This study includes data from participants that were tested in separate studies that investigated the effects of mild hypercapnia and acute BRJ consumption on renal hemodynamics (Chapman et al., 2020, 2021). These studies utilized similar measurements and protocols, but tested separate novel hypotheses. In the present study, thirteen healthy adults (age: 26 ± 4 years; height: 171 ± 9 cm; weight: 69 ± 11 kg; body mass index: 23 ± 3 kg/m²; 5 women) completed the study. All participants were fully informed of the experimental procedures prior to providing written, informed consent. Participants were excluded from the study if they self-reported to use tobacco or medications known to affect blood pressure or the autonomic nervous system (e.g., beta-blockers, angiotensin converting enzyme inhibitors); had a pre-existing autonomic, cardiovascular, metabolic, respiratory, or endocrine disorder; major depression; were pregnant or breastfeeding. Women were tested during the first 10 days of their self-identified menstruation (i.e., onset of menses) to control for the menstrual cycle (Favre and Serrador, 2019). The study was approved by the Institutional Review Board at the University at Buffalo and was performed in accordance with the standards set by the latest revision of the Declaration of Helsinki except for registration in a database.

Participants completed one study visit that consisted of oscillatory lower-body negative pressure (Oscillatory-LBNP) and static lower body negative pressure (Static-LBNP) before and 3 h after BRJ consumption to allow for plasma nitrite levels to peak (Webb et al., 2008; Casey et al., 2015). Participants were instructed to abstain from medications, alcohol, exercise, and caffeine for 12 h, and food for 2 h prior to reporting to the laboratory. Participants were also instructed to abstain from antibacterial mouthwash and chewing gum the morning of the study as this can hinder the breakdown of dietary nitrates (Govoni et al., 2008). Upon arrival, a urine sample was obtained to ensure euhydration (urine specific gravity <1.020) and negative pregnancy status (women only). Women also reported the start date of their last menstrual cycle.

Participants laid in the supine position for the duration of instrumentation and measurements. A neoprene skirt was used to seal the participants’ lower body at the level of the iliac crest into an airtight chamber for LBNP testing (described below). Participants were then instrumented with physiological monitoring equipment. Then, after >10 min of supine rest, 5 min of baseline data were recorded before initiating the LBNP protocols. To attenuate reductions in PETCO₂ that can be observed during LBNP, participants inhaled a mild hypercapnic gas (FiCO₂: 0.03, FiO₂: 0.21, FiN₂: 0.76) during baseline and throughout the duration of testing. Our lab has previously demonstrated that LBNP set to −40 mmHg for 5 min reduces PETCO₂ by 5 ± 3 mmHg in healthy, young adults (Worley et al., 2021), which likely influences cerebrovascular responses to LBNP. However, elevated FiCO₂ can also impair CA. This has been reported using an FiCO₂ of ≥0.05 (Panerai et al., 1999; Perry et al., 2014), thus we selected an FiCO₂ of 0.03 to attenuate potential reductions in PETCO₂ that may occur during LBNP, thereby reducing the risk of altering CA.

Following the 5-min baseline, the Oscillatory-LBNP protocol commenced. Oscillatory-LBNP was conducted at a frequency of ∼0.02 Hz, which consisted of 30 s on (−70 mmHg) and 30 s off (0 mmHg) for a total of six cycles. Then, a 3-min recovery period was allotted before the Static-LBNP protocol commenced. Static-LBNP consisted of −40 mmHg for a total of 7 min to assess static CA (Claassen et al., 2016; Panerai et al., 2022). Upon completion of Static-LBNP, participants were de-instrumented and consumed 500 mL of commercially available BRJ (listed as ∼1,500 mg/L of nitrate) (Biotta, Carmel, IN, USA) (Kenjale et al., 2011; Bond et al., 2013) within 10 min. The average concentration of nitrates measured in Biotta BRJ has been reported previously as 18,708 μmol/L, respectively (Casey et al., 2015). The testing protocol was then repeated 3 h after BRJ consumption. During this 3 h period, participants rested quietly in the laboratory (e.g., watched a non-stimulating video) and were allowed to consume 250 mL of water during the break.

The LBNP protocols had pre-defined termination endpoints, including (a) Sustained systolic blood pressure (SBP) <80 mmHg with a precipitous decrease in heart rate; (b) If participants self-reported any symptoms of pre-syncope (e.g., sweating, tunnel vision, nausea, dizziness); or (c) Upon participant request (Rickards et al., 2015). All participants completed the Oscillatory-LBNP protocol. One male participant did not complete the entire 7 min Static-LBNP protocol at pre-BRJ. This test was terminated at 6 min due to relative bradycardia with reduced SBP.

Height was measured with a stadiometer (Sartorius, Bohemia, NY, USA). Nude body weight was measured on an electronic scale to the nearest 0.01 kg in a private bathroom. Urine specific gravity was measured via a refractometer (Atago USA, Bellevue, WA, USA). Heart rate was continuously measured using 3-lead electrocardiogram (DA100C, Biopac Systems, Goleta, CA, USA). Beat-to-beat blood pressure was non-invasively measured using finger photoplethysmography on the left hand (middle finger) at heart level. Finger blood pressure was calibrated to brachial artery blood pressure in the supine position using the return-to-flow module (Finometer Pro, FMS, Amsterdam, Netherlands). Blood pressure was confirmed intermittently throughout the protocol using auscultation of the right brachial artery (electrosphygmomanometry; Tango M2; SunTech, Raleigh, NC, USA). End-tidal CO₂ tension (PETCO₂) was continuously measured to estimate arterial CO₂, from a mouthpiece using capnography while the nose was occluded with a nose clip (Nonin Medical, Inc., Plymouth, MN, USA). Left middle cerebral artery blood velocity (MCAv) was continuously measured using a 2 MHz probe (DWL USA, Inc., Germany, Europe) with insonation techniques outlined by Willie et al. (2012). The anatomical location for the left MCA was marked by indelible ink so that the transducer placement for pre- and post-BRJ testing was the same. Additionally, the settings for depth and gain were kept the same for pre- and post-BRJ testing.

Heart rate (1000 Hz), blood pressure (62.5 Hz), PETCO₂ (62.5 Hz), and MCAv (1000 Hz) were continuously recorded using a data acquisition system (Biopac MP150, AcqKnowledge 4.2.0, Goleta, CA, USA). Stroke volume was estimated using Modelflow (Wesseling et al., 1985). Cardiac output was calculated as the product of heart rate and stroke volume. Cerebrovascular conductance index (CVCi) was calculated as the quotient of MCAv and mean arterial pressure. Mean values were extracted for all variables during the last minute of supine rest preceding Oscillatory-LBNP (baseline Oscillatory-LBNP) and Static-LBNP (baseline Static-LBNP), every interval of Oscillatory-LBNP, and every minute of Static-LBNP. Data are presented as the mean with 95% confidence intervals.

Dynamic CA was assessed using Oscillatory-LBNP intervals of 30 s on (−70 mmHg) and 30 s off (0 mmHg) (∼0.02 Hz). This interval time was selected due to the cerebral vasculature’s ability to dampen changes in arterial pressure that occur slower than ∼30 s (0.03 Hz) (Brothers et al., 2009; Hamner et al., 2019). Static CA was assessed by applying static-LBNP (−40 mmHg) for 7 min (Claassen et al., 2016; Panerai et al., 2022). Raw data files of beat-to-beat blood pressure (input signal) and MCAv (output signal) were used to assess dynamic and static CA via transfer function analysis (WinCPRS, Absolute Aliens Oy, Turku, Finland). Please refer to the examples of typical tracings for oscillatory-LBNP and Static-LBNP protocols in Figure 1. These analog signals were integrated with the R-R intervals (RRIs) of the ECG and re-sampled at 5 Hz to align the beat-to-beat time series. Data were analyzed by fast Fourier transformation using a Hanning window with sliding bandwidth of 180 s and 50% overlap to reduce cross-spectral leakage. Using Welch’s method, coherence, phase, and gain were calculated across the 0.00−0.50 Hz frequency band (Panerai et al., 2022). Based on transfer function analysis recommendations, we calculated indices of CA in the standard bands defined as very-low-frequency (VLF; 0.02–0.07 Hz), low-frequency (LF; 0.07–0.20 Hz), and high-frequency (HF; 0.20–0.50 Hz) (Panerai et al., 2022) for both LBNP protocols. A critical coherence value was not implemented for the inclusion/exclusion of gain and phase estimates for the following reasons: (1) our oscillatory-LBNP was conducted in the VLF band (∼0.02 Hz), which can result in low coherence values due to blood pressure and MCAv exhibiting a non-linear relationship (Claassen et al., 2016; Panerai et al., 2022), and (2) low signal-to-noise ratio due to poor data quality was unlikely due to visual inspection and exclusion of data that had excessive artifact (see below). Additionally, within the VLF band, gain and phase estimates are not different when comparing low vs. high coherence data (Claassen et al., 2009). Transfer gain and phase reflect the magnitude and time delay between alterations in blood pressure and MCAv within a given frequency, respectively. A decrease in gain and increase in phase are indicative of an “improvement” in CA as this reflects a greater ability of the brain to buffer changes in arterial pressure. Negative phase values were excluded from averaging in frequency bands <0.10 Hz due to the “wrap-around” effect (Claassen et al., 2016). Additionally, raw data files of beat-to-beat blood pressure and MCAv were visually inspected for artifact. Participants were excluded from transfer function analysis if excessive artifact persisted beyond 3 beats [Oscillatory-LBNP: n = 1 (male); Static-LBNP; n = 3 (1 male)] (Claassen et al., 2016). One additional participant (male) was excluded from Static-LBNP analysis since pre-testing was terminated early due to pre-syncopal symptoms.

Raw data files of ECG and beat-to-beat blood pressure were used for offline analysis of cBRS (WinCPRS, Absolute Aliens Oy, Turku, Finland) during Oscillatory-LBNP (∼6 min) and Static-LBNP (7 min). The ECG and beat-to-beat blood pressure waveforms were visually inspected for artifact and ectopic beats (none were identified). cBRS was calculated using the sequence method (Bertinieri et al., 1985; Laude et al., 2009). Four successive increases or decreases in SBP (± 1 mmHg) that corresponded with appropriate changes in RRI (± 5 ms) were identified (Parati and Mancia, 2000). Sequences were deemed valid if the R² value calculated from individual linear regression analyses between SBP and RRI was ≥0.85 (Parati and Mancia, 2000). The time delay between SBP and RRI was set to 1 beat (Gross et al., 2002). Sequences were determined separately for Up-BRS (i.e., concurrent increases in SBP and RRI) and Down-BRS (i.e., concurrent decreases in SBP and RRI).

Data were analyzed using PRISM software (version 8, GraphPad Software, La Jolla, CA, USA). We used linear mixed-effects models to compare absolute mean values during Oscillatory-LBNP at PRE-BRJ and POST-BRJ. Linear mixed-effects models were also used to compare absolute mean values during Static-LBNP at PRE-BRJ and POST-BRJ. No statistical comparisons were made between LBNP protocols. If a linear mixed-effects model revealed a significant interaction or main effect (Tybout AaS, 2001), we used Bonferroni’s post hoc analyses to determine where differences occurred. Within LBNP protocols, we used two-tailed paired t-tests to determine if indices of dynamic CA and cBRS differed between PRE-BRJ and POST-BRJ. Lastly, Cohen’s d effect sizes were calculated for CA and cBRS between PRE-BRJ and POST-BRJ within each LBNP protocol (Lenhard and Lenhard, 2016). Cohen’s d effect sizes were interpreted as none (d = 0.0–0.1), small (d = 0.2–0.4), intermediate (d = 0.5–0.7), and large (d = 0.8 and above) (Cohen, 1988; Lakens, 2013). Statistical significance was set a priori at P ≤ 0.05. Data are presented as mean with 95% confidence intervals. Individual p-values are reported where possible.

We assessed dynamic CA using transfer function analysis during blood pressure oscillations induced at ∼0.02 Hz. Although the oscillatory LBNP protocol was not designed to elicit cardiovascular decompensation, all participants tolerated the oscillatory LBNP protocol before and after BRJ. Throughout the six cycles of oscillatory LBNP, we did not observe differences in the mean arterial pressure, PETCO₂, and MCAv responses between PRE- and POST-BRJ (Figures 3A–C). However, there was a condition effect for MCAv (P = 0.044), as MCAv was lower throughout oscillatory LBNP. Despite no statistical differences in the mean arterial pressure and MCAv responses to oscillatory LBNP, there were differences in some transfer function analysis indices of dynamic CA. In the VLF band, coherence was greater following BRJ (Table 1), which suggests a stronger relation between blood pressure and MCAv fluctuations. Despite a stronger relation, VLF phase and gain did not change after BRJ ingestion (Table 1). Thus, the cerebral arterioles buffered VLF oscillations in blood pressure similarly pre- and post-BRJ. We did find that LF gain decreased by ∼0.09 cm/s/mmHg after BRJ, but phase and coherence were not altered. A decrease in LF gain has been reported following 7-days of sodium nitrate supplementation in men, but not women, at rest (Fan et al., 2019). Although we did not examine sex differences, we controlled for menstrual cycle to avoid the potential confounding factor of estrogen on endothelial NOS activity (McNeill et al., 2002), whereas the other investigators did not as their intervention was conducted over 7 days. It is possible that the decrease in LF gain following dietary nitrate ingestion is an improvement in CA. But, it must be considered that the LF band (0.07–0.20 Hz) does not align with our blood pressure challenge (∼0.02 Hz) so these data must be interpreted cautiously. Future work should aim to determine if acute BRJ consumption alters CA during a LF dynamic blood pressure challenge (e.g., O-LBNP or squat-stand maneuvers performed at 0.10 Hz).

We also assessed CA during a constant pressure perturbation to lend insight on the static control of cerebral blood flow and blood pressure. Static LBNP decreased mean arterial pressure by ∼6 mmHg (range: 0 to −40 mmHg) at PRE-BRJ and ∼5 mmHg (range: 0 to −13 mmHg) at POST-BRJ. The wide range at pre-BRJ was contributed by one male participant that exhibited large reductions in blood pressure at minute 5 and 6 at −40 mmHg before LBNP was terminated due to experiencing relative bradycardia associated with a decreasing blood pressure. Interestingly, this participant tolerated static LBNP following BRJ. Regardless, mean arterial pressure throughout the 7 min of LBNP was not different between pre and post-BRJ (Figure 5A). We observed a similar response in PETCO₂, as BRJ did not alter PETCO₂ during LBNP compared to pre-ingestion. Utilizing an FiCO₂ of 0.03 attenuated potential reductions in PETCO₂ during LBNP, as PETCO₂ was ∼1–2 mmHg lower than baseline during every minute of LBNP (Figure 5B). More pronounced reductions in PETCO₂ may have occurred during LBNP if we did not add CO₂ to the inspirate. We previously observed ∼5 mmHg decrease in PETCO₂ during 5 min of LBNP in healthy men and women (Worley et al., 2021). Despite reductions in mean arterial pressure and PETCO₂, MCAv was maintained throughout LBNP at pre- and post-BRJ ingestion (Figure 5C). These mean values suggest that cerebral autoregulatory capacity was not altered by BRJ ingestion. Indices of static CA via transfer function analysis support the results of the analyses on the mean data as coherence, phase, and gain did not differ in any frequency band from PRE- to POST-BRJ (Table 3). Evidence is scant regarding the acute effects of dietary nitrate on CA during a sustained and acute hypotensive stimulus. Consuming 500 mL of BRJ (∼750 mg of nitrate) has been shown to reduce the cerebral augmentation index during aerobic exercise performed at 40% and 80% of peak oxygen consumption as well as lower SBP at rest 2 h following ingestion 66 (Curry et al., 2016). It is difficult to compare our results to this investigation as aerobic exercise poses a different challenge to cerebral vascular resistance and blood flow compared to LBNP. Our data align more with Horiuchi et al. (2022) who had participants ingest 140 mL of BRJ per day for 4 days. CA in the internal carotid artery (ICA) was not altered using the thigh-cuff deflation technique that induces a rapid decrease in blood pressure. Although we also observed no change in CA, the vessel of interest (ICA vs. MCA), analysis technique (rate of regulation vs. transfer function analysis), mode of test (thigh cuff deflation vs. O-LBNP), and BRJ dose (4 days of supplementation vs. acute and volume) were different. Nevertheless, our healthy participants maintained CA function following the acute ingestion of BRJ.

Utilizing the sequence method, we hypothesized that cBRS would decrease (i.e., worsen) following BRJ ingestion in healthy, young adults. Contrary to our hypothesis, we observed no change in Up-Up or Down-Down cBRS after BRJ during static or oscillatory LBNP. During oscillatory LBNP, we observed no change in blood pressure between conditions, but RRIs decreased from pre to post-BRJ (Table 2). Although RRIs during the static LBNP protocols were not statistically different between conditions, the RRI responses (P = 0.07) were similar to the oscillatory LBNP responses. Yet, blood pressure and cBRS during the static LBNP protocols were not different between conditions (Table 4). Although we did not observe a change in cardiovagal baroreflex sensitivity following BRJ consumption, the acute administration of a nitric oxide donor did increase baroreceptor sensitivity in rabbits (Matsuda et al., 1995), whereas sublingual nitroglycerin reduced spontaneous cBRS immediately following administration (∼10 min) in healthy participants (Hamaoka et al., 2021) and 6-days after administration (Gori et al., 2002). Our findings likely differ from others due to the inorganic nitrate administered, time of testing post-administration, and how baroreflex sensitivity was measured (e.g., spontaneous vs. perturbation).

There are several methodological considerations that should be taken into account when interpreting our data. First, we removed the transcranial Doppler ultrasound probe between testing sessions. This may alter the signal between testing time points. However, we utilized standard insonation techniques set forth by Willie et al. (2011) and recorded the depth, gain, and location (with indelible ink) to ensure the insonation angle was closely matched. Second, we did not establish baseline dynamic CA while participants were breathing room air. Lower body negative pressure increases ventilation (Koehle et al., 2010) and decreases end-tidal CO₂ content (Ahn et al., 1989; Bronzwaer et al., 2017; Rosenberg et al., 2021), which can exacerbate reductions in intracranial artery blood velocity. Therefore, we bled in a mild hypercapnic gas (3% CO₂, 21% O₂, balanced with N₂) to the inspirate in an attempt to attenuate reductions in PETCO₂ during LBNP protocols. The mild hypercapnic gas was inhaled for 5 min of supine baseline to allow for a new stable PETCO₂ and baseline to be achieved prior to LBNP initiation. It has previously been reported that hypercapnia impairs dynamic CA, but this has been shown when PETCO₂ is 5 mmHg greater than resting baseline (Aaslid et al., 1989; Zhang et al., 1998; Ainslie et al., 2005). On average, 3% CO₂ increased PETCO₂ by ∼3 mmHg (PRE-BRJ) and ∼4 mmHg (POST-BRJ) from room air supine rest. We speculate that CA was not impaired due to the modest increases in PETCO₂ that likely did not alter MCA diameter (Verbree et al., 2014; Kellawan et al., 2016), but we cannot confirm this without air breathing conditions. Nevertheless, PETCO₂ did not differ during supine rest, oscillatory LBNP, or static LBNP before or after BRJ consumption. Second, we did not have a placebo condition to more clearly determine the effects of BRJ ingestion on CA that may be influenced by time alone (i.e., tested 3 h apart). Daily fluctuations in CA are poorly understood, but it appears that spontaneous autoregulation remains stable during the daytime hours (8 am–8 pm) in healthy individuals (Guo et al., 2018). All of our experimental visits took place within this time frame. Third, we did not measure plasma nitrate or nitrite to determine the magnitude of increase after consuming 500 mL of BRJ. Using the same dose of a commercially available BRJ (500 mL of Biotta), Casey et al. (2015) reported a threefold increase in plasma nitrite 3 h following ingestion in a group of participants with similar demographics as our study. Thus, it is very likely that circulating nitrites were elevated in our study. Fourth, we did not randomize the LBNP protocols. There is a possibility that prolonged inhalation of hypercapnic gas may have influenced LBNP protocols differently. However, it was not our objective to compare CA or cBRS responses between LBNP protocols. Finally, females were tested during the early follicular phase to control for the influence of sex hormones on endothelial NOS activity (Miller and Duckles, 2008; Duckles and Miller, 2010). It is currently unclear if there is an interaction between menstrual cycle, dietary nitrates, and CA. Sex differences have been observed in CA, but this does not appear to be affected by menstrual cycle phase (Favre and Serrador, 2019). However, Fan et al. (2019) had participants consume sodium nitrate (10 mg/kg/day) for 7 days and observed a similar response in LF mean gain (no change) and LF phase (increased) between men and women. However, others have observed no sex differences in CA at rest and following 7 days of dietary nitrate supplementation (Fan et al., 2019). Fan et al. (2019) observed similar autoregulatory responses between men and women following dietary nitrate supplementation (sodium nitrate; 10 mg/kg/day for 7 days), with both sexes exhibiting contradictory alterations in CA (i.e., decrease in gain and phase).