Q Routledge Taylor &. Francis Group Journal of Sports Sciences ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/rjsp20 Effects of high-intensity interval training in men soccer player's physical fitness: A systematic review with meta-analysis of randomized-controlled and non-controlled trials Filipe Manuel Clemente , Rodrigo Ramirez-Campillo , Fabio Yuzo Nakamura & Hugo Sarmento To cite this article: Filipe Manuel Clemente , Rodrigo Ramirez-Campillo , Fabio Yuzo Nakamura & Hugo Sarmento (2021): Effects of high-intensity interval training in men soccer player's physical fitness: A systematic review with meta-analysis of randomized-controlled and non-controlled trials, Journal of Sports Sciences To link to this article: https://doi.Org/10.1080/02640414.2020.1863644 View supplementary material G? m Published online: 11 Jan 2021. Submit your article to this journal G? B>1 View related articles G? CrossMark View Crossmark data G? Journal of Journal of 1 \ ft Sports Sciences K Full Terms & Conditions of access and use can be found at https://www.tandfonline. com/action/journal lnformation?journalCode=rjsp20 JOURNAL OF SPORTS SCIENCES 1^ ROUtlödge https://doi.Org/10.1080/02640414.2020.1863644 M m Taylor & Francis Group SPORTS PERFORMANCE A Check for updates Effects of high-intensity interval training in men soccer player's physical fitness: A systematic review with meta-analysis of randomized-controlled and non-controlled trials Filipe Manuel Clemente©ab, Rodrigo Ramirez-Campillo©c, Fabio Yuzo Nakamura©d and Hugo Sarmento©6 aEscola Superior Desporto E Lazer, Institute Politecnico De Viana Do Castelo, Rua Escola Industrial E Comercial De Nun'alvares, Viana Do Castelo, Portugal; blnstituto De Telecomunicacoes, Delegacao Da Covilha, Lisboa, Portugal; cHuman Performance Laboratory, Quality of Life and Wellness Research Group, Department of Physical Activity Sciences, Universidad De Los Lagos, Osorno, Chile; dUniversidade Federal Da Paraiba: Joao Pessoa, Associate Graduate Program in Physical Education UPE/UFPB, Joao Pessoa, Brazil; eResearch Unit for Sport and Physical Activity, Faculty of Sport Sciences and Physical Education, University of Coimbra, Coimbra, Portugal ABSTRACT This systematic review with meta-analysis (SRMA) was conducted to assess the effects of high-intensity interval training (HUT) programmes on men soccer players' aerobic fitness (maximal oxygen uptake and aerobic performance), repeated sprint ability (RSA), vertical jump height (VJH), and linear sprinting time (ST). An electronic search yielded 1,714 articles, 33 of which were included in the present study. Meta-analyses revealed significant benefits of HUT compared to controls in maximal oxygen uptake (p = 0.018), AP (p = 0.041), and RSA (p = 0.049). No significant effects were found in terms of ST (p = 0.080). The meta-analyses of non-controlled studies revealed significant improvements after HUT in maximal oxygen uptake (p = 0.001), AP (p = 0.007), RSA (p = 0.001), and ST (p < 0.001). However, no significant improvements in VHJ were found (p = 0.063). Furthermore, no significant differences were found in sub-group analysis (comparisons between HUT types). In conclusion, HUT is effective for improving maximal oxygen uptake, AP, and RSA regardless of the HUT type. For VHJ and ST outcomes, it seems reasonable to complement the HUT since it might not be enough to achieve significant changes. ARTICLE HISTORY Accepted 10 December 2020 KEYWORDS Football; athletic performance; interval training; sprint interval training Introduction Soccer is characterized by long periods of low- to moderate-intensity activity interspersed with periods of high-intensity actions, such as single and repeated sprints (Buchheit & Laursen, 2013a; laia et al., 2009). Additionally, explosive actions, such as accelerations, decelerations, quick changes of direction, and jumps, often occur in key moments of the match (e.g., scoring situations) (Faude et al, 2012). Therefore, a well-prepared soccer player must be able to endure long periods of low- to moderate-intensity activity (i.e., endurance performance) and sustain periods of explosive match effort intensification (Stolen et al, 2005). Those qualities will help to express performance and guarantee an appropriate physical fitness level for worst-case scenarios during a match (Fereday et al, 2020). In order to improve endurance as well as single and repeated explosive performance among soccer players, a time-efficient method (i.e., compared to conventional continuous running) generically known as high-intensity interval training (HUT) has been proposed (Buchheit & Rabbani, 2014). Five main HUT types have been described (M Buchheit & PBLaursen, 2013b): (i) short-interval HUT (<45 s of high but not all-out intensity exercise); (ii) long-interval HUT (2-4 min of high not-maximal intensity exercise); (iii) repeated sprint training (RST: <10 s repeated all-out [or nearly all-out] short sprint sequences, with short rest intervals); (iv) sprint interval training (SIT: >20-30 s of long all-out sprints, with longer rest intervals (e.g., 3-min)); and (v) game-based training (e.g., small-sided games [SSGs], usually using similar regimens to long-interval HUT) (M Buchheit & PBLaursen, 2013b). Due to the high degree of neuromuscular and metabolic stress imposed by HUT, meaningful changes in human skeletal muscle can occur - namely, mitochondrial protein content, muscle oxidative capacity, and the maximal activity of key enzymes can change (Fransson et al, 2018; Gibala et al, 2006). Moreover, HUT can improve fitness variables such as maximal oxygen uptake (V02max), aerobic capacity, sprint time (ST), vertical jump height (VHJ), and repeated sprint ability (RSA) (Girard et al, 2011; Toh et al, 2011). Given the variant properties of each HUT type, different HUT types can have different effects on soccer players' physical fitness. Some randomized-controlled studies have attempted to compare the effects of one or two HUT types (Dellal et al, 2012; Helgerud et al, 2001; FM Impellizzeri et al, 2008; Macpherson & Weston, 2015; Ouerghi et al, 2014). However, such studies are usually limited by small sample sizes (eight to 14 participants per group) (Dellal et al, 2012; Helgerud et al, 2001; FM Impellizzeri et al, 2008; Macpherson & Weston, 2015; Ouerghi et al, 2014). This limitation can be largely overcome using a systematic review and meta-analysis (SRMA) approach. Although some SRMA have examined the effects of HUT on players' physical fitness (Engel et al, 2018; J Taylor et al, 2015), CONTACT Filipe Manuel Clemente @ filipe.clemente5@gmail.com © Rua Escola Industrial E Comercial De Nun'alvares, Viana Do Castelo 4900-347, Portugal Q Supplemental data for this paper can be accessed on the publisher's website. © 2021 Informa UK Limited, trading as Taylor & Francis Group 2 @ F. MANUEL CLEMENTEETAL. such studies combined athletes from sports other than soccer or were focused on one specific physical fitness outcome (i.e., endurance) (Moran et al., 2019) and did not compare the effects of different HUT types (Gist et al., 2014; Kunz et al., 2019). Further, previous SRMA have focused only on youth soccer players (Kunz et al., 2019; Moran et al., 2019). Therefore, the effects of different HUT types on a comprehensive battery of the key physical fitness traits for adult and youth soccer players are yet to be determined. Therefore, the purpose of this SRMA was to assess the effects of different HUT types on soccer players' physical fitness (i.e., V02max, aerobic performance assessed from maximal field based-tests [AP], RSA, VHJ, and ST). 2. Methods The present SRMA followed the Cochrane Collaboration guidelines (Green & Higgins, 2005). The systematic review strategy was conducted according to PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analyses) guidelines (Moher et al., 2009). The protocol was registered with the International Platform of Registered Systematic Review and Meta-Analysis Protocols with the number INPLASY202060006 and the DOI number 10.37766/inplasy2020.6.0006. 2.1. Information sources A comprehensive computerized search of the following electronic databases was performed: (i) Web of Science; (ii) Scopus; (iii) SPORTdiscus; and (iv) PubMed. The searching process for relevant publications had no restriction regarding year of publication and included articles retrieved until 16 May 2020. The following search strings were employed: "soccer" OR "football" AND "high-intensity interval training" OR "HUT" OR "high-intensity intermittent training" OR "interval training" OR "small-sided games" OR "sprint interval training" OR "repeated sprint training" OR "speed endurance training". 2.2. Eligibility criteria Considering that a non-controlled pre-post study design is the most prevalent for HIIT-related research (J Taylor et al., 2015), this SRMA will focus on both randomized controlled trials and non-controlled trials. The a priori inclusion criteria for this review were as follows: (i) randomized-controlled (active control; passive control) or non-controlled trials (parallel studies) conducted in men soccer players with no restriction of age or competitive level; (ii) isolated (i.e., not combined with other methods) soccer-specific (i.e., running-based) HUT interventions with no restrictions for duration (e.g., short-intervals, long-intervals, small-sided games, speed endurance training, repeated sprint training and sprint interval training); (iii) a pre-post outcome for physical fitness, including V02max, AP, VJH, RSA or ST; (iv) original peer-reviewed articles written in English that provided full-text. Studies were excluded on the basis that they: (i) were observational studies; (ii) included other sports; (iii) used HUT combined with other training methods; (iv) used combined HUT types (e.g., running-based long-interval HUT combined with small-sided games); (v) used other than running-based HUT (e.g., cycling, boxing, rowing); (v) were review articles, letters to the editor, errata, invited commentaries or conference abstracts. A posteriori, studies were excluded if they reported the inclusion of women. This exclusion criterion was adopted considering the low number of studies found during a pilot literature search, and the potential confounding factor of sex on the HIIT-physical fitness interaction. 2.3 Data extraction A data extraction sheet conceived in Microsoft Excel (Microsoft Corporation, Readmon, WA, USA) was made based on Cochrane Consumers and Communication Review Group's data extraction template (Group CCCR, 2016). The sheet was used to assess inclusion requirements and subsequently tested on ten randomly selected studies (i.e., pilot testing). The process was conducted by two of the authors (FMC and HS). Any disagreement regarding study eligibility was resolved in a discussion between the authors. Full text articles excluded, with reasons, were recorded. All the records were stored in the sheet. 2.4. Data items The outcomes chosen for this SRMA included V02max, AP, RSA, VHJ and ST. Both direct (i.e., gas analysis in a graded exercise test until exhaustion, n = 11) and indirect (e.g., equations applied to field-based tests, n = 4) measures of V02max were considered, usually assessed as maximal oxygen uptake, peak oxygen uptake, and expressed as mL-kg~1 -min-1. For the case of indirect measures of V02max (equation estimates from AP), three of the studies (see Table 1) used the Yo-Yo intermittent recovery test and one study used the 12-min Cooper test. The AP included progressive tests until exhaustion (e.g., multistage tests) or time-based tests (e.g., maximum distance covered at a given predefined time) in which the measures of total distance covered (m), maximal velocity achieved (km-h~1) or maximal aerobic speed (m-s~1) were collected. The VHJ (measured in cm) was usually assessed during a countermovement jump (CMJ) with or without arm swing. The RSA was collected based on the mean time (s) or total time (s) in a series of multiple sprints. The linear ST (s) at different distances was also collected, without including values of partial times. Additionally, the following information was extracted from the included studies: (i) number of participants (n), age (years), competitive level (if available) and design of the study (randomized-controlled trial or non-controlled trial); (ii) HUT type (e.g., short-interval HUT; long-interval HUT; SSGs; RST; SIT) following the classification of a previous study (M Buchheit & PBLaursen, 2013b); (iii) period of intervention (number of weeks) and number of sessions per week (n/w); and (iv) regimen of intervention (work duration, work intensity, modality, relief duration, relief intensity, repetitions and series, between-set recovery) (M Buchheit & PBLaursen, 2013b). 2.5. Assessment of methodological quality The Physiotherapy Evidence Database (PEDro) scale was used to assess the methodological quality of the randomized Table 1. Characteristics of the included studies and outcomes extracted. Study Mean age (y) CL Design Outcomes Tests used in the original studies Measure extracted from the tests in the original studies (Arslan et al., 2020) 14.2 Y NC AP; RSA; VHJ; ST AP: YYIRT RSA: 6 x 40-m all out/20s rest VHJ: CMJ ST: linear sprint 30-m AP: distance (m) RSA: total time during the sprints (s) VHJ: (cm) ST: time (s) (Beato et al., 2019) 21.0 AS NC AP; RSA; ST AP: YYIRT RSA: 6 x 40-m all out/20s rest ST: linear sprint 20-m AP: distance (m) RSA: mean time over the sprints (s) ST: time (s) (Bravo et al., 2008) 17.3 and 24.3 Y and AS NC V02max; VHJ; ST V02max: Incremental treadmill test VHJ: CMJ ST: linear sprint 10-m V02max: mL-kg^-min-1 VHJ: (cm) ST: time (s) (Buchheit et al., 2010) 14.5 Y NC RSA; VHJ; ST RSA: 6 x 30-m all-out/20s rest VHJ: CMJ ST: linear sprint 30-m RSA: mean time over the sprints (s) VHJ: (cm) ST: time (s) Chtara et al., 2017) 13.6 Y C V02max; RSA; ST RSA: 6 X 30-m all-out/20s rest ST: linear sprint 30-m RSA: mean time over the sprints (s) ST: time (s) (Dellal et al., 2012) 26.3 AS c AP AP: 30-15|ft AP: VIFT (km-fr1) (Dello lacono et al., 0000) 18.6 Y NC AP; RSA; VHJ; ST AP: YYIRT RSA: 6 x 25-m/20s rest VHJ: CMJ ST: linear sprint 20-m AP: distance (m) RSA: mean time over the sprints (s) VHJ: (cm) ST: time (s) (Eniseler et al., 2017) 16.9 Y NC AP; RSA AP: YYIRT RSA: 6 x 40 m all-out/10s rest AP: distance (m) RSA: mean time over the sprints (s) (Faudeetal., 2013) 15.9 Y NC AP; VHJ AP: Incremental running protocol VHJ: CMJ AP: Vpeak (km-rT1) VHJ: (cm) (Faudeetal., 2014) 16.5 Y NC AP; VHJ; ST AP: Multistage endurance test VHJ: CMJ ST: linear sprint 30-m AP: Vpeak (km-rT1) VHJ: (cm) ST: time (s) (Harrison et al., 2015) 13.9 Y NC V02max; AP; VHJ; ST V02max: Incremental treadmill test AP: 30-15|FT VHJ: CMJ ST: linear sprint 20-m V02max: mL-kg^-min-1 AP: V|FT (km-rT1) VHJ: (cm) ST: time (s) (Helgerud et al., 2001) 18.1 Y C V02max; ST V02max: Incremental treadmill test V02max: mL-kg^-min-1 ST: linear sprint 40-m ST: time (s) (Hill-Haas et al., 2009) 14.6 Y NC V02max; AP; RSA; ST V02max: Incremental treadmill test AP: YYIRT RSA: 12 x 20-m all-out/20s rest ST: linear sprint 20-m V02max: mL-kg^-min-1 AP: distance (m) RSA: total time during the sprints (s) ST: time (s) (Hostrup et al., 2019) 22 SPS NC AP; ST AP: YYIRT ST: linear sprint 30-m AP: distance (m) ST: time (s) (laia et al., 2015) 18.5 Y NC AP; RSA; ST AP: YYIRT RSA: 15 x 40-m all-out/30s rest ST: linear sprint 20-m AP: distance (m) RSA: total time during the sprints (s) ST: time (s) (laia et al., 2017) 17.0 y NC AP; RSA; ST AP: YYIRT RSA: 15 x 40-m all out/30s rest ST: linear sprint 20-m AP: distance (m) RSA: total time during the sprints (s) ST: time (s) (F Impellizzeri et al., 2006) 17.2 Y NC V02max V02max: Incremental treadmill test V02max: mL-kg^-min-1 (FM Impellizzeri et al., 2008) 17.8 Y C V02max; AP V02max: Incremental treadmill test AP: YYIRT V02max: mL-kg^min-1 AP: distance (m) (Jastrzebski et al., 2014) 15.8 Y NC V02max; ST V02max: Incremental treadmill test ST: linear sprint 30-m V02max: mL-kg^-min 1 ST: time (s) (Kavaliauskas et al., 2017) 22 AS C V02max; AP V02max: YYIRT AP: YYIRT V02max: mL-kg^-min-1 (estimated) AP: distance (m) (Continued) Table 1. (Continued). Study Mean age (y) CL Design Outcomes Tests used in the original studies Measure extracted from the tests in the original studies (Los Arcos et al., 2015) 15.5 Y NC AP; VHJ AP: Universite de Montreal TrAPt Test VHJ: CMJ AP: MAS (km-hT1) VHJ: (cm) (Macpherson & Weston, 2015) 25 SPS C V02max; AP V02max: Incremental treadmill test AP: YYIRT V02max: mL-kg^-min-1 AP: distance (m) (Mohr & Krustrup, 2016) 19 U NC AP; RSA AP: YYIRT RSA: 5 X 30-m all-out/25 rest AP: distance (m) RSA: mean time over the sprints (s) (Mujika et al., 2009) 18.3 Y NC VHJ VHJ: CMJ VHJ: (cm) (Ouerghi et al., 2014) 22.9 AS C V02max; AP V02max: YYIRT AP: YYIRT V02max: mL-kg^-min-1 (estimated) AP: MAS (km-hT1) (Radziminski et al., 2013) 15.1 Y NC V02max V02max: Incremental treadmill test ST: linear sprint 30-m V02max: mL-kg^-min-1 ST: time (s) (Rey et al., 2019) 12.3 Y NC RSA; ST RSA: 6 X 25-m all-out/25s rest ST: linear sprint at 20-m RSA: mean time over the sprints (s) ST: time (s) (Safania et al., 2011) 15.7 Y NC V02max V02max: estimated at 12-minutes running test V02max: mL-kg^-min-1 (Sanchez-Sanchez et al., 2019) 14.9 Y C V02max; AP; RSA V02max: estimated at YYIRT AP: YYIRT RSA: 8 X 30-m all out/25s rest V02max: mL-kg^-min-1 AP: distance (m) RSA: mean time over the sprints (s) (Slettalokken & Ronnestad, 2014) 18-26 SPS NC V02max; AP V02max: Incremental treadmill test AP: 20-m shuttle run test V02max: mL-kg^-min-1 AP: distance (m) (Sperlich et al., 2011) 13.5 Y NC V02max; VHJ; ST V02max: Incremental treadmill test VHJ: CMJ ST: linear sprint 40-m V02max: mL-kg^-min-1 VHJ: (cm) ST: time (s) (JM Taylor et al., 2016) 24.1 SPS NC AP; VHJ; ST AP: YYIRT VHJ: CMJ ST: linear sprint 20-m AP: distance (m) VHJ: (cm) ST: time (s) (Tonnessen et al., 2011) 16.4 Y C AP; RSA; VHJ; ST AP: Incremental running protocol RSA: 10 X 40-m all-out/60s rest VHJ: CMJ ST: linear sprint 40-m AP: final level (n) RSA: mean time over the sprints (s) VHJ: (cm) ST: time (s) N: number of participants in the study; Y: years; CL: competitive level; Y: youth; PS: professional senior; SPS: semi-professional senior; AS: amateur senior; U: universitarian; V02: maximal oxygen uptake; AP: aerobic performance assessed from maximal field based-tests; RSA: repeated-sprint ability; VHJ: vertical height jump; ST: sprinting time at linear trajectories; *: control group without intervention; CMJ: countermovement jump; YYIRT: Yo-Yo intermittent recovery test; 30-15|FT: 30-15 Intermittent Fitness Test;1: long-interval HUT was tested with one (1) session week and one session two in two weeks (0.5);$: the SIT training (described in the original work as speed endurance training and speed endurance maintenance) tested two versions with different recovery periods; RSAmean: mean during repeated sprint ability tes; V!ft: maximal velocity at 30-15 IFT; Vpeak: peak velocity at Multistage endurance test; RSAtotal: total time during the RSA test; MAS: maximal aerobic speed; s: seconds; cm: centimetres; m: metres; JOURNAL OF SPORTS SCIENCES © 5 controlled trials included in this SRMA. The scale scores the internal study validity in a range of 0 (high risk of bias) to 10 (low risk of bias). Eleven items are measured in the scale. The criterion 1 is not included in the final score. Points for items 2 to 11 were only attributed when a criterion was clearly satisfied. Two of the authors (FMC and HS) independently scored the articles. Disagreements in the rating between both authors were resolved through discussion. Aiming to control the risk of bias between authors, the Kappa correlation test was used to analyse the agreement level for the included studies. An agreement level of k = 0.84 was obtained. In the case of the non-randomized trials, the methodological index for non-randomized studies (MINORS) was used (Slim et al, 2003). Twelve items were analysed, in which zero represented cases of no report, one case of report but inadequate, and two in cases of report and adequate. Two of the authors (FMC and HS) independently scored the articles. Disagreements in the rating between both authors were resolved through discussion. Aiming to control the risk of bias between authors, the Kappa correlation test was used to analyse the agreement level for the included studies. An agreement level of k = 0.78 was obtained. 2.6. Summary measures The analysis and interpretation of results in this SRMA were only conducted in the case of at least three study groups provided baseline and follow-up data for the same measure (Garcia-Hermoso et al, 2019; Moran et al, 2018; Skrede et al, 2019). Means and standard deviations for a measure (V02max; AP; RSA; VHJ; ST) of pre-post HUT interventions were converted to Hedges's g effect size (ES). The inverse variance random-effects model for meta-analyses was used because it allocates a proportionate weight to trials based on the size of their individual standard errors (Deeks et al, 2008) and enables analysis while accounting for heterogeneity across studies (Kontopantelis et al, 2013). The ESs were presented alongside 95% confidence intervals (CIs) and interpreted using the following thresholds (Hopkins et al, 2009): <0.2, trivial; 0.2-0.6, small; >0.6-1.2, moderate; >1.2-2.0, large; >2.0-4.0, very large; >4.0, extremely large. All analyses were carried out using the Comprehensive Meta-Analysis programme (version 2; Biostat, Englewood, NJ, USA). 2.7. Synthesis of results To estimate the degree of heterogeneity between the included studies, the percentage of total variation across the studies due to heterogeneity was used to calculate the I2 statistic (Higgins, 2003). Low, moderate and high levels of heterogeneity correspond to I2 values of <25%, 25-75%, and >75%, respectively (Higgins, 2003; Higgins & Thompson, 2002). 2.8. Risk of bias across studies The extended Egger's test (Egger et al, 1997) was used to assess the risk of bias across the studies. To adjust for publication bias, a sensitivity analysis was conducted using the trim and fill method (Duval & Tweedie, 2000), with L0 as the default estimator for the number of missing studies (Shi & Lin, 2019). 3. Results 3.1. Study identification and selection The searching of databases identified a total of 1,714 titles. These studies were then exported to reference manager software (EndNote™ X9, Clarivate Analytics, Philadelphia, PA, USA). Duplicates (502 references) were subsequently removed either automatically or manually. The remaining 1,212 articles were screened for their relevance based on titles and abstracts, resulting in the removal of a further 811 studies. The full texts of the remaining 391 articles were examined diligently; 315 were rejected, as they did not satisfy the relevant criteria (e.g., HUT interventions in soccer). Following the screening procedure, 78 articles were selected for in-depth reading and analysis. After reading full texts, further 45 studies were excluded due to a number of reasons (Figure 1). The 33 studies included in the meta-analysis provided mean and standard deviation for pre- and post-interventions data for at least one main outcome. 3.2. Study characteristics The characteristics of the 33 studies included in the metaanalysis can be found in Table 1. Additionally, the details of the HUT programs can be found in Table 2. The included randomized-controlled studies involved 11 individual experimental groups and 107 participants, and 78 participants in the 9 control groups. The included non-controlled studies involved 39 individual groups and 400 participants. 3.3. Methodological quality Using the PEDro checklist, one study scored 5 points and was classified as being of "moderate" quality, while eight studies achieved 6 points and were therefore considered as being of "high" methodological quality (Table 3). In the case of nonrandomized trial, eleven studies were between 16 and 17 points, 12 studies between 18 and 19 points and one study achieved 20 points (Table 4). 3.4. Effects of HUT on maximal oxygen uptake A summary of the included studies and results of V02max reported before and after HUT programs are provided in Table 5. Six controlled studies provided data for V02max, involving seven experimental and six control groups (pooled n = 122). There was a significant effect of HUT on V02max (ES = 0.57; 95% CI = 0.10 to 1.03; p = 0.018; I2 = 40.1%; Egger's test p= 0.103; Figure 2; a funnel plot for bias assessment is available as Supplementary Figure 1). The relative weight of each study in the analysis ranged from 12.1% to 16.8%. Nine non-controlled studies provided data for V02max, involving 14 experimental groups (pooled n = 148). There was a significant effect of HUT on V02max (ES = 0.66; 95% CI = 0.27 to 1.05; p = 0.001; I2 = 86.8%; Egger's test p= 0.001; Figure 3, bottom). The relative weight of each study in the analysis ranged from 2.4% to 8.1%. After the trim and fill method was applied, the adjusted values remained equal as the observed values (a funnel plot for bias assessment is available as 6 @ F. MANUEL CLEMENTE ET AL. Records identified through database searching (Web of Science, Scopus, Sportdiscus, PubMed) [n = 1,714) Records excluded (n = 315) Full-text articles assessed for eligibility (n = 78) Studies included in qualitative synthesis [n = 33) Studies included in quantitative synthesis (meta-analysis) (n = 33) Full-text articles excluded: Observational designs {n - 5) Just women or men and women in the same study {n - 3) Combined HIITand other training method {n -6) Combined two formats of HIIT {n - 4) Non-running based HIIT {n - 2) Other sport {n - 7) Not related with the topic {n - 14) No outcomes related {n - 4) Figure 1. PRISMA flow diagram highlighting the selection process for the studies included in the systematic review and meta-analysis. Supplementary Figure 2). No significant sub-group difference (between-group p = 0.776) was found between long-interval HIIT (six study groups; ES = 0.69; 95% CI = -0.08 to 1.46; within-group I2 = 91.1%) and SSGs (6 study groups; ES = 0.83; 95% CI = 0.23 to 1.43; within-group I2 = 86.3%). Due to reduced number of studies available for robust meta-analyses, one RST study group and one short-interval HIIT study group were not included in the between-group comparison. 3.5. Effects of HIIT on aerobic performance assessed from maximal field based-tests A summary of the included studies and results of AP reported before and after HIIT programs are provided in Table 6. Seven controlled studies provided data for AP, involving nine experimental and seven control groups (pooled n = 160). There was a significant effect of HIIT on AP (ES = 0.52; 95% CI = 0.02 to 1.02; p = 0.041; I2 = 53.9%; Egger's test p= 0.719; Figure 4; a funnel plot for bias assessment is available as Supplementary Figure 3). The relative weight of each study in the analysis ranged from 8.8% to 13.1%. Fifteen non-controlled studies provided data for AP, involving 25 experimental groups (pooled n = 250). There was a significant effect of HIIT on AP (ES = 0.35; 95% CI = 0.10 to 0.60; p = 0.007; I2 = 83.5%; Egger's test p= 0.497; Figure 5, bottom; a funnel plot for bias assessment is available as Supplementary Figure 4). The relative weight of each study in the analysis ranged from 2.5% to 4.6%. No significant subgroup difference (between-group p = 0.129) was found between long-interval HIIT (three study groups; ES = -0.56; 95% CI = -1.59 to 0.47; within-group I2 = 89.7%), RST (6 study groups; ES = 0.54; 95% CI = 0.28 to 0.81; within-group I2 = 49.4%), short-interval HIIT (3 study groups; ES = 0.78; 95% CI = -0.46 to 2.01; within-group I2 = 93.8%), SIT (4 study groups; ES = 0.19; 95% CI = -0.06 to 0.44; within-group I2 = 0.0%), and SSGs (9 study groups; ES = 0.43; 95% CI = -0.07 to 0.94; within-group I2 = 86.8%). 3.6. Effects of HIIT on repeated-sprint ability A summary of the included studies and results of RSA reported before and after HIIT programs are provided in Table 7. Three controlled studies provided data for RSA, involving four experimental and three control groups (pooled n = 71). There was a significant effect of HIIT on RSA (ES = 0.47; 95% CI = 0.00 to 0.93; p = 0.049; I2 = 0.0%; Egger's test p= 0.907; Figure 6; a funnel plot for bias assessment is available as Supplementary Figure 5). The relative weight of each study in the analysis ranged from 17.2% to 31.6%. Ten non-controlled studies provided data for RSA, involving 17 experimental groups (pooled n = 182). There was a significant effect of HIIT on RSA (ES = 0.45; 95% CI = 0.19 to 0.71; p = 0.001; I2 = 79.4%; Egger's test p= 0.022; Figure 7, bottom). The relative weight of each study in the analysis ranged from 4.5% to 6.7%. After the trim and fill method was applied, the adjusted values remained equal as the observed values (a funnel plot for bias assessment is available as Supplementary Figure 6). No significant sub-group difference (between-group p = 0.631) was found between RST (eight study groups; ES = 0.39; 95% CI = 0.20 to 0.57; within-group I2 = 35.2%), SIT (3 study groups; ES = 0.07; 95% CI = -0.66 to 0.80; within-group I2 = 82.1 %), and SSGs (5 study groups; ES = 0.57; 95% CI = -0.17 to 1.30; within-group I2 = 89.6%). Due to reduced number of studies available for robust meta-analyses, one short-interval Table 2. Characteristics of HUT programs in the included studies. Study HUT Duration d/ Total Work Relief format (w) w* sessions duration* Work intensity duration Recovery between Recovery between SSG Relief intensity Series* Reps* sets (duration) sets (intensity) formats SSG pitch dimension (length x width) SSG area per player (m2) (Dellaletal., 2012) siHNT 1-2 10-30S 95-100% VlFT 10-30S NR 7-10 5-6 min Passive (Faude et al., 2013) siHNT 11 2-3 22-33 15-30S 140-125% I AT 15-30S NR 2 12-15 10 min Light activity (Faude et al., 2014) siHNT 4 2 8 15s 140% IAT 15s NR 2 12-15 10 min NR (Jastrzebski et al., siHNT 8 2 16 15s 85-90% 15s NR 7 6 90s Active recovery 2014) HRmax (Ouerghi et al., siHNT 12 2 24 15s 105-120% 15s Passive 2 20 3 min NR 2014) MAS (Arslan et al., 2020) siHNT 5 2 10 15s 90-95% V|FT 15s Passive 2 12-20 NR NR (Bravo et al., 2008) NHNT 7 2 14 4 min 90-95% HRmax 3 min Light activity (60-70% HRmax) 4 (Helgerud et al., NHNT 8 2 16 4 min 90-95% 3 min 50-60% HRmax - 4 - - 2001) HRmax (FM Impellizzeri NHNT 4 2-3 10 4 min 90-95% 3 min Jogging - 4 - - et al., 2008) HRmax (Los Arcos et al., NHNT 6 2 12 4 min 90-95% 3 min 50-60% HRmax - 3 - _ 2015) HRmax (Radziminski et al., NHNT 8 2 16 4 min >90% HRmax 3 min Light activity - 5 - - 2013) (Safania et al., NHNT 6 3 18 4 min 70-95% 3 min NR - 4 - _ 2011) HRmax (Slettalokken & NHNT 6 1 6 4 min 87-97% NR NR - 5 - _ Ronnestad, HRmax 2014) (Slettalokken & NHNT 6 0-1 3 4 min 87-97% NR NR - 5 - _ Ronnestad, HRmax 2014) (Sperlich et al., NHNT 5 2-3 13 30s-240s 90-95% 30s- Passive - 4-12 - _ 2011) HRmax 180s (Beato et al., 2019) RSTe 2 3 6 30-m All-out 20s NR 3 7 4 min NR (Beato et al., 2019) RSTf 2 3 6 20 + 20-m All-out 20s NR 3 7 4 min NR (Bravo et al., 2008) RST 7 2 14 40-m All-out 20s Passive 3 6 4 min Passive (Buchheit et al., RST 10 1 10 15-to 20 m All-out 14s or Passive or active 2-3 5-6 NR NR 2010) 23s (~2 m-s-1) (Chtara et al., 2017) RST 6 2 12 20-30-m All-out 20s NR 2-4 5-6 4 min NR (Eniseler et al., RST 6 2 12 40-m All-out 20s Passive 3 6 4 min NR 2017) (laia et al., 2017) Rsr 5 2 10 30-m All-out 15s Passive 1-3 6 2 min NR (laia et al., 2017) RSTd 5 2 10 30-m All-out 30s Passive 1-3 6 2 min NR (Kavaliauskas et al., RST£ 6 2 12 10s All-out 60s Light activity - 10 - - 2017) (Rey et al., 2019) RSTa 6 1 6 15 to 30-m All-out 20s Passive 2-6 4-6 NR NR (Rey et al., 2019) RSTb 6 2 12 15 to 30-m All-out 20s Passive 2-6 4-6 NR NR (Sanchez-Sanchez RST9 8 2 16 18-m All-out 18-m Low-intensity 3 10 4 min Passive et al., 2019) running (Sanchez-Sanchez RSTh 8 2 16 18-m All-out 18-m Low-intensity 3 10 4 min Passive et al., 2019) running (Continued) Table 2. (Continued). SSG pitch dimension HUT Duration d/ Total Work Relief Recovery between Recovery between SSG (length x SSG area per Study format (w) w* sessions duration* Work intensity duration Relief intensity Series* Reps* sets (duration) sets (intensity) formats width) player (m2) (JM Taylor et al. RST 2 3 6 30-m All-out 20s Passive 3-4 7 4 min Passive _ _ _ 2016) (Tonnessen et al. RST 10 1 10 40-m 95-100% 90-120s NR 2-5 4-5 10 min NR - - 2011) Maximal Speed (Hostrup et al. SIT 10 3 30 6s All-out 54s Passive 2-3 6 4 min NR - - 2019) (laia et al, 2015) SIT 3 3 9 20s All-out 120s Passive 6-8 - - - - (laia et al, 2015) SIT 3 3 9 20s All-out 40s Passive 6-8 - - - - - (Macpherson & SIT5 2 3 6 30s All-out 4 min Light activity 4-6 - - - - - Weston, 2015) (Mohr & Krustrup, SIT 4 2 8 30s All-out 150s NR 8-10 - - - - 2016) (Mujika et al, 2009) SIT 7 1 6 30 m NR 90s NR 2-4 4 180s NR - - - (Arslan et al, 2020) SSGs 5 2 10 2.5-4.5 min NR 2 min Passive 2 2 NR NR 2vs.2 20 x 15-m 75 m2 (Dellal et al, 2012) SSGs 6 1-2 9 90 to 150s NR 1.5 to 2 min NR 5 - - 1vs.1 to 2vs.2 15 x 10 to 20 x 20-m 150 to 100 m2 (Dello lacono et al. SSGs 8 2 16 4 min 89.7% HRmax 2 min Passive 3-5 - - 5vs.5 42 x 30-m 126 m2 0000) (Eniseler et al. SSGs 6 2 12 3 min 90-95% 4 min Passive 4 - - 3vs.3 18 x 30-m 90 m2 2017) HRmax (Faude et al, 2014) SSGs 4 2 8 4 min NR 4 min Technical activities 4 - - 3vs.3 to 4vs.4 35 x 25-m to 40 x 30-m 145 to 150 m2 (Harrison et al. SSGs 6 2 12 8-11 min NR 3 min Passive 2 - - 3vs.3 NR NR 2015) (Hill-Haas et al. SSGs 7 2 14 6-13 min NR 1-3 min NR 2-6 - - 2vs.2 to 20 X 15 to 75-220 m2 2009) 7vs.7 60 X 40-m (F Impellizzeri et al. SSGs 12 2 24 4 min 91.3% HRmax 3 min 60-70% HRmax 4 - - 3 vs. 3 to 5 25 X 40 to - 2006) vs. 5 35 x 50-m (Jastrzebski et al. SSGs 8 2 16 3 min NR 90s Active recovery 7 - - 3vs.3 18 x 30-m 90 m2 2014) (Los Arcos et al. SSGs 6 2 12 4 min NR 3 min Passive 3 - - 3vs.3 to NR 85 m2 2015) 4vs.4 (Mohr & Krustrup, SSGs 4 2 8 45 s NR 45s NR 8-10 - - 2vs.2 20 x 20-m 100 m2 2016) (Radziminski et al. SSGs 8 2 16 4 min >90% HRmax 3 min Light activity 5 - - 3vs.3 or 3VS.3 + 1 18 X 30-m 2013)2 77-90 m2 (Safania et al. SSGs 6 3 18 4 min NR 3 min NR 4 - - 2vs.2 to 10 x 15 to NR 2011)_4vs.4 40 X 50-m_ HUT: type of high-intensity interval training; siHNT: short-interval HUT; liHNT: long-interval HUT; RST: repeated sprint training; SIT: sprint interval training; SSGs: small-sided games; W: weeks; d/w: days per week; MAS: maximal aerobic speed; s: seconds; NR: not reported; *: the range between the programs; min: minutes; m: metres; V|FT: maximal velocity at 30-15|FT; I AT: individual anaerobic threshold;£: uphill sprint training (gradient of 7%);$: only the first two-weeks of intervention were considered (SIT development); MAS: maximal aerobic speed; HRmax: maximal heart rate; Passive: passive recovery; s: seconds; min: minutes; m: metres;a: one session week; b: two-session week;c: short rest; d: long rest;e: straight sprint;f: with change of direction; 9: group above 48 mL-kg^-min-1 of V02max;h: group below 48 mL-kg^-min-1 of V02max JOURNAL OF SPORTS SCIENCES © Table 3. Physiotherapy Evidence Database (PEDro) scale ratings. N.°1* N.°2 N.°3 N.°4 N.°5 N.°6 N.°7 N.°8 N.°9 N.°10 N.°11 Total** (Dellal et al., 2012) 0 1 0 1 0 0 0 1 1 1 1 6 (Chtara et al., 2017) 0 1 0 1 0 0 0 1 1 1 1 6 (Helgerud et al., 2001) 1 1 0 1 0 0 0 1 1 1 1 6 (FM Impellizzeri et al., 2008) 0 1 0 1 0 0 0 1 1 1 1 6 (Kavaliauskas et al., 2017) 0 1 0 0 0 0 0 1 1 1 1 5 (Macpherson & Weston, 2015) 1 1 0 1 0 0 0 1 1 1 1 6 (Ouerghi et al., 2014) 0 1 0 1 0 0 0 1 1 1 1 6 (Sanchez-Sanchez et al., 2019) 1 1 0 1 0 0 0 1 1 1 1 6 Tonnessen et al (Tonnessen et al., 2011) 1 1 0 1 0 0 0 1 1 1 1 6 *: PEDRro scale items number; **: the total number of points from a possible maximal of 10; N.°1: eligibility criteria were specified; N.°2: subjects were randomly allocated to groups; N.°3: allocation was concealed; N.°4: the groups were similar at baseline regarding the most important prognostic indicators; N.°5: there was blinding of all subjects; N.°6: there was blinding of all therapists who administered the therapy; N.°7: there was blinding of all assessors who measured at least one key outcome; N.°8: measures of at least one key outcome were obtained from more than 85% of the subjects initially allocated to groups; N.°9: all subjects for whom outcome measures were available received the treatment or control condition as allocated or, where this was not the case, data for at least one key outcome was analysed by "intention to treat"; N.°10: the results of between-group statistical comparisons are reported for at least one key outcome; and N.°11: the study provides both point measures and measures of variability for at least one key outcome. Table 4. Methodological index for non-randomized studies (MINORS). N.°1* N.°2 N.°3 N.°4 N.°5 N.°6 N.°7 N.°8 N.°9 N.°10 N.°11 N.12 Total** (Arslan et al., 2020) 2 1 2 2 0 2 2 0 1 2 2 2 18 (Beato et al., 2019) 2 1 2 2 0 2 0 0 1 2 2 2 16 (Bravo et al., 2008) 2 1 2 2 0 2 0 0 1 2 2 2 16 (Buchheit et al., 2010) 2 1 2 2 0 2 1 2 1 2 2 2 19 (Dello lacono et al.) 2 1 2 2 0 2 2 0 1 2 2 2 18 (Eniseler et al., 2017) 2 1 2 2 0 2 2 0 1 2 2 2 18 (Faude et al., 2013) 2 2 2 2 0 2 0 0 2 2 2 18 (Faude et al., 2014) 2 0 2 2 0 2 0 0 2 2 2 16 (Harrison et al., 2015) 2 1 2 2 0 2 2 0 1 2 2 2 18 (Hill-Haas et al., 2009) 2 1 2 2 0 2 0 0 1 2 2 2 16 (Hostrup et al., 2019) 2 1 2 2 0 2 1 0 1 2 2 2 17 (laia et al., 2015) 2 1 2 2 0 2 2 0 1 2 2 2 18 (laia et al., 2017) 2 1 2 2 0 2 1 0 1 2 2 2 17 (F Impellizzeri et al., 2006) 2 1 2 2 0 2 0 0 1 2 2 2 16 (Jastrzebski et al., 2014) 2 1 2 2 0 2 0 0 1 2 2 2 16 (Los Arcos et al., 2015) 2 1 2 2 0 2 2 0 1 2 2 2 18 (Mohr & Krustrup, 2016) 2 1 2 2 0 2 2 0 1 2 2 2 18 (Mujika et al., 2009) 2 2 2 2 0 2 0 0 1 2 2 2 17 (Radziminski et al., 2013) 2 1 2 2 0 2 0 0 1 2 2 2 16 (Reyetal., 2019) 2 1 2 2 0 2 2 2 1 2 2 2 20 (Safania et al., 2011) 2 1 2 2 0 2 2 0 1 2 2 2 18 (Slettalokken & Ronnestad, 2014) 2 2 2 2 0 2 2 0 1 2 2 2 19 (Sperlich et al., 2011) 2 0 2 2 0 2 1 0 2 2 2 2 17 (JM Taylor et al., 2016) 2 1 2 2 0 2 2 0 1 2 2 2 18 *: MINORS scale items number; N.°1: A clearly study aimed; N.° 2: Inclusion of consecutive patients; N.° 3: Prospective collection of data; N.° 4: Endpoints appropriate to the aim of the study; N.° 5: Unbiased assessment of the study endpoint; N.° 6: Follow-up period appropriate to the aim of the study; N.° 7: Loss to follow less than 5%; N.° 8: Prospective calculation of the study size; N.° 9: An adequate control group; N.°10: Contemporary groups; N.° 11: Baseline equivalence of groups; N.° 12: Adequate statistical analyses; **: the total number of points from a possible maximal of 24. HUT study group was not included in the between-group comparison. 3.7. Effects of HUT on vertical height jump A summary of the included studies and results of VJH reported before and after HUT programs are provided in Table 8. Sixteen experimental groups (pooled n = 167) provided data for VJH. There was a non-significant effect of HUT on VJH (ES = 0.25; 95% CI = -0.01 to 0.51; p = 0.063; I2 = 79.4%; Egger's test p= 0.000; Figure 8, bottom). The relative weight of each study in the analysis ranged from 1.7% to 7.3%. After the trim and fill method was applied, the adjusted values were ES = 0.35, 95% CI = 0.08 to 0.63 (a funnel plot for bias assessment is available as Supplementary Figure 7). No significant sub-group difference (between-group p = 0.665) was found between long-interval HUT (three study groups; ES = 1.07; 95% CI = -0.29 to 2.43; within-group I2 = 92.9%), RST (4four study groups; ES = 0.21; 95% CI = -0.01 to 0.44; within-group I2 = 0.0%), short-interval HUT (three study groups; ES = 0.14; 95% CI = -0.67 to 0.96; within-group I2 = 90.0%), and SSGs (five study groups; ES = 0.28; 95% CI = -0.18 to 0.74; within-group I2 = 77.0%). Due to reduced number of studies available for robust meta-analyses, one SIT group was not included in the between-group comparison. 3.8. Effects of HUT on sprint time A summary of the included studies and results of ST reported before and after HUT programs are provided in Table 9. Three controlled studies provided data for ST, involving three experimental and three control groups (pooled n =61). 10 © F. MANUEL CLEMENTE ET AL. Table 5. Summary of the included studies and results of V02max (mL-kg '-min reported before and after HUT programs. HUT Before After Before-after Study Format Design N Mean±SD Mean±SD (A%) (Jastrzebski et al., 2014) siHNT NC 11 55.7 ± 5.2 56.9 ± 5.6 2.2 (Ouerghi et al., 2014) siHNT C 8 53.3 ± 4.5 57.3 ± 4.0 7.5 (Bravo et al., 2008) NHIIT NC 13 52.8 ± 3.2 56.3 ± 3.1 6.6 (Helgerud et al., 2001) NHIIT C 9 58.1 ± 4.5 64.3 ±3.9 10.7 (FM Impellizzeri et al., 2008) NHIIT C 11 56.6 ± 2.5 58.9 ± 3.0 4.1 (Radziminski et al., 2013) NHIIT NC 11 56.2 ± 8.7 55.3 ± 6.1 -1.6 (Safania et al., 2011) NHIIT NC 10 34.0 ± 1.4 43.5 ± 1.4 27.9 (Slettalokken & Ronnestad, 2014f NHIIT NC 9 65.6 ± 2.1 64.3 ± 1.3 -2.0 (Slettalokken & Ronnestad, 2014)1 NHIIT NC 8 63.4 ± 5.9 64.0 ±5.9 0.9 (Sperlich et al., 2011) NHIIT NC 9 55.1 ± 4.9 58.9 ± 4.7 6.9 (Bravo et al., 2008) RST NC 13 55.7 ± 2.3 58.5 ± 4.1 5.0 (Kavaliauskas et al., 2017) RST C 7 48.8 ± 3.4 50.2 ± 3.2 2.9 (Sanchez-Sanchez et al., 2019) RST9 C 10 51.2 ± 2.8 51.6 ± 2.8 0.8 (Sanchez-Sanchez et al., 2019) RSTh C 10 44.1 ± 2.8 44.7 ± 2.8 1.3 (Macpherson & Weston, 2015) SIT C 14 52.7 ± 4.7 54.3 ±3.4 3.0 (Harrison et al., 2015) SSGs NC 10 55.9 ± 3.0 57.1 ± 3.5 2.1 (Hill-Haas et al., 2009) SSGs NC 10 59.3 ± 4.5 58.9 ± 5.5 -0.7 (F Impellizzeri et al., 2006) SSGs NC 14 57.7 ± 4.2 61.8 ± 4.5 7.1 (Jastrzebski et al., 2014) SSGs NC 11 52.5 ± 5.2 57.0 ± 5.4 8.6 (Radziminski et al., 2013) SSGs NC 9 58.6 ± 6.9 63.3 ± 8.0 8.0 (Safania et al., 2011) SSGs NC 10 34.2 ± 1.6 42.9 ± 1.4 25.4 @: one liHNT session every week; 1: one NHIIT session two every two weeks; HUT: type of high-intensity interval training; siHNT: short-interval HUT; NHIIT: long-interval HI IT; RST: repeated sprint training; SIT: sprint interval training; SSGs: small-sided games; C: controlled; NC: non-controlled; 9: group above 48 mL-kg"1-min"1 of V02 max;h: group below 48 mL-kg~1-min~1 of V02max; N: number of participants per group; SD: standard deviation. Hedaes'saand95%Cl Oiiet$ii et al, 2014) (siHUT) aHll Helgerud, et al, 2001) (liHUT) liHll Impellizzeri et al., 200S) (liHUT) liHll Kavaliau&as et al., 2017) (RST) RST Sanchez-Sanchez et al., 2019) (RST - hf)RST Sanchez-Sanchez et al., 2019) (RST - If) RST Macpherajn & Weston, 201 5) (SIT) SIT Hedges'* g 0.S96 0.379 -0.374 -0.309 0.675 0.566 0.499 0.47S 0.461 0.506 0.55S 0.519 0.424 0.239 0.249 0.229 0.212 0.256 0.312 0.269 0.179 0.057 2.105 2.140 0.720 0.707 0.750 -0.670 -0.596 -0.156 0.09S p-Value 0.072 0.015 0.007 0.454 0.503 0.551 Favours control Favours HIIT Figure 2. Forest plot of changes in maximal oxygen consumption, in men soccer players participating in high-intensity interval training (HIIT) compared to controls. Values shown are effect sizes (Hedges's g) with 95% confidence intervals (CI). The size of the plotted squares reflects the statistical weight of the study. There was a non-significant effect of HIIT on ST (ES = 0.48; 95% CI = -0.06 to 1.02; p = 0.080; I2 = 15.1 %; Egger's test p= 0.691; Figure 9; a funnel plot for bias assessment is available as Supplementary Figure 8). The relative weight of each study in the analysis ranged from 32.9% to 34.2%. Sixteen non-controlled studies provided data for ST, involving 25 experimental groups (pooled n = 267). There was a significant effect of HIIT on ST (ES = 0.30; 95% CI = 0.19 to 0.41; p < 0.001; I2 = 33.0%; Egger's test p= 0.011; Figure 10, bottom). The relative weight of each study in the analysis ranged from 3.0% to 5.6%. After the trim and fill method was applied, the adjusted values remained equal as the observed values (a funnel plot for bias assessment is available as Supplementary Figure 9). No significant sub-group difference (between-group p = 0.366) was found between long-interval HIIT (three study groups; ES = 0.15; 95% CI = -0.10 to 0.40; within-group F = 0.0%), RST (9 study groups; ES = 0.40; 95% CI = 0.25 to 0.55; within-group I2 = 7.6%), short-interval HIIT (3 study groups; ES = 0.43; 95% CI = -0.04 to 0.89; within-group I2 = 65.9%), SIT (3 study groups; ES = 0.23; 95% CI = -0.07 to 0.52; within-group F = 0.0%), and SSGs (7 study groups; ES = 0.20; 95% CI = -0.06 to 0.45; within-group I2 = 54.3%). 3.9. Adverse effects Among the included studies, none reported soreness, pain,fatigue, injury, damage or adverse effects related to the HIIT interventions. 4. Discussion 4.1. Effects of HIIT on maximal oxygen uptake Overall, HIIT had favourable effects on V02max. In the case of controlled trials, slight improvements were observed in comparison to control groups. Interestingly, among the controlled studies, those testing short or long-interval HIIT (Bravo et al., 2008; Helgerud et al., 2001; Ouerghi et al., 2014) had moderate to large benefits, while RST (Kavaliauskas et al., 2017) and SIT (Macpherson & Weston, JOURNAL OF SPORTS SCIENCES © 11 o Negative Positive Negative Positive Figure 3. Forest plot on the effects of high-intensity interval training (HUT) on maximal oxygen consumption in men soccer players from non-controlled trials. Upper figure: individual studies results (the size of the plotted squares reflects the statistical weight of the study). Bottom figure: overall results. Values shown are effect sizes (Hedges's g) with 95% confidence intervals (CI). Table 6. Summary of the included studies and results of aerobic performance assessed from maximal field based-tests reported before and after HUT programs. Study HUT Format Design Before Mean±SD After Mean±SD Before-after (A%) (Dellal et al., 2012) siHNT C 8 19.4 ± 0.5 20.7 ± 1.2 6.7 (Faude et al., 2013) siHNT NC 20 17.05 ± 1.1 17.30 ± 0.9 1.5 (Faude et al., 2014) siHNT NC 10 17.8 ± 1.0 17.3 ± 1.0 -2.8 (Ouerghi et al., 2014) siHNT C 8 16.0 ± 1.5 17.6 ± 1.0 10.0 (Arslan et al., 2020) siHNT NC 10 1240 ± 75 1484 ± 74 19.7 (FM Impellizzeri et al., 2008) NHIIT C 11 1890 ±180 2100 ± 200 11.1 (Los Arcos et al., 2015) NHIIT NC 8 16.8 ± 0.9 17.1 ± 1.0 1.8 (Slettalokken & Ronnestad, 2014)# NHIIT NC 9 2531 ± 106 2327 ± 96 -8.1 (Slettalokken & Ronnestad, 2014)1 NHIIT NC 8 2335 ± 390 2213 ± 345 -5.2 (Beato et al., 2019) RSTe NC 18 1642 ± 365 1822 ± 461 11.0 (Beato et al., 2019) RSTf NC 18 1686 ± 359 1811 ± 260 7.4 (Eniseler et al., 2017) RST NC 9 2307 ± 252 2480 ± 159 7.5 (laia et al., 2017) RSTC NC 9 1000 ±169 1111 ± 171 11.1 (laia et al., 2017) RSTd NC 10 1016 ± 217 1072 ±156 5.5 (Kavaliauskas et al., 2017) RST C 7 1468 ± 409 1643 ± 382 11.9 (Sanchez-Sanchez et al., 2019) RST9 C 10 1764 ± 334 1798 ± 335 2.0 (Sanchez-Sanchez et al., 2019) RSTh C 10 914 ± 330 985 ± 337 8.1 (JM Taylor et al., 2016) RST NC 8 1830 ± 274 2269 ± 201 24.0 (Tonnessen et al., 2011) RST C 10 12.0 ± 1.2 12.6 ± 1.2 5.0 (Hostrup et al., 2019) SIT NC 8 1910 ± 557 1940 ± 553 1.6 (laia et al., 2015)£ SIT NC 6 927 ±185 1020 ± 155 10.0 (laia et al., 2015)* SIT NC 7 989 ± 226 1026 ± 210 3.7 (Macpherson & Weston, 2015) SIT C 14 1523 ± 493 1799 ± 292 18.1 (Mohr & Krustrup, 2016) SIT NC 9 680 ± 68 693 ± 52 1.9 (Arslan et al., 2020) SSGs NC 10 1284±152 1472 ± 99 14.6 (Dellal et al., 2012) SSGs NC 8 19.5 ± 0.9 20.5 ± 1.2 5.1 (Dello lacono et al.) SSGs NC 10 1646 ±138 1990 ±176 20.9 (Eniseler et al., 2017) SSGs NC 10 2320 ± 388 2432 ± 336 4.8 (Faude et al., 2014) SSGs NC 9 17.5 ± 1.0 17.8 ± 0.7 1.7 (Harrison et al., 2015) SSGs NC 10 18.1 ± 1.3 19.1 ± 1.4 5.5 (Hill-Haas et al., 2009) SSGs NC 10 1488 ± 345 1742 ± 362 17.1 (Los Arcos et al., 2015) SSGs NC 7 17.0 ± 0.8 16.9 ± 0.8 -0.6 (Mohr & Krustrup, 2016) SSGs NC 9 978 ± 57 858 ± 48 -12.3 : 120s passive recovery; *: 40s passive recovery; @: one NHIIT session every week; 1: one NHIIT session every two weeks; siHNT: short-interval HUT; NHIIT: long-interval HUT; RST: repeated sprint training; SIT: sprint interval training; SSGs: small session per week;c: short rest;d: long rest;e: straight sprint;': with change of direction;9: group above 48 mL-kg^-min-1 of V02max; C: controlled; NC: non-controlled; N: number of participants per group; SD: standard deviation. HUT: type of high-intensity interval training; -sided games;a: one session per week; b: two of V02max;h: group below 48 mL-kg^-min-1 12 © F. MANUEL CLEMENTE ET AL. Statistics for each si Hedges's a and 95% CI g Stan dard Variance 'limit "E & Value p-Value (Oierghi et al., 2014) (siHllT) 0.S76 0.497 0.247 -0.099 1.S51 1.762 0.07S Delia] et al (2012) (siHllT) 1.02S 0.657 0.431 -0.259 2.314 1.565 0.1 IS (lmpellizzeri et al., 200S) (liHUT) 1.634 0.4S9 0.239 0.675 2.593 3.33S 0.001 (Tonnessen et al., 2011)(RST) 0.2S2 0.431 0.1 S5 -0.562 1.126 0.655 0.513 (Kavaliauskas et al., 2017) (RST) 0.3S5 0.506 0.256 -0.606 1.376 0.762 0.446 (Sanchez-Sanchez et al., 2019) (RST high-fitness) -0.721 0.570 0.325 -1.S39 0.397 -1.264 0.206 (Sanchez-Sanchez et al., 2019) (RST low-fitness) -0.619 0.52S 0.279 -1.654 0.415 -1.173 0.241 (Macpherson & Weston, 201 5) (SIT) 0.9SS 0.437 0.191 0.131 1.S44 2.261 0.024 Delia] et al (2012) (SSGs) 0.771 0.640 0.410 -0.4S5 2.026 1.203 0.229 0.522 0.255 0.065 0.022 1.021 2.047 0.041 0.00 1.00 2.00 Favours control Favours HHT Figure 4. Forest plot of changes in aerobic performance assessed from maximal field based-tests measured in men soccer players participating in high-intensity interval training (HUT) compared to controls. Values shown are effect sizes (Hedges's g) with 95% confidence intervals (CI). The size of the plotted squares reflects the statistical weight of the study. &F3niaai2014)(ll &F3]Tiaai.3]14)(0 m.KEll™%&»yEr.2014)(*Mn) fan Qffan. & DHff. 3317) SK Tm.KEll™%&»yEr.2014)(SS3s) Negative Positive iur Variance limit limit ZrValue p-Value 0.128 0.016 0.065 0.596 2.707 0.007 Hedpea'a °and95%CI 41.50 0.00 0.50 1.00 Negative Positive Figure 5. Forest plot on the effects of high-intensity interval training (HUT) on aerobic performance assessed from maximal field based-tests measured in men soccer players from non-controlled trials. Upper figure: individual studies results (the size of the plotted squares reflects the statistical weight of the study). Bottom figure: overall results. Values shown are effect sizes (Hedges's g) with 95% confidence intervals (CI). 2015) had smaller benefits. V02max improvements seem to be elicited by working at intensities closer to V02max, in which the high aerobic demands signal peripheral adaptations (e.g., increased skeletal muscle mitochondrial content and capillary density) and central adaptations (e.g., maximal stroke volume, maximal cardiac output, and blood volume) (Maclnnis & Gibala, 2017). However, the type of HUT used elicits different acute responses that may explain different levels of adaptations found in the results (Figures 2 and 3). For example, short-interval HUT seems to stimulate a large spectrum of responses, namely metabolic load (oxygen transport and utilization), anaerobic glycolytic energy contribution, neuromuscular strain, and load, while long-interval HUT seems to be more closely associated with the stress of the anaerobic system and neuromuscular strain and load (Buchheit & Laursen, 2013a). The JOURNAL OF SPORTS SCIENCES © 13 Table 7. Summary of the included studies and results of repeated sprint ability before and after HIIT programs. HIIT Before After Before-after Study Format Design N Mean±SD Mean±SD (A%) (Arslan et al., 2020) siHUT NC 10 38.2 ± 1.7 34.9 ± 1.5 -8.6 (Beato et al., 2019) RSTe NC 18 7.46 ± 0.19 7.40 ± 0.20 -0.8 (Beato et al., 2019) RSTf NC 18 7.50 ± 0.21 7.48 ± 0.21 -0.3 (Buchheit et al., 2010) RST NC 10 6.35 ± 0.20 6.18 ± 0.14 -2.7 (Chtara et al., 2017) RST C 12 6.53 ± 0.13 6.42 ± 0.14 -1.7 (Eniseler et al., 2017) RST NC 9 7.13 ± 0.17 7.13 ± 0.21 0.0 (laia et al., 2017) RSf NC 9 92.91 ± 4.66 90.47 ± 4.24 -2.6 (laia et al., 2017) RSTd NC 10 91.45 ± 4.35 88.22 ± 4.65 -3.5 (Reyetal., 2019) RSTa NC 14 4.20 ± 0.17 4.12 ± 0.20 -1.9 (Reyetal., 2019) RSTb NC 13 4.20 ± 0.20 4.08 ± 0.19 -2.9 (Sanchez-Sanchez et al., 2019) RST9 C 10 4.40 ± 0.33 4.26 ± 0.23 3.0 (Sanchez-Sanchez et al., 2019) RSTh C 10 4.77 ± 0.41 4.76 ± 0.40 0.3 (Tonnessen et al., 2011) RST C 10 5.42 ± 0.18 5.30 ± 0.14 -2.2 (laia et al., 2015)£ SIT NC 6 86.09 ± 6.30 83.97 ± 4.72 -2.5 (laia et al., 2015)* SIT NC 7 83.81 ± 2.37 84.65 ± 2.27 2.2 (Mohr & Krustrup, 2016) SIT NC 9 4.45 ± 0.05 4.36 ± 0.14 -2.0 (Arslan et al., 2020) SSGs NC 10 37.8 ± 1.5 35.6 ± 1.2 -5.8 (Dello lacono et al.) SSGs NC 10 5.48 ± 0.14 5.23 ± 0.10 -4.6 (Eniseler et al., 2017) SSGs NC 10 7.12 ± 0.17 7.22 ± 0.20 1.4 (Hill-Haas et al., 2009) SSGs NC 10 42.1 ± 1.1 42.0 ± 1.4 -0.2 (Mohr & Krustrup, 2016) SSGs NC 9 4.41 ± 0.07 4.35 ± 0.22 -1.4 £: 120s passive recovery; *: 40s passive recovery; HIIT: type of high-intensity interval training; siHUT: short-interval HIIT; I i HI IT: long-interval HIIT; RST: repeated sprint training; SIT: sprint interval training; SSGs: small-sided games;a: one session per week; b: two sessions per week;c: short rest; : long rest;e: straight sprint;f: with change of direction; 9: group above 48 mL-kg^-min-1 of V02max; h: group below 48 mL-kg^-min-1 of V02max; C: controlled; NC: non-controlled; N: number of participants per group; SD: standard deviation Statistics for each study Hedges's gand95%CI Hedges's g Standard Variance limit Upper ZrValue p-Value (Chtara etal, 2017) (RST) 0.621 0.422 0.178 -0.207 1.449 1.470 0.142 (TBnnessen etal, 2011) (RST) 0.356 0.432 0.187 -0.490 1.203 0.825 0.409 (Sanchez-Sanchez et al.,2019) (RST-hf) 0.765 0.572 0.328 -0.357 1.887 1.336 0.181 (Sanehez-Sanehez et al.,2019) (RST - If) 0.158 0.516 0.267 -0.854 1.169 0.305 0.760 0.468 0.237 0.056 0.003 0.933 1.973 0.049 Favours control Favours HIIT Figure 6. Forest plot of changes in repeated sprint ability, in men soccer players participating in high-intensity interval training (HIIT) compared to controls. Values shown are effect sizes (Hedges's g) with 95% confidence intervals (CI). The size of the plotted squares reflects the statistical weight of the study. RST that implies repeated all-out efforts also seem to elicit the oxygen system but primarily focuses on the anaerobic system (higher rate of anaerobic energy turnover) (laia et al., 2009) and neuromuscular participation due to the greatest power required during sprints (Mendez-Villanueva et al., 2008). The SIT is also more focused on neuromuscular load, power output, and the anaerobic system, considering the levels of blood lactate accumulation (MacDougall et al., 1998). These differences in acute responses could explain the different magnitudes of adaptations observed between the different types of training considered in the current SRMA. This justifies the greater improvements of short and long-interval HIIT on V02max than RST and SIT, which can be more appropriate for developing other physical qualities. The additional analysis made with the non-controlled studies (Figure 3) confirmed greater magnitudes of benefits regarding long-interval HIIT and SSGs when compared to other HIIT types (despite the small number of studies associated with short-interval HIIT and RST in this section). Interestingly, SSGs prescribed with similar regimens as long-interval HIIT (two to five bouts of 2-4 min of intense effort) also reveals moderate improvements in V02max, thus confirming previous findings that have compared the efficacy of running-based exercises and SSGs (Moran et al., 2019). In our comparisons, both HIIT types (long-interval HIIT and SSGs) were effective in improving V02max, and no significant differences were found between them. This could be interesting for those who want to choose the most effective HIIT types to implement in practical scenarios. Even though the majority of studies have been conducted on youth players (Table 1), it seems that coaches may choose between long-interval HIIT and SSGs to improve players' V02max taking into account the idea behind the training intervention, granting technical proficiency 14 © F. MANUEL CLEMENTE ET AL. Bidets's s and 95% CI (Reya; (Reya; al.,2019") fsttaiglit sp al., 2019) (with chan^ taal.,2010) ■, 7*31, CEcai, &Din .,2017) (short rest) .,2017) (long rest) n2019)(1 sessioiV™ ,2019)(2sessions/v (laiaetal.,2015) (120s passive recovery) (laia et al., 2015) (40s passive recovery) (Mohr&Ktusmip, 2016) (SIT) (Arsl an, Orer, & aerraite.2020) (SSGs) (Dell o Iacono, Beato, &Unnithai) SSGs (Eniseler, 7ahai, CEcai, &Dinler,2017) (Mi 11 -Haas, Coutts, Rowsel 1, & Dawson, 2009) (Mohr&Ktusmip, 2016) (SSGs) -0.056 -0.232 0319 0.434 0520 .145 1.005 Z622 1.954 -0.016 -2.029 -1.50 0.00 150 3D0 Negatíve Positive Statistics for each study Hedges's gand95%CI Standard Loner Upper error Variance limit limit ZrValue p-Value Negative Positive Figure 7. Forest plot on the effects of high-intensity interval training (HUT) on repeated sprint ability in men soccer players from non-controlled trials. Upper figure: individual studies results (the size of the plotted squares reflects the statistical weight of the study). Bottom figure: overall results. Values shown are effect sizes (Hedges's g) with 95% confidence intervals (CI). and tactical awareness while choosing SSGs, but finely controlling the physiological responses while implementing long-interval HUT. Additionally, although one study has included a group with small V02max levels at the baseline (-34 mL-Kg-min~1) (Safania et al., 2011), the remaining studies related to long-interval HUT and SSGs displayed players' baseline values between -53 and 65 mL-Kg-min~1, thus suggesting that even in cases of moderate to high V02max baseline levels, long-interval HUT and SSGs seem to be effective enough to elicit favourable adaptations. The only exceptions of no improvement were noted after long-interval intervention involving a single training session per week and very high players' baseline V02max level (65.6 mL-Kg-min~1) (Slettalokken & Ronnestad, 2014), and a study using SSGs in a sample with 59.3 mL-kg-min"1 as baseline V02max (Hill-Haas et al., 2009). Establishing the link for the practice, the analysis of training prescription (Table 2) revealed that long-interval HUT varied from four to eight weeks, with the majority of interventions applying one to two sessions/week and the regimen consisting of 3-5 bouts of work at 87-95% of HRmax or 90-95% V,FT. Relief intervals of 3 min with light activities (e.g., jogging) occurred in the majority of cases. Considering SSGs, the interventions had similar regimens to long-interval HUT, while one vs. one to four vs. four games were contested in most cases. These regimens seem to be effective enough to increase V02max in soccer players. 4.2. Effects of HUT on aerobic performance assessed from maximal field based-tests The synthesis of controlled studies revealed significant moderate benefits on field-based AP after HUT. Among the controlled studies (Figure 4), the exception (in which the intervention was not favourable compared to the control condition) was a study implementing RST (Sanchez-Sanchez et al., 2019). The remaining interventions (short-interval HUT, long-interval HUT, and SSGs) showed that the intervention was favourable, as did one study that examined SIT (Macpherson & Weston, 2015). Overall, these studies reveal that regardless of the HUT type, this type of training seems to effectively improve aerobic performance during field-based tests. Consistently with the observations related to V02max, RST does not appear to be as effective as short and long-interval HUT and SSGs to improve AP, and this fact deserves consideration while implementing this kind of training in soccer players. Nonetheless, more studies are still warranted on this topic. Considering the non-controlled studies, significant improvements in field-based aerobic performance were JOURNAL OF SPORTS SCIENCES © 15 Table 8. Summary of the included studies and results of vertical height jump before and after HUT programs. HUT Before After Before-after Study Format Design N Mean±SD Mean±SD (A%) (Faude et al., 2013) siHNT NC 20 38.0 ± 4.0 36.3 ± 5.0 -4.5 (Faude et al., 2014) siHNT NC 10 38.5 ± 4.0 37.3 ± 4.0 -3.1 (Arslan et al., 2020) siHNT NC 10 28.2 ± 2.0 30.6 ± 1.8 8.5 (Bravo et al., 2008) NHNT NC 13 48.5 ± 3.8 48.1 ± 3.8 -0.8 (Los Arcos et al., 2015) NHNT NC 8 42.8 ± 4.6 42.4 ± 4.8 -0.9 (Sperlich et al., 2011) NHNT NC 9 26.0 ± 0.5 29.0 ± 0.6 11.5 (Bravo et al., 2008) RST NC 13 46.1 ± 3.5 46.1 ± 3.0 0.0 (Buchheit et al., 2010) RST NC 10 35.5 ± 5.8 38.0 ± 7.0 7.0 (JM Taylor et al., 2016) RST NC 8 41.9 ± 3.8 42.5 ± 5.1 1.4 (Tonnessen et al., 2011) RST C 10 35.2 ± 3.9 37.9 ± 5.7 7.7 (Mujika et al., 2009) SIT NC 10 42.4 ± 6.0 42.7 ± 5.9 0.7 (Arslan et al., 2020) SSGs NC 10 28.5 ± 2.5 31.3 ± 1.9 9.8 (Dello lacono et al.) SSGs NC 10 42.1 ± 5.2 45.8 ± 5.2 8.8 (Faude et al., 2014) SSGs NC 9 38.1 ± 4.7 37.5 ± 4.6 -1.6 (Harrison et al., 2015) SSGs NC 10 33.2 ± 6.2 33.8 ± 6.4 1.8 (Los Arcos et al., 2015) SSGs NC 7 42.7 ± 2.4 42.0 ± 2.8 -1.6 : 120s passive recovery; : 40s passive recovery; HUT: type of high-intensity interval training; siHNT: training; SIT: sprint interval training; SSGs: small-sided games; C: controlled; NC: non-controlled; n: group; SD: standard deviation. short-interval HUT; NHNT: long-interval HUT; RST: repeated sprint number of participants per group; N: number of participants per KEll™i.&»yff.2014)(5i ln.KEUnram.&»yrr.2C14)(S; Negative Positive Negative Positive Figure 8. Forest plot on the effects of high-intensity interval training (HUT) on vertical jump height in men soccer players from non-controlled trials. Upper figure: individual studies results (the size of the plotted squares reflects the statistical weight of the study). Bottom figure: overall results. Values shown are effect sizes (Hedges's g) with 95% confidence intervals (CI). observed after HUT, although the magnitudes were rated as small. Additional sub-group analysis comparing different HUT types did not reveal significant differences. These subgroup comparisons included long-interval HUT, RST, SIT, and SSGs. Interestingly, some parallel studies included in this section tested different HUT types. In particular, comparisons of SSGs vs. short- and long-interval HUT were tested (Arslan et al., 2020; Dellal et al., 2012; Faude et al., 2014; Los Arcos et al., 2015). In one study conducted on amateur players, no significant differences were found between SSGs and short-interval HUT, even though both groups had improved their final velocity at 30-15,FT test (V,FT) by -5% (Dellal et al., 2012). Also, in a study that compared short-interval HUT to SSGs in youth players, the results revealed that both interventions improved aerobic performance in YYIRT (-13% for SSG group and 16% for 16 © F. MANUEL CLEMENTE ET AL. Table 9. Summary of the included studies and results of sprint time before and after HUT programs. HUT Before After Before-after Study Format Design N Mean±SD Mean±SD (A%) (Faude et al., 2014) siHUT NC 10 4.12 ± 0.13 4.09 ± 0.11 -0.7 (Jastrzebski et al., 2014) siHUT NC 11 4.66 ± 0.22 4.62 ± 0.22 -0.9 (Arslan et al., 2020) siHUT NC 10 5.00 ± 0.34 4.66 ± 0.29 -6.8 (Bravo et al., 2008) NHIIT NC 13 1.77 ± 0.06 1.77 ± 0.06 0.0 (Helgerud et al., 2001) NHIIT C 9 5.58 ± 0.16 5.56 ± 0.15 -0.4 (Radziminski et al., 2013) NHIIT NC 11 4.80 ± 0.28 4.77 ± 0.24 -0.6 (Sperlich et al., 2011) NHIIT NC 9 6.41 ± 0.37 6.23 ± 0.39 -2.8 (Beato et al., 2019) RSTe NC 18 2.94 ± 0.11 2.92 ± 0.11 -0.7 (Beato et al., 2019) RSTf NC 18 2.96 ± 0.12 2.90 ± 0.10 -2.0 (Bravo et al., 2008) RST NC 13 1.77 ± 0.06 1.76 ± 0.06 -0.6 (Buchheit et al., 2010) RST NC 10 4.70 ± 0.12 4.60 ± 0.19 -2.1 (Chtara et al., 2017) RST C 12 4.68 ± 0.09 4.57 ± 0.07 -2.4 (laia et al., 2017) RSTC NC 9 3.30 ± 0.09 3.25 ± 0.06 -1.5 (laia et al., 2017) RSTd NC 10 3.29 ± 0.08 3.21 ± 0.08 -2.4 (Reyetal., 2019) RSTa NC 14 3.31 ± 0.15 3.23 ± 0.21 -2.4 (Reyetal., 2019) RSTb NC 13 3.28 ± 0.15 3.23 ± 0.22 -1.5 (JM Taylor et al., 2016) RST NC 8 2.96 ± 0.10 2.85 ± 0.18 -3.7 (Tonnessen et al., 2011) RST C 10 5.21 ± 0.21 5.15 ± 0.20 -1.2 (Hostrup et al., 2019) SIT NC 8 4.34 ± 0.16 4.30 ± 0.12 -0.9 (laia et al., 2015)£ SIT NC 6 2.84 ± 0.08 2.83 ± 0.12 -0.4 (laia et al., 2015)* SIT NC 7 2.91 ± 0.09 2.87 ± 0.10 -1.4 (Arslan et al., 2020) SSGs NC 10 5.15 ± 0.32 4.81 ± 0.31 -6.6 Dello lacono et al.) SSGs NC 10 2.79 ± 0.09 2.76 ± 0.10 -1.1 (Faude et al., 2014) SSGs NC 9 4.13 ± 0.13 4.13 ± 0.11 0.0 (Harrison et al., 2015) SSGs NC 10 3.33 ± 0.27 3.24 ± 0.24 -2.7 (Hill-Haas et al., 2009) SSGs NC 10 3.26 ± 0.12 3.24 ± 0.17 -0.6 (Jastrzebski et al., 2014) SSGs NC 11 4.61 ± 0.25 4.67 ± 0.25 1.3 (Radziminski et al., 2013) SSGs NC 9 4.91 ± 0.29 4.89 ± 0.40 -0.4 £: 120s passive recovery; *: 40s passive recovery; HIIT: type of high-intensity interval training; siHUT: short-interval HI IT; NHIIT: long-interval HUT; RST: repeated sprint training; SIT: sprint interval training; SSGs: small-sided games;a: one session per week; b: two sessions per week;c: short rest; : long rest;e: straight sprint;f: with change of direction; C: controlled; NC: non-controlled; N: number of participants per group; SD: standard deviation. Slalisli s for each sludv Hedges's g (Helgsrud. Engen, WislofT, Sl HofT, 20011 (liHHT| 0.166 (Oitaraet al.. 201 7| (RST| 1.031 (Tonnessen et al.. 2011|(RST| 0.252 0.4S0 error Variance limil 0.440 0.193 -0.696 0.440 0.194 0.16S 0.430 0.1S5 -0.591 0.274 0.075 -0.057 Z-Value p-Value 0.37S 0.705 2.341 0.019 0.5S5 0.55S 1.753 0.0S0 Favours control Favours HIIT Figure 9. Forest plot of changes in sprint time, in men soccer players participating in high-intensity interval training (HIIT) compared to controls. Values shown are effect sizes (Hedges's g) with 95% confidence intervals (CI). The size of the plotted squares reflects the statistical weight of the study. short-interval HIIT) without any significant changes found between the two interventions (Arslan et al., 2020). Comparing short-interval HIIT and SSGs, significant improvements were also found regarding individual anaerobic thresholds in both groups, while no significant differences between interventions were found (Faude et al., 2014). Finally, one study that compared long-interval HIIT and SSGs in youth players found that neither intervention promoted significant benefits, and no significant differences were found between them (Los Arcos et al., 2015). Therefore, it seems reasonable to assume that, overall, SGGs have similar effects as short and long-interval HIIT on aerobic performance. Some parallel studies included in this section also tested different strategies that share the same HIIT type (Beato et al., 2019; laia et al., 2017). One study compared straight line vs. RST with changes of direction and reported no significant improvements for either group and no significant changes between them regarding the aerobic performance on the YYIRT (Beato et al., 2019). Comparing RST in players with high and low aerobic fitness levels, it was found that only RST was effective in low-aerobic fitness players, and none of the intervention groups performed better than a control group that engaged in soccer-specific training (Sanchez-Sanchez et al., 2019). Possibly, the fact that RST taxes more anaerobic systems and a greater neuromuscular load and strain than the oxygen system chain may explain the reduced effectiveness of this method to significantly improve aerobic capacity (M Buchheit & PBLaursen, 2013c). Also, a study that tested two RST interventions and compared variations in resting time revealed that JOURNAL OF SPORTS SCIENCES © 17 Negatíve Positive Figure 10. Forest plot on the effects of high-intensity interval training (HUT) on sprint time in men soccer players from non-controlled trials. Upper figure: individual studies results (the size of the plotted squares reflects the statistical weight of the study). Bottom figure: overall results. Values shown are effect sizes (Hedges's g) with 95% confidence intervals (CI). shorter rests (15 s) had greater efficacy than longer rests (30 s) in improving aerobic performance at YYIRT (laia et al., 2017). In fact, the period of rest might aid the general understanding of the adaptations promoted by training, particularly considering that shorter periods might elicit a greater taxing of both aerobic and anaerobic metabolisms (M Buchheit & PBLaursen, 2013c). Nonetheless, RST does not seem to be the most suited training strategy to promote gains in aerobic endurance in soccer players. 4.3. Effects of HUT on repeated-sprint ability Randomized controlled trials revealed a significant beneficial effect of HUT in RSA when compared to control groups. In this case, only RST interventions were included in the analysis (Chtara et al., 2017; Sanchez-Sanchez et al., 2019; T0nnessen et al., 2011). The RST interventions included in the randomized controlled trials varied from six to 10 weeks, and one to two sessions/week during which working intervals involved 18 to 40-m sprints performed 10 to 16 times per set. Because RST involves aerobic and anaerobic pathways, the capacity of improving RSA is expected, specifically when compared to control groups (Buchheit, 2012). Moreover, one possible explanation for the improvements is the similarity between RST and the specific RSA tests. Thus it is reasonable to expect improvements, and this should be highlighted and carefully interpreted (Buchheit, 2012). Nevertheless, it is important to note that some studies did not find any positive effect of performing RST on RSA in soccer (Haugen et al., 2014, 2015) and futsal players (Soares-Caldeira et al., 2014). Therefore, more controlled studies need to be conducted before concluding upon the effectiveness of RST on RSA, especially after considering that short-interval HUT showed superior effects in the improvement of RSA than RST in team sports players (Buchheit et al., 2008). The analysis of non-controlled studies revealed significant improvements in RSA from HUT No significant differences between RST, SIT, and SSGs were found, thus suggesting that all of them are equally effective in improving RSA. In nature, RST and SIT are highly similar to the RSA tests, and they are expected to improve sprinting speed (Buchheit, 2012). Interestingly, even considering that SSGs may not involve enough sprinting during typical external load demands (Clemente, 2020), significant improvements at RSA were also found. In a parallel study that compared SIT and SSGs (Mohr & Krustrup, 2016) over four weeks, similar benefits were found (-1.5-2%). However, the fatigue index was improved only to SIT in comparison to SSGs (Mohr & Krustrup, 2016). Oppositely, a study comparing RST and SSGs interventions revealed that SSGs led to significant decreases in the best time during RSA, while no significant changes were found for RST (Eniseler et al., 2017). However, in the same study, it was found that SSGs improved RSA decrements after the intervention, suggesting a greater capacity to recover between sprints (Eniseler et al., 2017). Finally, a parallel study that compared short-interval HUT 18 © F. MANUEL CLEMENTE ET AL. and SSGs revealed that both interventions contributed to improvements in RSA and that there were no significant differences between interventions (Arslan et al., 2020). Within the same HIIT type, some parallel studies tested the variations of training regimens in RST. Comparing RST using straight running or running with a change of direction, improvements in RSA were found only in the group exposed to changes of direction. This can be explained by the neuromuscular load and tension promoted by the change of direction, also contributing such stimulus for improving acceleration (Beato et al., 2019). In a different approach, a study comparing short vs. long rest intervals during RST revealed that longer rest periods might improve RSA total time, while unclear findings were found regarding decrements in RSA (laia et al., 2017). A study comparing the effects of resting periods (120 vs. 40 s) during SIT (laia et al., 2015) found that longer periods (120 s) led to meaningful improvements in RSA total time, while short periods led to very likely impairments in this outcome. Possibly, the decreased running performance was due to fatigue, which could justify the decreases associated with shorter periods (laia et al., 2015). 4.4. Effects of HIIT on vertical height jump The included HIIT studies indicated no significant effects on VHJ. Additionally, sub-group comparisons (long-interval HIIT vs. RST vs. SSGs) revealed no differences between HIIT types. Despite sprinting being associated with short contact time and dependence from reactive strength (Suchomel et al., 2016), it seems plausible that VHJ is not the main target of HIIT. However, and interestingly, comparing SIT with contrast training (heavy-light resistance with soccer-specific drills for speed and power development), no significant changes were found for VJH, and no differences were observed among interventions (Mujika et al., 2009). By comparing different HIIT types, parallel studies have revealed that short-interval HIIT and SSGs implemented in youth soccer had positive effects on VHJ, though there was no significant difference between them (Arslan et al., 2020). Comparing long-interval HIIT and SSGs in youth players, no significant changes in time were found among interventions (Los Arcos et al., 2015). Similar trends of no benefits and no differences between interventions were found in a study comparing RST and SSGs (Faude et al., 2014). Based on the trivial effects of HIIT (independent of the type), it seems reasonable to utilize more specific training that allows improvements in the stretch-shortening cycle and reactive strength, thus potentially complementing HIIT interventions. Even considering the considerable neuromuscular strain and load promoted by RST and SIT (Buchheit & Laursen, 2013a), no meaningful benefits were found in this SRMA - for that reason, it can be assumed that the inefficacy of HIIT to develop VHJ. 4.5. Effects of HIIT on sprint time The analysis of randomized controlled trials did not reveal significant benefits of HIIT in comparison to a control condition regarding sprinting time. Despite that, small effects were beneficial to HIIT groups. Among the included studies, two of them used RST (Chtara et al., 2017; Tonnessen et al., 2011), and one used long-interval HIIT (Helgerud et al., 2001). Because long-interval HIIT works at submaximal speeds, no meaningful changes in sprinting time are to be expected. However, in the RST, the work is centred on speed endurance at all-out intensity, and the results of the individual studies revealed significant improvements in 40-m maximum sprint (Tonnessen et al., 2011) and 30-m sprint (Chtara et al., 2017). More specifically, when comparing RST vs. plyometric training vs. control, the most significant improvements were found for RST (Chtara et al., 2017). Thus, it might be prudent to assume that HIIT can produce different effects on sprinting time, depending on the type and the level of players. The analysis of non-controlled studies revealed significant improvements in sprinting time after HIIT interventions. No significant differences were found between HIIT types (i.e., long-interval HIIT, RST, SIT, and SSGs). Some parallel studies compared different HIIT types. A study comparing short-interval HIIT and SSGs revealed the positive effects of both interventions on sprinting time, and no changes were found between the interventions (Arslan et al., 2020). Nevertheless, two studies comparing short-interval HIIT and SSGs in youth players revealed no significant improvements in any of the groups and no differences between them (Faude et al., 2014; Jastrzebski et al., 2014). Similarly, in a study that compared SSGs and long-intervals interventions in youth players, it was found that neither intervention yielded significant improvements in sprinting time, and there were no differences among them (Radziminski et al., 2013). In the particular case of SSGs, it may be unlikely to observe maximal sprint in smaller formats and pitch dimensions, and in cases of longer field dimensions, there is often a poor frequency of sprinting stimulus (Djaoui et al., 2017). Thus, it is recommended to use running-based drills to promote enough stimulus to achieve the maximal sprint (Castagna et al., 2017; Clemente, 2020). In a study that compared long-interval HIIT and RST interventions in youth soccer players, no significant improvements were found in sprinting performance for any of the groups, and no differences were observed between the groups (Bravo et al., 2008). Two RST interventions (one using straight sprinting and other using sprinting with a change of direction) revealed no significant improvements in 20-m sprint. However, RST with a change of direction yielded significant improvements in 10-m sprinting, which is mainly justified by the greater neuromuscular tension promoted by RST with a change of direction and the transfer for acceleration (Beato et al., 2019). Furthermore, a comparison of two different RST interventions (short vs. long resting intervals) revealed that the group exposed to longer rest intervals exhibited unclear small benefits, possibly because they were given more time for recovery, thus allowing them to maintain greater performance at each bout (laia et al., 2017). 4.6. Potential limitations, directions for future research, and practical implications There are some potential limitations to the current SRMA. One of them is the limited number of studies per each HIIT type. Another is the limited number of randomized controlled studies. These facts limited the sub-group analysis and did not permit an examination of the effects of controlled interventions JOURNAL OF SPORTS SCIENCES © 1 on VHJ. Another potential limitation relates to the relatively reduced number of participants among the included studies. Considering reduced sample sizes are common in the sports science literature (Abt et al., 2020), it was not surprising to find that most included studies comprised between 8 and 10 participants per experimental (or control) group. In this sense, some of our meta-analyses were conducted with a relatively reduced sample size (e.g., n = 61), potentially limiting inferences towards other soccer players. Future studies in the field of HUT in soccer are encouraged to follow recently published recommendations (Abt et al., 2020) to increase the sample size. From the results of this SRMA, it can be highlighted that V02max, field-based aerobic performance, or RSA can be elicited by different HUT types and - with SSGs among them. This may allow coaches to decide the best type of HUT for each moment of the season. The majority of the studies involved one or two sessions a week. Thus, we may suggest that this schedule should be maintained during regular weeks to ensure positive adaptations or maintenance. However, HUT does not seem to elicit positive adaptations in VHJ and, in some cases, in sprinting time. Therefore, it could be beneficial to add complementary training to improve reactive strength/plyometrics (for improving VHJ and the stretch-shortening cycle), sprinting training (closer to maximal velocity), and other strength and power training methods that help to improve sprinting. 5. Conclusions The current SRMA indicates that randomized controlled trials have revealed the significant beneficial effects of HUT (overall) on V02max, field-based aerobic performance, and RSA. No significant benefits were found in terms of sprinting time. Considering the analysis of non-controlled trials (parallel studies), HUT (overall) was found to produce significant improvements in V02max, field-based aerobic performance, RSA, and sprinting, although no significant improvements were found for VHJ. Sub-group analysis (i.e., comparisons between HUT types) revealed no significant differences in any of the outcomes. However, more research should be conducted to test this idea and, possibly, greater analysis to responder profile of players is necessary, aiming to adjust HUT type to human variability. Nevertheless, coaches can use HUT methods as part of their regimens and should perhaps vary the HUT type throughout the season to foster stimulus variability, but also keep in mind that some specificities exist across the HUT strategies in terms of adaptation. Acknowledgments FMC: This work is funded by FCT/MCTES through national funds and when applicable co-funded EU funds under the project UIDB/EEA/50008/2020. HS gratefully acknowledges the support of a Spanish government sub-project Integration ways between qualitative and quantitative data, multiple case development, and synthesis review as main axis for an innovative future in physical activity and sports research [PGC2018-098742-B-C31] (Ministerio de Economia y Competitividad, Programa Estatal de Generacion de Conocimiento y Fortalecimiento Cientifico y Tecnologico del Sistema l+D+i), that is part of the coordinated project New approach of research in physical activity and sport from mixed methods perspective (NARPAS_MM) [SPGC201800X098742CV0]. Disclosure statement No conflicts of interest. Funding This work was supported by the Fundacao para a Ciencia e a Tecnologia [This work is funded by FCT/MCTES through national funds and when applicable co-funded EU funds under the project UIDB/50008/2020]; Hugo Sarmento gratefully acknowledges the support of a Spanish government subproject Integration ways between qualitative and quantitative data, multiple case development, and synthesis review as the main axis for an innovative future in physical activity and sports research [PGC2018-098742-B-C31 ] (Ministerio de Economia y Competitividad, Programa Estatal de Generacion de Conocimiento y Fortalecimiento Cientifico y Tecnologico del Sistema l+D+i), that is part of the coordinated project New approach of research in physical activity and sport from mixed methods perspective (NARPAS_MM) [SPGC201800X098742CV0]. ORCID Filipe Manuel Clemente http://orcid.org/0000-0001-9813-2842 Rodrigo Ramirez-Campillo http://orcid.org/0000-0003-2035-3279 Fabio Yuzo Nakamura 0 http://orcid.org/0000-0002-5336-3652 Hugo Sarmento 0 http://orcid.org/0000-0001-8681-0642 References Abt, G., Boreham, C, Davison, G., Jackson, R, Nevill, A., Wallace, E., & Williams, M. (2020). Power, precision, and sample size estimation in sport and exercise science research. Journal of Sports Sciences, 38(17), 1933-1935. https://doi.Org/10.1080/02640414.2020.1776002 Arslan, E., Orer, G., & Clemente, F. (2020). Running-based high-intensity interval training vs. small-sided game training programs: Effects on the physical performance, psychophysiological responses and technical skills in young soccer players. Biology of Sport, 37(2), 165-173. https:// doi.org/10.5114/biolsport.2020.94237 Beato, M., Coratella, G., Bianchi, M., Costa, E., & Merlini, M. (2019). Short-term repeated-sprint training (Straight Sprint vs. Changes of Direction) in soccer players. Journal of Human Kinetics, 70(1), 183-190. https://doi. org/10.2478/hukin-2019-0040 Bravo, D., Impellizzeri, F., Rampinini, E., Castagna, C, Bishop, D., & Wisloff, U. (2008). Sprint vs. Interval Training in Football. International Journal of Sports Medicine, 29(8), 668-674. https://doi.org/10.1055/s-2007-989371 Buchheit, M. (2012). Should we be recommending repeated sprints to improve repeated-sprint performance? Sports Medicine, 42(2), 169-172. https://doi.Org/10.2165/11 598230-000000000-00000 Buchheit, M., & Laursen, P. B. (2013a). High-intensity interval training, solutions to the programming puzzle: Part I: Cardiopulmonary emphasis. Sports Medicine, 43(5), 313-338. https://doi.org/10.1007/s40279-013-0029-x Buchheit, M., & Laursen, P. B. (2013b). High-intensity interval training, solutions to the programming puzzle : Part II: Anaerobic energy, neuromuscular load and practical applications. Sports Medicine, 43(10), 927-954. https://doi.Org/10.1007/S40279-01 3-0066-5 Buchheit, M., & Laursen, P. B. (2013c). High-intensity interval training, solutions to the programming puzzle: Part I: Cardiopulmonary emphasis. Sports Medicine, 43(5), 313-338. https://doi.org/10.1007/s40279-013-0029-x Buchheit, M, Mendez-Villanueva, A., Delhomel, G., Brughelli, M, & Ahmaidi, S. (2010). Improving repeated sprint ability in young elite soccer players: Repeated Shuttle Sprints Vs. Explosive Strength 20 © F. MANUEL CLEMENTE ET AL. Training. Journal of Strength and Conditioning Research, 24(10), 2715-2722. https://doi.Org/10.1519/JSC.0b013e3181 bf0223 Buchheit, M, Millet, G. P., Parisy, A., Pourchez, S„ Laursen, P. B„ & Ahmadi, S. (2008). Supramaximal training and postexercise parasympathetic reactivation in adolescents. Medicine & Science in Sports & Exercise, 40(2), 362-371. https://doi.org/10.1249/mss.0b013e31815aa2ee Buchheit, M., & Rabbani, A. (2014). The 30-15 intermittent fitness test versus the Yo-Yo intermittent recovery test level 1: Relationship and sensitivity to training. International Journal of Sports Physiology and Performance, 9 (3), 522-524. https://doi.org/10.1123/ijspp.2012-0335 Castagna, C, Francini, L., Povoas, S. C. A., & D'Ottavio, S. (2017). Long-sprint abilities in soccer: Ball Versus Running Drills. International Journal of Sports Physiology and Performance, 12(9), 1256-1263. https://doi.org/ 10.1123/ijspp.2016-0565 Chtara, M., Rouissi, M., Haddad, M., Chtara, H., Chaalali, A., Owen, A., & Chamari, K. (2017). Specific physical trainability in elite young soccer players: Efficiency over 6 weeks' in-season training. Biology of Sport, 34 (2), 137-148. https://doi.org/10.5114/biolsport.2017.64587 Clemente, F. M. (2020). The threats of small-sided soccer games. Strength & Conditioning Journal, 42(3), 100-105. https://doi.org/10.1519/SSC. 0000000000000526 Deeks, J. J., Higgins, J. P., & Altman, D. G. (2008). Analysing data and undertaking meta-analyses. In J. P. Higgins & S. Green (Eds.), Cochrane Handb. Syst. Rev. Interv. Cochrane Collab (pp. 243-296). The Cochrane Collaboration. Dellal, A., Varliette, C, Owen, A., Chirico, E. N„ & Pialoux, V. (2012). Small-sided games versus interval training in amateur soccer players. Journal of Strength and Conditioning Research, 26(10), 2712-2720. https://doi.org/ 10.1519/JSC.0b013e31824294c4 Dello lacono, A., Beato, M., & Unnithan, V. Comparative effects of game profile-based training and small-sided games on physical performance of elite young soccer players. Journal Of Strength And Conditioning Research / National Strength & Conditioning Association, https://doi.org/ 10.1 519/JSC.0000000000003225 Djaoui, L., Chamari, K., Owen, A. L., & Dellal, A. (2017). Maximal sprinting speed of elite soccer players during training and matches. Journal of Strength and Conditioning Research, 31(6), 1509-1517. https://doi.Org/10. 1 519/JSC.0000000000001642 Duval, S., & Tweedie, R. (2000). Trim and fill: A simple funnel-plot-based method of testing and adjusting for publication bias in meta-analysis. Biometrics, 56(2), 455-463. https://doi.Org/10.1111/j.0006-341X.2000. 00455.x Egger, M, Smith, G. D„ Schneider, M„ & Minder, C. (1997). Bias in meta-analysis detected by a simple, graphical test. BMJ, 315(7109), 629-634. https://doi.Org/10.1136/bmj.315.7109.629 Engel, F. A., Ackermann, A., Chtourou, H, & Sperlich, B. (2018). High-intensity interval training performed by young athletes: A systematic review and meta-analysis. Frontiers in Physiology, 9. https://doi.org/10. 3389/fphys.2018.01012 Eniseler, N., Sahan, C., Ozcan, I., & Dinler, K. (2017). High-intensity small-sided games versus repeated sprint training in junior soccer players. Journal of Human Kinetics, 60(1), 101-111. https://doi.org/10. 1515/hukin-2017-0104 Faude, O., Koch, T, & Meyer, T. (2012). Straight sprinting is the most frequent action in goal situations in professional football. Journal of Sports Sciences, 30(7), 625-631. https://doi.org/10.1080/02640414.2012. 665940 Faude, 0„ Schnittker, R, Schulte-Zurhausen, R, Muller, F, & Meyer, T. (2013). High intensity interval training vs. high-volume running training during pre-season conditioning in high-level youth football: A cross-over trial. Journal of Sports Sciences, 31(13), 1441-1450. https://doi.org/10.1080/ 02640414.2013.792953 Faude, O., Steffen, A., Kellmann, M., & Meyer, T. (2014). The effect of short-term interval training during the competitive season on physical fitness and signs of fatigue: A crossover trial in high-level youth football players. International Journal of Sports Physiology and Performance, 9(6), 936-944. https://doi.Org/10.1123/ijspp.201 3-0429 Fereday, K, Hills, S. P., Russell, M, Smith, J„ Cunningham, D. J„ Shearer, D, McNarry, M., & Kilduff, L. P. (2020). A comparison of rolling averages versus discrete time epochs for assessing the worst-case scenario locomotor demands of professional soccer match-play. Journal of Science and Medicine in Sport, 23(8), 764-769. https://doi.0rg/IO.IOI6/j. jsams.2020.01.002 Fransson, D., Nielsen, T. S., Olsson, K., Christensson, T, Bradley, P. S., Fatouros, I. G., Krustrup, P., Nordsborg, N. B., & Mohr, M. (2018). Skeletal muscle and performance adaptations to high-intensity training in elite male soccer players: Speed endurance runs versus small-sided game training. European Journal of Applied Physiology, 118(1), 111-121. https://doi.Org/10.1007/S00421 -01 7-3751-5 Garcia-Hermoso, A., Ramirez-Campillo, R., & Izquierdo, M. (2019). Is muscular fitness associated with future health benefits in children and adolescents? A systematic review and meta-analysis of longitudinal studies. Sports Medicine, 49(7), 1079-1094. https://doi.org/10.1007/s40279-019-01098-6 Gibala, M. J„ Little, J. P., Van Essen, M, Wilkin, G. P., Burgomaster, K. A., Safdar, A., Raha, S., & Tarnopolsky, M. A. (2006). Short-term sprint interval versus traditional endurance training: Similar initial adaptations in human skeletal muscle and exercise performance. The Journal of Physiology, 575 (3), 901-911. https://doi.org/10.1113/jphysiol.2006.112094 Girard, O., Mendez-Villanueva, A., & Bishop, D. (2011). Repeated-sprint ability - Part I: Factors contributing to fatigue. Sports Medicine, 41(8), 673-694. https://doi.Org/10.2165/11 590550-000000000-00000 Gist, N. H„ Fedewa, M. V, Dishman, R. K„ & Cureton, K. J. (2014). Sprint interval training effects on aerobic capacity: A systematic review and meta-analysis. Sports Medicine, 44(2), 269-279. https://doi.org/10.1007/ S40279-013-0115-0 Green, S., & Higgins, J. (2005). (Cochrane handbook for systematic reviews of interventions). Cochrane library. Group CCCR. (2016). Data extraction template for included studies. Cochrane library. Harrison, C, Kinugasa, T., Gill, N., & Kilding, A. (2015). Aerobic fitness for young athletes: Combining game-based and high-intensity interval training. International Journal of Sports Medicine, 36(11), 929-934. https://doi.Org/10.1055/S-0034-1396825 Haugen, T., Tonnessen, E., Leirstein, S., Hem, E., & Seiler, S. (2014). Not quite so fast: Effect of training at 90% sprint speed on maximal and repeated-sprint ability in soccer players. Journal of Sports Sciences, 32(20), 1979-1986. https://doi.org/10.1080/02640414. 2014.976248 Haugen, T., Tonnessen, E., Oksenholt, 0., Haugen, F. L., Paulsen, G., Enoksen, E., & Seiler, S. (2015). Sprint conditioning of junior soccer players: Effects of training intensity and technique supervision. PLoS One, 10(3), e0121827. https://doi.org/10.1371/journal.pone.0121827 Helgerud, J„ Engen, L. C, Wisloff, U„ & Hoff, J. (2001). Aerobic endurance training improves soccer performance. Medicine and Science in Sports and Exercise, 33(11), 1925-1931. https://doi.org/10.1097/00005768-200111000-00019 Higgins, J. P. T. (2003). Measuring inconsistency in meta-analyses. BMJ, 327 (7414), 557-560. https://doi.org/10.1136/bmj.327.7414.557 Higgins, J. P. T., & Thompson, S. G. (2002). Quantifying heterogeneity in a meta-analysis. Statistics in Medicine, 21(11), 1539-1558. https://doi.org/ 10.1002/sim.1186 Hill-Haas, S. V., Coutts, A. J„ Rowsell, G. J„ & Dawson, B. T. (2009). Generic versus small-sided game training in soccer. International Journal of Sports Medicine, 30(9), 636-642. https://d0i.0rg/l 0.1055/S-0029-1220730 Hopkins, W. G„ Marshall, S. W, Batterham, A. M„ & Hanin, J. (2009). Progressive statistics for studies in sports medicine and exercise science. Medicine & Science in Sports & Exercise, 41(1), 3-13. https://doi. org/10.1249/MSS.0b013e31818cb278 Hostrup, M., Gunnarsson, T. P., Fiorenza, M., Morch, K., Onslev, J., Pedersen, K. M., & Bangsbo, J. (2019). In-season adaptations to intense intermittent training and sprint interval training in sub-elite football players. Scandinavian Journal of Medicine & Science in Sports, 29(5), 669-677. https://d0i.0rg/l 0.1111/sms.l 3395 laia, F. M, Fiorenza, M„ Larghi, L, Alberti, G, Millet, G. P., & Girard, O. (2017). Short- or long-rest intervals during repeated-sprint training in soccer? PLoS One, 12(2), eOl 71462. https://doi.org/10.1371/journal.pone. 0171462 laia, F. M„ Fiorenza, M, Perri, E, Alberti, G„ Millet, G. P., & Bangsbo, J. (2015). The effect of two speed endurance training regimes on performance of JOURNAL OF SPORTS SCIENCES © 21 soccer players. PLoS One, 70(9), eOl 38096. https://doi.org/10.1371/jour nal.pone.0138096 laia, F. M., Rampinini, E., & Bangsbo, J. (2009). High-intensity training in football. International Journal of Sports Physiology and Performance, 4(3), 291-306. https://doi.Org/10.1123/ijspp.4.3.291 Impellizzeri, F., Marcora, S., Castagna, C, Reilly, T, Sassi, A., laia, F., & Rampinini, E. (2006). Physiological and performance effects of generic versus specific aerobic training in soccer players. InternationalJournal of Sports Medicine, 27(6), 483-492. https://doi.org/10.1055/s-2005-865839 Impellizzeri, F. M., Rampinini, E., Maffiuletti, N. A., Castagna, C, Bizzini, M., & Wisloff, U. (2008). Effects of aerobic training on the exercise-induced decline in short-passing ability in junior soccer players. Applied Physiology, Nutrition, and Metabolism, 33(6), 1192-1198. https://doi.org/ 10.1139/H08-111 Jastrzebski, Z., Barnat, W., Dargiewicz, R., Jaskulska, E., Szwarc, A., & Radzimihski, t. (2014). Effect of in-season generic and soccer-specific high-intensity interval training in young soccer players. International Journal of Sports Science & Coaching, 9(5), 1169-1179. https://doi.org/ 10.1260/1 747-9541.9.5.1169 Kavaliauskas, M., Kilvington, R., & Babraj, J. A. (2017). Effects of in-season uphill sprinting on physical characteristics in semi-professional soccer players. The Journal of Sports medicine and Physical Fitness, 57(3), 165-1 70. https://doi.Org/10.23736/S0022-4707.16.06066-7 Kontopantelis, E., Springate, D. A., & Reeves, D. (2013). A re-analysis of the Cochrane library data: The dangers of unobserved heterogeneity in meta-analyses. PLoS One, 8(7), e69930. https://doi.org/10.1371/journal. pone.0069930 Kunz, P., Engel, F. A., & Holmberg H-C, S. B. (2019). A meta-comparison of the effects of high-intensity interval training to those of small-sided games and other training protocols on parameters related to the physiology and performance of youth soccer players. Sports Medicine - Open, 5(1), 7. https://doi.Org/10.1186/S40798-019-0180-5 Los Arcos, A., Vazquez, J. S., Martin, J., Lerga, J., Sanchez, F., Villagra, F., & Zulueta, J. J. (2015). Effects of small-sided games vs. interval training in aerobic fitness and physical enjoyment in young elite soccer players. PLoS One, 10(9), eOl 37224. https://doi.org/10.1371/journal.pone. 0137224 MacDougall, J. D, Hicks, A. L, MacDonald, J. R„ McKelvie, R. S„ Green, H. J„ & Smith, K. M. (1998). Muscle performance and enzymatic adaptations to sprint interval training. Journal of Applied Physiology, 84(6), 2138-2142. https://d0i.0rg/l 0.11 52/jappl.l 998.84.6.21 38 Maclnnis, M. J., & Gibala, M. J. (2017). Physiological adaptations to interval training and the role of exercise intensity. The Journal of Physiology, 595 (9), 2915-2930. https://doi.org/10.1113/JP273196 Macpherson, T. W., & Weston, M. (2015). The effect of low-volume sprint interval training on the development and subsequent maintenance of aerobic fitness in soccer players. International Journal of Sports Physiology and Performance, 10(3), 332-338. https://doi.org/10.1123/ijspp.2014-0075 Mendez-Villanueva, A., Hamer, P., & Bishop, D. (2008). Fatigue in repeated-sprint exercise is related to muscle power factors and reduced neuromuscular activity. European Journal of Applied Physiology, 103(4), 411-419. https://doi.Org/10.1007/S00421 -008-0723-9 Moher, D., Liberati, A., Tetzlaff, J., & Altman, D. G. (2009). Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA statement. PLoS Medicine, 6(7), el 000097. https://doi.org/10.1371/jour nal.pmed.1000097 Mohr, M., & Krustrup, P. (2016). Comparison between two types of anaerobic speed endurance training in competitive soccer players. Journal of Human Kinetics, 51(1), 183-192. https://d0i.0rg/l 0.1515/hukin-2015-0181 Moran, J., Blagrove, R. C, Drury, B., Fernandes, J. F. T., Paxton, K., Chaabene, H, & Ramirez-Campillo, R. (2019). Effects of small-sided games vs. conventional endurance training on endurance performance in male youth soccer players: A meta-analytical comparison. Sports Medicine, 49(5), 731-742. https://doi.org/10.1007/s40279-019-01086-w Moran, J., Ramirez-Campillo, R., & Granacher, U. (2018). Effects of jumping exercise on muscular power in older adults: A meta-analysis. Sports Medicine, 48(12), 2843-2857. https://d0i.0rg/l 0.1007/s40279-018-1002-5 Mujika, I, Santisteban, J., & Castagna, C. (2009). In-season effect of short-term sprint and power training programs on elite junior soccer players. Journal of Strength and Conditioning Research, 23(9), 2581 -2587. https://d0i.0rg/l 0.1519/JSC.0b013e3181 bcl aac Ouerghi, N., Feki, M., Kaabachi, N., Khammassi, M., Boukorraa, S., & Bouassida, A. (2014). Effects of a high-intensity intermittent training program on aerobic capacity and lipid profile in trained subjects. Open Access Journal of Sports Medicine, 5, 243-248. https://doi.org/10.2147/ OAJSM.S68701 Radziminski, L., Rompa, P., Barnat, W., Dargiewicz, R., & Jastrzebski, Z. (2013). A comparison of the physiological and technical effects of high-intensity running and small-sided games in young soccer players. International Journal of Sports Science & Coaching, 8(3), 455-466. https://doi.org/10. 1260/1747-9541.8.3.455 Rey, E„ Padron-Cabo, A., Costa, P. B, & Lago-Fuentes, C. (2019). Effects of different repeated sprint-training frequencies in youth soccer players. Biology of Sport, 36(3), 257-264. https://doi.org/10.5114/biolsport.2019. 87047 Safania, A. M., Alizadeh, R, & Nourshahi, M. (2011). A comparison of small-side games and interval training on same selected physical fitness factors in amateur soccer players. Journal of Social Sciences, 7(3), 349-353. https://thescipub.com/abstract/jssp.2011.349.353 Sanchez-Sanchez, J., Ramirez-Campillo, R., Petisco, C, Gonzalo-Skok, O., Rodriguez-Fernandez, A., Mihano, J., & Nakamura, F. Y. (2019). Effects of repeated sprints with changes of direction on youth soccer player's performance: Impact of initial fitness level. Journal of Strength and Conditioning Research, 33(10), 2753-2759. https://doi.org/10.1519/JSC. 0000000000002232 Shi, L., & Lin, L. (2019). The trim-and-fill method for publication bias. Medicine (Baltimore), 98(23), el 5987. https://doi.org/10.1097/MD. 0000000000015987 Skrede, T., Steene-Johannessen, J., Anderssen, S. A., Resaland, G. K., & Ekelund, U. (2019). The prospective association between objectively measured sedentary time, moderate-to-vigorous physical activity and cardiometabolic risk factors in youth: A systematic review and metaanalysis. Obesity Reviews, 20(1), 55-74. https://doi.org/10.1111/obr. 12758 Slettalokken, G., & Ronnestad, B. R. (2014). High-intensity interval training every second week maintains V[Combining dot above]02max in soccer players during off-season. Journal of Strength and Conditioning Research, 28(7), 1946-1951. https://doi.org/10.1519/JSC.0000000000000356 Slim, K., Nini, E, Forestier, D, Kwiatkowski, F, Panis, Y., & Chipponi, J. (2003). Methodological index for non-randomized studies (MINORS): Development and validation of a new instrument. ANZ Journal of Surgery, 73(9), 712-716. https://doi.Org/10.1046/j.l 445-2197.2003.02748.x Soares-Caldeira, L. F, de Souza, E. A, de Freitas, V. H, de Moraes, S. M. F, Leicht, A. S, & Nakamura, F. Y. (2014). Effects of additional repeated sprint training during preseason on performance, heart rate variability, and stress symptoms in futsal players. Journal of Strength and Conditioning Research, 28(10), 2815-2826. https:// doi.org/10.1519/JSC0000000000000461 Sperlich, B„ De Marees, M, Koehler, K„ Linville, J„ Holmberg, H.-C, & Mester, J. (2011). Effects of 5 weeks of high-intensity interval training vs. volume training in 14-year-old soccer players. Journal of Strength and Conditioning Research, 25(5), 1271-1278. https://doi.org/10.1519/JSC. 0b013e3181d67c38 Stolen, T, Chamari, K„ Castagna, C, & Wisloff, U. (2005). Physiology of soccer: An update. Sports Medicine, 35(6), 501-536. https://doi.org/10. 2165/00007256-200535060-00004 Suchomel, T. J., Nimphius, S., & Stone, M. H. (2016). The importance of muscular strength in athletic performance. Sports Medicine, 46(10), 1419-1449. https://doi.Org/10.1007/S40279-016-0486-0 Taylor, J., Macpherson, T., Spears, I., & Weston, M. (2015). The effects of repeated-sprint training on field-based fitness measures: A meta-analysis of controlled and non-controlled trials. Sports Medicine, 45(6), 881-891. https://doi.org/10.1007/s40279-015-0324-9 Taylor, J. M., Macpherson, T. W., McLaren, S. J., Spears, I., & Weston, M. (2016). Two weeks of repeated-sprint training in soccer: To turn or not to turn? International Journal of Sports Physiology and Performance, 11(8), 998-1004. https://doi.Org/10.1123/ijspp.2015-0608 Toh, S. H., Guelfi, K. J., & Wong, P. (2011). Energy expenditure and enjoyment of small-sided soccer games in overweight boys. Human 22 © F. MANUEL CLEMENTE ET AL. Movement Science, 30(3), 636-647. https://doi.Org/10.1016/j.humov. 2010.12.001 Tonnessen, E., Shalfawi, S. A., Haugen, T., & Enoksen, E. (2011). The effect of 40-m repeated sprint training on maximum sprinting speed, repeated sprint speed endurance, vertical jump, and aerobic capacity in young elite male soccer players. Journal of Strength and Conditioning Research, 25(9), 2364-2370. https://doi.org/10.1519/JSC. 0b013e3182023a65