The oral administration of L-arginine has been related to improved physical performance due to a likely reduction in muscle fatigue, resulting from the vasodilator effect of nitric oxide on skeletal muscle. However, there is no precise and quantitative analysis of the information in the literature. Objective: The main objective was to assess the effectiveness of L-arginine supplementation on physical performance in strength training with a systematic review and meta-analysis. Methods: Systematic review and meta-analysis, the PRISMA® (Preferred Reporting Items for Systematic Reviews and Meta-Analysis), PICOS statement and methodology guidelines, the quality of evidence by Classification of Assessment, Development and Assessment Recommendations (GRADE), Multiple Systematic Review Tool Assessment (AMSTAR2), and to assess the risk of bias of randomized clinical trials, the RoB 2.0 tool (Risk of Bias - Cochrane risk of bias tool revised for randomized trials). The inclusion process established articles with well-designed human experiments that included only L-arginine supplementation testing the effects of L-arginine supplementation on sports performance related to strength training; identical experimental conditions in placebo or control group; and publications in the last ten years (until December 31, 2020). Three studies were included that compared L-arginine supplementation with placebo in anaerobic performance tests. Results: There was no significant heterogeneity (p>0.05) in the analysis of the three selected articles and the effects of L-arginine supplementation in muscular endurance; performance had a mean of 0.26 (95% CI = -0.129; 0.649; p = 0.190), peak torque with a mean of -0.002 (95% CI = -0.531; 0.527; p = 0. 99) of the third series of exercises and, furthermore, when comparing the integrated effect (resistance rate with the peak torque) there was no difference with a mean of 0.168 (95% CI = -0.145; 0.481; p = 0.292).

Conclusions: Acute L-arginine supplementation provides no ergogenic effect on strength training performance.

This systematic review and meta-analysis were performed according to PRISMA® (Preferred Reporting Items for Systematic Reviews and Meta-Analysis) PICOS declaration guidelines and methodology, which defines the “Population,” “Intervention,” “Comparisons,” “Results,” and “Study Design.” The review research was carried out considering the following phases:

1. Population: Adults who practice physical activity (male and female);
2. Intervention: Impact of L-arginine supplementation by strength training practitioners;
3. Comparisons: Changes or alterations in the results in the same sample of participants, when there is a comparison of variables related to the use of L-arginine compared to placebo in strength training;
4. Outcomes: Any measures of physical performance and variables that relate to improvement in strength training (ex, peak strength, fatigue, number of repetitions, muscle performance);
5. Study design: Clinical trial.

The identification of the article selection process in the first phase was carried out in January 2021 in the Cochrane, PubMed, MEDLINE, LILACS, and SciELO databases. For the selection of keywords, the descriptors in Health Sciences-DeCS were used, obtained through advanced searches in the union of search terms, including the “Mesh.” The descriptors used were: “L-arginine”; “resistance exercises” and “strength training,” a cognitive logical operator “AND,” and “NOT,” were used for the effective tracking of publications.

The inclusion process established that articles with the following characteristics would be used: (1) well-planned experiments that included L-arginine supplementation only (without any additional compounds); (2) studies with English and Spanish languages[BN1] ; (3) identical experimental conditions in the placebo or control group; (4) test the effects of L-arginine supplementation on sports performance related to strength training; (5) clinical trial; (6) publications in the last ten years (until December 31, 2020); (7) investigations carried out on humans; (8) results related to the use of L-arginine; (09) clear information about funding sources; and (10) absence of conflict of interest.

Texts with the following characteristics were excluded: (1) L-arginine supplementation with some other compound; (2) investigations in which the topic was not essentially the use of L-arginine; (3) different samples of the effects of using the L-arginine not mentioning its effects (articles that do not show L-arginine results, in particular); (4) data from studies without Qualis status and surveys that did not show results; (5) other types of outcomes exclusively from reviews, pneumatic, contagious, cancerous and transmissible diseases.

Two authors independently screened and agreed with the selected studies for eligibility (CP, ROG, and TRL). Likewise, after the inclusion/exclusion criteria are applied to each study, the source study data (including authors and year of publication), sample size, participant characteristics (level, race, and gender), supplement administration (dose and time), and final results of interventions were extracted independently (CP and ROG) using an Excel spreadsheet (Microsoft Inc, Seattle). Then, possible disagreements were resolved through discussion until a consensus was reached, or by third-party judgment (MJK). In this sense, the Kappa coefficient, which indicates the reliability between raters, among the authors was above 90 with a level of agreement "almost perfect".

The Assessment of Multiple Systematic Reviews (AMSTAR2) tool was applied in the systematic review to assess the methodological quality of the studies. The reviewers searched the studies for data that answered each question, and each reviewer's final ratings were placed on a spreadsheet to discuss agreements and disagreements, with a score assigned to each assessed study. To assess the quality of the evidence and the strength of the recommendations, the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) system was used, in which the research question was structured considering the domains of the acronym PICOS.

According to the Cochrane Collaboration Guidelines, to assess the risk of bias of randomized clinical trials, the RoB 2.0 tool (Risk of Bias - Cochrane risk of bias tool revised for randomized trials) was used and evaluated by two authors independently, and disagreements were resolved by discussion and/or a third author. The list was separated into six different domains: selection bias (random sequence generation, allocation concealment); performance bias (blinding of participants and researchers); detection bias (masking the outcome assessment); friction bias (incomplete result data); reporting bias (selective reporting); and other sources of bias.

Domains were considered as 'low' if the criteria met a low risk of bias (probable bias that would not seriously alter the results) or 'high' if the criteria presented a high risk of bias (probable bias that seriously weakens confidence in the results), or it was considered "not clear" (a plausible bias that raises some doubts about the results), whether the risk of bias was unknown. Full details of each article and domain.

The Clinical Trials database was queried using the terms "descriptor" and "descriptors" to locate potential studies with L-arginine. Phase 2, 3, and 4 clinical studies enrolled in the aforementioned database were considered. For this review, approval by the ethics committee was not required due to the use of secondary data.

Statistical analyses

The meta-analysis was performed using the Comprehensive Meta-Analysis (CMA) software, version 3.3.070, including the three selected articles, one for each exercise series (first, second and third). The effect size was calculated based on the difference in performance between groups L-arginine and control group. We calculated the effect standardized by the standard deviation (SMD), since the variables ​​were not presented in the same unit in all studies, for example, total work of the series and total work presented in Joule, peak torque in Nanometer, number of maximum repetitions and rate of resistance).

Some studies presented performance measures for more than one muscle group, each subgroup of these studies was included in the analysis as if it were a different study. To test the effect of L-arginine with no sample overlap, combined analyzes with all subgroups from each study were also presented and included in the analysis as a single effect per study. Sensitivity analyzes were also performed for strength endurance measurements (total set work and total work, peak torque, maximum repetition number, and endurance rate) and peak torque (Nanometer). In the case of peak torque, as all studies presented data in the same measurement unit (Nanometer), we present the data in gross mean difference.

The inconsistency statistic (I²) was calculated as an indicator of the percentage of the total variation observed within the studies due to real heterogeneity rather than chance, to avoid errors when using the Q statistic in the evaluation of heterogeneity [14]. I² values ​​were included from 0 to 100%, representing a small amount of inconsistency between 25% and 50%, the average amount of heterogeneity between 50% and 75%, and a large amount of heterogeneity when the I² value was greater than 75 % [15] In this sense, low, moderate and high adjectives would be accepted referring to I2 values ​​of 25%, 50%, and 75%, respectively, although a restrictive categorization was not appropriate in all circumstances.

Publication bias was analyzed using the Egger test and a p-value ≤0.05 was considered significant. Data were presented as standardized mean difference and 95% confidence interval, and we considered significant p ≤ 0.05.