Non-invasive measurement of the haemodynamic effects of inhaled salbutamol, intravenous L-arginine and sublingual nitroglycerin

Abstract

AIMS

To examine the effects of salbutamol and L-arginine, two compounds acting largely on the endothelium, and the endothelium-independent agent nitroglycerin on blood pressure, arterial compliance, cardiac function and vascular resistance.

METHODS

Continuous radial pulse wave analysis, whole-body impedance cardiography, and plethysmographic blood pressure from fingers in the supine position and during head-up tilt were recorded in nine healthy subjects. Data were captured before and after L-arginine (10 mg mg⁻¹ min⁻¹) or saline infusion, salbutamol (400 µg) or placebo inhalation, and sublingual nitroglycerin (0.25 mg) or placebo resoriblet.

RESULTS

The results of all measurements were comparable before drug administration. The effects of inhaled salbutamol were apparent in the supine position: systemic vascular resistance (−9.2 ± 2.6%) and augmentation index (−4.0 ± 1.5%) decreased, and heart rate (8.6 ± 2.5%) and cardiac output (8.8 ± 3.1%) increased. L-arginine had no clear effects on supine haemodynamics, but during head-up tilt blood pressure was moderately decreased and reduction in aortic reflection time prevented, indicating improved large arterial compliance. Nitroglycerin reduced supine vascular resistance (−6.7 ± 1.8%) and augmentation index (−7.4 ± 1.6%), and increased cardiac output (+9.2 ± 2.7%). During head-up tilt, nitroglycerin increased cardiac output (+10.6 ± 5.6%) and heart rate (+40 ± 7.5%), decreased vascular resistance (−7.8 ± 5.8%) and augmentation index (−18.7 ± 3.2%), and prevented the decrease in aortic reflection time.

CONCLUSIONS

Inhaled salbutamol predominantly changed supine haemodynamics, whereas the moderate effects of L-arginine were observed during the head-up tilt. In contrast, small doses of nitroglycerin induced major changes in haemodynamics both supine and during the head-up tilt. Altogether, these results emphasize the importance of haemodynamic measurements in both the supine and upright positions.

WHAT IS ALREADY KNOWN ABOUT THIS SUBJECT

• Haemodynamic effects of endothelial stimuli induced by salbutamol and L-arginine in humans have been previously studied predominantly at rest in selected vascular beds.

• We studied the effects of salbutamol, L-arginine and nitroglycerin on cardiac and vascular function by continuous recording of pulse wave analysis and impedance cardiography both in the supine position and during passive head-up tilt.

WHAT THIS STUDY ADDS

• The divergent effects of the research drugs in supine position and during head-up tilt indicate that human haemodynamics should also be studied in the upright position.

• Since inhaled salbutamol induced more pronounced changes in haemodynamics, it provides a clinically more applicable tool than infused L-arginine for the assessment of endothelial function in humans.

Introduction

The endothelium controls vascular tone via various mechanisms including the release of nitric oxide (NO), prostacyclin, and several pathways causing vascular smooth muscle cell hyperpolarization [1]. In addition, the endothelium actively regulates vascular permeability, platelet and leucocyte adhesion and aggregation, and thrombosis [2, 3]. Impairments in these mechanisms maintaining vascular homeostasis lead to functional manifestations, which has emphasized the importance of studying endothelial function in vivo. The present treatment strategies of hypertension with lower treatment goals than ever, especially in high-risk patients with complicated diabetes or kidney disease, stress the importance of the measurement of blood pressure (BP), haemodynamics and endothelial function also in the upright position [4, 5]

Peripheral endothelial function can be measured invasively by venous occlusion plethysmography, and non-invasively by measuring arterial diameter by ultrasound systems after ischaemia or warming-induced increase in blood flow [flow-mediated dilatation (FMD)][6, 7]. Repeatability of the plethysmographic measurements is good, but as an invasive technique it is not convenient for multiple measurements. FMD requires good technical skills, and the lack of universal standards in the measurement protocol makes comparison between different laboratories difficult [8]. The responsiveness of coronary arteries to different endothelium-dependent stimuli has been evaluated by observing changes in artery diameter [9] or coronary blood flow [10, 11]. However, these methods are invasive and laborious, and demand high technical expertise.

Pulse wave analysis (PWA) is a non-invasive, repeatable technique that analyses the arterial pulse wave form providing information about arterial compliance [12]. Recently, the changes in PWA-derived measure of wave reflection and arterial stiffness, the augmentation index (AIx), after β₂-adrenoceptor agonist-induced endothelial stimulation have been used to assess the role of endothelium in vascular responsiveness [13, 14]. Another possibility to stimulate the endothelium is the administration of the semi-essential amino acid L-arginine, which serves as a precursor for endogenous NO synthesis. L-arginine administration has been shown to improve endothelium-dependent vasodilation in animals [15, 16] and humans [17, 18]. However, the effect of endothelial stimulation on systemic vascular resistance or cardiac function has seldom been studied [19, 20].

The purpose of this study was to evaluate the haemodynamic effects of inhaled salbutamol and infused L-arginine in comparison with the endothelium-independent vasodilator nitroglycerin. For a comprehensive analysis we applied continuous PWA combined with the measurement of cardiac output and systemic vascular resistance using whole-body impedance cardiography both in the supine position and during passive orthostatic challenge. The results suggest that inhaled salbutamol induces changes in supine haemodynamics, the moderate effects of L-arginine are observed during head-up tilt, whereas already a small dose nitroglycerin influences haemodynamics in both the supine and upright positions.

Methods

Study subjects

The study population consisted of nine healthy (five male, four female), normotensive individuals aged 25–44 years. All participants gave written informed consent and thereafter underwent a physical examination performed by a physician. During the interview, lifestyle habits, family history of cardiovascular disease, and medical history were documented. The study was approved by the Ethics Committee of Tampere University Hospital and the National Agency of Medicines, Finland, and it complies with the Declaration of Helsinki.

Laboratory analyses

Blood and urine samples were obtained after a minimum of 12 h fast in the morning for laboratory analyses, and a standard 12-lead electrocardiogram was recorded. Plasma sodium, potassium, calcium, glucose, creatinine, triglyceride, and total, high-density and low-density lipoprotein cholesterol concentrations were determined by Cobas Integra 700/800 (F. Hoffmann-LaRoche Ltd, Basel, Switzerland), and blood cell count by ADVIA 120 or 2120 (Bayer Health Care, Tarrytown, NY, USA). Creatinine clearance was estimated using the Cockroft–Gault formula [21].

Haemodynamic measurement protocol

Haemodynamic measurements were performed in a quiet, temperature-controlled research laboratory by a trained nurse on five separate days. The study subjects had refrained from caffeine-containing products, smoking and heavy meals for at least 4 h and from alcohol for at least 24 h prior to the investigation. The subjects were resting supine on a tilt-table, and the electrodes for impedance cardiography were placed on the body surface, the tonometric sensor for PWA on the radial pulsation to the left wrist, and oscillometric brachial cuff for BP calibration to the right upper arm. Before the actual measurement, an introductory head-up tilt was performed to familiarize the study subject with the method.

The actual measurement consisted of six consecutive 5-min intervals, during which haemodynamic data were captured continuously. For the first 5 min, subjects were resting supine on the tilt table, followed by 5 min of head-up tilt to 60°, then the tilt table was returned to the horizontal position for another 5 min. During this period no research drugs were given, but saline was infused before the head-up tilt with L-arginine. After that the research drug (L-arginine, salbutamol, placebo inhalation, nitroglycerin or placebo resoriblet) was administered, the same protocol was repeated (5 min supine–5 min head-up tilt–5 min supine). The results of the haemodynamic measurements during the first 15 min on the five separate recording days did not differ either in the supine position or during the orthostatic challenge before the test drug administration (data not shown).

Research drugs

Administration of L-arginine, salbutamol, placebo inhalation, nitroglycerin and placebo resoriblet was performed on five separate measurements on different days. Placebo inhalation and 400 µg salbutamol inhalation (placebo for Ventolin® and Ventolin®, respectively; GlaxoSmithKline, Uxbridge, UK) were given with a spacer device (Volumatic; Allen & Hanbury's, Uxbridge, UK) in a blinded fashion. The 400-µg salbutamol dose was chosen on the basis of test experiments with the present study protocol and previously published work on the effects of inhaled salbutamol on haemodynamics [13, 14, 22].

Sublingual placebo resoriblet and 0.25 mg nitroglycerin (Nitro resoriblet; Orion Pharma, Espoo, Finland) were also administrated in a blinded fashion. The placebo resoriblets were professionally manufactured by the University Pharmacy, Helsinki, Finland, to resemble very closely the commercial nitroglycerin resoriblets in both appearance and taste. The nitroglycerin dose was chosen on the basis of test experiments with the present study protocol, and a previous report using the same dose [13].

In the fifth measurement, an intravenous (i.v.) 20-G cannula was placed in a brachial vein and a slow saline infusion was started. L-arginine hydrochloride 20 mg ml⁻¹ (B. Braun Melsungen Ag, Melsungen, Germany) was diluted in 100 ml of saline, and infusion at the dose 10 mg kg⁻¹ min⁻¹ was started at 15 min and continued for 10 min. This administration protocol resulted in a relatively high dose of infused L-arginine in 10 min. This dose can be considered sufficient to induce cardiovascular changes, as i.v. L-arginine infusion at the rate of (i) 3.33 mg kg⁻¹ min⁻¹ has induced vasodilation in the renal vasculature in humans [23], (ii) 0.3 g min⁻¹ has enhanced the acute haemodynamic effects of 50 mg of losartan [24], and (iii) 0.5 g min⁻¹ has increased skeletal muscle glucose clearance in humans, probably via increased NO production [25]. On the basis of previous work, L-arginine has rarely caused side-effects at infusion rates that do not exceed 1 g min⁻¹[26], and the dose of 1 g min⁻¹ of L-arginine has been used in many studies [27–29].

Pulse wave analysis

Radial BP and pulse wave form were continuously determined from the radial pulsation by a tonometric sensor (Colin BP-508T; Colin Medical Instruments Corp., San Antonio, TX, USA), which was fixed on the radial pulse with a wrist band. The radial BP signal was calibrated every 2.5 min by a brachial BP measurement. Continuous aortic BP was derived with the SphygmoCor pulse wave monitoring system (SpygmoCor PWMx; AtCor Medical, West Ryde, Australia) using the previously validated generalized transfer function [12]. Ejection duration, aortic reflection time and AIx (augmented pressure/pulse pressure × 100) were determined.

Whole-body impedance cardiography

A whole-body impedance cardiography device (CircMon^R; JR Medical Ltd., Tallinn, Estonia), which records the continuous changes in body electrical impedance during a cardiac cycle, and plethysmographic BP recordings from fingers (Finapres; Ohmeda, Englewood, CO, USA) were used to determine beat-to-beat heart rate, stroke volume (ml), cardiac output (l min⁻¹), systemic vascular resistance index (SVRI, systemic vascular resistance/body surface area, dyn × s cm⁻⁵× m²) and pulse wave velocity (PWV) [30–32]. The cardiac output values measured with CircMon^R whole-body impedance cardiography are in good agreement with the values measured by the thermodilution method, both in the supine position and during head-up tilt [32]. A detailed description of the method and electrode configuration has been previously reported [30–32]. PWV was not assessed during the head-up tilt due to less accurate timing of left ventricular ejection during reduced stroke volume.

Exhaled alveolar nitric oxide measurement after salbutamol inhalation

To verify the salbutamol-induced production of NO, we measured NO concentration from the alveolar air in six of the study subjects using a Sievers NOA 280 analyser (Sievers Instruments, Boulder, CO, USA) at three exhalation flow rates (100, 200 and 300 ml s⁻¹) [33, 34]. The measurements were performed three times before (0, 10 and 20 min) and twice after (10 and 20 min) the placebo and salbutamol 400-µg inhalations, and alveolar NO concentration was calculated as previously described [33, 34]. Briefly, exhaled NO output (concentration × flow rate) was plotted against exhalation flow rate and a linear regression was set. The slope and intercept of the regression line are approximates of alveolar NO concentration and bronchial NO flux, respectively. The exhalation flow rates were computer-controlled using an adjustable flow restrictor, and the subjects maintained the exhalation pressure between 5 and 20 cmH₂O, as described previously [34, 35].

Statistical analysis

Values are expressed as mean ± standard error of the mean (SEM) and the data were analysed using SPSS 11.5 for Windows (SPSS Inc., Chicago, IL, USA). To compare haemodynamic measurements, one-way analysis of variance (anova) and anova for repeated measurements were applied (ranova). P-values < 0.05 were considered statistically significant. The study power was analysed using the PS 3.0.2 power and sample size calculations program (Vanderbilt Biostatistics, Nashville, TN, USA). A change in SVRI > 400 dyn × s cm⁻⁵× m² in the supine position was chosen as the main outcome variable (such a change was observed after both salbutamol and nitroglycerin). When using the variables of the salbutamol measurements in the power calculations, the present number of subjects (n= 9) was found to have a power of 92% to detect a significant difference (δ= 400 dyn × s cm⁻⁵× m²) between the salbutamol and the placebo effect (α-level 0.05).

Results

Study population

The basic characteristics and laboratory values of the study population were all within the normal range. The mean age was 32 ± 2.5 years, body mass index 24.5 ± 1.0 kg m⁻², and waist circumference 80 ± 5 cm in women and 91 ± 3 cm in men. None of the study subjects had a medical history of cardiovascular disease or elevated BP, and none of the subjects was a present smoker, while two subjects had a previous smoking history. Plasma lipid profile, fasting glucose, electrolytes, and kidney function were all within the normal range. None of the study subjects reported any adverse effects related to the research drug administration.

Haemodynamic effects of L-arginine, salbutamol and nitroglycerin in supine position

The average values of the first 5 min of the haemodynamic recordings following drug administration in the supine position are shown in Tables 1–3 for L-arginine, salbutamol and nitroglycerin, respectively. The average results of the second and fifth minute during each recording phase (5 min supine–5 min tilt–5 min supine) are depicted in Figures 1 and 2.

Table 1.

Average haemodynamic effects of L-arginine (10 mg kg⁻¹ min⁻¹) and saline infusion during the first 5 min in the supine position and 5 min of head-up tilt (mean ± SEM)

	Saline infusion		L-arginine infusion
Tonometry	Supine	Head-up tilt	Supine	Head-up tilt
Radial SBP (mmHg)	131 ± 3	127 ± 2	127 ± 3	121 ± 2*
Aortic SBP (mmHg)	116 ± 3	111 ± 2†	110 ± 3	105 ± 2*
Radial DBP (mmHg)	76 ± 2	79 ± 1	73 ± 3	74 ± 2*
Aortic DBP (mmHg)	77 ± 2	80 ± 2	74 ± 3	74 ± 2*
Augmentation index (%)	11.8 ± 3.3	4.2 ± 2.8†	9.5 ± 3.2	3.2 ± 2.6†
Aortic reflection time (ms)	171 ± 6	162 ± 4†	169 ± 8	170 ± 4*
Ejection duration (ms)	342 ± 4	277 ± 8†	347 ± 5	290 ± 9†
Impedance cardiography
Stroke volume (ml)	100 ± 9	68 ± 5†	98 ± 8	71 ± 5†
Heart rate (beats min−1)	58 ± 3	68 ± 3†	55 ± 2	65 ± 2†
Cardiac output (l min−1)	5.71 ± 0.44	4.58 ± 0.27†	5.30 ± 0.39	4.55 ± 0.24†
Systemic vascular resistance index (dyn × s cm−5× m2)	2385 ± 191	3052 ± 174†	2521 ± 162	2920 ± 166†
Pulse wave velocity (m s−1)	8.66 ± 0.17	–	8.63 ± 0.16	–

^*P < 0.05 compared with corresponding value during saline infusion.

^†P < 0.05 compared with corresponding supine values.

SBP, systolic blood pressure; DBP, diastolic blood pressure.

Table 3.

Average haemodynamic effects of 0.25 mg nitroglycerin resoriblet during the first 5 min in the supine position and 5 min of head-up tilt (mean ± SEM)

	Placebo resoriblet		Nitroglycerin resoriblet
Tonometry	Supine	Head-up tilt	Supine	Head-up tilt
Radial SBP (mmHg)	122 ± 3	121 ± 2	124 ± 1	116 ± 3†
Aortic SBP (mmHg)	106 ± 3	105 ± 2	107 ± 2	99 ± 2†
Radial DBP (mmHg)	68 ± 3	75 ± 2†	70 ± 2	69 ± 3
Aortic DBP (mmHg)	69 ± 3	76 ± 2†	71 ± 2	71 ± 2
Augmentation index (%)	8.1 ± 3.9	−0.9 ± 5.3†	3.0 ± 3.5	−15.8 ± 1.6*†
Aortic reflection time (ms)	178 ± 10	156 ± 6	173 ± 7	175 ± 7*
Ejection duration (ms)	344 ± 5	269 ± 9†	330 ± 6*	247 ± 10†
Impedance cardiography
Stroke volume (ml)	90 ± 6	66 ± 5†	94 ± 7	76 ± 6†
Heart rate (beats min−1)	55 ± 2	70 ± 3†	62 ± 3*	85 ± 3*†
Cardiac output (l min−1)	4.89 ± 0.26	4.64 ± 0.47	5.74 ±0.36*	6.34 ± 0.48*
Systemic vascular resistance index (dyn × s cm−5× m2)	2438 ± 148	2663 ± 155	2089 ± 119	1955 ± 189*
Pulse wave velocity (m s−1)	8.47 ± 0.12	–	7.89 ± 0.28	–

^*P < 0.05 compared with corresponding value during placebo resoriblet.

^†P < 0.05 compared with corresponding supine values.

SBP, systolic blood pressure; DBP, diastolic blood pressure.

Figure 1.

Mean values for aortic mean arterial pressure (a–c), cardiac output (d–f) and systemic vascular resistance index (SVRI) (g–i) after L-arginine infusion (10 mg kg⁻¹ min⁻¹), saline infusion, salbutamol inhalation (400 µg), placebo inhalation, nitroglycerin resoriblet (0.25 mg) and placebo resoriblet. Research drug was administrated at measurement time 0 and head-up tilt was performed from 5 to 10 min. *P < 0.05 vs. saline/placebo, anova

Figure 2.

Mean values for heart rate (a–c), augmentation index (d–f) and aortic reflection time (g–i) after L-arginine infusion (10 mg kg⁻¹ min⁻¹), saline infusion, salbutamol inhalation (400 µg), placebo inhalation, nitroglycerin resoriblet (0.25 mg) and placebo resoriblet. Research drug was administrated at measurement time 0 and head-up tilt was performed from 5 to 10 min. *P < 0.05 vs. saline/placebo, anova

L-arginine infusion did not have any significant haemodynamic effects in the supine position before the head-up tilt (Table 1, Figures 1 and 2). In contrast, before the head-up tilt salbutamol induced a 9.2 ± 2.6% decrease in SVRI (Table 2) and during the fifth minute an 8.6 ± 2.5% increase in heart rate (Figure 2) (P < 0.05 for both compared with placebo inhalation). During the fifth minute following nitroglycerin administration, supine cardiac output was increased (+9.2 ± 2.7%) and SVRI was reduced (−6.7 ± 1.8%) (P < 0.05 for both compared with placebo resoriblet) (Figure 1).

Table 2.

Average haemodynamic effects of 400 µg inhaled salbutamol during the first 5 min in the supine position and 5 min of head-up tilt (mean ± SEM)

	Placebo inhalation		Salbutamol inhalation
Tonometry	Supine	Head-up tilt	Supine	Head-up tilt
Radial SBP (mmHg)	126 ± 3	129 ± 2	123 ± 2	124 ± 2
Aortic SBP (mmHg)	111 ± 3	112 ± 2	105 ± 2	106 ± 2*
Radial DBP (mmHg)	73 ± 2	80 ± 2†	70 ±2	75 ± 2†
Aortic DBP (mmHg)	74 ± 2	82 ± 2†	70 ± 2	75 ± 2†
Augmentation index (%)	12.7 ± 3.7	−0.4 ± 4.4†	6.0 ± 2.7	−1.9 ± 2.5†
Aortic reflection time (ms)	172 ± 10	158 ± 5	168 ± 8	158 ± 4†
Ejection duration (ms)	342 ± 7	270 ± 7†	343 ± 5	272 ± 9†
Impedance cardiography
Stroke volume (ml)	91 ± 6	66 ± 5†	90 ± 5	71 ± 4†
Heart rate (beats min−1)	54 ± 2	71 ± 3†	61 ± 2*	71 ± 2†
Cardiac output (l min−1)	4.78 ± 0.18	4.65 ± 0.38	5.44 ± 0.26	5.08 ± 0.34
Systemic vascular resistance index (dyn × s cm−5× m2)	2449 ± 61	2769 ± 176	2148 ± 96*	2448 ± 139†
Pulse wave velocity (m s−1)	8.40 ± 0.29	–	8.54 ± 0.18	–

^*P < 0.05 compared with corresponding value during placebo inhalation.

^†P < 0.05 compared with corresponding supine values.

SBP, systolic blood pressure; DBP, diastolic blood pressure.

During the second minute in the supine position after the head-up tilt aortic mean BP was lower during L-arginine infusion than during saline infusion, but L-arginine infusion did not have any other effects on BP values or supine haemodynamics (Table 1, Figure 1). In the supine position after the head-up tilt AIx was significantly lower after salbutamol (AIx value 5.5 ± 2.5%) and nitroglycerin (AIx −1.4 ± 2.5%) when compared with placebo inhalation (AIx 14.1 ± 3.2%), placebo resoriblet (AIx 12.1 ± 3.7%) and with L-arginine infusion (AIx 11.4 ± 2.6%) (P < 0.05 for all comparisons) (Figure 2). SVRI in the supine position after the second head-up tilt was reduced after salbutamol when compared with placebo inhalation (5-min averages 2105 ± 89 and 2546 ± 73 dyn × s cm⁻⁵× m², respectively, P= 0.026) (Figure 1).

Supine PWV values were not significantly different after L-arginine infusion, salbutamol inhalation or nitroglycerin administration when compared with the respective controls (Tables 1–3). However, nitroglycerin resoriblet induced a small but significant reduction in PWV (−5.7 ± 2.4%) when compared with placebo resoriblet (0.9 ± 1.7%) (P < 0.05).

Haemodynamic effects of L-arginine, salbutamol and nitroglycerin during the head-up tilt

The average values of the haemodynamic recordings during the 5-min head-up are shown in Tables 1–3, and the average results of the second and fifth minute in Figures 1 and 2.

During the head-up tilt aortic and radial BP were reduced with L-arginine when compared with saline infusion (Table 1). Aortic mean BP was also lower during L-arginine infusion (85 ± 2 mmHg) than during saline infusion (90 ± 1 mmHg, P= 0.02). During the head-up, aortic reflection time was numerically but not statistically significantly longer during L-arginine than saline infusion (P= 0.162, Table 1, Figure 2), and the respective changes in aortic reflection time were not significantly different either (8.0 ± 8.8% vs.−6.8 ± 2.4%, respectively, P = 0.147). However, in additional comparisons aortic reflection time was longer during the head-up tilt with L-arginine infusion (170 ± 4 ms) than during placebo inhalation (153 ± 4 ms, P= 0.02), placebo resoriblet (155 ± 6 ms, P= 0.045) and salbutamol inhalation (158 ± 4 ms, P= 0.047, Table 3). These results suggest that the reduction in aortic reflection time during head-up tilt was prevented during L-arginine infusion.

During the head-up tilt aortic systolic BP after salbutamol inhalation was lower than after placebo inhalation (Table 2, P= 0.024), while there was no significant difference in radial systolic, aortic and radial diastolic, or mean BP (Table 2). Although SVRI was numerically lower during the head-up tilt with salbutamol than with placebo inhalation, the difference was not significant (P= 0.196) (Table 2, Figure 1).

During the head-up tilt, aortic and radial diastolic and systolic BP were not reduced with nitroglycerin when compared with placebo resoriblet (Table 3), but during the fifth minute of the head-up tilt aortic mean BP was lower with nitroglycerin than with placebo resoriblet (80 ± 2 vs. 91 ± 2 mmHg, P= 0.002, Figure 1). However, during the head-up tilt the other effects of 0.25 mg sublingual nitroglycerin on haemodynamics were very clear (Table 3, Figures 1 and 2): cardiac output was higher in comparison with placebo resoriblet, and also higher than during L-arginine infusion and placebo inhalation (P < 0.05 for all), SVRI was lower when compared with placebo resoriblet, placebo inhalation and L-arginine infusion (P < 0.05), heart rate was higher and AIx was lower in comparison with all other measurements (P < 0.05), and aortic reflection time was longer than with placebo resoriblet (P= 0.023).

Exhaled nitric oxide concentrations after salbutamol

Alveolar NO concentration in six of the study subjects was 1.55 ± 0.17 parts per billion before salbutamol inhalation, while the concentration was 1.81 ± 0.19 parts per billion before placebo inhalation. After salbutamol inhalation alveolar NO concentration increased by 19%, whereas after placebo inhalation the concentration decreased by 10% (P= 0.01).

Discussion

Here we evaluated the use of two pharmacological compounds acting on the endothelium and the endothelium-independent agent nitroglycerin on non-invasive haemodynamics in healthy volunteers. The study has shown that inhaled salbutamol decreased systemic vascular resistance and AIx and increased heart rate and cardiac output, whereas L-arginine only resulted in a decrease of BP during the head-up tilt. In contrast, nitroglycerin markedly decreased systemic vascular resistance, AIx and BP, and increased heart rate, cardiac output and aortic reflection time.

Although the observed haemodynamic changes with L-arginine and salbutamol were moderate, the combination of PWA and impedance cardiography provided data that would have remained uncovered when using only one of these methods. The present non-invasive measurement protocol provides continuous haemodynamic information about central wave reflection and BP, arterial compliance, systemic vascular resistance and cardiac function in both the supine and upright positions. Thus, beat-to-beat changes in haemodynamics can be revealed, which provides benefits when compared with single tonometric measurements using a pen-like sensor, the approach of which has been applied in the majority of recent PWA studies [14, 36]. The present approach also enables more thorough assessment of vascular responsiveness than methods like FMD in the upper arm arteries, which examine only a section of the cardiovascular system. It is important to notice that the effects of endothelial stimulation on arterial tone may depend on the vascular bed studied [37, 38].

The semi-essential amino acid L-arginine serves as a precursor for endothelial nitric oxide synthase (NOS). L-arginine administration has been reported to induce vasodilation in healthy subjects and patients with cardiovascular disease [17–19]. L-arginine has rarely caused side-effects at doses <30 g during a 30-min infusion (i.e. at a maximum rate of 1 g min⁻¹) [26], which favours its use in clinical research. However, the beneficial effects of L-arginine do not seem to be due to extracellular substrate supply for endothelial NOS, as the enzyme should be saturated with physiological L-arginine levels [39]. In contrast, excess substrate availability can overcome the inhibition of endothelial NOS activity due to accumulation of false L-arginine derivatives during cardiovascular disease states, the mechanism of which can explain the vasodilatory effects of L-arginine. Furthermore, L-arginine has been reported to improve vasodilation in healthy humans with low concentrations of endothelial NOS-inhibitory substances [40, 41], and in this case other endocrine mechanisms, i.e. stimulation of growth hormone and insulin secretion, may play a role [42, 43].

In the present study the haemodynamic effects of L-arginine infusion were modest. The BP-lowering influence was seen only during the head-up tilt, during which the normal decrease in aortic reflection time was abolished by L-arginine. As the compliance of the arteries increases, the aortic reflection time lengthens and the reflected wave shifts towards diastole, resulting in decreased systolic BP. Since L-arginine did not decrease systemic vascular resistance, the BP effect probably resulted from increased compliance in large arteries. In spite of the reduced BP and prolonged aortic reflection time induced by L-arginine, AIx decreased correspondingly during the head-up tilt in response to L-arginine, saline infusion, salbutamol and placebo. The reduction in AIx in the upright position cannot be attributed to changes in large arterial compliance, but may result from a more pronounced decrease in the augmentation pressure than pulse pressure. Thus, AIx is not always a reliable indicator of arterial compliance but merely an indicator of central wave reflection.

Although β₂-adrenoceptors mediate vasorelaxation at the level of vascular smooth muscle, the stimulation of β₂-adrenoceptors is known to increase endothelial release of NO and cause largely endothelium-mediated vascular relaxation [44, 45]. Therefore, the effect of the β₂-adrenoceptor agonist salbutamol on the PWA-derived measure of wave reflection and arterial stiffness, the AIx, has been applied as a method to evaluate the influence of endothelial stimulation in the whole arterial tree [13, 46]. Inhaled salbutamol has induced an 8–12% decrease in the AIx in healthy subjects, but in many studies the effects on other haemodynamic variables have not been determined [13, 14, 46]. As a standard methodological approach, the AIx has been derived from 10 consecutive heart beats every 5 min for up to 20 min, which provides a relatively narrow window of observation when compared with continuous PWA recording. In the present study, the inhalation of 400 µg salbutamol reduced systemic vascular resistance and BP and increased heart rate already before the decrease in AIx was observed. These results favour the use of continuous recording of haemodynamics in the assessment of vascular responsiveness, and indicate that additional methods besides PWA increase the reliability of the analysis.

Administration of sublingual nitroglycerin was included in the study as an endothelium-independent vasodilator, and even a small 0.25-mg dose of nitroglycerin resulted in major haemodynamic changes in comparison with salbutamol and L-arginine. Traditionally, venodilation and venous pooling of blood into the lower extremities and splanchnic vasculature, and subsequently reduced left ventricular preload, have been regarded as major haemodynamic effects of nitroglycerin in humans, whereas arterial dilation is thought to occur to a lesser extent [47, 48]. However, in the present study nitroglycerin induced a clear decrease in systemic vascular resistance and an increase in cardiac output and aortic reflection time, strongly suggesting reduced arterial resistance.

In the present study, alveolar NO concentration was increased by 19% after salbutamol inhalation. Alveolar NO concentration could be increased via three mechanisms: increased NO production in the alveolar epithelium, decreased NO diffusion from alveolar air to capillaries, or increased NO release from the capillary walls. Previously, increased alveolar NO concentration has been found in diseases affecting pulmonary parenchyma, probably due to increased inducible NOS expression in the alveolar epithelium [34, 49]. Increased alveolar NO concentration has also been reported in subjects with liver cirrhosis [50], presumably as a result of increased endothelial NO production [51]. In addition, the administration of the ACE-inhibitor enalapril has increased NO level in exhaled air, which was thought to reflect enhanced NO release from the capillary endothelium [52]. Since inhaled salbutamol is not likely to influence NO production in the alveolar epithelium or change NO permeability of the alveolar wall, the present findings suggest that salbutamol inhalation increased NO release from the capillary endothelium. This supports the view that salbutamol stimulates the endothelial cells in the arterial tree.

The purpose of the present study was also to evaluate the use of three pharmacological tools in the haemodynamic screening of humans, and not directly to compare the effects of L-arginine, salbutamol and nitroglycerin. Therefore, a double-blind placebo infusion was not included in the study protocol. The crucial role of the endothelium in different cardiovascular disorders, and as a target of pharmacological interventions, underlies the importance of assessing endothelial function in vivo. The more pronounced cardiovascular effects of salbutamol than L-arginine, together with low-cost and easy administration via inhalation, imply that salbutamol provides a clinically more applicable tool for the assessment of the role of endothelium in haemodynamic responsiveness. Finally, the present study underscores the need for comprehensive methods in the measurement of haemodynamics in humans.

Competing interests

None to declare.