The effect of β-alanine and NaHCO₃ co-ingestion on buffering capacity and exercise performance with high-intensity exercise in healthy males

Eight apparently healthy, recreationally active males (26.2 ± 1.9 year; 79.8 ± 2.11 kg; 179.0 ± 2.2 cm; VO_2peak 51.0 ± 2.5 ml kg⁻¹ min⁻¹) volunteered to take part in this study, which was approved by the Victoria University Human Research Ethics Committee (HRETH 11/11) and performed in accordance with the ethical standards set out in the 1964 Declaration of Helsinki. After completing a medical questionnaire, each participant signed informed consent forms and presented for preliminary testing.

Participants were asked to refrain from consuming caffeine and alcohol, and from undertaking strenuous exercise 24 h prior to all experimental trials. Participants recorded their dietary intake 24 h before the first day of the first experimental trial, and were asked to consume a similar dietary intake the day prior to all subsequent trials. Experimental trials were conducted in the morning, approximately 10–12 h after the last meal. A standardised warm-up period of 5 min of cycling at 80 W was performed on an Excalibur Lode Cycle ergometer (Netherlands) before each trial, followed by 5 min rest. All tests were followed by 60 min passive recovery in a supine position.

Peak oxygen consumption (VO_2peak) was determined approximately 1 week before the beginning of each chronic supplementation period. A standard graded exercise protocol on a Excalibur Lode Cycle ergometer (Netherlands) of 3 × 3 min sub-maximal workloads at 50, 100 and 150 W followed by successive 1-min workload increments of 25 W until volitional exhaustion. Participants were encouraged to maintain a pedal frequency between 80 and 100 RPM and the test was terminated when RPM could not be maintained >80 RPM for a period of 5 s. Expired air was directed by a Hans Rudolph valve via a ventilometer (Moxus; AEI Technologies, Pennsylvania, USA) into a mixing chamber and analysed for oxygen and carbon dioxide content. Prior to each VO_2peak test, the gas analyser was calibrated using commercially prepared gas mixtures (BOC Gases, Australia). Following preliminary testing, participants were familiarised with the exercise protocols in which they would be participating in during the experimental trials.

MRS was used in this study as a non-invasive quantification of human muscle carnosine before and after the β-alanine and placebo chronic supplementation periods in n = 5 participants (due to access and availability of MRS). These measurements were made on a whole body proton 3 tesla scanner (Siemens Trio Tim, Germany) equipped with a flexible knee coil at The Royal Children’s Hospital (Melbourne, Australia). Muscle carnosine concentration was quantified in the gastrocnemius and soleus muscles of the right leg throughout the study using a voxel of dimension 40 mm × 30 mm × 12 mm. Data were acquired using a standard PRESS sequence with an echo time of 30 ms, a relaxation delay of 2,000 ms and 176 transients. Muscle carnosine concentration was determined by comparing data recorded from the gastrocnemius and soleus muscles to a standard curve generated from external carnosine reference phantoms for absolute quantification. External carnosine reference phantoms contained 6.32, 30.00 and 58.93 mM carnosine (Sigma) with sodium azide (NaN₃) as an antibacterial agent. No correction was made for differences in relaxation time.

A line broadening of 1 Hz was applied to the data prior to Fourier transformation to reduce any noise which may have altered the accuracy of results. Manual phasing was then applied before line fitting the spectra of each scan with an AMARES algorithm (jMRUI, version 4.0, build 113) (Naressi et al. 2001; Vanhamme et al. 1997). Peak amplitude (arbitrary scale) and linewidth (lw) results attained through this algorithm were then used to determine the signal intensity (area of each peak) using a Lorentzian equation; (I, Intensity; w, 0.5 × lw at 1/2 maximum height; 2θ ₀, position of peak max):

Each spectrum contained two peaks, corresponding to the C₂–H and C₄–H imidazole protons of carnosine, and therefore the signal intensity of each peak was combined after adjusting to account for the number of protons in the structures giving rise to the peaks. A standard curve was then generated based on the concentration of each phantom scan and the signal intensity (area calculated from the phantom peak). The apparent concentration of carnosine from the gastrocnemius and soleus muscle scans was then determined with reference to the standard curve.

The dual supplementation, double-blind, crossover experimental design was randomised for both β-alanine vs. placebo and NaHCO₃ vs. placebo (Fig. 1) and all participants received the four supplement combinations. Specifically, there were two periods of 6-week chronic supplementation of capsulated β-alanine (Musashi, Australia) or the placebo calcium carbonate (CaCO₃) (E170, Melbourne Food Ingredient Depot, Australia) which were separated by a minimum washout period of 6 weeks (no supplementation). The β-alanine supplementation procedure consisted of 4.8 g day⁻¹ (6 × 800 mg) for 4 weeks and then 6.4 g day⁻¹ (8 × 800 mg) for 2 weeks. CaCO₃ was administered in equal mass amounts as β-alanine. Participants were asked to spread their consumption of daily doses of β-alanine/CaCO₃ evenly throughout the day to avoid any possible paraesthesia. To facilitate compliance, a supplement diary was provided to participants upon commencing each chronic supplementation bout.

To investigate the superimposition of NaHCO₃ with β-alanine, the acute administration of NaHCO₃ (Sodibic, Aspen Pharmacare, Australia) occurred following each of the 6-week periods of β-alanine and placebo supplementation. This required two trials of either 300 mg kgbw⁻¹ NaHCO₃ or placebo (CaCO₃) supplementation in a randomised crossover design. This acute supplement ingestion was administered 90 min prior to the exercise bouts of the respective trials and was split into 6 equal doses over the first 50 min of the 90-min pre-exercise period. Participants were asked to consume 300 ml of water with each dose to assist in avoiding potential gastrointestinal side effects. These trials were separated by 7 days, performed in a randomised order, and β-alanine or respective placebo supplementation was extended until the second acute supplementation protocol was completed. Ultimately, the following four supplementation combinations were generated: β-alanine + CaCO₃ (BAl-Pl), β-alanine + NaHCO₃ (BAl-SB), CaCO₃ + CaCO₃ (Pl–Pl), and CaCO₃ + NaHCO₃ (Pl-SB).

Participants were asked to complete 2 exercise tests, over consecutive days, at the end of each of the aforementioned four co-supplement periods (Fig. 1). These tests were designed to produce high acid loads whilst allowing examining the effect of supplementation on both intermittent and continuous HIE. The first test, a repeated sprint ability test (RSA), consisted of 5 repeats of 6 s maximal effort cycling bouts separated by 24 s rest (1:4 work-to-rest ratio). Peak and average power output were recorded after each sprint. This test was performed using a Wattbike Pro (British Cycling, Nottingham) and the same level of resistance was set for all trials (Air Brake 7.0). All sprints were performed in the standing position beginning with level pedals and the right leg forward, and with strong verbal encouragement. During the 24 s recovery intervals between sprints, participants were instructed to rest passively.

The second exercise test was a cycling capacity test (CCT_110 %) performed at 110 % of the workload achieved at VO_2peak prior to each chronic supplementation period. Participants were required to cycle in a seated position at 110 % of their peak power output (W _max) (335.9 ± 16.6 W) between 80 and 100 rpm for as long as possible. The clock was stopped at the point where their cadence could not be sustained above 80 rpm (time to exhaustion; TTE). To overcome the inertia of the electromagnetic resistance of initial exercise at low cadence, the protocol on the electrically loaded bike the beginning of the CCT_110 % was modified to begin with 15 s of 80 % W _max and 15 s of 90 % W _max before 110 % W _max was attained.

Blood was sampled from a vein in the antecubital space and the cannula was kept patent with isotonic saline (0.9 % NaCl, Pfizer). Blood was sampled at rest (pre- and post-NaHCO₃/placebo supplementation), immediately after exercise and during a 60 min post exercise recovery period. One portion of the blood was immediately placed into lithium heparin (BD Vacutainer) tubes and centrifuged at 12,000 rpm for 2 min. Plasma was decanted and stored at −80 °C for later analysis of lactate concentration (YSI 2300 STAT; Yellow Springs Instruments, Ohio, USA). Blood pH and HCO₃ ⁻ were analysed immediately using a Radiometer ABL800 FLEX Blood Gas Analyser (ABL800 FLEX; Radiometer Medical, Brønshøj, Denmark).

Results are expressed as mean ± standard error of the mean (SEM). Two-way analysis of variance (ANOVA) with repeated measures was performed (GraphPad Prism 6, Windows XP) one within group factor (time or sprint) and one between group factor (treatment), for blood, plasma and RSA data. Where there was an interaction between the factors, Tukey’s post hoc analysis was performed. One-way AVOVA with repeated measures was performed for CCT_110 % and muscle carnosine data (individual × treatment), with paired two-tailed t tests when interactions between factors were found. The level of probability required to reject the null hypothesis was set at p < 0.05.

The carnosine concentration was elevated in the gastrocnemius and soleus muscles following 6 weeks of supplementation of β-alanine (Fig. 2). In the gastrocnemius muscle (Fig. 2a), the absolute carnosine was 8.08 ± 0.68 and 8.73 ± 1.08 mM in the β-alanine and placebo groups, before supplementation, respectively, and increased by 5.03 ± 1.44 mM (+62 %, to 13.11 ± 1.97 mM; p = 0.03) after supplementation, whereas it remained stable at 7.42 ± 0.74 mM (p = 0.20) in the placebo group. Six weeks of chronic supplementation also caused the carnosine content of the soleus muscle (Fig. 2b) to increase by 4.92 ± 1.28 mM (+88 %, 5.57 ± 0.25 to 10.48 ± 1.35 mM; p = 0.02) before and after supplementation, respectively, in the β-alanine group whilst the carnosine concentration of the placebo group did not change (+7 %, 5.94 ± 0.56 to 6.33 ± 0.89 mM; p = 0.6). Carnosine concentrations in both muscle groups were also significantly elevated compared to that of the placebo (Pl) condition following 6 weeks of supplementation [gastrocnemius, +5.69 ± 2.08 (p = 0.05); soleus +4.15 ± 0.78 (p = 0.006)].

The washout protocol of carnosine from the muscle for participants who received β-alanine during the first chronic supplementation period (n = 3) was modified at the end of the experimental procedures, once the supplementation order randomisation was de-coded. This was done to ensure a minimum 12-week washout period upon cessation of β-alanine supplementation, as estimated from previous research (Baguet et al. 2009; Derave et al. 2007). In these participants, the post-supplementation sample in the second of the 6-week supplementation periods (placebo) was unmodified; however, values observed via MRS pre-β-alanine supplementation (initial MRS) were used for baseline data in these participants when examining the effect of chronic placebo supplementation. This was due to the second chronic supplementation period beginning during the 6 week of the washout period, and hence data collected here did not accurately reflected basal muscle carnosine concentrations. After a washout of 6 weeks, carnosine concentrations had decreased to 10.03 ± 1.34 mM (−19 %) in the gastrocnemius and 6.54 ± 0.60 mM (−23 %) in the soleus, at a rate of −0.3 mM week⁻¹ and −0.4 mM week⁻¹, respectively. Twelve weeks following cessation of β-alanine supplementation (end of second chronic supplementation period), muscle carnosine concentrations in both the gastrocnemius and soleus had returned to baseline concentrations, with the average rate of carnosine consistently decreasing in the gastrocnemius by 0.5 ± 0.03 mM week⁻¹ and in the soleus by 0.4 ± 0.02 mM week⁻¹. This information demonstrates that a washout period of 12 weeks was adequate to employ a randomised crossover experimental design, with an overall average rate of carnosine concentration decrease following the cessation of β-alanine supplementation.

There were significant increases in blood HCO₃ ⁻ between baseline and the post-supplementation period in both trials involving NaHCO₃ supplementation (Fig. 3, BAl-SB and Pl-SB) (average increase of 3.67 ± 0.46 mmol L⁻¹; p ≤ 0.0001). The placebo group showed no changes. The blood HCO₃ ⁻ in these groups was also significantly higher than that of the placebo condition (Pl–Pl) throughout both exercise bouts (p ≤ 0.01). No change was observed in blood pH between supplement groups or compared to baseline values prior to either exercise (Fig. 4). There was an increase in both blood HCO₃ ⁻ concentration and blood pH at the end of the recovery period following BAl-SB and Pl-SB compared to baseline values and the Pl–Pl condition in both exercise bouts (p ≤ 0.01), with the exception of BAl-SB blood HCO₃ ⁻ between baseline and 60 min following RSA (p = 0.07)). BAl-SB was the only treatment group to show significantly higher blood pH values (up to 0.07 ± 0.01 pH units; p ≤ 0.0001) throughout the first 10 min of recovery compared to Pl–Pl, in both exercise tests.

No differences in peak power output (PPO) or average power output (APO) were observed between supplement groups during each of the 56-s sprints of the RSA test (Table 1). There were also no differences in the total work done between either or combined supplement groups with a comparison of total work done (TW) (BAl-Pl, 22,136.25 ± 1,757.01 J; BAl-SB, 22,487.00 ± 1,724.30 J; Pl–Pl, 23,586.00 ± 1,831.58 J; Pl-SB, 23,068.50 ± 1,749.20 J).

A greater TTE by 18.13 ± 6.75 s (+14 %) (p = 0.005) was measured in the BAl-Pl compared to Pl–Pl (147 ± 13.16 and 129 ± 10.94, respectively) during the CCT_110 % (Fig. 5). Although the addition of NaHCO₃ to β-alanine further increased TTE by 2.75 ± 4.61 s (+2 %) (p = 0.02 compared to Pl–Pl), there was no statistical difference between BAl-Pl and BAl-SB (p = 0.57). No change in TTE was determined in Pl-SB treated group (132 ± 13.48 s)(+2 %) compared to the Pl–Pl condition (p = 0.51).

Plasma lactate concentrations (mmol L⁻¹) were elevated in all treatment groups following the completion of exercise in both the RSA (data not shown) and CCT_110 % (Fig. 6), and remained significantly elevated compared to baseline at the end of the recovery period. A difference was observed at the 5th min of the recovery period following the CCT_110 % protocol, with BAl-SB higher than Pl–Pl by 2.66 ± 1.20 mmol L⁻¹ (p = 0.04). Although the Pl-SB treated groups also showed an apparent increase in plasma lactate (+2.28 ± 1.61 mmol L⁻¹) following this test, this was non-significant (p = 0.18).