. 2023 Feb 13;5:1116293. doi: 10.3389/fspor.2023.1116293

Table 2.

Summary of studies accompanied by study design, subjects, high-speed running metrics reported and details from the studies.

References	Participants	HSR metrics	Details
Scott et al., 2013 (25) Observational study	Professional male soccer players (n = 15) Individual training sessions (n = 97)	HSR > 14.4 km·h⁻¹ VHSR > 19.8 km·h⁻¹	Absolute and % of total distance values recorded during training: HSR = 544 ± 255 m (12.0 ± 3.8%), range 106–1,343 m (4.9–23.3%) VHSR = 132 ± 101 m (2.8 ± 1.9%), range 7–541 m (0.2–8.8%)
Wehbe et al., 2014 (26) Observational study	Elite male adult soccer players from Australian-league (A-League) soccer (Sydney Football Club) (n = 19) Preseason matches (n = 8)	HSR > 19.7 to ≤25.1 km·h⁻¹ Sprint > 25.1 km·h⁻¹ Putting together thresholds: HIR > 14.3 km·h⁻¹ VHIR > 19.7 km·h⁻¹	Positional comparison: midfielders covered 28% more HIR distance than defenders. Match half comparison: HIR and VHIR decreased from the first to the second half by 10 and 11%, respectively. Match status analysis: when the team was winning, average speed was 4% lower than when the team was drawing (p ≤ 0.05, d = 0.32). Pre- and post-goal analysis: scoring or conceding goals did not appear to affect HIR. In the 5-minute intervals before and after a goal was scored, 5-minute HIR distance was 140 and 128 m, respectively (p = 0.464). In the 5-minute intervals before and after a goal was conceded, 5-minute HIR distance was 144 and 110 m, respectively (p = 0.015). Average and peak 5-minute HIR distance during the whole match was 123 and 237 m, respectively.
Malone et al., 2015 (27) Observational study	Professional male players from English Premier League (Liverpool) (n = 30) Preseason weeks (n = 6) In-season weeks (n = 36) Microcycles (n = 3)	HSD > 19.8 km·h⁻¹	Higher total distances covered in the early stages of the competitive season and the highest HR response occurring at the midpoint of the season. HSD 1-week in-season microcycles (daily means): early-season = 243 ± 229 m, mid-season = 225 ± 213 m, late-season = 146 ± 104 m. Wide midfielders covered a higher amount of HSD across the different microcycles than central defenders (94 [43–145] m, ES = 0.47 [0.22–0.73], small). Periodization of training load was typically confined to MD-1 (regardless of mesocycle), whereas no differences were apparent during MD-2 to MD-5.
Anderson et al., 2016 (28) Observational study	English Premier League male players (n = 12) Training sessions (n = 10) + matches (n = 6) (1-, 2-, 3-game weeks)	HSR = 19.8–25.1 km·h⁻¹ Sprint > 25.1 km·h⁻¹	The majority of distance during specific training sessions was completed in the low-to moderate speed zones, whereas the distance completed in high-intensity zones were largely completed in the game itself. HSR: match demand = 706 m; training stimulus = 156 m (1-game week), 192 m (2-game week), 81 m (3-game week). Sprinting: match demand = 295 m; training stimulus = 8 m (1-game week), 16 m (2-game week), 7 m (3-game week).
Carling et al., 2016 (29) Observational study	French League 1 male players (n = 12) Matches (n = 31)	HSR = 19.8–25.2 km·h⁻¹ Sprint > 25.2 km·h⁻¹ Total HSR (THSR, ≥19.8 km·h⁻¹);	Math demand: HSR = 587 ± 133 m; Sprint = 184 ± 87 m; THSR = 770 ± 206 m.
Chmura et al., 2017 (10) Observational study	International male soccer players from 32 teams (n = 340) Single observations during 2014 World Cup (n = 905)	HIR = 19.9–25.2 km·h⁻¹ (% of TD) N° of sprints >25.2 km·h⁻¹	The mean distance covered by players at high intensity was 8.83 ± 2.11%. It was significantly longer between the quarter-finals and the semi-finals (p ≤ 0.01). In the semi-finals the percentage values of TD covered at HI were the greatest. Individually, the greatest percentage achieved was 17% by 2 midfielders. The mean number of sprints performed was 33 ± 11, 1 every 173 s. The greatest number of performed sprints was 68, 1 every 82 s, in a semi-final match. Winning a soccer championship requires players to run longer mean total distances and longer distances at high intensity during a single match.
Mara et al., 2017 (30) Observational study	Elite female players from the Australian national league (W-League) (n = 12) Matches (n = 7)	HSR = 12.24–19.0 km·h⁻¹ Sprint > 19 km·h⁻¹ High Speed Runs and Sprints (n)	Match demand: HSR = 2,452 ± 636 m; Sprint = 615 ± 258 m; high-speed runs = 376; sprints = 70. A large proportion of high-speed runs (81–84%) and sprints (71–78%) were performed over distances less than 10 m, with 14 s between high-speed runs and 87 s between sprints. The characteristics of high-speed runs and sprints differed between repeat and nonrepeat efforts, and the activity profiles of players varied according to positional groups and period of the match.
Miñano-Espin et al., 2017 (31) Observational study	Real Madrid matches (n = 149): data from Real Madrid and opposing teams’ male players	HIR = 21.1–24.0 km·h⁻¹ Sprint > 24 km·h⁻¹ High Speed Runs and Sprints (n)	Match demand: HIR distance = 269 m Real Madrid vs. 285 m opposing team; Sprint distance = 245 m vs. 248 m; High Intensity Runs = 11; Sprints = 20. Players from Real Madrid covered shorter distances in HIR and Sprint and executed less sprints than players from the opposing team. No differences were revealed in the HIR and Sprint distances or the number on high intensity runs and sprints performed by players from Real Madrid depending on the quality of the opposition.
Abbott et al., 2018 (32) Observational study	Premier League 2 under 23 professional male players (Brighton and Hove Albion) (n = 46) Matches (n = 22) LSG, MSG, SSG (n = 39)	VHSR = 100% MAS – 30% ASR Sprint >30% ASR Mean and 1-min peak values	Despite eliciting significantly higher average total distances compared with competition, LSGs produced significantly lower peak total distance relative to the competition. For VHSR and sprinting, LSGs elicited similar average intensities to competition; however, peak intensities were significantly lower than competition. VHSR and sprinting distances increased with game format, with LSGs (>7v7) producing the highest intensities. Only LSGs were able to replicate competitive demands, with SSGs and MSGs significantly below competitive values for all positions.
Baptista et al., 2018 (33) Observational study	Professional male soccer players (Tromsø Idrettslag) (n = 18) Official matches (n = 23)	HIR ≥ 19.8 km·h⁻¹ Sprint ≥ 25.2 km·h⁻¹ Number of HIR and sprint efforts of various length (1–5, 6–10, 11–15, 16–20, 21–25, 26–30, 31–35, 36–40, 41–45, 46–50 m) CoD counts	CB had the lowest values of all positions in both variables but especially pronounced in Sprint (1 m·min⁻¹) when compared with CF (2.5 m·min⁻¹). HIR analysis: CF presented higher values in 26-30 m than all the other positions, while distances of 36–40 and 46–50 m were covered more times by FB. CB were the players with lowest values in these longer distances (36–40 and 46–50). Sprint analysis: CB, FB, CM and WM performed higher number of 1–5 m sprints, while CF covered higher number of 6–10 m sprints. The most common distance covered in HIR for CB, CM, WM and CF was 1–5 m, but for FB was 6–10 m.
Malone et al., 2018 (34) Longitudinal observational study	Professional male soccer players (Benfica) (n = 37) Weeks (n = 48)	HSR > 14.4 km·h⁻¹ Sprint > 19.8 km·h⁻¹	When HSR and SR distances are considered independently of aerobic fitness and previous training load history, a U-shaped association exists for distance completed at these speeds and subsequent injury risk. Players with higher aerobic fitness were able to complete increased weekly HSR and SR distances with a reduced injury risk. Higher 21-day chronic sRPE-TL (≥2,584 AU) allow exposure to greater volumes of HSR and SR, which in turn offers a protective effect against injury. 1-week safer zone: HSR = 700–750 m, SR = 200–350 m. Absolute weekly change safer zone: HSR < 100 m, SR < 50 m 3:21 ACWR safer zone: HSR < 0.85, SR = 0.71–0.85
Scott and Lovell, 2018 (35) Observational study	International women's soccer players (n = 22)	HSR > 12.67 km·h⁻¹ (HRDP) VHSR > 17.82 km·h⁻¹ (MAS)	In this approach, each players running speed corresponding to HRDP, together with their MAS determined from the VAM-EVAL, were used as the entry-points to the HSR and VHSR zones. Individualised speed thresholds for external load monitoring were not able to better quantify the dose-response of football training during a 21-day training camp in players representing the highest level of women's football. Quantifying the external load using players’ peak sprinting speed demonstrated a lower capacity to determine the dose-response of training, with consistently lower associations with heart rate and RPE.
Martín-García et al., 2018 (36) Observational study	Professional male soccer players (Barcelona 2nd team) (n = 24) Matches (n = 37) + training weeks (1 game per week) (n = 42)	HSR > 19.8 km·h⁻¹ Sprint > 25.2 km·h⁻¹	When comparing starters and non-starters at MD + 1, thanks to the SSG approach used in players with limited game time, non-starters demonstrated greater external loads for TD, HMLD, AMP, ACC, and DEC, but not for HSR or SR. The session that produced the greatest HSR (43%) and SR (45%) distances relative to competition was MD-4. HSR and SR distances are the metrics illustrating the most variability within the microcycle (>80%), which is consistent with the variability found in SSG formats (60–140%), but lower than competition variability (20–30%).
Martín-García et al., 2018 (37) Observational study	Professional male soccer players (Barcelona 2nd team) (n = 23) Official matches (n = 37)	HSR > 19.8 km·h⁻¹ Sprint > 25.2 km·h⁻¹ 1′, 3′, 5′ and 10′ MIP using TD, HMLD e AMP as the criterion variables	HSR: FB covered the greatest distance, reaching values of 47.2 ± 24.0 m·min⁻¹ in the 1′ period. 1′ MIP demand using TD as the criterion variable (positions’ average): TD = 191.6 ± 19.7, HSR = 38.3 ± 23.1, Sprint = 10.6 ± 15.6, ACC > 3 m·s² = 2.8 ± 1.6, DEC < −3 m·s⁻² = 3.5 ± 1.6 1′ MIP demand using HMLD as the criterion variable (positions’ average): TD = 173.5 ± 26.0, HSR = 49.9 ± 19.8, Sprint = 16.6 ± 17.4, ACC > 3 m·s² = 3.5 ± 1.7, DEC < −3 m·s⁻² = 3.6 ± 1.7
Soroka, 2018 (38) Observational study	2010 World Cup male players (n = 599)	HIR = 19.9–25.2 km·h⁻¹ Sprint > 25.2 km·h⁻¹	The largest amount of HIR and Sprint distance was found in midfielders, which did not correspond to studies carried out on players of the Premier League and Primera Division in 2006–2007 (strikers covered the largest sprint distance) (Carling 2008).
Clemente et al., 2019 (39) Observational study	Professional male soccer players (Portuguese Second League) (n = 23) 5v5 + GK in 40 × 31 m (124 m²) 6v6 + GK in 45 × 32 m (120 m²) 9v9 + GK in 70 × 50 m (194 m²)	Running = 14–20 km·h⁻¹ Sprinting > 20 km·h⁻¹	Greater values for sprinting distance were found in the full match compared to 5vs5 + GK (d = 3.673, strong effect), 6vs6 + GK (d = 2.606, moderate effect) and 9vs9 + GK (d = 1.903, moderate effect) sided games. MSG are not appropriate for simulating the sprinting conditions of official full matches. LSG (9vs9 + GK) simulate official full matches more accurately than the other sided-games that were studied (5vs5 + GK and 6vs6 + GK).
Clemente et al., 2019 (40) Observational study	Professional male soccer players (Sporting Lisbona) (n = 27) Training weeks (with 3-4-5 training sessions + 1 game) (n = 22)	RD = 14.0–19.9 km·h⁻¹ HSR = 20.0–24.9 km·h⁻¹ Sprint > 25.0 km·h⁻¹ TMr = Training/Match ratio	It was observed that specific variables (e.g., HSR distance and sprinting distance) were associated with substantially lower ratios than other variables. The TMr for RD and HSR distance were 1.2 ± 0.7 and 1.1 ± 0.8, respectively, in 3-days week and 2.3 ± 1.3 and 2.3 ± 1.5, respectively, in 5-days week. This suggests that the number of training sessions tend to emphasize the stimuli of overall distance and that the demand of three days of training is very similar to the demand of one match. Some determinant external load measures (e.g., HSR or sprinting) are clearly undertrained comparing with more prevalent measures (e.g., TD, ACC or DEC): SSG increase the frequency of ACC/DEC while decreasing opportunities to perform HSR or sprinting.
Dalen et al., 2019 (41) Observational study	Male soccer players from an elite Norwegian league team (n = 26) Matches (n = 18) SSGs (28 4vs4 + 28 6vs6) (n = 56)	HIR > 19.8 km·h⁻¹ Sprint > 25.2 km·h⁻¹	HIR (m·min⁻¹) in match peak (5 min most demanding period), match mean, 4v4 and 6v6 = 19 ± 3.5, 8.3 ± 2.1, 2.7 ± 0.9, 3.7 ± 2.1. Sprint = 8.8 ± 4, 1.7 ± 0.7, 0.1 ± 0.1, 0.2 ± 0.5. The smaller pitch used for SSGs may lead to a different work pattern from match play, which is supported by the relatively low HIR and sprint distances observed during SSGs in this study. 4vs4 games are a good method of training acceleration and player load tolerance, but SSGs do not represent a good method of training HIR.
Hills et al., 2019 (42) Observational study	Championship male soccer players (Hull City Tigers) (n = 17) Matches (35 single observations) (n = 13)	MSR > 14.4 ≤ 19.8 km·h⁻¹ HSR > 19.8 ≤ 25.2 km·h⁻¹ Sprint > 25.2 km·h⁻¹	Relative TD (+13.4 m·min⁻¹) and HSR (+0.4 m·min⁻¹) distances covered during rewarm-ups increased with proximity to pitch-entry. Very few HSR and no sprint distance were performed during each warmup or rewarm-up bout. Substitutes covered greater TD (+67 to +93 m) and HSR (+14 to +33 m) distances during the first 5 min of match-play versus all subsequent epochs.
Jones et al., 2019 (43) Observational study	Professional male soccer players (English Football League One) (n = 37) Matches partitioned in 3 fixture congestion scenarios (n = 79)	HID = 19.9–25.2 km·h⁻¹ Sprint > 25.2 km·h⁻¹	The Linear Mixed Model did not identify significant interactions between position, fixture congestion scenario and time period (p = 0.549), position and fixture congestion scenario (p = 0.481), nor fixture congestion scenario and time period (p = 0.162).
Modric et al., 2019 (44) Observational study	Professional male soccer players from Croatian Soccer League (6th of 10) (n = 101) Matches (n = 14)	RD = 14.4–19.7 km·h⁻¹ HSR = 19.8–25.1 km·h⁻¹ Sprint > 25.2 km·h⁻¹ InStat technical index	Math demand: HSR = 462 ± 160 m; Sprint = 156 ± 97 m. Association between the running performance of players involved in certain playing positions and overall game performance (InStat index). Specifically, it seems that CD distance in the running zone and number of high-intensity accelerations, FB number of decelerations, and FW sprinting distance are crucial physical requirements of team success.
Oliveira et al., 2019 (45) Observational study	Elite male soccer players participating in UEFA Champions League (n = 19) Weeks (n = 39) + matches (n = 50)	HSD > 19 km·h⁻¹ Hooper Index	Although there are some significant differences between mesocycles, there was minor variation across the season for the internal and external TL variables used. MD-1 presented a reduction of external TL during in-season match-day-minus training comparison.
Park et al., 2019 (46) Observational study	International female players (n = 27) International matches (n = 52)	HSR: ≥ 12.5 km·h⁻¹ VHSR ≥ 19 km·h⁻¹ Sprint ≥ 22.5 km·h⁻¹	PS in elite women = 29.0 ± 1.5 km·h⁻¹ k-means clustering and Gaussian mixture modelling were not appropriate for soccer given the limited instances in which players move at velocities associated with sprinting, which are often considered key physical performance indicators. A spectral Clustering technique with application of a β = 0.1 smoothing factor derived new thresholds featuring both logical validity and analysis rigor. Similar analyses may be warranted to determine appropriate velocity zones for other sports and youth populations.
Rago et al., 2019 (47) Observational study	Italian Serie B male soccer players (n = 13)	MSR = arbitrary 14.4–19.8 km·h⁻¹ or individualised 80–99% MAS HSR = 19.9–25.1 km·h⁻¹ or 100% MAS – 29% ASR Sprint = ≥25.2 km·h⁻¹ or ≥30% ASR	Perceptual responses (RPE) were moderately correlated to MSR and HSR quantified using the arbitrary method (p < 0.05; r = 0.53–0.59). However, the magnitude of correlations tended to increase when the individualised method was used (p < 0.05; r = 0.58–0.67). Distance covered by sprinting was moderately correlated to perceptual responses only when the individualised method was used (p < 0.05; 0.55 [0.05; 0.83] and 0.53 [0.02; 0.82]). The magnitude of the relationships between ETL and RPE parameters appear to slightly strengthen when ETL are adjusted to individual fitness capacities, with special emphasis on cardiorespiratory fitness (MAS).
Ramos et al., 2019 (48) Observational study	Under 17 (n = 14), Under 20 (n = 14) and adult (n = 17) international women soccer players	High intensity (HID) = 15.6–20 km·h⁻¹ Sprint > 20 km·h⁻¹	Likely to almost certainly differences among all age brackets for the HID and sprint were found (adult > U20 > U17, ES varying from 0.41 [20.23–1.06] to 3.69 [2.63–4.76]), except for the comparison between U17 and U20 for sprint where the differences were rated as unclear. HID: adult (756 m) > U20 (688 m) > U17 (485 m). Sprint: adult (307 m) > U20 (223 m) ≈ U17 (192 m).
Asian-Clemente et al., 2020 (49) Observational study	Under 19 professional male soccer players from an elite Spanish first division soccer club (n = 17) SSGs (5c5c5 + 2) in 1 single 35 × 35 m pitch or in 2 28.5 × 28.5 m contiguous pitches (n = 4)	HSD = 18–21 km·h⁻¹ VHSD > 21 km·h⁻¹	VHSD (m·min⁻¹): 2.5 ± 1.8 in 35 × 35 m, 12.8 ± 6.3 using 2 contiguous 28.5 × 28.5 m pitches, 4.6 ± 2.3 in official matches. When soccer is played in smaller relative areas than those used for official games, the ACC and DEC will be increased. Similarly, forcing players to change spaces quickly during SSGs promotes greater running activity, with higher HSD and VHSD covered per player. Although most of the running demands during matches were simulated with the proposed SSGs, it may be necessary to design other types of tasks to train for peak speed and distance covered at sprint speed.
Kelly et al., 2020 (50) Observational study	English Premier League male players (Manchester United) (n = 26) Entire season (n = 1)	HSD > 14.4 km·h⁻¹ VHSD = 19.8–25.2 km·h⁻¹	HSD was greater 3 days before a game (MD-3) vs MD-1 (95% CI, 140–336 m) while VHSD was greater on MD-3 and MD-2 than MD-1 (95% CI range, 8–62 m; p < 0.001). HSD was similar between mesocycles during the whole season suggesting that training schedules employed in elite soccer may be highly repetitive likely reflecting the nature of the competition demands.
Scott et al., 2020 (51) Observational study	Elite female players from National Women's Soccer League (NWSL, United States) (n = 36) Match observations (n = 208, 11 ± 6 per player)	HSR ≥ 12.5 km·h⁻¹ or 60% vIFT (50% PS) VHSR ≥ 19 km·h⁻¹ or 80% vIFT (65% PS) Sprint ≥ 22.5 km·h⁻¹ or 30% ASR (80% PS)	Subjective ratings of fatigue and wellness are not sensitive to substantial within-player changes in match physical performance. HSR, VHSR, and SR thresholds customized for individual players athletic qualities did not improve the dose-response relationship between external load and wellness ratings. PS in elite women = 30.5 ± 1.8 km·h⁻¹ (mean of 5 different roles). Match demand (ABS): HSR = 2,401 ± 454 m; VHSR = 398 ± 143 m; SR = 122 ± 69 m.
Altmann et al., 2021 (52) Observational study	German Bundesliga male players (n = 25) Match observations (n = 163)	HID = 17.0–23.99 km·h⁻¹ Sprint ≥ 24.0 km·h⁻¹	CM showed both the largest total (11.66 ± 0.92 km, ES = 0.68–1.86) and HID (1.57 ± 0.83 km, ES = 0.08–0.84) compared to all other positions, WM demonstrated the largest sprinting distance (0.42 ± 0.14 km, ES = 0.34–2.39). Some professional soccer players will likely incur differences in the composition of physical match performance when switching positions and therefore should pay special consideration for such differences in the training and recovery process of these players.
Oliva-Lozano et al., 2022 (53) Observational study	Spanish LaLiga male players (n = 277) Match observations (n = 1,252)	Maximal Intensity Sprint: when an acceleration occurred from 14 km·h⁻¹ and the player got to exceed 30 km·h⁻¹ for 0.2 s.	Professional soccer players need to be prepared for maximal intensity sprints in the first period of the match as well as maximal intensity sprints under high fatigue conditions given the frequency of sprints in the last period of the match. Training drills should be designed with a special focus on non-linear sprints without possession of the ball, based on the main tactical purpose of each position (e.g., CD: interceptions; CM: recovery runs; FB, WM and FW: run the channel).

ABS, absolute thresholds; ACC, accelerations; ACWR, acute:chronic workload ratio; AMP, average metabolic power; ASR, anaerobic speed reserve [MSS – MAS]; AU, arbitrary units; CB, central backs; CD, central defenders; CF, central forwards; CM, central midfielders; CoD, change of direction; DEC, decelerations; ES, effect size; ETL, external training load; FB, full-backs; FW, forwards; GK, goalkeepers; HID, high-intensity distance; HIR, high-intensity running; HMLD, high metabolic load distance; HR, heart rate; HRDP, heart rate deflection point; HSD, high-speed distance; HSR, high-speed running; LSG, large sided game; MAS, maximal aerobic speed; MD −/+ n, match day minus/plus n days, i.e., n days before/after the match; MIP, maximum intensity period; MSG, medium sided game; MSR, moderate speed running; PS, peak speed; RD, running distance; RPE, rating of perceived exertion; SR, sprint running; sRPE-TL, session rating of perceived exertion training load; SSG, small sided game; TD, total distance; TMr, training/match ratio; UEFA, Union of European Football Associations; VAM-EVAL, a modified version of the Montreal track test; VHSD, very high-speed distance; VHSR, very high-speed running; VHIR, very high-intensity running; vIFT, final velocity of the 30:15 intermittent fitness test; WM, wide midfielder.