Acute inorganic nitrate supplementation and the hypoxic ventilatory response in patients with obstructive sleep apnea

Abstract

Patients with obstructive sleep apnea (OSA) have increased cardiovascular disease risk largely attributable to hypertension. Heightened peripheral chemoreflex sensitivity (i.e., exaggerated responsiveness to hypoxia) facilitates hypertension in these patients. Nitric oxide blunts the peripheral chemoreflex, and patients with OSA have reduced nitric oxide bioavailability. We therefore investigated the dose-dependent effects of acute inorganic nitrate supplementation (beetroot juice), an exogenous nitric oxide source, on blood pressure and cardiopulmonary responses to hypoxia in patients with OSA using a randomized, double-blind, placebo-controlled crossover design. Fourteen patients with OSA (53 ± 10 yr, 29.2 ± 5.8 kg/m², apnea–hypopnea index = 17.8 ± 8.1, 43%F) completed three visits. Resting brachial blood pressure and cardiopulmonary responses to inspiratory hypoxia were measured before, and 2 h after, acute inorganic nitrate supplementation [∼0.10 mmol (placebo), 4.03 mmol (low dose), and 8.06 mmol (high dose)]. Placebo increased neither plasma [nitrate] (30 ± 52 to 52 ± 23 μM, P = 0.26) nor [nitrite] (266 ± 153 to 277 ± 164 nM, P = 0.21); however, both increased following low (29 ± 17 to 175 ± 42 μM, 220 ± 137 to 514 ± 352 nM) and high doses (26 ± 11 to 292 ± 90 μM, 248 ± 155 to 738 ± 427 nM, respectively, P < 0.01 for all). Following placebo, systolic blood pressure increased (120 ± 9 to 128 ± 10 mmHg, P < 0.05), whereas no changes were observed following low (121 ± 11 to 123 ± 8 mmHg, P = 0.19) or high doses (124 ± 13 to 124 ± 9 mmHg, P = 0.96). The peak ventilatory response to hypoxia increased following placebo (3.1 ± 1.2 to 4.4 ± 2.6 L/min, P < 0.01) but not low (4.4 ± 2.4 to 5.4 ± 3.4 L/min, P = 0.11) or high doses (4.3 ± 2.3 to 4.8 ± 2.7 L/min, P = 0.42). Inorganic nitrate did not change the heart rate responses to hypoxia (beverage-by-time P = 0.64). Acute inorganic nitrate supplementation appears to blunt an early-morning rise in systolic blood pressure potentially through suppression of peripheral chemoreflex sensitivity in patients with OSA.

NEW & NOTEWORTHY The present study is the first to examine the acute effects of inorganic nitrate supplementation on resting blood pressure and cardiopulmonary responses to hypoxia (e.g., peripheral chemoreflex sensitivity) in patients with obstructive sleep apnea (OSA). Our data indicate inorganic nitrate supplementation attenuates an early-morning rise in systolic blood pressure potentially attributable to blunted peripheral chemoreflex sensitivity. These data show proof-of-concept that inorganic nitrate supplementation could reduce the risk of cardiovascular disease in patients with OSA.

INTRODUCTION

Since the 1990s, the prevalence of obstructive sleep apnea (OSA) has consistently increased (1) and may greatly exceed previous estimations (2). Accumulation of body mass is strongly associated with OSA (3), which, when considered with increased rates of obesity (4), indicates that the prevalence will continue to rise. Patients with OSA experience cyclical complete (apnea) or partial (hypopnea) upper airway collapse, concomitant hypoxemia, and hypercapnia, followed by rapid arousal during sleep (5). This facilitates various pathologies [e.g., atrial fibrillation (6) stroke (7)] and largely explains the significant prevalence in patients with heart disease (8) as well as heart failure (9).

Evidence suggests autonomic imbalance contributes to many of the aforementioned conditions associated with OSA (10, 11). More specifically, patients with OSA have elevated sympathetic activity during waking hours (12) with attenuated vagal tone (11). Seminal data from Narkiewicz et al. (13) report that exaggerated peripheral chemoreflex sensitivity greatly contributes to these autonomic imbalances and high rates of hypertension in patients with OSA. Principally stimulated during hypoxemia, the peripheral chemoreflex is a complex and integrative response responsible for increasing ventilation as well as sympathetic outflow (14). Of the numerous signaling molecules regulating the peripheral chemoreflex, nitric oxide (NO) plays a multifactorial role (15). Outlined by Campanucci et al. (15), NO suppresses peripheral chemoreflex activity by directly hyperpolarizing glomus cells (primary oxygen sensors), efferent inhibition via the glossopharyngeal nerve, or by promoting intercellular adenosine triphosphate signaling. Regardless of the specific mechanism, patients with OSA have low endogenous NO production (16, 17) potentially explaining their potentiated responses to hypoxia (15, 18).

Dietary consumption of inorganic nitrate serves as an exogenous source of NO through a pathway independent of the classical synthase enzyme family. More directly, consumption of nitrate-rich foods (e.g., green leafy vegetables) increases plasma [nitrate] and [nitrite]. Previously thought to be inert by-products of NO metabolism, these molecules can be reduced to NO through the enterosalivary circulation, deoxyhemoglobin, and xanthine oxidoreductase (19). Acute supplementation of beetroot juice, a common vehicle for inorganic nitrate, increases circulating NO bioavailability and reduces blood pressure (20) as well as sympathetic nerve activity (21). Our recent data (22) suggest regularly consuming beetroot juice blunts the ventilatory response to hypoxia (e.g., peripheral chemoreflex sensitivity) in healthy older adults. Taken together, these works suggest that consumption of inorganic nitrate could provide multifaceted benefits to patients with OSA. Thus, the present study investigated whether acute inorganic nitrate supplementation, via beetroot juice, reduces resting blood pressure in patients with OSA and whether this could be attributable to blunted peripheral chemoreflex sensitivity.

METHODS

Inclusion criteria were diagnosis of mild-to-moderate OSA [apnea–hypopnea index (AHI) 5–30] based on in-home sleep tests (WatchPAT 200, Itamar Medical Inc., Atlanta, GA) interpreted by a licensed sleep specialist (M.E.D.). Exclusion criteria were heart disease, smoking (≤6 mo since cessation), diabetes mellitus, autonomic disorders, AHI (≥30 indicating severe OSA), central sleep apnea, kidney disease, or a body mass index ≥40 kg/m². Potential subjects using CPAP (≤1 yr) or other treatments for OSA (e.g., oral appliances) were excluded. All women were postmenopausal and not taking hormone therapy. Subjects completed four visits: One to provide written informed consent (before data collection) followed by three identical experimental visits. For experimental visits, subjects arrived following an overnight (∼10 h) fast while refraining from exercising or consuming alcohol and caffeine for 24 h. Subjects also withheld from taking prescription medications on the morning of all experimental study visits, to minimize the acute effects on study measurements. Additionally, subjects followed a low-nitrate diet (i.e., minimizing green leafy vegetables, well water, and processed meats) for 48 h preceding study visits. Experimental protocols were approved by the Institutional Review Board at the University of Iowa and are part of a clinical trial (clinicaltrials.gov NCT: 03930563).

Experimental Protocol

Experimental visits began with venous catheter placement and blood sampling followed by brachial blood pressure measurement. Subjects were then outfitted with a facemask, pneumotachometer, pulse oximeter, and electrodes and connected to a self-regulated partial rebreathe circuit breathing compressed normoxic gas. Following a 2-min baseline, subjects were exposed to inspiratory hypoxia for 3–5 min. Subjects then consumed one of three beetroot juice beverages in a randomized, double-blind, crossover manner and waited 2 h before repeating the protocol, concluding the experimental visit. All study visits were separated by a minimum of 72 h to ensure adequate washout of the supplement (23, 24).

Brachial Blood Pressure Measurements

Subjects voided their bladder before resting supine for 15 min in a dimly lit, temperature-controlled, quiet room. Blood pressure measurements were made on subjects’ left arm at heart level, in duplicate, using an automated cuff (23- to 33-cm adult cuff, Cardiocap/5, Datex-Ohmeda, Louisville, CO) during pre- and postsupplementation baselines. If either systolic or diastolic values differed by ±5 mmHg, additional measurements were taken until this criterion was met. All measurements were separated by ∼120 s, and reported data reflect the average of two measurements within 5 mmHg of one another.

Cardiopulmonary Measurements

Heart rate and pulse oxygen saturation (%SpO₂) were continuously recorded via three-lead electrocardiogram and oximetry (Cardiocap/5), respectively. Tidal volume (Vt) and respiratory rate (RR), measured via pneumotachography (Linear Pneumotach Flow Sensors and PA-1 Pnemotach Amplifiers, Hans Rudolph Inc., Shawnee Mission, KS), were multiplied to calculate minute ventilation (V_E). Peak end-expiratory carbon dioxide (P_EECO₂) was measured via capnography (Cardiocap/5).

Acute Inspiratory Hypoxia

Subjects began trials by breathing from a normoxic gas cylinder (20.9%FiO₂, 0.03%FiCO₂, balance nitrogen) for 2 min. Then, subjects’ SpO₂ was titrated to ≤80% using an anesthesia gas blender mixing the normoxic and hypoxic cylinders (10.0%FiO₂, 0.03%FiCO₂, balance nitrogen), over 3–5 min (variability due to interindividual SpO₂ responsiveness). Peripheral chemoreflex sensitivity was quantified as the absolute changes (Δ) from baseline to the final 30 (heart rate) and 60 s (V_E) of hypoxia. Once ≤80% SpO₂ was achieved, subjects were disconnected from the rebreathe circuit leaving the facemask intact, whereas heart rate and oxygen saturation were monitored until returning to baseline. During analysis, we observed nearly all peak ventilatory responses occur before the final minute of hypoxia. Therefore, we also calculated the absolute change in V_E from baseline to subjects’ individual peak response similar to previous works (22, 25–27).

Inorganic Nitrate Supplementation

Beetroot powder (Superbeets, HumanN, Inc., Austin, TX) dissolved in ∼240-mL bottled water served as vehicles for inorganic nitrate supplementation as follows: placebo [BR_P; 5–10 mg nitrate (0.08–0.16 mmol), no detectable nitrite], low dose [BR_L; 250 mg nitrate (4.03 mmol), 20 mg nitrite (0.29 mmol)), and higher dose [BR_H; 500 mg nitrate (8.06 mmol), 40 mg nitrite (0.58 mmol)). The amount of nitrate and nitrite in each beverage was confirmed by the manufacturer before data collection. Subjects consumed each beverage, one per experimental visit, in random order, using a double-blinded approach.

Plasma Analysis

Venous blood samples were obtained from an intravenous catheter before, and 2 h after, beetroot juice consumption for the determination of plasma [nitrate] and [nitrite] as previously described (22). Briefly, blood was collected in lithium heparin tubes, centrifuged, aliquoted into Eppendorf Tubes, and frozen at −80°C for later analysis. Quantification of plasma [nitrate] and [nitrite] was performed using a Sievers chemiluminescence NO analyzer (NOA 280i, Sievers Instruments, Boulder, CO) in accordance with previously established recommendations (28).

Data Analysis

All data, with the exception of resting brachial blood pressure, were sampled at 250 Hz and analyzed offline using signal processing software (WinDaq, DATAQ Instruments, Akron, OH) and Microsoft Excel (Excel2016, Microsoft, Redmond, WA). Presupplementation data were compared across study visits to identify potential between-day changes using a one-way repeated-measures analysis of variance (RMANOVA). Outcome variables were analyzed across time (pre- vs. postsupplementation) and between supplements (BR_P, BR_L, and BR_H) using a two-way RMANOVA. Main effects were interpreted using Tukey’s post hoc test if significant F ratios were detected. Covariate analyses were conducted to determine whether sex and/or disease severity (mild (AHI 5–14) or moderate (AHI 15–29) OSA) (29) influenced findings from original analyses. All statistical analyses were deemed significant a priori at α < 0.05 and were completed using SigmaPlot version 11.0 (Systat Software Inc., San Jose, CA) or SPSS (version 25, IBM, Armonk, NY).

RESULTS

Subjects

Table 1 contains demographic data representing 14 of the 18 subjects who completed all study visits. Two subjects did not complete any experimental visits due to complications with scheduling. Another subject became claustrophobic when outfitted with the mask and voluntarily dropped out of the study. A fourth subject was unable to complete data collection as research activities were ceased due to COVID-19. One of the remaining 14 subjects did not adequately desaturate on two of their experimental visits; therefore data, with the exception of resting brachial blood pressure, reflect 13 subjects (6 F).

Table 1.

Subject demographics

	Study Cohort (n = 14)
AHI, events/h	17.8 ± 8.1
Nadir SpO2, %	86 ± 3
Age, yr	53 ± 10
Females, n (%)	6 (43)
BMI, kg/m2	29.2 ± 5.8
Coconditions, n (%)
Hypercholesterolemia	4 (29)
Hypertension	1 (7)
Prescriptions, n (%)
Statin	4 (29)
β-blocker	2 (14)
Losartan	1 (7)
PPI	2 (14)

Data are presented as means ± SD or n (%) where noted. AHI, apnea–hypopnea index; nadir SpO₂, lowest saturation recorded during diagnostic sleep study; PPI, proton pump inhibitor.

Plasma [Nitrate] and [Nitrite]

Figure 1 shows plasma [nitrate] (A) and [nitrite] (B) at baseline as well as 2 h post supplementation. Baseline [nitrate] and [nitrite] were similar across all experimental visits (one-way RMANOVA, P = 0.14 and 0.11, respectively). A two-way RMANOVA revealed a significant beverage-by-time interaction in concert with significant main effects for beverage and time for both plasma [nitrate] and [nitrite] (P < 0.05 for all). Post hoc analyses found plasma [nitrate] and [nitrite] increased following both BR_L and BR_H relative to their respective presupplementation values as well as postsupplementation BR_P concentrations (P < 0.01 for all).

Fig. 1.

Plasma [nitrate] (A) and [nitrite] (B) before (darker bars) and after (lighter bars) supplementation of beetroot juice containing a negligible amount of nitrate and no nitrite (BR_P), beetroot juice containing 4.03 mmol nitrate and 0.29 mmol nitrite (BR_L), and beetroot juice containing 8.06 mmol nitrate and 0.58 mmol nitrite (BR_H). Group data (n = 14) were compared using a two-way repeated-measures analysis of variance and are shown as means ± SE, whereas individual responses are shown with filled circles. *P < 0.05 pre- vs. postsupplementation within beverage; †P < 0.05 vs. post-BR_P; ‡P < 0.05 vs. post-BR_L.

Resting Brachial Blood Pressure

Resting, presupplementation systolic and diastolic brachial pressures were similar between visits (one-way RMANOVA P = 0.90 and 0.95, Figs. 2, A and B, respectively). A significant beverage-by-time interaction was observed as was a significant main effect of time for systolic blood pressure (P < 0.05 for all). Post hoc analyses found resting systolic blood pressure was elevated 8 ± 7 mmHg relatively to presupplementation measurements for BR_P (P < 0.01), with no change observed for BR_L (3 ± 8 mmHg, P = 0.19) and BR_H (0 ± 8 mmHg, P = 0.96). Neither a significant beverage-by-time interaction nor main effects for beverage or time were observed for diastolic blood pressure (P = 0.15, 0.48, and 0.07, respectively). Covariate analyses revealed neither sex nor disease severity contributed to changes in resting systolic (P = 0.38 and 0.34, respectively) or diastolic (P = 0.85 for both) blood pressure during the study.

Fig. 2.

Resting systolic (A) and diastolic (B) blood pressure before (darker bars) and after (lighter bars) supplementation of beetroot juice containing a negligible amount of nitrate and no nitrite (BR_P), beetroot juice containing 4.03 mmol nitrate and 0.29 mmol nitrite (BR_L), and beetroot juice containing 8.06 mmol nitrate and 0.58 mmol nitrite (BR_H). Group data (n = 14) were compared using a two-way repeated-measures analysis of variance and are shown as means ± SE, whereas individual responses are shown with filled circles. *P < 0.05 pre- vs. postsupplementation within beverage.

Cardiopulmonary Responses to Hypoxia

Vt, RR, SpO₂, and P_EECO₂ during the final minute of hypoxia as well as at the peak ventilatory response are shown in Table 2. In response to subjects’ first bout (presupplementation) of hypoxia, V_E during the final 60 s (Δ1.9–3.0 L/min, one-way RMANOVA P = 0.36) and the peak ventilatory response (Δ3.6–4.5 L/min, one-way RMANOVA P = 0.35) were similar across experimental visits (Fig. 3). A two-way RMANOVA identified nonsignificant P values for a beverage-by-time interaction, as well as main effects for beverage and time (P = 0.16, 0.09, and 0.10, respectively) for the V_E response (Δ from rest) to the final minute of hypoxia (Fig. 3A). Although a significant beverage-by-time interaction was not observed for ΔV_E from baseline to peak responsiveness (P = 0.49), there were significant main effects for beverage and time (P = 0.04 for both) with post hoc analyses revealing an increase following BR_P (P < 0.05 vs. presupplementation, Fig. 3B). The increase (Δ) in heart rate to subjects’ first bout of hypoxia was similar across experimental visits (Δ11–13 beats/min, one-way RMANOVA P = 0.71). Similarly, a two-way RMANOVA reported no significant beverage-by-time interaction (P = 0.74) along with nonsignificant main effects for beverage and time (P = 0.41 and 0.08, respectively). Neither sex nor disease severity (mild or moderate OSA) appeared to influence findings from either ventilatory analysis (P = 0.48 and 0.81, respectively) or the heart rate responses (P = 0.84 for both analyses) as determined by analyses of covariance. Importantly, the reduction (Δ from baseline) in %SpO₂ to the final minute of hypoxia, as well as to peak ventilatory responses, was similar across beverages and time (Table 2). The duration subjects were administered hypoxic gas, calculated as total time or time to peak response, was similar across all study visits (beverage-by-time interactions P = 0.38 and 0.26), main effects of beverage (P = 0.13 and 0.10) and time (P = 0.45 and 0.30). Likewise, P_EECO₂ during the final minute as well as at the peak ventilatory response was comparable across beverages and time (Table 2).

Table 2.

Changes in respiratory variables during hypoxia

		BRP		BRL		BRH		P Values
		Pre	Post	Pre	Post	Pre	Post	BxT	B	T
Vt,mL	Hy	225 ± 195	261 ± 180	306 ± 195	327 ± 209	321 ± 240	349 ± 269	0.69	0.09	0.08
	Pk	308 ± 224	358 ± 245	407 ± 224	447 ± 240	341 ± 245	367 ± 269	0.60	0.20	0.06
RR,f	Hy	−1 ± 2	1 ± 2	−1 ± 2	1 ± 2	0 ± 2	1 ± 2	0.21	0.36	0.85
	Pk	0 ± 2	0 ± 1	−1 ± 2	1 ± 2	−1 ± 2	1 ± 2	0.89	0.58	0.45
SpO2,%	Hy	−17 ± 3	−18 ± 2	−17 ± 2	−18 ± 2	−17 ± 2	−18 ± 2	0.32	0.67	<0.01
	Pk	−15 ± 5	−14 ± 6	−14 ± 2	−15 ± 5	−16 ± 3	−15 ± 5	0.79	0.62	0.55
PEECO2,mmHg	Hy	−2 ± 2	−2 ± 2	−2 ± 1	−2 ± 2	−2 ± 2	−2 ± 2	0.64	0.79	0.37
	Pk	−2 ± 1	−1 ± 2	−2 ± 2	−2 ± 2	−2 ± 2	−2 ± 2	0.35	0.98	0.21

Data represent the change (Δ from baseline) to the final minute of hypoxia (Hy) or peak ventilatory response (Pk), are presented as means ± SD, and were compared using a two-way repeated-measures analysis of variance. B, beverage (BR_P, BR_L, BR_H); BR_H, beetroot juice containing 8.06 mmol nitrate and 0.58 mmol nitrite; BR_L, beetroot juice containing 4.03 mmol nitrate and 0.29 mmol nitrite; BR_P, beetroot juice containing a negligible amount of nitrate and no nitrite; P_EECO₂, peak end-expiratory carbon dioxide; pre, prior to beverage consumption; post, following beverage consumption; RR, respiratory rate; SpO₂, arterial oxygen saturation measured via pulse oximetry; T, time (pre and post); Vt, tidal volume.

Fig. 3.

Increase (Δ from baseline) in minute ventilation to the final minute of hypoxia (A) and peak ventilatory response (B) to hypoxia before (darker bars) and after (lighter bars) supplementation of beetroot juice containing a negligible amount of nitrate and no nitrite (BR_P), beetroot juice containing 4.03 mmol nitrate and 0.29 mmol nitrite (BR_L), and beetroot juice containing 8.06 mmol nitrate and 0.58 mmol nitrite (BR_H). Group data (n = 13) were compared using a two-way repeated-measures analysis of variance and are shown as means ± SE, whereas individual responses are shown with filled circles. *P < 0.05 pre- vs. postsupplementation within beverage.

DISCUSSION

Data in the present study suggest increasing NO bioavailability, via acute inorganic nitrate supplementation, attenuates an early-morning rise in systolic blood pressure in patients with mild-to-moderate OSA (Fig. 2). Additionally, this beneficial effect could be mediated by a reduction in peripheral chemoreflex sensitivity as evidenced by a blunted ventilatory response to a second bout of hypoxia; however, this does not appear to be dose-dependent (Fig. 3). Importantly, these findings are independent of positive airway pressure therapy. Interestingly, and similar to our previous work in healthy older adults (22), the heart rate response to hypoxia was unchanged following inorganic nitrate supplementation.

Patients with OSA are at greater risk of developing several pathologies including cardiovascular disease (30) and stroke (31) as well as coconditions such as atrial fibrillation (32) and cognitive impairment (33). Although these conditions involve multiple organ systems, they share several commonalities, particularly hypertension (34). Several lines of work indicate patients with OSA have an exaggerated surge in blood pressure during the morning hours (35, 36), which likely contributes to the aforementioned pathologies. In contrast, other works report patients with OSA have an attenuated fall in blood pressure during sleep (i.e., nondippers) (37, 38). Although mathematically nondippers may not have a true diurnal rise, their blood pressure is nevertheless elevated during early-morning hours. Interestingly, data from the present study quantified an early-morning rise in systolic blood pressure as to the degree of 8 mmHg over ∼3 h (BR_P, Fig. 2). Importantly, consumption of dietary inorganic nitrate, in similar amounts used in the present study (4.0–8.0 mmol), has been shown to reduce resting systolic blood pressure by 5 mmHg (39), which is more pronounced in hypertensives (40). Although resting systolic blood pressure was not reduced in the present study, our data indicate acute inorganic nitrate supplementation may blunt the diurnal rise in systolic blood pressure (Fig. 2). That is, systolic blood pressure appeared to only increase 3 mmHg following BR_L (4.03 mmol nitrate and 0.29 mmol nitrite) and did not change following BR_H (8.03 mmol nitrate and 0.58 mmol nitrite). Interestingly, the majority of subjects in the present study (92%) did not have diagnosed hypertension (Table 1). When considered along with data showing half of patients with OSA are indeed hypertensive (41), inorganic nitrate supplementation may provide even greater benefit to this patient population as a whole.

Several lines of evidence indicate a 2 mmHg reduction in systolic and/or diastolic blood pressure is considered the minimal clinically important difference (MCID) (42). That is, a 2 mmHg lowering is associated with reduced risk of developing hypertension, stroke, and heart disease (42). Although the MCID for resting blood pressure provides a defined benchmark, there is no such metric for the diurnal rise in blood pressure. Over the specific window of measurements (∼7:00–10:00 AM), Jones et al. (43) quantified the morning rise in systolic blood pressure as 20–30 mmHg in hypertensives (10 mmHg is common in normotensives) (44), which is largely attributable to physical activity and eating. Importantly, subjects in the present study remained relatively sedentary and did not consume anything but their beetroot beverage and water during experimental visits. Despite these controls, subjects still demonstrated a substantial increase in blood pressure (Fig. 2). In a prospective study, Kario et al. (45) found after adjusting for medication use and differences in ambulatory blood pressure, a 10 mmHg surge in morning blood pressure was associated with 24% greater risk of stroke. Although the present study was not designed to quantify morning blood pressure surges, the increase in systolic blood pressure observed following BR_P was attenuated following inorganic nitrate supplementation (Fig. 2). Collectively, findings from the present study provide a proof-of-concept inorganic nitrate supplementation could help prevent pathological developments in patients with OSA by blunting the diurnal rise in blood pressure.

It is well known that exaggeration of peripheral chemoreflex sensitivity is a root cause of sympathetic-mediated comorbidities (e.g., hypertension) associated with OSA (46), a notion supported by Figs. 2 and 3. That is, following BR_P, systolic blood pressure and ΔV_E increased 8 mmHg and 1.3 L/min, respectively, effects which were prevented following BR_L and BR_H. Although sympathetic activity was not quantified in the present study, diurnal elevations in sympathetic tone lead to commensurate elevation in blood pressure (47). Given patients with OSA have heightened sympathetic outflow during waking hours (12) and exaggerated peripheral chemoreflex sensitivity (13), it is possible the underlying mechanism by which inorganic nitrate blunts the diurnal rise in systolic blood pressure is attenuation of tonic peripheral chemoreflex activity. Indeed, we observed a similar parallel previously (22) where 4 wk of daily inorganic nitrate supplementation reduced resting systolic blood pressure by 5 mmHg and the ventilatory response to hypoxia by 63% in healthy older adults.

Despite the promise of our present findings, and those from previous experiments (22), we recognize changes in the ventilatory response to hypoxia may not mirror sympathetic responses. Although the carotid body is responsible for stimulating both responses (14), the pathways which govern the ventilatory and sympathetic responses to acute hypoxia differ centrally (48). Data from the present study support this notion, whereby differences were observed in the V_E (Fig. 3), but not heart rate, response to hypoxia. As such, it is possible the beneficial effects of inorganic nitrate on systolic blood pressure (Fig. 2) may not be attributable to attenuation of sympathetic nerve activity. However, Notay et al. (21) found acute beetroot juice supplementation (6.4 mmol inorganic nitrate) lowered resting systolic blood pressure and muscle sympathetic nerve activity (MSNA) in young healthy adults. Unfortunately, neither peripheral chemoreflex sensitivity nor plasma [nitrite] were measured in Notay’s work (21) limiting comparability. Using a murine model, Guimarães and colleagues (49) found hypertension induced via angiotensin II (AngII) and N(G)-nitro-l-arginine methyl ester infusion increased angiotensin receptor 1 (AT-1) in the rostral ventrolateral medulla (RVLM) and renal sympathetic nerve activity. Inorganic nitrate supplementation reversed the attenuated NO bioavailability not only in plasma but also in cerebrospinal fluid and reduced blood pressure. This experimental design is particularly relevant to patients with OSA as they too demonstrate elevated levels of AngII (50) and reduced NO bioavailability (51). These highly sophisticated experiments led the authors to conclude resting sympathetic nerve activity can be reduced by augmenting NO bioavailability, via inorganic nitrate supplementation, through a central (e.g., RVLM) AngII/AT-1 mechanism. When these data (21, 49) are considered along with findings in Figs. 2 and 3, it appears acute consumption of inorganic nitrate likely reduced basal sympathetic outflow in patients in the present study.

Although the present study provides proof-of-concept data for inorganic nitrate supplementation in patients with mild-to-moderate OSA, there are some experimental considerations worth addressing; chiefly, SpO₂ was used in surrogate of Pao₂. Physiologically, a fall in the partial pressure of oxygen, not oxygen saturation, stimulates the peripheral chemoreflex; however, our protocol would have required three brachial artery catheters or numerous radial artery punctures within a 2-wk window for determination of arterial blood gases during hypoxia, before and after inorganic nitrate supplementation. It is also worth noting our approach to analyzing the data differs from other studies (52, 53) which may explain why we observed differential ventilatory responses. Importantly, our data, and analysis (25–27), align with previous studies reporting the peak ventilatory response occurs between the second and third minutes of hypoxia (13, 26). Additionally, MSNA was not measured during our experiments which would have improved the overall study design and mechanistic insight. However, previous works have shown the ventilatory and sympathetic arms of the peripheral chemoreflex follow similar patterns (54, 55). Although an attenuated heart rate response to hypoxia could indicate a commensurate reduction in MSNA, co-activation of pulmonary afferents may have masked these findings due to their sympathoinhibitory effects (48). Therefore, the addition of a voluntary end-expiratory apnea at 80%SpO₂ may have better reflected cardiac responsiveness (56). The direct vasodilatory effects of inorganic nitrate via NO should also be considered when interpreting the current findings (20, 57). This is important to note as augmented perfusion to the carotid body (58) or central tissues (e.g., RVLM) (59) could have influenced our conclusions. Our brachial blood pressure data are also limited in scope, as the gold-standard assessment for diurnal changes in blood pressure is done over a 24-h period (60) as opposed to “office” measurements completed in the current study.

We also recognize data in Fig. 2 conflict with a large body of the literature illustrating a blood pressure-lowering effect with comparable amounts of inorganic nitrate (4–8 mmol) (39). Interestingly, similar findings were reported in patients with heart failure (61), a population with elevated peripheral chemoreflex sensitivity (62). Thus, patients with exaggerated peripheral chemoreflex sensitivity (e.g., OSA and heart failure) may need to consume inorganic nitrate for a longer period (weeks) to see blood pressure reductions described in other works (39). We also recognize patients who completed the present study may not represent all patients with OSA due to our relatively small sample size, exclusion of patients with severe OSA, and those currently using positive airway pressure therapies. Our a priori power analysis was based upon data collected in older adults who demonstrated a 62% reduction in their hypoxic ventilatory response following 4 wk of 4.03 mmol inorganic nitrate supplementation (22). Since the previous study was considerably longer than the present (4 wk vs. acute, respectively), we anticipated a smaller 40% reduction corresponding to an eta-squared of 0.10, which would require 15 subjects to be properly powered. Nevertheless, we observed statistically significant, beneficial effects on both blood pressure (Fig. 2) and the ventilatory response to hypoxia (Fig. 3) in patients with mild-to-moderate OSA. The synergistic effects of inorganic nitrate supplementation with concomitant positive airway pressure therapy would be interesting to study as this arguably has the greatest clinical implications. However, the benefits of treatment with positive airway pressure on both blood pressure (63) and peripheral chemoreflex sensitivity (64) may induce a therapeutic ceiling.

Findings from this study provide proof-of-concept data that augmented circulating NO bioavailability via inorganic nitrate supplementation exerts beneficial effects on blood pressure in patients with mild-to-moderate OSA. More specifically, increase in early-morning systolic blood pressure was blunted following inorganic nitrate supplementation. Interestingly, this finding was paralleled by changes in the peak ventilatory response to hypoxia (i.e., peripheral chemoreflex sensitivity). Collectively, these data show promise for future works investigating the longitudinal effects of inorganic nitrate supplementation on autonomic regulation of blood pressure regulation in patients with OSA.

GRANTS

Core ESR facilities were supported by the National Institutes of Health P30 CA086862.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

J.M.B., M.E.D., and D.P.C. conceived and designed research; J.M.B., B.E.H., T.F.A., A.J.F., S.H., A.W.A., and D.P.C. performed experiments; J.M.B., T.F.A., and D.P.C. analyzed data; J.M.B. and D.P.C. interpreted results of experiments; J.M.B. prepared figures; J.M.B. drafted manuscript; J.M.B., B.E.H., T.F.A., A.J.F., S.H., A.W.A., M.E.D., and D.P.C. edited and revised manuscript; and J.M.B., B.E.H., T.F.A., A.J.F., S.H., A.W.A., M.E.D., and D.P.C. approved final version of manuscript.