Background

Inorganic nitrate (NO3) is found in high abundance in various green leafy vegetables and roots. Although NO3 itself is thought to be relatively inert, it is converted via entero-salivary bacterial reduction in the oral cavity and gut to nitrite (NO2) [1, 2]. Subsequently, NO2 can be converted to nitric oxide (NO) in conditions of low oxygen tension and mildly acidic pH in a process facilitated by deoxyhemoglobin and enzymes such as xanthine oxidase [3]. Therefore, exercise at a relatively high-intensity results in conditions favorable for increased reduction of NO2 to NO. NO has been shown to play a potentially beneficial role in several physiological processes linked to exercise performance, including increased blood flow and O2 delivery [4, 5], increased microvascular PO2 [6], improved muscular contractility [7, 8], and reduced O2 cost of exercise [9,10,11] by enhancing mitochondrial function [12]. This suggests a potential for enhanced exercise training benefits, as NO may increase acute exercise responses which may in turn accumulate into an increased training response. Contrarily, these acute findings are equivocal, and it is possible that factors such as the decreased exercise-induced muscle perturbations found after NO3 supplementation [13] may result in lessened adaptation in response to exercise training.

The role of exogenous NO3 supplementation in human exercise performance has generated a great deal of academic interest in the last decade [14]. Several high-quality meta-analyses have been performed, and most recently an expert consensus paper using the modified Delphi technique concluded that acute and chronic NO3 supplementation was likely safe up to 16 mmol/day when consumed over several weeks, and that it is likely to produce ergogenic benefits during acute exercise in individuals with lower and more moderate aerobic fitness (i.e., those with VO2peak > 60 ml/kg/min have shown generally less benefit) [15]. Despite this, whether NO3 supplementation causes acute improvements in factors such as exercise endurance in each exercise training session, subsequently resulting in an accumulated larger training volume and greater adaptation for outcomes such as VO2peak, remains unclear.

VO2peak is considered the criterion measure of cardiorespiratory adaptation to exercise training [16, 17]. VO2peak improves with exercise training via multifaceted improvements in oxygen delivery (e.g., stroke volume, blood volume) and oxygen utilization (e.g., a-vO2 difference), and is a predictor of both endurance capacity and mortality/morbidity [18]. Although VO2peak primarily quantifies aerobic fitness, there is considerable variability in the endurance capacity of individuals with similar VO2peak [19]. More practical measures such as exercise time-to-exhaustion (TTE) may better represent competitive endurance performance (as this is likely to be more closely related to the percent of VO2peak associated with the lactate threshold and/or the lactate turn point) and offer additional insight into exercise training adaptations. As such, many exercise training studies evaluate changes to both VO2peak and TTE in a relatively consistent manner, making comparison across studies possible.

The purpose of this systematic review and meta-analysis is to explore whether NO3 supplementation can provide additional benefits when combined with chronic exercise training, and to determine whether factors such as baseline fitness, sex, health status, NO3 dosage, route of NO3 administration, and training conditions may moderate the effects of NO3 on training outcomes.

Methods

This systematic review and meta-analysis was performed in accordance with PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines [20] but was not pre-registered. Covidence (Covidence systematic review software, Veritas Health Innovation, Melbourne, Australia) was used for title, abstract, and full-text screening.

Literature Search

Electronic databases (PubMed, Medscape, and Web of Science) were searched by two authors in tandem, ACH and JOZ, with articles published from inception of the databases through June 2022 included. The search used the following terms: (((((("Dietary nitrate") OR ("Inorganic nitrate")) AND (Training)) NOT (Acute)) NOT (Rat)) NOT (rodent)) NOT (mouse)). Reference lists of all relevant studies along with reviews and book chapters were also examined. Articles were limited to randomized controlled trials (RCTs) in the English language.

Article Selection

For this meta-analysis, the term ‘article’ is used synonymously with ‘study’, and ‘trial’ is the unit included in the meta-analysis. Articles sometimes contained multiple eligible trials that comprised an intervention group and a comparable control.

First, the titles and abstracts of the articles were screened for eligibility. The following criteria were determined a priori for article inclusion: (1) the study was a RCT, (2) exercise training (i.e., repeated bouts of exercise multiple times per week) lasted at least 3 weeks, (3) the placebo and NO3 group received identical exercise training, (4) the placebo and NO3 had matched VO2peak at baseline. Full texts were reviewed of the remaining articles to determine eligibility. Two authors (ACH and JOZ) independently completed the study selection.

Data Extraction and Bias Assessment

Articles meeting inclusion criteria had the following data extracted and systematically organized: (1) author and publication year; (2) continuous variables: change in VO2peak, change in time to exhaustion, sample size, baseline VO2peak, age, body weight (kg), duration of training intervention, and NO3 dosage both in mmol and in mmol/kg/day; (3) categorical variables: sex, route of NO3 administration, environmental oxygenation during training, training intensity, health status, training modality, and whether an acute dose of NO3 was taken prior to the final post-training testing. If NO3 mmol/kg/day data were not provided, they were manually calculated based on mmol and mean kg data. If data were not available in the articles, authors were contacted for data.

Study quality was assessed using the Cochrane risk-of-bias tool for randomized trials (RoB 2) for both outcomes which includes the following domains: randomization, deviations from interventions, missing outcome data, measurement of outcome data, and results [21]. In each domain are signaling questions, where the risk of bias calculated from each domain is generated from an algorithm. Each study is scored as either “low risk”, “high risk” or “some concern” of bias based on the answers to the signaling questions. Two authors (ACH and JOZ) independently answered the signaling questions. Additionally, the Grading of Recommendations Assessment, Development and Evaluation (GRADE) using GRADEPro was performed to assess the certainty of evidence for the outcomes and sensitivity analyses. Certainty of evidence for each outcome was assessed using the following scale: “high,” “moderate,” “low,” or “very low” certainty. Certainty of each outcome was downgraded due to (1) risk of bias/study limitations, (2) inconsistency in results, (3) indirectness of results, (4) imprecision, or (5) reporting/publication bias. Two authors (ACH and CP) independently assessed the certainty of evidence.

Statistical Analysis

The meta-analysis and subgroup analysis was performed using Open Meta-Analyst, whereas funnel plots, Cook’s distance, studentized residuals, and sunset plots [22] were calculated and plotted using ‘metafor’ [23] in RStudio (Version 1.3.1073). Data were considered statistically significant a priori at p < 0.05 and data are presented as standardized mean difference (SMD).

A random effects model with restricted maximum likelihood estimation was utilized. The SMD of both VO2peak and TTE between the NO3 groups and the placebo groups of each trial were utilized in the model to determine the pooled effect. SMD was utilized due to inconsistent reporting of VO2peak in absolute or relative units. The SMDs are expressed as Hedges' g to account for any bias due to small sample sizes within trials. The Hedges' g values are interpreted as follows: ≤ 0.2, 0.2, 0.5, and 0.8 are considered to represent trivial, small, moderate, and large effect sizes, respectively [24].

The robustness of the pooled results was examined via funnel plots for small study effects. Sensitivity analysis was performed using Cook's distance and studentized residuals to identify potentially influential or outlying trials, respectively [25]. If a study was identified as being potentially influential or outlying, the robustness of the overall analysis was tested by removing the identified trial(s). A second sensitivity analysis was performed on only studies administering NO3 in the form of beetroot juice (BRJ) in healthy participants only. Statistical heterogeneity of the overall model was assessed with using Cochrane’s Q and I2, where < 25% indicates low risk of heterogeneity, 25–75% indicates moderate risk of heterogeneity, and > 75% indicates considerable risk of heterogeneity [26]. A sunset funnel plot was used to visualize the statistical power of each trial [27].

Moderator analysis was performed to determine whether the following subgroups impacted pooled effects: sex, health status, route of NO3 administration, and whether participants trained under normoxic or hypoxic conditions. While training intensity and modality were of interest, the included trials that had healthy populations all trained both at a high intensity while cycling, whereas all trials included that had clinical populations trained at a moderate intensity on a treadmill. Because of this, subgroup analysis for health status was the same subgroup for exercise intensity and modality. Meta-regressions were also performed to determine whether the following continuous variables impacted pool effects: age, weight, baseline VO2peak, NO3 dosage in mmol per day, and NO3 dosage in mmol/kg/day. The regression coefficient (ß) is reported along with 95% confidence intervals (CI).

Results

Literature Search

The PRISMA flow diagram outlining the literature search is presented in Fig. 1. In total, 144 references were imported for screening, with 47 duplicates removed leaving 97 studies screened against title and abstract. Following screening of titles and abstracts (ACH and JOZ), 84 studies were excluded leaving 13 studies assessed for full-text eligibility. Of these, 4 studies were excluded, leaving 9 studies [28,29,30,31,32,33,34,35,36] which met inclusion criteria containing 11 eligible trials. The trial participant and supplement characteristics are shown in Table 1, while the trial exercise training characteristics are shown in Table 2.

Fig. 1
figure 1

Flowchart of article and trial selection

Table 1 Trial participant and supplementation characteristics
Table 2 Trial exercise training characteristics

Risk of Bias

Both outcomes produced similar results on the RoB analysis. Of the 9 included studies, 7 were considered to have a low risk of bias, with 2 having some concern due to the nature of their single-blinded supplement design [33] or lack of explicit mentioning of double-blinding [31] (Additional file 1: Figure S1).

Pooled Effects

The overall model indicated that NO3 supplementation did not improve exercise training responses on VO2peak beyond exercise alone (n = 11 trials, SMD = 0.18; 95% confidence interval (CI): − 0.09, 0.44; p = 0.19; Fig. 2). There was no significant statistical heterogeneity present within this analysis (Q = 6.7, df = 10, p = 0.76, I2 = 0%, p = 0.76). Sensitivity analysis identified one trial [31] as potentially outlying/influential (Additional file 1: Figure S2), but removal of this trial had no significant effect on the observed effect. For the secondary outcome, the overall model indicated that NO3 supplementation did not improve exercise training responses on TTE beyond exercise alone (n = 9 trials, SMD = 0.08; 95% CI: − 0.21, 0.37; p = 0.58, Fig. 3). There was no significant statistical heterogeneity present within this analysis (Q = 3.7; df = 8; p = 0.88; I2 = 0%, p = 0.88).

Fig. 2
figure 2

Forest plot of VO2peak. [32a] Heart failure trial. [32b] Hypertension trial. [35a] Beetroot juice trial. [35b] Potassium nitrate trial

Fig. 3
figure 3

Forest Plot of Time to Exhaustion (TTE). [32a] Heart failure trial. [32b] Hypertension trial. [35a] Beetroot juice trial. [35b] Potassium nitrate trial

Outliers, Influence, and Power

Examination of studentized residuals for trials included in the VO2peak model were analyzed, revealing no indication of outliers in this model. One study was deemed influential according to Cook’s distance for VO2peak [31]. Removal of this study did not change the lack of significant effect of NO3 supplementation on training for the pooled outcome for VO2peak, although it did greatly reduce the effect size (SMD = 0.08; 95% CI: − 0.19, 0.36; p = 0.55). This study had large improvements in VO2peak in the NO3 group compared to the placebo group, but this study was identified as having “some concerns” in the risk of bias assessment due to the potential single-blinded nature of the study design. Visual inspection of the funnel plot for VO2peak also revealed that the one study that was detected as influential by Cook’s distance also had a large effect size (Additional file 1: Figure S2), although neither the rank correlation nor the regression test indicated any funnel plot asymmetry (p = 0.76 and p = 0.96, respectively). All other studies removed during sensitivity analysis had no substantial effect on the overall model.

Examination of studentized residuals for studies included in the TTE model were analyzed, revealing no indication of outliers in this model. According to Cook’s distance, none of the studies were considered influential. Visual inspection of the funnel plot for TTE did not reveal any asymmetry (Additional file 1: Figure S3), and neither the rank correlation nor the regression test indicated any funnel plot asymmetry (p = 0.76 and p = 0.85, respectively). All studies removed during sensitivity analysis had no substantial effect on the overall model.

Examination of the sunset funnel plot (Additional file 1: Figure S4) shows that at the pre-determined α = 0.05, the median power of all trials included in this meta-analysis was 6.7%, with an average effect size of 0.68 and 1.07 required for statistical power levels of 33% and 66%, respectively.

Subgroup Analysis

Moderator subgroup analyses for VO2peak and TTE are shown in Tables 3 and 4, respectively. Subgroup analyses showed no significant moderation of sex, health status, route of NO3 administration, level of training oxygenation, or the presence of an acute NO3 dose prior to post-testing on either VO2peak or TTE (all p > 0.05). Similarly, all meta-regressions performed did not reveal baseline VO2peak, age, body weight, duration of training intervention, NO3 dose, or NO3 dose normalized to bodyweight to be significant moderators on either VO2peak or TTE (Table 5; all p > 0.05).

Table 3 Moderator subgroup analysis for VO2peak
Table 4 Moderator subgroup analysis for TTE
Table 5 Meta-regression analysis for VO2peak and TTE

Sensitivity Analysis

As there is evidence in the literature suggesting that BRJ may provide more favorable benefits than other forms of NO3 supplementation, a separate model was performed including only healthy participants who consumed NO3 via BRJ (Additional file 1: Figure S5). Although this model revealed a trend for BRJ benefiting VO2peak improvement beyond exercise training alone (SMD = 0.35; 95% CI: − 0.04, 0.75; p = 0.08), removal of the study previously identified as being influential [31] eliminated the observed trend (SMD = 0.16; 95% CI: − 0.01, 0.88; p = 0.46). Moderator analysis on this subset of data did not reveal any significant moderators (Additional file 1: Table S1; all p > 0.05). The sensitivity analysis model performed with TTE as an outcome showed no significant effect of NO3 consumed as BRJ in healthy participants (SMD = 0.26; 95% CI: − 0.13, 0.65; p = 0.19) (Additional file 1: Figure S6). Moderator analysis on this subset did not reveal any significant moderators (Additional file 1: Table S2).

GRADE Assessment

Using the GRADE Assessment, all outcomes ranged from “low” to “high” certainty of evidence (Additional file 1: Figure S7). VO2peak for the overall pooled effects, as well as the sensitivity analysis for TTE in healthy participants consuming BRJ, were both downgraded to “moderate” certainty because of a potentially influential study and risk of publication bias. Sensitivity analysis on the outcome of VO2peak in healthy participants consuming BRJ was downgraded to “low” because of both a potentially influential study and because two studies (of five total) with the largest effect were not described as double-blinded.

Discussion

Overall model

The use of inorganic NO3 supplementation to improve physical function and exercise performance has increased over the last decade. To date, data demonstrating the benefits of acute and short-term NO3 supplementation on exercise performance are equivocal, but the consensus is that NO3 likely confers an overall small beneficial effect in individuals with low to moderate fitness. A summary of these data can be found in several meta-analyses and a recent expert consensus [15, 37,38,39,40]. Given that most athletes and clinical populations engaged in NO3 supplementation are likely involved in chronic exercise training/rehabilitation to improve function/performance, it is important to examine the outcomes from exercise training and NO3 supplementation in combination compared with training alone.

The results of this meta-analysis suggest that the addition of NO3 supplementation to exercise training does not enhance VO2peak or TTE beyond normal exercise training responses (Figs. 2 and 3, respectively), although it is important to note the limited number of studies and small sample sizes associated with studies in this area. As visualized in the sunset funnel plot (Additional file 1: Figure S4), existing studies are not sufficiently powered to detect the trivial-small effect observed in this meta-analysis. Further, while the present data suggest trivial improvements in VO2peak, these data are heavily influenced by a trial that was potentially single-blinded [31]. Removal of this study further weakened the effects of NO3 supplementation. The limited existing data suggest that any improvement in VO2peak and TTE observed after NO3 supplementation were likely trivial, and that more studies with longer training durations and with larger sample sizes are needed.

Subgroup Data

It has been hypothesized that NO3 supplementation may produce greater benefits in hypoxic or low oxygen conditions [41], as well as in individuals with lower VO2peak at baseline [42] who may have greater capacity to improve, in clinical populations with inhibited endogenous NO production via endothelial nitric oxide synthase (eNOS) [43], and in biological males (although this may be partly due to a lack of studies in females) [37]. Furthermore, evidence suggests that NO3 may be most effective when delivered in more moderate doses (both absolute mmol and mmol/kg) [44,45,46], and when delivered in the form of beetroot juice rather than another form such as a nitrate salt [47, 48]. Accordingly, we sought a priori to determine the potential influence of these variables on the exercise training responses to NO3 via a subgroup moderation. These results showed that there were no statistically significant moderators (Tables 2, 3, and 4).

General Discussion

While these moderator variables have been proposed to predict acute exercise responses to NO3 supplementation, it is possible that these acute changes are not large enough to result in an accumulated fitness benefit over time. Additionally, each moderator variable has considerations that may have impacted the lack of significant findings. For example, one moderator variable was baseline VO2peak, as those who are less aerobically fit initially have a lower VO2peak and may be more likely to have limitations in their baseline endogenous production of NO [49, 50]. Theoretically, these individuals will have greater responses to training and to NO3 supplementation, but this was not the case. For example, the study with the highest improvement in VO2peak following NO3 supplementation compared to control (∆ 5.5 vs. 3.0 ml/kg/min, respectively), was also the group with the highest baseline fitness (~ 60 mL/kg/min) [31], whereas the study group that experienced the largest difference in TTE following NO3 supplementation compared to control (∆ -2 vs. -95 s, respectively), had the second highest baseline fitness (~ 56 mL/kg/min) [33]. Further, the VO2peak improvements were low in the clinical populations, averaging only 0.55 ml/kg/min across all conditions, although this may be due to the more moderate exercise intensity used in these studies compared to the healthy population studies which all employed high intensity or sprint interval training [51].

While data have also suggested that moderate absolute doses of NO3 have generally shown similar benefits to larger doses in exercise responses [45], there is increasing scientific interest in the effects of NO3 dose normalized relative to body mass [44, 52]. Because of this, a moderator analysis was performed for dose of NO3 expressed both as mmol/day as well as mmol/kg/day, but neither were shown to be significant predictors of outcomes (Table 5). Finally, biological sex has also been proposed as a factor impacting the beneficial effects of NO3, suggesting a potential lack of impact in females compared to males [37, 53, 54], although there is a paucity of data examining the impact of NO3 in females. Indeed, none of the studies involved in this meta-analysis assessed only females, and those that included both sexes [32, 34,35,36] did not explore sex-differences in responses. Whether NO3 dose should be normalized to bodyweight has key implications in females due to the typically lower body mass compared to males and higher basal NO2 levels [55]. As there appears to be a possible hormesis response to NO3 supplementation in which an optimal dose outperforms a low or a high dose [46], it seems plausible that females are at risk of exceeding an optimal dose of NO3 if given the same absolute dose shown to benefit males. Thus, the normalization of the dose of NO3 to bodyweight, especially in females, merits further exploration.

Despite the lack of significant moderators, a trend was seen in improvement of VO2peak in the BRJ subgroup (p = 0.10; Table 3). This was perhaps due to the presence of biologically active polyphenols, antioxidants, etc. that are found in BRJ [56] that may facilitate conversion of NO2 to NO [57,58,59,60]. Indeed, BRJ has been shown to improve plasma NO3, submaximal VO2, and TTE all to a greater extent than equimolar sodium nitrate [47, 48]. While not reaching statistical significance, this meta-analysis suggests that BRJ may offer greater physiological outcomes than other forms of NO3 for improving the responses to exercise training. Further, trials included in this meta-analysis (all of which used BRJ) that included participants with clinical conditions had varying severity of disease (hypertension, peripheral arterial disease, and heart failure). These trials reported poor fitness improvements in the training interventions (NO3: 0.75 ml/kg/min vs. placebo: 0.93 ml/kg/min average improvements; data not shown). While this was potentially due to the maladaptive and continually worsening pathology of some of these diseases, inclusion of these clinical populations in the meta-analysis may have weakened an overall training effect of the combined interventions. This prompted an additional sensitivity analysis in which a meta-analysis was performed on only studies using BRJ in healthy participants. This analysis revealed a stronger trend for improvement with BRJ supplementation in VO2peak (p = 0.08; Additional file 1: Figure S5) compared with TTE (p = 0.19; Additional file 1: Figure S6). Further subgroup and meta-regression analysis revealed that there were no significant moderators (Additional file 1: Tables S1 and S2). While this subgroup involved only 5 trials (total n = 103), 1 of which was identified as being potentially influential (Fig. 2), this finding may suggest that NO3 when administered in the form of BRJ could potentially improve exercise training responses to VO2peak in healthy participants. While these data are underpowered as they stand, and any potential improvements in VO2peak appear to be small, this suggests that additional research is needed to determine if BRJ supplementation may serve as an ergogenic aid in healthy individuals. While many collegiate and professional athletes are taking supplemental NO3 (often via BRJ) in hopes of performance enhancement, existing data suggest these recommendations may be speculative and premature in nature.

Why chronic NO3 supplementation does not appear to clearly improve exercise training benefits despite acute benefits on exercise is unknown. It appears plausible that improvements of each acute exercise bout throughout a training intervention would accumulate into a greater improvement in responses. VO2peak is improved with exercise training by factors such as increases in stroke volume, arterio-venous oxygen extraction, and oxygen carrying capacity. Given the shorter duration of training in many of these studies, many training-induced improvements in VO2peak are attributed to improvements in blood volume and the associated improvement in venous return and stroke volume [61, 62]. To our knowledge, there are no studies examining the effects of NO3 supplementation on blood volume in humans. It is also possible that exercise training outcomes were not improved in this analysis because the adaptations induced by exercise training alone far exceed enhancements seen by simply supplementing with NO3. Ultimately, more studies are needed to address these gaps in knowledge.

Limitations

A major limitation of this meta-analysis is the low number of trials (n = 11) and the small sample sizes of trials included (average n = 20 per trial, 10 per group). The authors feel that this is important to point out as it highlights the need for more and larger studies addressing the ability of NO3 supplementation to augment exercise training responses. This also applies to the subgroup analyses performed in this meta-analysis, as the subgroups will have less statistical power than the entire model. Of the 9 studies, 11 total trials were included in the meta-analysis as two of the studies had multiple study populations as well as a placebo group (BRJ and potassium nitrate (KNO3) [35]; heart failure and hypertension [32]). We recognize that including multiple trials from one study may contribute to analytical issues such as “double counting.” Because the placebo groups were the only groups with trials used twice, and these groups experienced expected training responses, this appears unlikely to have impacted the outcomes. Further, the NO3 normalized to bodyweight data for each trial was calculated based on absolute values of NO3 supplemented and mean bodyweight for the study. Although this is meant for exploratory purposes, caution should be warranted interpreting these data as they present risk of error due to the method of calculation. This meta-analysis was not pre-registered, although all outcomes and subgroup analyses were determined a priori.

Additionally, the inclusion criteria included a minimum of 3 weeks of exercise training with an average training time of 5 weeks (only 1 study beyond 6 weeks [36]). This relatively short duration is less likely to induce large changes in VO2peak [63, 64] and thus may not be sufficient to provide an accurate representation of potential differences between treatments. On average, VO2peak improved 4.0% and 6.4% for placebo and NO3 supplemented groups after exercise training, respectively. It is possible that larger improvements as well as larger disparities between supplemented groups would be seen with longer supplementation and training time. Despite this, duration of training intervention was not a statistically significant moderator for any outcomes. The lack of additional exercise measures that are known to impact performance, such as lactate threshold, is another limitation to the results. However, large inconsistencies between studies (i.e., lactate samples taken during submaximal vs maximal workloads) did not allow for accurate analyses of these data. Additionally, NO3 combined with exercise training may have an impact on measures of cardiovascular health which was not explored in this meta-analysis.

Finally, a limitation not of this meta-analysis, but rather of the trials themselves, is that the majority of the studies that observed exercise training with NO3 supplementation also provided an acute dose of NO3 prior to post-training testing [31,32,33,34,35]. This makes it impossible to determine whether any differences in training responses stemmed from the chronic training alongside NO3, or whether the acute effects of NO3 were simply the cause of any observed differences. Indeed, subgroup analysis showed this to be a significant moderator in the BRJ and healthy subset of data (Additional file 1: Table S1), and studies which included acute supplementation overall showed a general trend for greater improvements. Future studies in this field should abstain from an acute NO3 dose prior to post-testing to determine a true training difference.

Conclusions

The results of this systematic review and meta-analysis suggest that, based on the limited data available, NO3 supplementation in addition to exercise training does not appear to improve VO2peak or time to exhaustion above and beyond that of exercise training alone. Additionally, there were no statistically significant moderators observed (i.e., sex, health status, training oxygenation, route of NO3 administration, baseline VO2peak, age, bodyweight, or NO3 dose). A subset of studies revealed a trend for improvement in VO2peak beyond exercise alone (p = 0.08) in healthy participants using BRJ as the mode of NO3 administration, although these improvements were greatly impacted by a trial that was deemed influential and concerning in terms of bias. Ultimately, more studies with longer training duration, larger sample sizes, and the addition of examining sex as a biological variable are needed to determine whether NO3 supplementation can improve exercise training responses, regarding VO2peak and TTE, compared to exercise training alone. Because of this, caution is warranted for individuals supplementing with NO3 in hopes of greater exercise training responses as the current data suggest trivial and non-significant improvements in these outcomes.