Participants.

Nine healthy subjects (four male: mean ± SD, age 30 ± 6 years; body mass 77 ± 11 kg; stature 1.78 ± 0.06 m; and five female: age 30 ± 6 years; body mass 58 ± 4 kg; stature 1.66 ± 0.02 m) volunteered to participate in the study. The participants were all recreationally active, but not highly trained. Prior to testing, participants were informed of the protocol and risks and gave written consent to participate in the study. All procedures were approved by Swansea University Ethics Committee and were conducted in accordance with the Declaration of Helsinki. Participants were asked to arrive at the exercise physiology laboratory at Swansea University in a rested state, at least 2 h postprandial and to avoid strenuous exercise in the 24 h preceding each testing session. Participants were also asked to refrain from caffeine and alcohol for 6 and 24 h before each test, respectively. The participants also refrained from the use of antibacterial mouthwash throughout the duration of the study because this has been shown to markedly attenuate the oral bacteria which are necessary for the conversion of nitrate to nitrite (26). All tests were performed at the same time of day (± 0.5 h).

Procedures.

Participants were required to visit the laboratory on seven occasions over a 4-week period. On the first visit, participants completed a ramp incremental exercise test for determination of the V̇o_{2 peak} and GET. The test included 3 min of baseline cycling at 15 W, after which the work rate was increased at a rate of 20 W/min for females and 30 W/min for males until the limit of tolerance. The participants were asked to maintain a cadence of 70–80 rpm. Breath-by-breath pulmonary gas-exchange data were collected continuously during the incremental tests and averaged over consecutive 5-s periods (Oxycon Pro, Jaeger, Germany). The V̇o_{2 peak} was taken as the highest 10-s mean value attained before the subject's volitional exhaustion in the test. The GET was determined using the V-slope method (9) as the first disproportionate increase in CO₂ production (V̇co₂) relative to the increase in V̇o₂, and subsequently verified by an increase in the ventilatory equivalent for V̇o₂ (V̇e/V̇o₂) with no increase in V̇e/V̇co₂. The work rates that would require 90% of the GET (moderate-intensity exercise) and 70% of the difference (Δ) between the GET and V̇o_{2 peak} (severe-intensity exercise, Δ70%) were subsequently determined, with account taken of the mean response time for V̇o₂ during ramp exercise [i.e., two-thirds of the ramp rate was deducted from the work rate at the GET and peak V̇o₂ (65)].

Following the ramp incremental test, participants were randomly assigned in a crossover, double-blind design to receive 6 days of dietary supplementation with NO₃⁻-rich beetroot juice (BR) (140 ml/day; ∼8 mmol NO₃⁻; Beet It, James White Drinks, Ipswich, UK) or NO₃⁻-depleted BR as a placebo (PL; 140 ml/day; 0.0034 mmol NO₃⁻; Beet It, James White Drinks, Ipswich, UK). The placebo NO₃⁻-depleted BR beverage was identical in color, taste, smell, and texture to the experimental NO₃⁻-rich BR beverage. The PL beverage was created by passage of the juice, before pasteurization, through a column containing Purolite A520E ion exchange resin, which selectively removes NO₃⁻ ions (44). Five participants began with the BR condition, and the other four participants began with the PL condition. The subjects were instructed to consume the beverages (70 ml in the morning and afternoon) on days 1–3 of the supplementation period. On days 4–6, the subjects were instructed to consume the beverages over a 10-min period, 2 h prior to the start of the exercise test (see below), based on recent evidence that plasma [NO₂⁻] peaks at ∼2–2.5 h postadministration of BR containing 8.4 mmol NO₃⁻ (71). A 7-day washout period separated each supplementation period. Throughout the study, subjects were instructed to maintain their normal daily activities and food intake.

On days 4, 5, and 6 of the supplementation periods, subjects completed a series of step exercise tests for the determination of V̇o₂ and muscle [HHb] kinetics. The protocol, which was performed on three consecutive days, consisted of 3-min “unloaded” pedaling at 15 W, followed by 4-min of moderate-intensity cycling (U→M), and then 6-min of severe-intensity cycling (M→S). The tests were performed on separate days because it is known that prior exercise can alter the V̇o₂ response to exercise (3). A schematic illustration of the experimental protocol is shown in Fig. 1. On day 6 of each supplementation period, the M→S bout was continued until task failure. The participants were blinded to the elapsed exercise time in both the BR and PL conditions. The time to task failure was used as a measure of exercise tolerance and was recorded when the pedal rate fell by >10 rpm below the required pedal rate. In total, the participants completed three bouts of U→M and M→S exercise following BR and PL ingestion, with the V̇o₂ data being subsequently ensemble-averaged prior to curve-fitting to enhance the signal-to-noise ratio.

Measurements.

Venous blood samples (∼4 ml) were drawn into lithium-heparin tubes (7.5 ml Monovette lithium heparin; Sarstedt, Leicester, UK), which have very low levels of NO₃⁻ and NO₂⁻, on each of days 4–6. Within 3 min of collection, the samples were centrifuged at 2,700 g and 4°C for 10 min. Plasma was extracted and immediately frozen at −80°C for later analysis of [NO₂⁻] using a modification of the chemiluminescence technique (7). All glassware, utensils, and surfaces were rinsed with deionized water to remove residual NO₂⁻ prior to analysis. Following defrosting at room temperature, the [NO₂⁻] of the undiluted (nondeproteinized) plasma was determined by its reduction to NO in the presence of glacial acetic acid and 4% (wt/vol) aqueous NaI. The spectral emission of electronically excited nitrogen dioxide product, from the NO reaction with ozone, was detected by a thermoelectrically cooled, red-sensitive photomultiplier tube housed in a Sievers gas-phase chemiluminescence nitric oxide analyzer (Sievers NOA 280i; Analytix, Durham, UK). The [NO₂⁻] was determined by plotting signal (mV) area against a calibration plot of 100 nM to 1 μM of sodium nitrite.

Throughout all exercise tests, participants wore a face mask and breathed through a low dead space (90 ml), low-resistance (0.75 mmHg·l⁻¹·s⁻¹ at 15 l/s) impeller turbine assembly (Jaeger Triple V, Hoechberg, Germany). The inspired and expired gas volumes and gas concentration signals were continuously sampled at 100 Hz, the latter using paramagnetic (O₂) and infrared (CO₂) analyzers (Jaeger Oxycon Pro, Hoechberg, Germany) via a capillary line connected to the mouthpiece. These analyzers were calibrated before each test with gases of known concentration, and the turbine volume transducer was calibrated using a 3-liter syringe (Hans Rudolph, Kansas City, MO). The volume and concentration signals were time aligned by accounting for the delay in capillary gas transit and analyzer rise time relative to the volume signal. Breath-by-breath fluctuations in lung gas stores were corrected for by computer algorithms (8). A Reynolds Lifecard CF digital Holter recorder (Spacelabs Medical, Hertford, UK) was used to record a three-lead ECG continuously throughout the tests. The ECG leads were positioned in the modified V5, CC5, and modified V5R electrode configuration. This system provided ECG data with a sample accuracy of 2.5 μV and 1,024-Hz sampling frequency. During one of the U→M and M→S transitions, for both supplementation periods, a blood sample was collected from a fingertip into a capillary tube over the 20 s preceding the step transition in work rate and within the last 20 s of exercise. A capillary blood sample was also collected at the limit of tolerance for the M→S bout performed on day 6 of each supplementation period. The blood samples were subsequently analyzed to determine [lactate] (YSI 1500; Yellow Springs Instruments, Yellow Springs, OH) within 30 s of collection. Blood lactate accumulation was calculated as the difference between blood [lactate] at end-exercise and blood [lactate] at baseline.

NIRS was used to monitor changes in oxygenation status of the musculus vastus lateralis of the right leg during step exercise (NIRS; OxiplexTS; ISS, Champaign, IL). The NIRS probe was affixed over the midway point between the greater trochanter and lateral epicondyle of the right leg using adhesive tape and secured by elastic Velcro strapping to ensure the device remained stationary and to minimize the interference of extraneous light during exercise. Source (NIR) light was emitted into the muscle at wavelengths of 690 and 830 nm, and detection was sampled at 2 Hz to measure absolute concentrations (μM) of oxyhemoglobin (HbO₂) and deoxyhemoglobin (HHb) within the microcirculation of the interrogated muscle region. Light source-detector separation distances of 1.50–3.04 cm for each wavelength were used with cell water concentration assumed to be constant at 70%. The NIRS probe was calibrated before each testing session using a calibration block of known absorption and scattering coefficients. Calibration was then cross-checked using a second block of known, but distinctly different, absorption and scattering coefficients. Each of these procedures was performed according to the manufacturer's recommendations. The contribution of myoglobin (Mb) to the NIRS signal is generally accepted to be relatively small (50, 62) but is currently unresolved. The [HHb] signal reported herein should, therefore, be considered to reflect the combined concentrations of both deoxygenated Hb and Mb.

Data analysis procedures.

The breath-by-breath V̇o₂ data from each step exercise bout were initially examined to exclude “errant” breaths by removing values lying more than four standard deviations from the local mean determined using a 5-breath rolling average. Filtered V̇o₂ data were subsequently linearly interpolated to provide second-by-second values and, for each individual, identical repetitions of each exercise condition were time aligned to the start of exercise and averaged together to form a single data set for analysis.

For each step transition, the first 20 s of data after the onset of exercise were deleted to remove the phase I (cardiodynamic) response and a monoexponential model with time delay (Eq. 1) was then fitted to the averaged V̇o₂ data.

A monoexponential model was ultimately used for both moderate and severe-intensity exercise because, for the M→S transition, a biexponential model (Eq. 2) produced an inferior and ambiguous fit based on analysis of the model residuals.

Given the failure of the biexponential model to adequately describe the V̇o₂ response during M→S, the onset of the V̇o₂ slow component was determined using purpose-designed LabVIEW software, which iteratively fits a monoexponential function to the V̇o₂ data until the window encompasses the entire response. The estimated τ for each fitting window was plotted against time and the onset of the V̇o₂ slow component was identified as the point at which the estimated τ consistently deviated from the previously “flat” profile (61). The amplitude of the V̇o₂ slow component was subsequently determined by calculating the difference between the end exercise V̇o₂ and the sum of the primary amplitude and baseline V̇o₂. This was expressed both in absolute terms and relative to the end-exercise V̇o₂. The functional gain of the primary V̇o₂ response during U→M and M→S was also calculated by dividing the primary phase amplitude by the change in work rate. Finally, the mean response time (MRT) for both U→M and M→S was calculated by fitting a single exponential curve to the data with no time delay from the onset to the end of exercise.

The NIRS-derived [HHb] response to exercise was also modeled to provide information on muscle oxygenation. The responses to each transition were interpolated to 1-s intervals, time aligned, and averaged to produce a single data set. Since the [HHb] signal increased after a short delay in response to step exercise, the time of onset for the exponential-like rise in [HHb] was defined as a 1 SD increase in [HHb] above the mean baseline value (19). The model in Eq. 1 was then used to resolve the [HHb] τ after omitting data points preceding the exponential-like increase. For M→S, the model fitting window was constrained to the onset of the [HHb] slow component determined using the iterative curve-fitting procedure, as described for V̇o₂ above. The primary [HHb] amplitude was divided by the phase II V̇o₂ asymptote to determine the Δ[HHb]/ΔV̇o₂ ratio as an index of the change in fractional muscle O₂ extraction required to elicit a given ΔV̇o₂ during the primary phase. In addition, we assessed changes in total blood volume by summing the [HbO₂] and [HHb] signals to provide an estimate of the total [Hb_tot] in the area under investigation. Specifically, we determined the mean value at baseline (30 s preceding each transition), at 60-s intervals throughout exercise (15-s bins centered on each time point), and at end exercise (final 30 s) to facilitate comparisons between conditions. Finally, heart rate (HR) kinetics was modeled for each condition with the TD parameter in Eq. 1 fixed to t = 0 s (i.e., monoexponential with no delay) and with the fitting window constrained to the onset of the V̇o₂ “slow component.”

Statistics.

Gaussian distribution was confirmed by the Shapiro-Wilks test. Following this, the pulmonary V̇o₂, HR, and NIRS-derived variables were analyzed using two-way repeated-measures ANOVA with “exercise intensity” (U→M and M→S) and “supplement” (BR vs. PL) included as within-subject factors. Differences in BP and plasma [NO₂⁻] were determined using two-way (supplement × time) repeated-measures ANOVA. Subsequent paired samples t-tests were employed as appropriate to identify the location of statistically significant effects. Pearson product moment correlation coefficients were used to analyze the degree of association between key variables. All statistical analyses were conducted using PASW Statistics 18 (SPSS, Chicago, IL). Data are presented as means ± SD. Statistical significance was accepted when P ≤ 0.05.

Plasma [NO₂⁻].

There was a main effect for supplement on plasma [NO₂⁻] at rest over the last 3 days of the supplementation period [F_(1,8) = 21.59, P = 0.01]. Follow-up paired comparisons revealed that plasma [NO₂⁻] was elevated (P < 0.02) at each sample point following BR compared with PL ingestion on day 4 (PL: 64 ± 36 vs. BR: 300 ± 141 nM), day 5 (PL: 66 ± 35 vs. BR: 374 ± 149 nM), and day 6 (PL: 65 ± 32 vs. BR: 348 ± 170 nM).

Muscle oxygenation.

The [Hb_tot] and [HHb] values derived from NIRS interrogation are presented in Table 1. There was no significant main effect for supplement on the [Hb_tot] during U→M and M→S exercise. The [HHb] response during step exercise for a representative subject is illustrated in Fig. 2. Two-way ANOVA revealed a significant interaction effect between exercise intensity and supplement on [HHb] kinetics following the onset of exercise [F_(1,6) = 15.30, P = 0.01]. Specifically, compared with PL, the [HHb] τ was speeded during M→S following BR supplementation (PL: 20 ± 9 vs. BR: 10 ± 3 s, P = 0.05), but there were no differences between PL and BR during U→M (PL: 7 ± 3 vs. BR: 10 ± 5 s, P = 0.17). The [HHb] τ was significantly slower for M→S compared with U→M in PL (P = 0.01), but there was no difference between the upper and lower step in BR (P = 0.94) There was no significant main effect for supplement on the primary [HHb] amplitude when normalized per unit change in V̇o₂ during the fundamental exponential phase [F_(1,6) = 4.81, P = 0.07].

HR kinetics.

The HR responses to step exercise are presented in Table 2. There were no differences in the primary HR τ between PL and BR for U→M or M→S [F_(1,8) = 0.10, P = 0.77]. During M→S, the relative change in the V̇o₂ τ_p was not correlated with the relative change in HR kinetics between conditions (r = 0.42, P = 0.27). There were no significant differences in blood [lactate] between conditions.

V̇o₂ kinetics and exercise tolerance.

The V̇o₂ kinetic parameters derived from the monoexponential fit are presented in Table 3, and the V̇o₂ response of a representative subject to U→M and M→S is shown in Fig. 2. The group mean V̇o₂ profile during M→S is presented in Fig. 3. Two-way ANOVA revealed a significant interaction effect between exercise intensity and supplement on phase II V̇o₂ kinetics following the onset of exercise [F_(1,8) = 18.54, P = 0.01]. Compared with PL, the τ_p was shorter during M→S following BR ingestion (PL: 46 ± 13 vs. BR: 36 ± 10 s, P = 0.01), but there were no differences during U→M (PL: 25 ± 4 vs. BR: 27 ± 6 s, P = 0.25). For the PL condition, the τ_p was greater in M→S compared with U→M (P = 0.001), but there were no significant differences between U→M and M→S in the BR condition (P = 0.12). During M→S, the speeding of V̇o₂ τ_p was not correlated with the speeding of the primary [HHb] τ after BR compared with PL (r = −0.16, P = 0.76).

There was no significant main effect for supplement on the primary V̇o₂ amplitude [F_(1,8) = 0.01, P = 0.91] or primary V̇o₂ gain [F_(1,8) = 0.05, P = 0.83] during U→M or M→S. The emergence of a slow phase in V̇o₂ during M→S occurred after a similar time delay, and there were no differences in the absolute or relative amplitude of the V̇o₂ slow component between PL and BR (both P = 0.44). For M→S, there were no differences between PL and BR in the V̇o₂ amplitude at end-exercise [F_(1,8) = 0.60, P = 0.46] or the total V̇o₂ gain [F_(1,8) = 0.14, P = 0.72].

The V̇o₂ attained at task failure (PL: 3.12 ± 0.51 vs. BR: 3.09 ± 0.51 l/min) was not different between conditions or compared with the peak V̇o₂ obtained during the initial ramp incremental test (P > 0.66). Compared with PL, the exercise time to task failure was significantly increased during M→S following BR supplementation (PL: 521 ± 158 vs. 635 ± 258 s, P = 0.02). The time to task failure was greater in every participant after BR compared with PL (range = 3% to 54%; Fig. 4). The increased time to task failure was not correlated with the reduction in the V̇o₂ τ_p after BR compared with PL (r = 0.03, P = 0.95).