the handgrip dynamometer (TSD121C, Biopac Systems, Goleta, CA). HG exercise was performed at three relative exercise intensities based on each patient’s respective MVC (15, 30, and 45% of MVC) at a contraction rate of 1 Hz, assisted by the sound of a metronome, and real-time force displayed to provide visual feedback. Each exercise intensity was performed for 3 min, to allow the achievement of steady-state hemodynamics, with a 2 min recovery period between each workload. The HG protocol consisted of 30 s of baseline data acquisition prior to exercise and subsequent data acquisition during the final min of each exercise stage.

Passive leg movement (PLM) was performed, as previously described, to assess microvascular function of the lower limb [30–33]. Briefly, patients initially rested in an upright seated position, with legs extended and supported at ~180 degrees, for ~20 min. Ultrasound imaging was performed 2–3 cm proximal to the bifurcation of the common femoral artery after steady state hemodynamics were determined. Following baseline steady state measurements, the 1 min bout of passive leg flexion and extension was performed by a member of the research team moving the lower leg through a 90° range of motion at 1 Hz (starting at a 180° angle of the knee joint), while the contralateral leg remained extended and supported. Patients were encouraged to relax and remain motionless during the protocol. Patients were made aware that PLM would take place in ~1 min, but to minimize the chance of an anticipatory response, they were not informed of exactly when this movement would begin. The peak change in leg blood flow (LBF) and leg vascular conductance (LVC) (ΔLBF_PEAK and ΔLVC_PEAK, respectively) were calculated as peak minus baseline and cumulative area under the curve (AUC) for these variables was calculated as the summed second-by-second response above baseline during the first 60 s of passive movement.

Statistical Analyses

Statistical analysis was performed using commercially available software (SigmaPlot 11.0; Systat Software, Point Richmond, CA). The impact of LS and RS on vascular function assessed by FMD and PLM was compared using a paired t-test or a nonparametric signed rank test for non-normally distributed data and reported as mean±SD or median [interquartile range], respectively. A 2 × 3 repeated-measures ANOVA (α < 0.05 for all tests) (group: LS vs. RS) (intensity: 15, 30, and 45% MVC) was performed to determine differences in diameter and shear responses during HG exercise. Stata mixed (Stata 17, StataCorp LLC, College Station, TX, USA) was used to a conduct linear mixed-effects analysis of the relationship between the change in brachial artery diameter and the change in shear rate. The model consisted of a dependent variable (Δ diameter), independent variables (Δ shear rate, salt condition, and Δ shear rate × salt condition interaction) as fixed effects, and a random effect for subject (intercept) to account for repeated measures within subjects. Pearson correlations were used to evaluate the relationship between blood pressure and vascular function. Data were analyzed for potential sex differences and the impact of BMI. Based on the results of this initial analysis, no significant sex differences or effect of BMI, the data for men and women were pooled and BMI was not included as a covariate.

Results:

Patient Characteristics and Blood Pressure

Initial patient characteristics and blood chemistries, assessed in the LS condition, are presented in Table 1. Notably, as a consequence of sodium restriction, body weight significantly decreased (LS: 99±24, RS: 97±22 kg, p < 0.001). Resting systolic blood pressure (LS: 141±14, RS: 124±12 mmHg), diastolic blood pressure (LS: 85±9, RS: 79±9 mmHg), and mean arterial pressure (LS: 103±9, RS: 94±9 mmHg) were decreased during RS compared with LS (Figure 1, all p < 0.001). Based on laboratory blood pressure assessments during the LS phase, 13 patients presented with stage 2 hypertension, 8 patients with stage 1 hypertension, and 1 with elevated blood pressure, according to current guidelines [34]. On average, sodium restriction improved blood pressure by 2 clinical categories, from stage 2 hypertension to elevated blood pressure. Therefore, following the RS phase, 7 patients presented with normal blood pressure (<120/80), 4 with elevated blood pressure, 6 with stage 1 hypertension, and 4 with stage 2 hypertension.

Table 1.

Patient characteristics

	Liberal sodium	Reference range
Age, y	51 ± 10
Sex, n	14 M / 8 F
Height, cm	173 ± 11
Weight, kg	99 ± 23
BMI, kg/m2	33 ± 6
Handgrip MVC, kg	20 ± 4
Hemoglobin (g/dL)	15 ± 1	14.6 – 17.8
Hematocrit (%)	45 ± 4	40.8 – 51.9
RBC, M/μL	5.0 ± 0.6	4.69 – 6.07
Platelet, k/μL	251 ± 75	177 – 406
WBC, k/μL	6.4± 2.0	3.20 – 10.60
Sodium, mmol/L	141 ± 3	136 – 144
Potassium, mmol/L	4.4 ± 0.4	3.3 – 5.0
Chloride, mmol/L	103 ± 4	102 – 110
Carbon Dioxide, mmol/L	23 ± 2	20 – 26
BUN, mg/dL	15± 3	8 – 24
Creatinine, mg/dL	0.94 ± 0.15	0.72 – 1.25
Glucose, mg/dL	99 ± 12	64 – 128
Calcium, mg/dL	9.6 ±0.6	8.4 – 10.5
Protein, Total, g/dL	7.1 ± 0.5	6.5 – 8.4
Albumin g/dL	4.5 ± 0.2	3.5 – 5.0
Hgb A1C, %	5.7 ± 0.4	< 6.0

BMI, body mass index; MVC, maximal voluntary contraction force; RBC, red blood cell; WBC, white blood cell; BUN, Blood urea nitrogen; Hgb A1C, glycated hemoglobin; Data are mean ± SD.

Figure 1.

Systolic (SBP, 1A), diastolic (DBP, 1B), and mean arterial blood pressure (MAP, 1C) during liberal sodium (LS, 200 mmol/day) and restricted sodium (RS, 10 mmol/day) conditions. Paired t-tests were used to evaluate pre vs. post salt restriction differences in blood pressure variables (N=22). Data are presented as individual (◯) and mean ± SD (bars). * Indicates P < 0.05.

Vascular Function

Baseline brachial artery diameter, prior to FMD testing, was unchanged following sodium restriction (LS: 4.6 ± 0.6mm vs. RS: 4.6 ± 0.7mm, p = 0.063). Additionally, sodium restriction did not alter conduit artery function as measured by the absolute (LS: 0.20 ± 0.08mm vs. RS: 0.22 ± 0.09mm, p = 0.528) or relative (LS: 4.6 ± 1.8% vs. RS: 5.1 ± 2.1%; p = 0.265, Figure 2A) FMD of the brachial artery. Normalizing this dilation for the shear stimulus had no impact on this finding (LS: 0.15 ± 0.07mm/s vs. RS: 0.12 ± 0.05 mm/s; p = 0.399 Figure 2B). Reactive hyperemia was not different between conditions (LS: 548±201ml vs. RS: 615±206ml; p = 0.080, Figure 2C). Of note, participants, on average, exhibited persistently poor vascular function, as assessed by FMD, relative to published norms [35], in both the liberal sodium (36^th percentile) and the sodium restriction phases (41^st percentile).

Figure 2.

Vascular function assessed by flow-mediated dilation (FMD) during liberal sodium (LS, 200 mmol/day) and restricted sodium (RS, 10 mmol/day) conditions. A: FMD measured as percent change in brachial artery diameter following cuff occlusion. B: FMD response normalized for shear stimulus up to peak dilation. C: the reactive hyperemia response following cuff occlusion. Paired t-tests were used to evaluate pre vs. post salt restriction differences in FMD, FMD/shear, and reactive hyperemia (N=22). Data are presented as individual (◯) and mean ± SD (bars).

During progressive HG exercise, continuing dilation of the brachial artery (Figure 3A) was evident in both LS (change from baseline diameter: 15%: 0.06 ± 0.09mm, 30%: 0.19 ± 0.16mm, 45%: 0.36 ± 0.19 mm) and RS (15%: 0.07 ± 0.11mm, 30%: 0.23 ± 0.16mm, 45%: 0.42 ± 0.18 mm), which was not different between conditions (p > 0.05 for all pairwise comparisons). The progressive increase in shear rate during HG exercise was also not different between conditions (p > 0.05) (Figure 3B). Delta shear rate was significantly related to delta diameter (beta = 0.0005, SE = 0.0001, p<0.0001, 95% confidence interval (CI) = 0.0003, 0.0007), however there was no effect of sodium restriction (beta = −0.093, SE = 0.067, p=0.166, 95% CI = −0.22, 0.04) or delta shear by sodium restriction interaction (beta = 0.0001, SE = 0.0001, p=0.469, 95% CI = −0.08, 0.15; Figure 3C).

Figure 3.

Brachial artery (BA) vasodilatory responses to handgrip (HG) exercise during liberal sodium (LS, 200 mmol/day; (●) and solid line, mean ± SD) and restricted sodium (RS, 10 mmol/day; (◯) and dotted, mean ± SD, conditions. A 2 × 3 repeated-measures ANOVA (group: LS vs. RS) (intensity: 15, 30, and 45% MVC) was performed to determine differences in brachial artery vasodilation and shear rate during HG exercise (N=22). A: Percent change in BA diameter in response to progressive (15, 30, and 45% of maximal voluntary contraction, MVC) HG exercise. B: Shear rate in response to progressive HG exercise. C: The relationship between the change in shear rate and brachial artery vasodilation for each individual (LS, 200 mmol/day, solid lines; RS, 10 mmol/day, dotted lines) and the linear mixed effects slope for each condition (LS, red line; RS, blue line). A linear mixed-effects model was used to evaluate the relationship between the change in brachial artery diameter and the change in shear rate. The model consisted of a dependent variable (Δ diameter), independent variables (Δ shear rate, salt condition, and Δ shear rate × salt condition interaction) as fixed effects, and a random effect for subject (intercept) to account for repeated measures within subjects (N=22).

Femoral artery diameter was unchanged, prior to PLM, with sodium restriction (LS: 9.5 ± 1.2mm vs. RS: 9.5 ± 1.2mm; p = 0.516). Baseline LBF was significantly lower during RS compared to LS (LS: 408 ± 151ml/min vs. RS: 314 ± 103ml/min; p = 0.002) (Figure 4). However, accounting for the effects on blood pressure, baseline LVC was not different between LS and RS (LS: 4.0 ± 1.6 vs. RS: 3.4 ± 1.2ml/min/mmHg, p = 0.054) (Figure 5). Vascular function, assessed in the thigh by PLM, was not altered by sodium restriction as there was no difference in ΔLBF_PEAK (LS: 489 ± 225 vs. RS: 499 ± 257ml/min, p = 0.797) or LBF_AUC (LS: 217 [96–264] vs. RS: 166 [97–196] ml/min, p = 0.820) (Figure 4). Furthermore, there was no difference in ΔLVC_PEAK (LS: 5.3 ± 2.5 vs. RS: 5.8 ± 3.6 ml/min/mmHg; p = 0.297) or LVC_AUC (LS: 2.0 [0.9–2.5] vs. RS: 1.7 [1.1–2.6] ml/min/mmHg; p = 0.495) between the two conditions (Figure 5).

Figure 4.

Passive leg movement (PLM)-induced hyperemia during liberal sodium (LS, 200 mmol/day) and restricted sodium (RS, 10 mmol/day) conditions. A: Absolute leg blood flow (LBF). B: Change in LBF, normalized for the baseline reduction in LBF in the RS condition. C: Peak change in LBF. D: LBF area under the curve (AUC) calculated as the summed second-by-second response above baseline. The impact of LS and RS on PLM variables was compared using a paired t-test (LBF, ΔLBF_peak, and ΔLBF) or a nonparametric signed rank test (LBF_AUC) (N=22). Values are mean ± SD. * Indicates P < 0.05.

Figure 5.

Passive leg movement (PLM)-induced leg vascular conductance (LVC) during liberal sodium (LS, 200 mmol/day) and restricted sodium (RS, 10 mmol/day) conditions. A: Absolute LVC. B: Peak change in LVC, normalized for baseline LVC. C: Change in peak LVC. D: LVC area under the curve (AUC) calculated as the summed second-by-second above baseline. The impact of LS and RS on PLM variables was compared using a paired t-test (LVC, ΔLVC_peak, and ΔLVC) or a nonparametric signed rank test (LVC_AUC) (N=22). Values are mean ± SD.

There were no significant correlations between the RS phase-induced changes in blood pressure (systolic, diastolic, or mean arterial blood pressure) and the change in brachial artery FMD, brachial artery vasodilation in response to handgrip exercise, or PLM-induced hyperemia (Figure 6). Moreover, lowering blood pressure below 120/80 mmHg (n=7) was not associated with improved vascular function (data not reported). When sex was added as a factor, there was no significant effect of sex or significant interaction of sex with blood pressure. Additionally, there was no significant effect of BMI or significant interaction of BMI with blood pressure.

Figure 6.

Relationship between the changes in mean arterial blood pressure (MAP) and changes in vascular function. Correlation between MAP change and A: flow-mediated dilation of the brachial artery (BA), B: BA vasodilation during progressive handgrip exercise, C: passive leg movement (PLM)-induced hyperemia, measured as peak leg blood flow (LBF), and D: PLM-induced hyperemia, measured as LBF area under the curve (AUC). Pearson correlations were used to evaluate the relationship between the change in mean arterial pressure and vascular function (FMD, BA vasodilation, ΔLBF_peak, LBF_AUC) (N=22). *Indicates the correlation results with the outlier removed for panel D.

Discussion

This study provides novel evidence that both conduit artery and microvascular function remain impaired in patients with hypertension following a substantial non-pharmacological reduction in blood pressure, evoked by acute dietary sodium restriction. By utilizing a comprehensive battery of both conventional (FMD) and innovative assessments of vascular function (HG and PLM), these findings support and extend recent findings that blood pressure improvements, accomplished by a non-pharmacological approach, do not obligate a concomitant change in vascular function in patients with hypertension [5]. Identifying that vascular dysfunction is a persistent consequence of hypertension, despite significant and clinically relevant improvements in blood pressure, has important public health implications as endothelial dysfunction is associated with increased risk of cardiovascular events.

Vascular Dysfunction Independent of Blood Pressure Status in Hypertension

Understanding the impact of hypertension on vascular function is critically important as vascular dysfunction precedes the development of overt cardiovascular disease, is independently associated with the pathogenesis of atherosclerosis, and predicts future cardiovascular events [4,6,36,37]. In the context of hypertension, vascular dysfunction is clearly present as evidenced by reports of impaired vasodilation in response to acetylcholine and FMD in the coronary and peripheral vasculature, respectively [1–3,5,38,39]. This diminished vascular function has been largely attributed to attenuated NO bioavailability, thought to be initiated by elevated oxidative stress. In the current study, vascular function assessed by FMD was, on average, 4.6% during the liberal sodium phase, which falls in the 36^th percentile of published norms [35], and agrees with several other reports, confirming the presence of impaired vascular function in the hypertensive population [5,21,40]. Sodium restriction was highly effective at improving blood pressure status as patients, classified as hypertension stage 2 (systolic blood pressure >140 or diastolic blood pressure >90) during the liberal sodium phase, on average, presented in the elevated range (~123/79 mmHg) after just 5 days of sodium restriction. These findings are very similar to several other studies utilizing comparable dietary sodium restriction interventions [23,24]. Despite these significant and clinically meaningful improvements in blood pressure following sodium restriction, brachial artery FMD was not significantly improved (5.1%, 41^st percentile) (Figure 2A and B).

The assessment of vascular function by FMD has been considered a bioassay of NO bioavailability [41–44]. This perceived NO-dependency of FMD is important as NO is cardioprotective and antiatherogenic. However, data from our group and others have challenged this notion by providing evidence that FMD may only be minimally NO-dependent (~30%) [42,45,46]. To address this concern several non-invasive assessments of vascular function have been developed, namely, PLM and progressive HG. During PLM, the robust hyperemic and vasodilatory response in the microvasculature is largely NO-dependent (~60–80%) in healthy adults and is significantly attenuated by aging and disease [32,47–49]. Likewise, during HG exercise brachial artery vasodilation is attenuated in the presence of NO synthase inhibition by N^G-monomethyl-L-arginine (L-NMMA) and is augmented in the presence of ascorbic acid [50,51]. Collectively, evidence indicates that both PLM and HG provide highly NO-dependent assessments of vascular function at the microvascular and conduit level, respectively. The patients with hypertension in the current study displayed a ~30–40% lower PLM-induced change in hyperemia and vasodilation compared to a previous report in healthy adults [32]. Interestingly, microvascular function assessed by PLM was not improved by the lowering of blood pressure induced by sodium restriction (Figure 4) and this is in agreement with microvascular function assessed by the reactive hyperemia, induced by tissue ischemia, during the FMD measurements (Figure 2C). In terms of conduit artery function, brachial artery dilation was not different between LS and RS conditions during progressive HG exercise (Figure 3), which is also in agreement with lack of an effect measured by FMD (Figure 2A and B). Overall, this comprehensive assessment of vascular function by four unique approaches provides strong evidence that an improvement in blood pressure in hypertensive patients does not oblige a concomitant improvement in vascular function.

Blood Pressure and Vascular Function: Partitioning Cause and Effect

Examining the relationship between blood pressure status and vascular function in hypertensive patients is important to determine if an elevated risk of CVD, due to vascular dysfunction, persists despite improved blood pressure control. Antihypertensive medications are effective at lowering blood pressure, however, the impact of these medications on vascular function is not clear and appears to be drug-class dependent [9]. Of note, little information is available regarding blood pressure status and vascular function in patients receiving treatment for hypertension. Interestingly, in untreated, drug naïve hypertensive patients, the severity of hypertension has been directly linked to the magnitude of vascular dysfunction [5,7,8]. Indeed, a recent large-scale study of hypertensive patients reported an inverse relationship between blood pressure and vascular function, measured by FMD, in drug naïve patients, however, vascular dysfunction remained unaltered in treated hypertensive patients [5]. Therefore, an inverse correlation between hypertension and vascular function may only be evident in a select cohort of patients who are unmedicated. Although the patients with hypertension in the current study had withdrawn from antihypertensive medications for at least 2 weeks prior to assessments, there was no evidence a relationship between blood pressure improvement and measures of vascular function (Figure 6).

To circumvent the complex and potentially cofounding impact of antihypertensive medications on both blood pressure and vascular function, this study employed acute dietary sodium restriction (5 days) to lower blood pressure. Of note, intervening with dietary sodium is not without important caveats, especially when examining vascular function. Modifying dietary sodium intake, either through high sodium loading or sodium restriction, has been linked to alterations in vascular function as several studies report direct effects of sodium on the vasculature [20,52–54]. In both sodium-sensitive and sodium-resistant normotensive individuals, high dietary sodium intake has been reported to induce vascular dysfunction [20,53]. Conversely, sodium restriction in middle-aged and older adults with elevated blood pressure improved vascular function [21,40]. Importantly, and particularly relevant to the current study, these improvements in vascular function, induced by more chronic sodium restriction (2 to 4 weeks), were independent of changes in blood pressure and were mechanistically linked to diminished oxidative stress leading to increased NO bioavailability. In terms of the current study, the failure to see an improvement in PLM and HG, as these assessments are highly NO dependent, suggests that the current 5-day sodium restriction did not improve NO bioavailability.

Sodium intake and vascular function

The level of sodium intake during the restricted sodium phase of this study was quite low (10 mmol/day), albeit not without precedent [23,24], when compared to several other studies evaluating the impact of a low sodium diet on blood pressure and vascular function (20–60 mmol/day) [21,22,53]. Importantly, this level of sodium intake was selected based on previous studies reporting marked reductions in blood pressure in just 5 to 6 days [23,24] and the linear relationship between sodium intake and blood pressure [55]. A low sodium intake has been associated with elevated sympathetic nervous system activity which may offset improved vasodilatory capacity [56]. However, this is unlikely given the large reduction in blood pressure in the current study which was not observed previously [57]. Additionally, sodium restriction may have decreased blood volume causing increases in Vasopressin and slight vasoconstriction, which may have offset any vascular function improvements. Sodium has been reported to have vascular effects independent of changes in blood pressure that are linked to changes in oxidative stress [40,54]. Such effects were not observed in the present study, which may be due to the relatively short duration of sodium restriction that was utilized. A longer sodium restriction intervention lasting several weeks may be required to evoke changes in oxidative stress and vascular function. Direct markers of oxidative stress were not assessed in the current study. However, as already noted, the novel bioassays of NO bioavailability employed in this study (PLM and HG) were not improved following sodium restriction suggesting that the oxidative stress and NO axis was not altered with sodium restriction. Additionally, the direct effect of sodium on vascular function may only be evident when undergoing a dramatic increase in sodium intake [20,53]. Specifically, studies reporting a decrease in vascular function with high sodium intake had participants consume 300–350 mmol/day of sodium [20,53]. In contrast, in the current study, to ensure homogeneity among patients, the liberal sodium phase was achieved by a relatively modest increase in sodium intake of just 200 mmol/day. Furthermore, we did not see a significant effect or interaction of sex on the vascular function response to salt restriction-induced reductions in blood pressure, however this study may not be adequately powered to explore sex differences.

Perspectives

Recent changes in the guidelines for the therapeutic treatment of hypertension, prompted by findings from the Systolic Pressure Intervention Trial (SPRINT), recommend intensively lowering systolic blood pressure to below 120 mmHg [14]. SPRINT determined that targeting a systolic blood pressure of less than 120 mmHg was associated with reduced cardiovascular events (myocardial infarction, acute coronary syndromes, and heart failure) and improved mortality. Vascular function was not assessed as part of this trial and it remains unknown if patients receiving intensive antihypertensive therapy exhibited any changes in vascular function. As already noted, a large-scale trial assessing the impact of antihypertensive treatment on vascular function, Maruhashi et al. reported persistent vascular dysfunction independent of blood pressure status [5]. This finding may provide a rationale for the findings of several epidemiological studies which reported higher cardiovascular risk in treated patients with hypertension compared to those without treatment even at a given blood pressure level [10–13]. Together, these findings indicate that an assumption of improved vascular function coinciding with improved blood pressure status is likely incorrect and, therefore, an elevated risk of cardiovascular disease likely remains in patients with hypertension, despite receiving effective medical therapy. In light of these recent advances and large-scale changes in the guidelines for the treatment of blood pressure, the clinical need to assess vascular function becomes imperative to further assess and minimize risks of cardiovascular disease and improve mortality in patients with hypertension.

Conclusions

The lowering of blood pressure through sodium restriction did not improve conduit artery or microvascular function in patients with hypertension following a 2-week antihypertensive medication withdrawal. Hence, the assumption that improved blood pressure in patients with hypertension, by either pharmacological or non-pharmacological means, confers improved vascular function is not supported. Vascular dysfunction, acutely, remains a consequence of hypertension, independent of blood pressure status. Strategies aimed at directly improving vascular function should be included in therapies aimed at minimizing the risk of cardiovascular disease in patients with hypertension.