Blood Bicarbonate (HCO₃-) is part of the acid base homeostatic bicarbonate buffer system, which is critical in regulating blood pH concentrations and supporting metabolic functions. If blood pH is too alkaline a proton (H+) dissolves from carbonic acid (H2CO₃) forming HCO3-. On the contrary, HCO₃- binds a proton if blood pH is too acidic, resulting in H2CO3 again dissociating into water (H₂O) and carbon dioxide (CO₂). This leads to increased breathing rate in an attempt to exhale CO₂ and restore acid-base balance. The additional buffering capacity of exogenous HCO3- led to increased blood lactate after exercise, which is suspected to be due to greater efflux of H+ (Close et al., 2016; Peart et al., 2012; Felippe et al., 2016; Talbott et al., 1931). Also, adding sodium bicarbonate (NaHCO₃) appears to influence phosphocreatine degradation (PCr) and inorganic phosphate (Pi) accumulation [(PCr/ (Pi)], glycolytic intermediates (muscle protons and lactate), intra- and extracellular distribution of metabolites and other strong ions (Na+, K + and Cl) which appear to contribute to a performance enhancing effect (Siegler et al., 2016).

Consequently, in recent years NaHCO₃ has attracted a lot of attention as it was shown to significantly improve performance by up to 3% in swimming and cycling athletes (Carr et al., 2011a; Mueller et al., 2013; Mero et al., 2013). Yet, studies have shown that NaHCO₃ may not only be an ergogenic, but also an ergolytic substance (Deb et al., 2018). The reasoning behind this may lie within the mechanism of action, as the bicarbonate buffer system is not solely responsible for blood pH but is also vital in other systems, such as the stomach and duodenum by neutralizing gastric acid. As a sensitive contributor to those multifactorial processes, a potential limitation of performance is possible. Participants supplementing NaHCO₃ reported gastrointestinal (GI) upset including nausea, stomach pain, diarrhea and vomiting (Saunders et al., 2014; Carr et al., 2011b). On the one hand, dosage is suspected to have a large effect on ergogenic potential and GI distress, whereas 0.2 up to 0.4 g/kg body mass has prevailed as being tolerable (Burke, 2013; Lindh et al., 2008). On the other hand, the effects of type (fluid solution or capsule) and time of consumption are also unclear as results of acute (0.5 – 4 h prior exercise) or chronic loading of NaHCO₃ (several days prior exercise) led to diverse results (Joyce et al., 2012; Zajac et al., 2009).

Also, a large proportion of empirical research and systematic reviews have approached the complex topic of alkalizing ergogenic aids, especially NaHCO₃ (Close et al., 2016; Siegler et al., 2016; Burke, 2013). Yet, during exercise trials neither administering NaHCO₃ via fluid solutions, tablets or gelatin capsules, nor combining it with creatine (CR), beta-alanine (BA) or caffeine (CAF) led to the desired incontestable results (Pruscino et al., 2008; Kilding et al., 2012; Barber et al., 2013; Painelli et al., 2013). On that account, it can be suspected that there are more factors influencing the effect of NaHCO₃ supplementation such as training condition, athletes’ class (college, elite. Olympic etc.), gender or duration of exercise. A recent review summarized that NaHCO₃ is effective in trained athletes across a range of exercises such as supramaximal exercise, high intensity intermittent activity as well as skill-based sports (McNaughton et al., 2016). This review included also randomized controlled trials (RCT) which have tested NaHCO₃ via numerous diverse exercise tasks (e.g. bench press, judo throw test, upper body Wingate test, rugby sprint test) which lasted between 60 seconds up to 60 minutes (McNaughton et al., 2016).

In opposition to the aforementioned literature review, we would like to approach this topic differently by focusing less on the type of exercise tasks, but more on the exercise duration which we believe might be a major contributor to the effect on the performance of NaHCO₃. This has previously been a subject of investigation in a study by McNaughton in 1992 investigating the effect of anaerobic exercise over various durations (McNaughton, 1992). However, a major issue occurs when exercise trials are analyzed and interpreted regarding their metabolic impact (e.g. increase in lactate), because most trials conduct multiple exercise bouts and either investigate them individually or in sum. A good example for this is the 4x4 method which uses 4 min bouts of exercise corresponding to 90-95% of the maximum heart rate (HRmax) interspersed with 3 min of active recovery at 70% HRmax (Helgerud et al., 2007; Wisløff et al., 2007). This protocol is considered aerobic, yet other studies were unable to reproduce a lactate steady state (LaSS) when this protocol was applied (Tschakert et al., 2015; Hofmann and Tschakert, 2010). Since we did not assume that all studies included in this review would determine metabolic responses to exercise (e.g. lactate levels) we have decided to set the exercise duration cut-off at ≤ 4 minutes when prescribed exercise intensity was considered ‘High’ (Table 1). The reason behind this was to ensure that the previously shown mitigating effects of NaHCO₃ on metabolic acidosis (i.e. decline in pH) induced by predominantly anaerobic high-intensity exercise represents a potentially performance enhancing effect (Gough et al., 2017).

As such, the aim of this review is to assist in the quantification of how NaHCO₃ acts during ‘short’ and predominantly anaerobic exercise tasks providing practitioners, coaches and athletes a valid tool for implementing this supplement in their methods. Since various exercise tests were analyzed in this review, active and passive resting periods were also taken into account.

A systematic search of literature was conducted in the electronic database of PubMed. To examine the latest studies, the study publication time was restricted from December 2006 to December 2016. Due to additional valuable publications in this field, the search was updated in July 2018. Only RCT’s were included. The search terms used were “sodium bicarbonate” according to MeSH (Medical Subject Headings) AND “athlete AND performance OR exercise OR recovery”. Two authors (MLE; MH) performed the literature search independently; disagreements were discussed and solved with total consistency. The search process included screening titles, abstracts and eligible full texts. In addition, reference lists of excluded systematic reviews, reviews and included and excluded articles were manually screened for studies of relevance. A flowchart for screening and selection of studies is shown in Figure 1.

The inclusion criteria of the present study consisted of: (a) field or laboratory RCT’s that were either single- or double-blinded; (b) trials that included participants that were referred to as athletes or fulfilled the definition made by Araújo & Scharhag (Araújo and Scharhag, 2016); (c) all possible sports respecting both genders; (d) placebo-controlled designs involving NaHCO₃ or any combination of NaHCO₃ with other substances like beta alanine (BA), caffeine (CAF) or creatine (CR); (e) exercise/performance as a main outcome. We excluded studies that investigated sodium citrate or sodium pyruvate as a main outcome, did not investigate exercise related performance parameters or did not give a detailed explanation about participation in competitions or comparable training status of the participants.

Data extracted from each eligible full text were: the first author’s last name, publication year, study design, country, study duration, performance level and type of sport, subject information (sample size, sex, age and weight), supplementation (type, dose, time of ingestion), exercise performed (name, type, frequency) and outcome parameters including exercise duration. Overall, the results of the included studies are presented in different units or parameters. Therefore, to determine the effectiveness of NaHCO₃ we summarized the effect as “Yes”, “No” or “Yes/No”, based on the study results and significance values (p < 0.05).

The Cochrane Collaborations’ risk of bias assessment tool was used to evaluate the internal validity of the included RCTs (Higgins et al., 2011). Each included study was examined independently by the two authors (MLE, MH) by the defined sources of bias, which are: selection bias (sequence generation and allocation concealment), performance bias (blinding of participants and personnel), detection bias (blinding of outcome assessors), attrition bias (incomplete outcome data), selective reporting bias and other potential bias. Discrepancies were resolved by consensus.

Initial search was performed in December 2016 and updated in July 2018. In the end 775 studies could be identified. After screening titles, a total of 617 were excluded, because they did not include NaHCO₃ or were not a RCT. When screening abstracts, further 64 studies were excluded because of irrelevant outcome, animal study or no RCT. The assessment of the remaining 94 full text articles led to the exclusion of further 59 studies because of non-athletic population or no RCT. Consequently 35 articles were available for final evaluation.

Detailed information about study characteristics, results as well as task duration and supplementation can be found in Table 1, Table 2 and Table 3. The average sample size across all included studies was 15 athletes (range = 6-49; NTotal = 507). Overall athlete’s age ranged from 15 to 34 years.

The majority of studies used 0.3 g/kg body mass (BM) of NaHCO₃, whereas only three studies used 0.4 g/kg BM, three studies used 0.125 g/kg BM or less and one study used 0.5 g/kg BM. Delivery of NaHCO₃ was conducted differently throughout the studies. Most of the studies [n=21] used gelatin capsules for supplementation, whereby 12 studies supplemented with fluid solutions. Two studies used tablets.

Exercise tasks included in this review were rowing [n=5] (Christensen et al., 2014; Hobson et al., 2014; Hobson et al., 2013; Kupcis et al., 2012; Driller et al., 2013),sport specific trials [n=8] e.g. rugby sprint (Cameron et al., 2010), water polo (Tan et al., 2010), boxing (Siegler and Hirscher, 2010), basketball (Afman et al., 2014), judo (Felippe et al., 2016), taekwondo (Lopes-Silva et al., 2018), wrestling (Durkalec-Michalski et al., 2018), tennis (Wu et al., 2010), cycling tasks [n=6] (Mueller et al., 2013; Kilding et al., 2012; Egger et al., 2014; Thomas et al., 2016; Zabala et al., 2011; Zabala et al., 2008), swimming [n=7] (Mero et al., 2013; Lindh et al., 2008; Joyce et al., 2012; Zajac et al., 2009; Pruscino et al., 2008; Painelli et al., 2013; Siegler and Gleadall-Siddall, 2010), running [n=6] (Krustrup et al., 2015; Marriott et al., 2015; Stöggl et al., 2014; Ducker et al., 2013; McClung and Collins, 2007; Freis et al., 2017), upper body tasks (Tobias et al., 2013; Oliveira et al., 2017) and resistance training (Duncan et al., 2014). Regarding the duration of exercise, 14 trials lasted over four minutes and 17 trials lasted four or less than four minutes using acute supplementation, which means the supplement was taken in one to five equal doses, whereby the first dose was taken either 150, 120, 90, 75 or 60 minutes before exercise. These doses had to be consumed throughout 60 or 30 minutes, every 10 or every 15 minutes depending on the study design (Table 1 and Table 2).

Chronic supplementation was seen in one study including an exercise longer than four minutes and four studies with a trial duration equal to or less than four minutes. The duration of the chronic supplementation varied across the studies. The only study with a trial duration over four minutes ingested the supplement on two days per week over a period of eight weeks (Driller et al., 2013). One study with exercise duration below four minutes ingested the supplement three times per day over a period of three days (Joyce et al., 2012), while the other two studies expanded this protocol to four times a day over five (Oliveira et al., 2017) and seven (Tobias et al., 2013) days, respectively. The most recent study used a progressive protocol over 10 days starting with 0.025 g/kg BM on days one and two and increasing the dose until the last day to 0.1 g/kg BM to ensure no GI stress (Durkalec-Michalski et al., 2018). One of the 35 studies included in this review investigated acute and chronic supplementation at the same time and is therefore treated as two separate studies in further procedure (Joyce et al., 2012) (Table 1; Table 3).

Figure 2 is summarizing the results of the methodological quality assessment across all included studies. An appropriate procedure for a randomly generated sequence was described in all 35 RCTs. Furthermore 30 studies concealed the allocation. A high risk of bias was found in 2 trials regarding blinding of participants and personnel (Afman et al., 2014; Krustrup et al., 2015). One trial was unable to display complete outcome data (McClung and Collins, 2007), 13 out of 35 trials showed high risk of “other bias” not controlling for nutrition intake or GI adverse events during supplementation of NaHCO₃.

Seventeen out of 35 studies showed performance enhancing effects after supplementing NaHCO₃, of which 15 studies supplemented acutely and two chronically. Additionally, six studies showed diverse results by NaHCO₃, only increasing one of multiple exercise tasks or outcomes or NaHCO₃ only being effective combined with beta-alanine or caffeine.

Regarding the effects of acute supplementation of NaHCO₃ under the aspect of exercise duration, 9 studies under four minutes showed favorable effects. Exercise tasks included in these studies were cycling [n=2], swimming [n=4], running [n=2] and boxing [n=1]. In contrary, the results of five studies including swimmers [n=2], cyclist [n=2] and runners [n=1] showed no performance improvements. In addition, three studies demonstrated diverse results containing resistance, judo and swimming exercises (Table 1).

Furthermore, six studies over four minutes showed positive effects of acute NaHCO₃ supplementation. Involved trials were cycling [n=2], running [n=2], tennis [n=1] and taekwondo [n=1]. On the other side, six studies with exercise duration of more than four minutes showed no effects while supplementing NaHCO₃ acutely. Those were rowing [n=3], basketball [n=1], water polo [n=1], and rugby [n=1]. Two studies using a test lasting six and more than 40 minutes demonstrated diverse results by enhancing rowing performance only in combination with caffeine (Christensen et al., 2014) and significantly improving higher maximum running speed in exhaustive graded exercise while supplementing NaHCO₃ (Freis et al., 2017), respectively (Table 2).

Referring to chronic supplementation of NaHCO₃, two studies (exercise duration ≤ 4 minutes) showed positive effects. Both included specific upper body exercises (Oliveira et al., 2017; Tobias et al., 2013). In contrast, one study including swimming exercise less than four minutes showed no effect of NaHCO₃ (Joyce et al., 2012) and one study including wrestlers showed diverse results by only improving the time-to-peak power in the second Wingate test after NaHCO₃ supplementation, but not for the first test (Durkalec-Michalski et al., 2018).

In addition, another study involving rowing exercise with a duration over four minutes resulted in no effect as well (Tobias et al., 2013) (Table 3).

The purpose of this systematic review was to investigate the effects of acute or chronic supplementation of NaHCO₃ on exercise performances lasting less or more than four minutes in trained athletes. In general, nearly 50% of all RCT’s reported positive effects after supplementation. However, we cannot recommend the use of NaHCO₃ for enhancing performance in athletes because of the inconsistent and eventually ergolytic results throughout the studies, which are discussed below.

Out of the 35 studies included in this systematic review seventeen RCT’s were acutely supplementing with NaHCO₃ during an exercise duration of ≤ 4 minutes. Nine of these studies were clearly showing performance enhancing results after NaHCO₃ supplementation. In addition, three trials were showing divergent results even when using the same protocol (Table 1). In these studies NaHCO₃ resulted in enhanced performance after supplementation only for one exercise task (Painelli et al., 2013; Duncan et al., 2014) or only in combination with beta-alanine (Painelli et al., 2013) or caffeine (Felippe et al., 2016). While various exercises were tested, greatest results were shown during swimming freestyle (~2.4% performance increase) (Mero et al., 2013; Siegler and Gleadall-Siddall, 2010). Indeed, 2.4% appears small, yet represents 1.1 seconds on a world class level. In reference to the latest FINA world aquatics championships the 1st and the 8th place in the final race of the 100m freestyle were exactly separated by 1.1 seconds.

However, results from these studies must be interpreted carefully, since the participants in the presented studies were elite swimmers but not at a world class level. Thus, the results may not be translatable to top-level athletes and real competition conditions. In addition, it is not clear if NaHCO₃ is responsible for these performance improvements because of the inconsistent results throughout other studies even though the same protocols were used.

In contrast to the aforementioned successful results in swimming, two different trials with elite swimmers were unable to show significant improvements, although they used the same supplementation protocol in dosage and delivery of NaHCO₃ (Joyce et al., 2012; Pruscino et al., 2008). Furthermore, two other trials involving cycling showed no improvements (Zabala et al., 2011; Zabala et al., 2008). No difference regarding amount, ingestion, type or task of exercise was detected also. Hence, there is an assumption that the level of training (elite- world class athlete) might be influential on the outcome of detecting performance enhancing effects. An earlier systematic review showed that the possible effect of NaHCO₃ on exercise performance is lower in trained athletes (Peart et al., 2012). Consequently, it is possible that NaHCO₃ is just balancing the deficits between the athletes at different levels. Better trained athletes might have a higher buffering capacity, whereas the lesser trained are more dependent on the additional buffering capacity of the supplement. Therefore, an improved muscle buffering capacity might be more effective than NaHCO₃ administration (Peart et al., 2012). In contrast, a more recent review concluded that the benefits of NaHCO₃ might be present to a greater extent within trained individuals (McNaughton et al., 2016). This demonstrates the substantial controversy in the literature and the diverse effects of this supplement.

The included studies in this review involved primarily highly trained athletes, whereby their training level is difficult to compare, mostly because of the lacking information or different definitions of athletes in the studies.

Indeed, the huge difference in the studies could provoke the unequal outcomes, even if the same procedures are used. Many studies had a small and heterogeneous group including male and female athletes. Admittedly, the number of female athletes was small but may have influenced the results of the studies, as the potential effect of the menstrual cycle on performance was ignored in all studies involving female athletes (Janse de Jonge, 2003). Another possible influencing factor for the different results may be the presence of a ‘high-responders’ and ‘low-responders’ phenomena, which has been reported previously (Tobias et al., 2013; Wu et al., 2010). Furthermore, every study group used different exercises under different conditions which makes comparisons difficult. Another influencing factor causing differences between the studies is the combination of blinding and GI discomfort. In some studies, the blinding efficacy was not proven or could not be guaranteed because of GI distress. When the participants noticed stomach pain or other symptoms, they assumed that they had supplemented NaHCO₃. Consequently, this could have led to blinding bias and psychological effects on the outcome. These blinding biases can be seen in the risk of bias assessment (Figure 2). McClung and Collins showed that only believing one had taken the supplement resulted in similar effects as those taking the substance (McClung and Collins, 2007). Finally, all these factors influence the outcome of a study and should be considered carefully when interpreting results.

Out of 14 RCT’s six trials showed a performance enhancement after supplementation and two showed diverse results. Yet, two of these trials detected impressive results by improving running and cycling performance up to 14% (Krustrup et al., 2015) and 23% (Mueller et al., 2013), respectively. Despite these results, six trials amongst rowing, rugby, water polo and basketball did not show performance improvements. In comparison with exercise duration lasting less than four minutes, extended duration (> 4 minutes) appeared to be less successful in increasing performance via NaHCO₃, whereby some exceptional cases exist (Krustrup et al., 2015; Tobias et al., 2013; Wu et al., 2010). It appears that the best exercises for performance enhancing effects ought to be short to middle distance high intensity exercises up to a duration of four minutes e.g. swimming (200m), cycling (up to 4km) and running (up to 1500m). These results are in contrast to other authors in the literature claiming benefits of NaHCO₃ to be from one to seven minutes (Burke, 2013).

In the current review only two of five studies with an exercise duration of over four to seven minutes resulted in positive effects (Christensen et al., 2014; Lopes-Silva et al., 2018), whereby one of these studies showed significant effects by combining NaHCO₃ with caffeine (Christensen et al., 2014). Nevertheless, some cases showed that durations of over 10 (Egger et al., 2014; Mueller et al., 2013; Marriott et al., 2015) to over 50 minutes (Wu et al., 2010; Afman et al., 2014) could be affected by acute NaHCO₃ supplementation. Above all, the influence of exercise duration on the performance enhancing effect of NaHCO₃ supplementation remains unclear. Because of these inconsistencies in the literature the evidence of exercise duration having an influence on the positive effect of NaHCO₃ is not given.

The serial supplementation of NaHCO₃ is based on the assumption that chronic manipulation of pH-level may supply a protective effect on mitochondria, which is leading to improved mitochondrial function and consequently improved performance (Driller et al., 2013; Edge et al., 2006). These performance improvements could be seen in multiple events over the same day (Burke, 2013). Furthermore, chronic supplementation of NaHCO₃ should reduce the side effects of NaHCO₃ ingestion, e.g. GI distress, stomach pain and vomiting because of consuming small doses over several days before the competition (Burke, 2013; Durkalec-Michalski et al., 2018; McNaughton et al., 1999). Two out of five trials of chronic loading of NaHCO₃ improved performance and one showed diverse results. In concordance with other results, performance enhancing trials had exercise durations of less than four minutes. Paradoxically, the amounts of consumed NaHCO₃ varied massively. It seems that the right amount of NaHCO₃ for chronic supplementation is not fully elucidated by present literature. In the current review, the dose varied between 0.025 and 0.5 g/kg BM. Moreover, the time of ingestion and the duration of supplementation were very different. The dosage per day of NaHCO₃ varied between one and four times. Furthermore, the duration of supplementation lasted between three to ten days. Consequently, comparisons between these studies are limited.

In 2012 Joyce et al. could not find any improvements in acute nor chronic supplementation protocol in highly-trained swimmers (Joyce et al., 2012). The authors reported similar GI upset under both, acute and chronic supplementation. Consequently, which supplementation protocol (acute vs. chronic) is more effective, cannot be answered, as the results of supplementation with NaHCO₃ are ambiguous.

The included RCT’s displayed overall good quality. An inherent but unavoidable issue in trials involving supplements is a high risk of “other Bias”. Nutrition is influenced by multiple factors which might influence the effect of exogenous bicarbonate supplementation. In some trials blinding of personnel has not been explained in detail or has not been conducted at all. In addition, food intake before the measurement day was not always reported.

The definition of how the included studies were divided in ≤ 4 or > 4 minutes is an uncommon approach which might have influenced the overall analysis of this systematic review. Nevertheless, due to the general inconsistencies in the implementation of exercise studies and overall heterogeneity in training status of the included individuals between the studies we are convinced that this method was suitable to classify all studies accordingly.

The definition of athletes was based on the description of the participants (national, international, university, elite, college, highly trained or well trained) in the included studies following the definition by Araújo and Scharhag (2016). Unfortunately, the majority of all included studies gave insufficient information about competition or training amount to differentiate between athletes and recreationally trained participants, which might have biased the inclusion- and overall results of this review. Overall, the effect sizes of the included studies were very small and therefore the results should be interpreted carefully when transferred into daily praxis. Furthermore, more databases could have been included in the literature search to increase the numbers of the included studies.

Athletes and practitioners would benefit if future trials were to focus on elite-athletes to reduce potential day to day variability in performance compared to lower level athletes. Also, the shortcomings in research regarding evidence in disciplines lasting less than 20-25 seconds have to be overcome. Therefore, future trials including elite athletes investigating e.g. Olympic and non-Olympic sprint disciplines are mandatory. Accompanying sprint disciplines, resistance training trials involving maximum repetitions need additional research, as only one trial involving this kind of exercise within the last decade met the inclusion criteria of our search strategy. Finally, more research in acute vs. chronic supplementation should be conducted to find out a correct supplementation protocol.

The duration of exercise might be essential for the supplementation of NaHCO₃. However, this cannot be confirmed as the results in the literature are very inconsistent and difficult to compare as different disciplines and varying exercise protocols were applied in the included studies. In addition, results differed even if the same exercise and supplementation protocols were applied. For this reason, small sample sizes and lack of studies showing the exact effects of NaHCO₃, substantiated evidence is not given according to the included literature. Also, NaHCO₃-induced side effects might have had an ergolytic effect, as few participants showed a decrease in performance following supplementation. Chronic loading protocols may overcome these circumstances yet did not induce mentionable differences in comparison to acute supplementation (Table 1, Table 2 and Table 3 ). Consequently, it cannot be clarified if supplementation with NaHCO₃ results in performance enhancement or simply balancing deficits (buffering capacity) between athletes. Therefore, to which extent the duration of the exercise has on the effects of NaHCO₃ remains unclear.