RECOVERY BETWEEN SETS IN STRENGTH TRAINING: SYSTEMATIC REVIEW AND META-ANALYSIS

Introduction: The recovery interval (RI) between sets and exercises has received attention from strength training (ST) researchers, to understand the relationship of rest on performance maintenance, especially the total load in a training session. It is known that each individual responds in a specific way to the training stimulus. So, what would be the effect of the different recovery interval strategies on the strength performance? Objective: Compare the different recovery intervals in strength training volume, considering the number of repetitions in healthy adults. Methods: We conducted a systematic review and meta-analysis based on methodological criteria, comparing fixed and self-selected RI on training volume, identified by the number of repetitions performed in a weight training program. Three electronic databases (Pubmed, VHL Virtual Health Library, Ebsco Sportdiscus) were analyzed, combining the expressions “resistance training”, “resistance exercise”, “strength exercise”, “recovery interval”, “rest interval”, “interval auto suggested”, “auto range selected” with “AND” and “OR” combination. Results: Pooled data from five studies showed a large significant effect in favor of the experimental group (>2 minutes) (MD: 1.24; 95%-CI [0.78; 1.71]; z: 5.25, Q:1.08; p < 0.01), since in the studies, recovery interval allowed a greater training volume. Conclusion: Longer RI seems be better, for maintaining total training volume, although there is no consensus for different training objectives against the self-selected RI. Thus, we imagine that this strategy may be important in the organizing a bodybuilding exercise program. Level of Evidence I


INTRODUCTION
Strength training (ST) has become one of the most popular physical activities in improving muscle strength, hypertrophy and power, 1,2 resulting in different health and performance benefits, such as improved body composition, improved performance in sports, strengthening of tendons 1 and can be used in cardiovascular and pulmonary rehabilitation programs or in the management of metabolic diseases, 3 as well as the effectiveness of ST in improving strength, hypertrophy. 4Skeletal muscle consists of muscle fibers, classified as Type I and Type II.Type I muscle fibers, or commonly known as red fibers, are slow to contract, generating small amounts of strength but with a long duration.Type II muscle fibers (IIA and IIB), or white fibers, of rapid contraction, with a high capacity in the production of strength, speed or power. 5These physiological responses to exercise are highly variable and depend on genetic predisposition.It can also be affected by sex, physical status, nutrition, type of exercise or protocol and training period. 6However, neuromuscular adaptations are maximized by manipulating ST variables, such as volume, intensity, training frequency, choice and order of exercises, execution speed, muscle actions, range of motion and RI. 2,7In the past two decades, the RI between the series has received a lot of attention from ST researchers. 8he RI between sets and exercises is an important variable in the acute ST program, 7,9 in addition to being used for different training purposes, directly affecting the number of repetitions during the progression of series, the total number of repetitions per exercise and the total repetition of the session. 10The fact is that intervals equal to or less than 1 min limit the recovery of creatine phosphate (CP) and ATP (Adenosine Triphosphate) reserves.It is estimated that the total recovery of ATP lasts, on average, from 3 to 5 min after strenuous exercise, while the CP for total recovery needs, on average, 8 min. 11Another important factor that can influence recovery between sets is the increase in lactate levels during intense ST. 4,12 The time required to decrease lactate after ST performed at high intensity should be between 4 and 10 min; times shorter than the aforementioned range lead to a high concentration of hydrogen ions (H +), decreasing the intracellular pH, resulting in muscle fatigue. 13he recommendations of the American College of Sports Medicine (ACMS), 2 involve times between 2 to 3 minutes in the IR between series in multiarticular exercises and between 1 to 2 minutes for monoarticular exercises. 2However, even with the recommendation in the definition of RI, we know that each individual responds to the training stimulus in a specific way.So, why have a fixed break time for different people?
Thus, there is a gap in the literature considering which would be the most recommended IR for each individual.However, in the last few years, a recovery perspective has emerged between series that considers the self suggested duration, which we call the Selected Auto Interval.In this case, the individual chooses the duration of the interval that will rest between sets and exercises, 14 which can favor the final performance in training. 15owever, we still do not know the self-selected IR imposes on the result of strength and hypertrophy associated with ST.In this way, we understand that it is necessary to analyze the literature in search of the results and effects of this new methodology to the ST, which may contribute to future studies and even to practitioners of this modality.Therefore, the objective of this review is to compare different recovery intervals in strength training in the training volume, identified by the number of repetitions in weight training in healthy adults.

METHODS
The systematic review was carried out according to the recommendations of Khan et al. 16 considering: 1) framing the questions for a literature review; 2) identify relevant research; 3) evaluate the quality of the studies; 4) summarize the evidence; 5) interpret the results.The research questions were defined by the PICOS model according to the PRISMA guidelines, as follows:

Database search method
During the period from March to July 2020, the records of 3 electronic databases were analyzed (Pubmed, Virtual Health Library BVS, Ebsco Sportdiscus).
The keywords were obtained using the query by PubMed "mesh terms".The research was conducted with the terms in English for: strength training, recovery interval and self-selected with combination "AND" and "OR".We follow the path: "resistance training Program "OR" Exercise Program, Weight-Bearing "OR" Exercise Programs, Weight-Bearing "OR" Weight Bearing Exercise Program "OR" Weight-Bearing Exercise Programs "AND" recovery periods "OR" interval "OR" recovery "OR "rest intervals" OR "rest periods" AND "self-selected" OR "self-selected" OR "self-selection" OR "self-selection" OR "rest self-selection".

Inclusion and exclusion criteria
The inclusion criteria for the articles were: 1) studies that compare different recovery intervals between the series; 2) used strength training as a training method; 3) individuals practicing strength training with a minimum experience of 1 year; 4) research published in a peerreviewed journal.The exclusion criteria were: 1) Studies that contained aerobic training as an intervention; 2) used less time of experience in strength training; 3) Articles that had women, elderly or adolescents in their sample; 4) articles that used supplementation in the intervention

Review process
The analysis and categorization of each article was carried out, separating the data in the excel spreadsheet following the order: names of the authors and year of publication, description of the sample, description of the intervention, results and conclusion.The first stage of the research resulted in 2789 articles and by reading only the titles and removing duplicate articles, 114 papers were selected.In this phase, all abstracts were read, observing the objectives, interventions with strength training and different interval times and experience in strength training.If the abstract did not provide these details, the article was separated for full reading.51 articles were selected in full, but only 18 articles entered for qualitative review and only 5 for quantitative.In addition, only 5 articles were found with a self-selected recovery interval.

Statistical analysis
Articles were selected that had the supine exercise as intervention and that used recovery intervals <2 minutes and> 2 minutes.They were grouped and these data were compared.The random effect of the metaanalysis was driven by a variable of the recovery interval of> 2 minutes (experimental group) and <2 minutes (control group) associated with training load.The present results were analyzed as differences in standardized means (DMP) ± standard deviation (SD) and 95% confidence interval (CI).Therefore, the effect of the recovery interval was determined by the DMP value and then calculated by the inverse of the variance. 17,18eterogeneity was estimated by the estimator (the DerSimonian-Laird estimator) and incorporated up to the standard error and estimated for the mean of the effect corresponding to the confidence interval.Heterogeneity was assessed using the Cochran Q test and the I2 index, which indicates the percentage of variance between studies, with corresponding cutoff points for low (0-25%), moderate (26-50%) and high (51 -100%). 19The funnel and cut and fill charts were used to assess publication bias using Egger regression tests in which non-significant asymmetry did not indicate bias. 20In addition, we conducted an adjusted cut and fill analysis 21 to remove the small studies from the positive side of the funnel graph, and recalculated the effect size (ES) in each iteration, until the funnel graph was symmetrical over the (new) ES.Finally, the fail-safe number of negative studies that would be needed to cancel (that is, make p> 0.05), the TE was calculated. 22All analyzes were performed using the meta package in version R 1.0.

RESULTS
The present study compared different recovery intervals in strength training in training volume, identified by the number of repetitions in weight training in healthy adults.The results of the meta analysis demonstrated a positive effect in the groups that used RI> 2 minutes, as this generated a greater number of repetitions and, consequently, a greater volume of training, which can stimulate physiological responses associated with signaling pathways of muscle hypertrophy, resulting in an increase of strength and muscle mass. 23,24These results are consistent with the study by Senna et al, 2016, 8 which compared several RIs between sets with the supine and crucifix exercise and found a greater number of repetitions in the 2-minute intervals (12.60 ± 2.35 repetitions; p = 0.027), 3 minutes (13.66 ± 1.84 repetitions; p = 0.001) and 5 minutes (12.93 ± 2.25 representatives; p = 0.001) vs. 1-minute protocol (10.33 ± 2.60 repetitions).From the results in the present study, it is clear that the longer RI promotes greater training volume.(Figure 1)

DISCUSSION
Considering that both the volume and the intensity of training are variables that are directly related to the stimulation of neuromuscular adaptations, 25 it was speculated that longer rests could provide greater volumes of training compared to 2 minutes, contrary to the data obtained in the review by GRGIC. 26Like AHTIAINEN, 23 who used RI of 2 and 5 minutes, but found no significant changes in the total training load during the training period.However, there was a significant increase of 6.8 ± 8.7% (from 3,370 ± 748 to 3,613 ± 949 N) (p, 0.05) in the maximum strength of the extensor chair in isometry in the total group of subjects.The IR between series is still a major unknown in the literature, as there is still no consent among the authors, and perhaps there will not be, on how long the individual will need to rest in order to obtain better results for hypertrophy or maximum strength.What is known is that the recovery interval guidelines presented by RATAMES, 2 indicated by the American College of Sports Medicine (ACMS) are between 2 to 3 minutes for multiarticular exercises and between 1 to 2 minutes for monoarticulars.In a study with monoarticular exercises, 27 short (30 seconds) and long (3 minutes) IR were compared under the hormonal responses of GH and cross section with high loads.The results showed that the group with short intervals demonstrated significant increases in GH (7704.20 ± 11833.49%,P <0 05) immediately after training.Regarding the cross section, there were significant increases in both groups [Short: 9.93 ± 4.86% (P <001), Long: 4.73 ± 3.01% (P <0 05)].For this study, the RI was not sufficient to enable muscle hypertrophy in these muscle groups.
Although the interval between series may not alter the hypertrophic conditions of individuals, it is possible that it contributes to the increase in maximum strength, as shown in the study by Villanueva. 28The author found no differences in body composition, but there were differences in the intensity of the 1-RM test in the groups that trained with intervals between 1 minute compared to 4 minutes in the leg press and bench press exercises.
MIRANDA; RAHIMI; TIBANA [29][30][31][32] compared different intervals in strength training, and like the studies already presented here, obtained similar results, with a higher number of repetitions and a greater total volume, less fatigue, as a response when the recovery interval was over 3 minutes, confirming the physiological hypothesis associated with the recovery of ATP concentrations, CP 11 that takes place between 3 and 5 minutes and the removal of lactate, after 4 minutes, after high intensity exercise. 13(Chart 2 and Figure 2) The studies in this review showed that short intervals promote a higher concentration of GH and lactate, which is associated with the hypertrophy process.On the other hand, lactate can be associated with reduced performance in strength training.Considering the longer intervals, a greater number of repetitions, less fatigue, greater training volume and improved performance were found.Among these results, there is only one question: What is the best recovery interval?Although it is a complex question, there is no consensus in the literature.Thus, this review aimed to compare fixed IR and self-selected RI, as an alternative that promotes autonomy and uses the practitioner's perception when choosing the RI that he deems necessary.
Five studies were found that used the RI Auto Selected methodology as a recovery strategy between series.De Salles 33 compared the selected auto with a fixed interval of 2 minutes and found no differences in the selected auto RI in relation to the total number of repetitions keeping the average and total training volume.On the other hand, Ibbott, 34 showed that the RI AS increased according to series progression, remaining close to 5 minutes (207.52s>277.71 s; p = 0.01).However, power (210 W; 8.03%) and speed (0.03 m.s-1; 6.73%) decreased as the sets progressed to all conditions (p <0.001), regardless of the IR used.
Similar to the study by De Salles, Sosciarelli and Polito 33,35 compared the self-selected RI with recovery ranges (1 to 2; 2 to 3 minutes).Statistical analysis showed that 1-minute RI improved the maintenance of the ideal pushing technique, evidenced by the reduction of falls in the vertical peak displacement bar (2.74%; p = 0.03), maximum bar speed (2.89% %; p = 0.03) and peak of the knee (1.61%; p = 0.03) and hip extensions (1.59%; p = 0.03) during the push movement phase.In addition, the 3 min IR led to the maintenance of the ideal elevation strategy, reducing the increase in horizontal displacement during the descending (3.85%; p = 0.04) and ascending (5.42%; o = 0 , 02).The total volume of training completed (sets x load x repetitions per set) for all exercises was significantly higher for the 3-minute resting condition versus the 1-minute resting condition (p <0.05;).Within each resting condition, there were significant differences in the repetitions completed for each series of exercises (p <0.05;).In addition, there were significant differences between resting conditions in completed repetitions for most sets of exercises (p <0.05)

minutes, self selected
There were no significant differences between a 2-min recovery interval and self-selected for total repetitions in all exercises.There were significant differences between the intervals for leg press, bench press and biceps curl (p <0.05).No significant differences were found between different RI for all exercises.For squats, the suggested RI results in a significant reduction in the number of repetitions from the first set to the second and third In general, the performance was better and the fatigue was lower in the IR 3 min, supporting all hypotheses, except Hypothesis 4. Overall, the performance was better and the fatigue was lower in the 3 min.IR, supporting all hypotheses, except Hypothesis 4. The 3-minute interval protocol showed higher values for power, training volumes, average speed and peak power .There were no significant differences in sEMG activity between the protocols.The efficiency (kg $ min 1) of P30 (633.6 ± 144) was significantly higher compared to P60 (397.4 ± 82.2), P90 (288.5 ± 58.4) and P120 (219.5 ± 38.5) significantly less total work in protocols P30 versus P60, P90 and P120 for all exercises, (b) lower total volume of training observed in P30 versus protocols P60, P90 and P120; and (c) a similar total volume was observed for the P60, P90 and P120 protocols.However, higher efficiency (TTV / time) was observed at P30 versus protocols P60, P90 and P120.The analysis of the results showed that the interval range from 1 to 2 minutes presented a lower number of repetitions (28.9 ± 4.7 min) only in relation to the self-selected RI (34.0 ± 7.2 min; P = 0.02), while the range of 2 to 3 minutes there were no significant differences (33.8 + 4.1).There was no difference in the subjective perception of effort.Ibbott, 34 compared two situations of self-selected RIs with high loads.It was found an average duration of RI of 283 ± 101 for session 01 and 249 ± 76 for session 02.Even in the case of the same individual, the responses to the same training condition were different.Only one individual was unable to conclude performing the 5 predetermined repetitions, possibly the RI was insufficient to recover the energy sources. 11he rest period increased significantly after series 3 and 4 in relation to series 1.There was no significant difference between the sessions for PSE.During the selection of studies that would be part of this review, we can notice the difference in the presentation of the data of the articles.This determined the choice of only 5 articles for the meta analysis, because in these studies the information was clearer with similar data that could be compared and analyzed.

CONCLUSION
The studies presented here have shown that the recovery interval between 3 and 4 minutes is effective between sets for training with high intensities.This applies to trained adults who are already familiar with the methodology.The benefits of the self-selected interval for strength training performance are not yet clear.Studies show that when individuals had this methodology available, they used times close to 3 and 4 minutes between sets.The subjective perception of effort can be a strategy to this recovery methodology, indicating which subjective intensity of the practitioner.
All authors declare no potential conflict of interest related to this article AUTHORS' CONTRIBUTIONS: SWJ: writing, revising and searching the databases, preparing the entire research project; BDS: data analysis and writing; CEC: statistical analysis and review; JGA: statistical analysis and review; AIP: writing and proofreading; JAF: writing, statistical analysis and intellectual concept.Only one participant was unable to complete the necessary repetitions until the end.The average duration of rest was 283 ± 101 for session 01 and 249 ± 76 for session 02.The average rest was shorter in AS2 than in AS1.The rest period increased significantly after series 3 and 4 in relation to series 1 There was no significant difference between sessions for PSE 4.4 -© 2009-2016 RStudio, Inc (The R Foundation for Statistical Computing, Vienna, Austria).An α level of p <0.05 was used to determine statistical significance.(Chart 1)

Figure 1 .
Figure 1.Flowchart shows the methodology used in the selection of studies for systematic review and meta-analysis.

Figure 2 .
Figure 2. Forest plot of studies used for meta analysis.

Chart 1 .
Methodological quality and strength of evidence for meta-analysis.
Chart 1. PEDro Scale scores of articles selected for systematic review.* Y = yes; N = no -Scores of 6 or more are considered indicative of high quality; scores below 6 considered to indicate low quality.-PEDro scale items: (a) eligibility criteria and source of participants; (b) random allocation; (c) hidden allocation; d) Baseline comparability; (e) blind subjects; (f ) blind therapists; (g) blind advisors; (h) adequate follow-up; (i) intention to treat; (j) comparisons between groups; k) point estimates and variability.

Chart 2 .
Data from articles selected for systematic review.
No significant changes occurred in body mass (from 83.9 ± 11.7 kg to 84.6 ± 12.9 kg) or body fat percentage (from 14.8 ± 3.9% to 15.3 ± 3.6 %) during the 6-month experimental training period in the total group of subjects.There were no statistically significant differences in the total training load.In 6 months, a significant increase of 6.8 ± 8.7% (from 3,370 6 748 to 3,613 6 949 N) (p, 0.05) was recorded in the maximum isometric extension of the leg strength in the total group of subjects