The interplay between bicarbonate kinetics and gastrointestinal upset on ergogenic potential after sodium bicarbonate intake: a randomized double-blind placebo-controlled trial

High-intensity functional training (HIFT) is derived from (1) well-studied high-intensity interval training (HIIT), in which relatively short bouts of high-intensity exercise separated by a recovery using low-intensity or inactivity are used as an alternative to traditional aerobic training to promote metabolic conditioning^1,2, and (2) strength training paradigms³. HIFT training stresses both aerobic and anaerobic energy pathways, and develops power, strength, flexibility, speed, endurance, agility and coordination¹. HIFT is executed at a high intensity and emphasizes functional, multi-joint movements via endurance, strength, power and speed stimulating exercises^2,4.

The majority of studies focusing on HIFT methodology have used a cross-training template^4,5. Cross-training’s specific modes of exercise (such as CrossFit’s programs) include power/Olympic lifting (i.e., squats, cleans, deadlifts, bench press, and push-presses), gymnastics (i.e., pull-ups, lunges, knees to elbows, handstand push-ups, push-ups, and sit-ups), and aerobic exercise/metabolic conditioning (i.e., swimming, running, and rowing). These high intensity exercise movements are often performed for specific lengths of time with little or no rest⁶.

Despite the growing popularity of HIFT and CrossFit training, limited interventional data exist on dietary and supplementation strategies to enhance performance in CrossFit^7,8. So far, there are only two studies investigating the effectiveness of sodium bicarbonate (SB)—which in fact is one of the most widely studied ergogenic aids—on CrossFit-like performance^9,10. According to the latest International Society of Sport Nutrition (ISSN) position stand¹¹, supplementation with SB (doses from 0.2 to 0.5 g·kg⁻¹ body mass < BM >) improves performance in muscular activities, various combat sports (e.g., boxing, judo, karate, taekwondo, wrestling), and in high-intensity cycling, running, swimming and rowing. The ergogenic effects of SB are mostly established for high-intensity exercise tasks that last between 30 s and 12 min¹¹. With the latter in mind, SB may exert great ergogenic potential in HIFT such as CrossFit. During short-term high-intensity exercise, the intramuscular ATP demands exceed the maximum rate of ATP resynthesis by mitochondria. In these circumstances, ATP production heavily relies on anaerobic systems—phosphorylcreatine (ATP-PCr) hydrolysis or glycolysis. Intense exercises that are more reliant on glycolysis (i.e., those lasting from ~ 30 s to ~ 5 min) results in greater H⁺ generation¹¹. Accumulation of H⁺ (leading to muscle acidosis and a decrease in muscle pH) has been shown to be an important contributor to fatigue due to its inhibitory effects on key glycolytic enzymes¹¹, depressing Ca²⁺ sensitivity, as well as direct effects on cross-bridge cycling¹¹. SB ingestion elicits ergogenic effect by increasing the concentration of blood bicarbonate, a buffer that contribute to maintain extra- and (indirectly) intracellular pH¹¹. In the stomach, SB rapidly reacts with hydrochloric acid to form NaCl, CO₂, and H₂O; excess bicarbonate that does not neutralize gastric acid rapidly empties into the small intestine and is absorbed¹². The reaction of NaHCO₃ and HCl in gastric lumen increases the pH of the gastric lumen, which stimulates gastric parietal cells to secrete more HCl into the gastric lumen, thus resulting in the extrusion of HCO₃¯ into the perigastric capillaries and eventually into the systemic circulation (one HCO₃¯ ion on each H⁺ ion used in the reaction in the gastric lumen). The surplus HCO₃¯ is eliminated via the kidney¹³ and carbon dioxide formed during this process is eliminated via the lungs¹⁴.

Numerous different protocols of SB supplementation have been investigated, however, so far only two protocols have been implemented in CrossFit-trained athletes^9,10. One⁹ was a 10 day progressive-dose regime implemented in male and female CrossFit practitioners. The protocol resulted in overall improvement in discipline-specific performance (as measured during Fight Gone Bed test) by ~ 6.1% and enhancement in selected indices of aerobic capacity (as measured during incremental cycling). The chronic progressive supplementation regime prevented the occurrence of gastrointestinal (GI) side effects. The second dosing protocol¹⁰, was acute supplementation of 0.3 g·kg⁻¹_BM of SB ingested in three doses with 10 min intervals between doses, with the last dose ingested 60 min before exercise. This protocol neither improved discipline specific performance (as measured during ‘Cindy’ workout), nor altered hemodynamic and perceptual parameters. Moreover, no GI side effects were observed.

Although the performance benefits with SB are well documented, data linking the kinetics of blood HCO₃^¯, GI occurrence and anaerobic performance after ingestion of high doses of SB in conjunction with carbohydrate (CHO) in HIFT-trained athletes is lacking. Thus, it was hypothesized that the ingestion of 0.4 g·kg⁻¹_BM of SB + 15 g CHO ninety min before repeated-bouts of highly intensive exercise would result in the elevation of blood HCO₃^¯ above the almost certain ergogenic threshold (> 6 mmol·L⁻¹)^15,16 and minimize decrements of performance in consecutive bouts of the WAnT in CrossFit-trained athletes. It was also assumed that the conjunction of SB with small dose of CHO would minimize GI side effects.

All the performed and presented results’ analyses of both primary and secondary outcomes covered 19 participants and the analyses were performed according to originally assigned groups.

In this study, mean peak blood bicarbonate (HCO₃^¯) occurred after 75 min following ingestion of SB + CHO, with 79% of participants peaking at this time point. There were clear decrements in performance during subsequent WAnTs, regardless of study condition. There was a significant main effect between conditions for PP, however it lost significance after including treatment order or gender in the models. Mean PP across the six WAnTs was higher with SB + CHO and PLA + CHO compared to CTRL, but there was no difference between SB + CHO and PLA + CHO. There were no associations between the level of elevation of blood HCO₃^¯ and WAnT performance. Severity of GI side effects were not significantly different between SB + CHO and PLA + CHO, but substantially higher compared to CTRL. However, there was an inverse interaction between GI side effects and performance indices (PP, AP, MP, PD) with SB + CHO, meaning GI worsened performance in WAnT.

In the current study after ingestion of 0.4 g·kg⁻¹_BM SB combined with 15 g CHO an effective and desirable plateau in an increased HCO₃^¯ concentration was achieved and maintained from − 30′_EX (60 min post-ingestion) until WAnT_1_POST, fluctuating between 31.4 and 33.0·mmol L⁻¹. There was a non-significant reduction in blood HCO₃^¯ between 75 and 90 min after SB + CHO (− 1.6 ± 1.7 mmol·L⁻¹) and PLA + CHO (− 1.6 ± 1.3 mmol·L⁻¹) ingestion (or after baseline blood draw < − 1.9 ± 1.6 mmol·L⁻¹ >) due to a standard 10 min warm-up (which took place between minutes 77 and 87). The reduction seems to be a physiological response to exercise, however, rarely taken into account with respect to its impact on buffering potential of ingested SB. The maintenance of increased level of blood HCO₃^¯ after the first bout of WAnT may arise from the fact that the metabolic perturbations elicited by maximal all-out exercises lasting less than 30 s are distinct from those occurring during intense exercise lasting from ~ 30 s to 5 min¹⁷.

Jones et al.¹⁵ observed that 75 min after ingestion of 0.3, 0.2 or 0.1 g·kg⁻¹_BM of SB, mean blood HCO₃^¯ was 31.7 ± 1.9, 30.1 ± 1.0, and 27.9 ± 1.3 mmol·L⁻¹, while the mean blood HCO₃^¯ peaks occurred at 120 (~ 32.4 mmol·L⁻¹), 90–105 (~ 30.5 mmol·L⁻¹) and 105 (~ 28.1 mmol·L⁻¹) min post-ingestion. In the study by Gough et al.¹⁸ the mean time to peak blood HCO₃^¯ after ingestion of 0.3 and 0.2 g·kg⁻¹_BM of SB ranged from 87 to 89 and 77 to 83 min. In general, it could be stated that the higher the dose of SB, the higher the peak HCO₃^¯ and the longer the time to reach this peak. Both the aforementioned studies^15,18 investigated resting kinetics of blood HCO₃^¯ after SB intake. However, little is known about the kinetics of blood HCO₃^¯ when doses higher than 0.3 g·kg⁻¹_BM are ingested or protocols in which SB ingestion was combined with the intake of small amounts of CHO. Protocols which tested acute single SB doses equal to 0.4 g·kg⁻¹_BM are relatively scarce^19–27, and unlike our protocol, most of them provided the acute dose of 0.4 g·kg⁻¹_BM split into smaller doses before exercise (e.g., 2 × 0.2 g·kg⁻¹_BM ingested 90 and 20 min^23,25 or 90 and 60 min²⁴ pre-exercise^23–25) or over 1 h period from starting until the end of SB ingestion¹⁹), and no information on blood HCO₃^¯ kinetics is given. It seems reasonable to verify, if higher doses of ingested SB would also result in a longer time of sustaining a desirable and effective blood HCO₃^¯ elevation. In these circumstances, peak blood HCO₃^¯ might not be perceived as a single time point, but as a time frame or ‘peak bicarbonate window’. Indeed, in this study, HCO₃^¯ was elevated and remained stable for a 30 min period. It has been previously shown that in a rested condition ingestion of 0.3 g·kg⁻¹_BM of SB resulted in HCO₃^¯ increase > 6 mmol∙L⁻¹ between 90 and 225 min²⁸. Regarding practical applications, the individualized ingestion timing might be more important when implementing lower doses of SB (e.g., ≤ 0.2 g·kg⁻¹_BM). It is also worth mentioning, that when the HCO₃^¯ load exceeds the amount of acid in the stomach, HCO₃^¯ enters the intestine and reaches the jejunum, where it can be absorbed¹¹. Thus, this phenomenon may partially explain the longer time to obtain peak HCO₃^¯ after consumption of high dose of SB compared to lower doses.

In the current study, SB was co-ingested with 15 g of CHO, which may have impacted the kinetics of blood HCO₃^¯. The addition of CHO was implemented based on previous literature, as a means to alleviate possible GI symptoms, and improving SB absorption. Carr et al.²⁹, compared 8 different protocols of SB ingestion (0.3 g·kg⁻¹_BM), which differed in the form of supplement (solution or capsule), ingestion period (30 or 60 min), volume of fluid (7 or 14 mL·kg⁻¹_BM) and SB intake with a meal (containing 1.5 g CHO·kg⁻¹_BM). The most effective protocol for eliciting increases in blood HCO₃^¯ was SB supplementation in the form of capsules in conjunction with a CHO-rich meal, and when participants ingested the supplement within 30 min from commencement to completion of ingestion. Peak blood HCO₃^¯ (~ 30.9 mmol·L⁻¹) was reached 150 min after ingestion, and the HCO₃^¯ elevation compared to baseline was ~ + 6.6 mmol·L⁻¹. The analogous model which, however, did not include co-ingestion of the meal, resulted in achieving earlier (~ 120 min) peak blood HCO₃^¯ after SB intake. Thus, the co-ingestion of SB with CHO may delay reaching the peak of blood HCO₃^¯, and as a result separates the moment of peak HCO₃^¯ (which presumably refers to highest ergogenic effect) and the moment of peak GI symptoms (which results in performance decrements). However, the absolute dose of CHO provided in the current study was equal for each participant (15 g) and when expressed in relation to BM, was substantially lower than in the study by Carr et al.²⁹. Thus, CHO impact on bicarbonate kinetics after ingestion of SB in the current study might be less pronounced compared to the study by Carr et al.²⁹, and additionally affected by the warm-up which preceded the test exercise. Furthermore, in the current study, the GI side effects were evaluated only at a single time point, which was 120 min after SB + CHO or PLA + CHO intake. In the study by Gough et al.¹⁸, the time to peak GI upset was reached 36 and 29 min after ingestion of 0.3 and 0.2 g·kg⁻¹_BM. However, Carr et al.²⁹, observed the greatest incidence of GI side effects 90 min after ingestion of 0.3 g·kg⁻¹_BM SB in a solution form. On the one hand, the great discrepancy with regard to peak GI upset after ingestion of the same dose of SB suggests the need for more precise investigation of that aspect in the future studies. On the other hand, due to its subjective nature GI symptoms are difficult to measure accurately.

It is also worth considering that not the absolute value of peak blood HCO₃^¯, but the actual increase (change) in blood bicarbonate from baseline to peak HCO₃^¯, or to the moment exercise is initiated, might be the crucial factor for extra-cellular buffering support and the ergogenic efficiency of SB. According to a recent meta-analysis¹⁶, greater effects in exercise were obtained when blood bicarbonate increases were medium (4–6 mmol·L⁻¹) and large (> 6 mmol·L⁻¹) compared to small (≤ 4 mmol·L⁻¹). In the current study, the mean increase in blood bicarbonate from baseline to blood HCO₃^¯ peak after SB + CHO ingestion was 7.8 ± 1.9 mmol·L⁻¹; 16% of participants reached an increase in blood HCO₃^¯ of 4–6 mmol·L⁻¹, which according to the previous literature corresponds to a zone of a potential ergogenic effect, and 84% of participants reached the increase of > 6 mmol·L⁻¹, which corresponds to an almost certain ergogenic effect. However, what seems to be more important, is the mean increase of blood bicarbonate at 0′_EX which was 6.1 ± 1.9 mmol·L⁻¹. In 9 participants (47.4%) the blood bicarbonate increase fell within the range of 4–6 mmol·L⁻¹ at 0′_EX (90 min post-ingestion) and in the remaining 10 participants (52.6%) it was > 6 mmol·L⁻¹. However, in the current study, in accordance with previous findings³⁰ no significant relationships between HCO₃^¯ increase and WAnTs performance were observed. The lack of simultaneous data on blood HCO₃^¯ time to peak and/or peak concentration increase and time to peak GI symptoms upset, could be perceive as the limitation of the current study.

Based on the results of the current study and assuming that the greatest ergogenic potential of SB is noted when blood HCO₃^¯ reach maximum and the elevation is > 6 mmol·L⁻¹, the implemented dosage of SB + CHO should have been ingested at least 75 min before the exercises started, or individualized ingestion times should be established based on individual blood HCO₃^¯ peak increases. However, it should also be emphasized, that the slight depletion in HCO₃^¯ concentration is unavoidable, as a consequence of the warm-up, which in turn is indispensable and essential to perform before staring competition or main training³¹. Furthermore, the exclusion of occurrence of GI side effects (impairing SB-induced exercise performance benefits) presumably requires extending the time between SB intake and exercise.

The ingestion of 0.4 g·kg⁻¹_BM of SB + 15 g CHO ninety min before exercise did not prevent decrements in performance during consecutive bouts of 15-s WAnTs. To our knowledge, this the first study to implement such a demanding exercise protocol, comprising of 6 bouts of the WAnT with 30-s rest between bouts. In previous studies on SB supplementation, single³², two^33–35, three^36–38 or four^39–41 bouts of 30-s WAnT were implemented. Similarly to our study, Zabala et al.³⁷ found no impact of 0.3 g·kg⁻¹_BM of SB 90 min before exercise on performance during 3 bouts of 30-s WAnT separated by 30-min or 15-min rest periods in elite cyclists. Only one study utilizing 4 bouts of the WAnT in judoists following acute SB supplementation (0.3 g·kg⁻¹_BM of SB ingested 120 min before exercise) showed significantly higher AP (3rd and 4th bouts) and PP (4th bout) of upper limb WAnTs³⁹. This SB supplementation timing could be entirely sufficient to avoid any GI disorders while effectively increasing the buffering capacity at the same time. This idea may be supported by the study by de Oliveira and colleagues⁴⁰, where 5-day supplementation with 0.5 g·kg⁻¹_BM of SB (split into four equal doses per day) resulted in the overall improvement (as measured by total mechanical work) in four bouts of upper limb WAnTs. The last dose of SB was ingested 4 h before the performance assessment. Similarly, Tobias et al.⁴¹ found that 7 days supplementation of 0.5 g·kg⁻¹_BM of SB (no data of the last dose intake time before exercise) increased total work done in 4 bouts of 30-s upper-body WAnTs in combat sport athletes.

The undeniable and urgent need for controlling and quantifying the impact of GI side effects on performance after SB ingestion has been raised in two recent extensive papers on SB efficacy^11,16. In general, the severity of GI side effects is dose-dependent, with higher doses resulting in more severe GI upset^11,21, while studies indicate a variability in time to peak of GI upset^18,29. Siegler et al.⁴² noted significant increases in the incidence of GI symptoms after ingestion of 0.3 g·kg⁻¹_BM of SB at 30, 60, 90, 120 and 150 min, but not 180 min post-ingestion compared to baseline. Simultaneously, the individual susceptibility to GI side effects and their severity should be taken into thoughtful consideration. In the current study, based on the IQR of total GI symptoms score (Table 2), large variability in occurrence and severity of GI upset between study participants was observed. Furthermore, there were some participants whose total GI symptoms score was higher (n = 2) or equal (n = 1) in CTRL compared to SB + CHO. This suggests that there may be some individuals who exhibit high levels of anxiety, which naturally predisposes them to GI distress in laboratory or field-based conditions, regardless of supplementation or familiarization procedures. Saunders and colleagues³⁰ found a significant overall performance improvement after SB ingestion only when excluding participants (4 individuals out of 21 participants) who experienced GI discomfort. In the current study, based on regression analysis, we found a significant adverse impact of GI symptoms severity on WAnT performance. Thus, the impact of GI distress must be taken into consideration when analyzing the ergogenic potential of SB, and particularly when protocol assuming implementation of (1) single (2) acute (3) high (≥ 0.4 g·kg⁻¹_BM) doses of SB. Highly-motivated elite athletes may overcome GI upset and still benefit from high doses of SB supplementation, though this should be experimentally confirmed.

One of the suggested strategies to minimize GI side effects after SB ingestion, is to ingest SB alongside a high-carbohydrate meal^11,29. In the current study we combined SB and PLA ingestion with a small CHO (15 g) dose, in an attempt to maximize tolerance to the relatively high SB dose. Though the implemented strategy did not avoid GI symptoms, adopted protocol did not intend to provide greater amounts of CHO, to avoid achieving ergogenic effects arising from consumption of CHO. The source of CHO was an orange juice (150 mL) added to the tested preparations just before intake. It must to be mentioned, that the organic acids present in the juice (e.g., citric, ascorbic or malic) were available to react with H₂O or NaHCO₃. These weak organic acids partially dissociate in water. They react with sodium bicarbonate, meaning sodium salts, water and carbon dioxide are produced. Thus, one the hand, the addition of orange juice might slightly decrease the overall buffering potential of the implemented dose of SB. On the other hand, it improved substantially the taste of the preparations, and eventually might lead to greater palatability of supplementation. Moreover, the addition of the juice did not prevent from obtaining effective increase in blood bicarbonate.

One of the limitations of the current study is evaluation of the incidence and severity of GI symptoms at one solitary time point. Knowing the results of the study, and the fact that the severity of GI symptoms negatively impacted the WAnTs, it would be reasonable to evaluate GI symptoms at additional time points e.g., pre-supplementation and during exercise. This approach might help to identify the time of peak GI symptoms, and in the future studies to establish time of SB ingestion according to individual bicarbonate kinetics and GI symptoms occurrence. Another limitation of the study is that mean peak increase in blood HCO₃^¯ (7.8 ± 1.9 mmol·L⁻¹) was achieved 15 min before the exercise test started. At the time point, right after 10 min standardized warm-up and before the exercise started (0′_EX), the mean increase of blood HCO₃^¯ was just slightly above the threshold of almost certain ergogenic effect (~ 6.1 mmol·L⁻¹). The strength of our study is a simultaneous comparison of blood bicarbonate kinetics and anaerobic performance. Most of the previous studies, which considered indices related to physical performance and high doses of SB (≥ 0.4 g·kg⁻¹_BM) controlled blood HCO₃^¯, lactate or pH (or other indices related to acid–base blood balance) solely at baseline, before and after exercise. In our study we insightfully controlled blood indices at 12 distinct time points. Thus, the current study must be perceived as highly innovative, and its undeniable strength is the ability to identify the peak blood HCO₃^¯, as well as peak increase (changes) in blood HCO₃^¯ and an increase (changes) in blood HCO₃^¯ directly before exercise was initiated, and to demonstrate there was a difference between these. Additionally, these blood collections clearly indicated the effect of warm-up causing a slight blood HCO₃^¯ depletion. The phenomenon is unavoidable, and has been raised up in one previous study⁴³, in which similarly to current results, a warm-up (regardless its intensity) led to decrease in blood bicarbonate, but did not affect the ergogenic potential of ingested SB. That seems to be valuable in the perspective of the future studies, to control if the time of starting the exercise and time of reaching peak increase in blood HCO₃^¯ overlap and/or are located in the assumed and desirable time gap (e.g. in case of longer-lasting exercises). It is possible that in some previous studies the time for the exercise tests to initiate following supplementation was not optimal from the point of view of peak HCO₃^¯ increases in blood, which may affect conclusions. To the best of our knowledge, this is the first study to investigate the kinetics of blood bicarbonate (and other acid–base blood indices) after ingestion of a single acute dose of 0.4 g·kg⁻¹_BM of SB. Finally, the undeniable strength of our study is also proving that to have a better insight into factors determining the buffering and ergogenic potential of SB, it is crucial to monitor blood HCO₃^¯ not only at baseline (pre-ingestion) and right before the exercise start, but also in between the ingestion and start of the exercise.

In this study in CrossFit trained athletes, the acute ingestion of single dose of 0.4 g·kg⁻¹_BM of sodium bicarbonate combined with a small amount of carbohydrate resulted in peak blood bicarbonate about 75 min after ingestion and significant increase in extra-cellular buffering capacity of plasma persisting from 60 until 90 min after intake. There was a significant negative impact of GI symptoms distress on mean indices of anaerobic performance of WAnT bouts. This may partially explain lack of improvement and/or lack of preserving from the decrements in performance in the six following bouts of the WAnT after SB ingestion. Thus, monitoring the time course of GI symptoms occurrence and severity post-SB intake seems to be crucial factors to control when implementing high, single, and acute doses of SB. Future studies should aim to investigate individualization of time of ingesting SB, combining the simultaneous elevation of blood bicarbonate referring to the zone of almost certain ergogenic action and lowest possible level of GI upset.

We thank the athletes for their participation and their coaches for help in conducting this research. We thank the company Sanum Polska sp. z o. o. for providing the alkalizing powder for the study.

K.D.-M.: conception, design of the work, and funding acquisition. K.D.-M., J.K., D.C., and T.P.: investigation and resources. K.D.-M., and P.M.N.: analysis, and interpretation of data for the work. K.D.-M., P.M.N., J.K., B.S., I.Ł., D.C., M.S., and T.P.: drafting of the work and revising it critically for important intellectual content. All authors approved the final version of the manuscript and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. All persons designated as authors qualify for authorship, and all those who qualify for authorship are listed.

This study was partially funded by the National Science Centre, Poland. K.D.-M has received research Grant from the Polish National Science Centre (Grant number 2018/02/X/NZ7/03217). Furthermore, K.D.-M. has participated in the Exchange Programme for Scientists as part of bilateral cooperation financed by The Polish National Agency for Academic Exchange (NAWA: BPN/BIL/2021/1/00108/U/00001). B.S. (2016/50438-0 and 2021/06836-0) acknowledges a personal research Grant from Fundação de Amparo à Pesquisa do Estado de São Paulo. B.S. also receives a Grant from Faculdade de Medicina da Universidade de São Paulo (2020.1.362.5.2).

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.

I.Ł. is a shareholder of an alkali products distributor and has received honoraria from the Sanum Polska sp. z o. o. However, the content of this study was not constrained by this fact in absolutely no extend. K.D.-M., P.M.N., J.K., B.S., D.C., M.S., and T.P. report no conflicts of interest. The results of the present study do not constitute endorsement by the Scientific Reports. The authors declare that the results of the study are presented clearly, honestly, and without fabrication, falsification, or inappropriate data manipulation.

The online version contains supplementary material available at 10.1038/s41598-023-34343-0.