Biomechanical Performance Factors in the Track and Field Sprint Start: A Systematic Review

In athletics sprint events, the block start performance can be fundamental to the outcome of a race. This Systematic Review aims to identify biomechanical factors of critical importance to the block start and subsequent first two steps performance. A systematic search of relevant English-language articles was performed on three scientific databases (PubMed, SPORTDiscus, and Web of Science) to identify peer-reviewed articles published until June 2021. The keywords “Block Start”, “Track and Field”, “Sprint Running”, and “Kinetics and Kinematics” were paired with all possible combinations. Studies reporting biomechanical analysis of the block start and/or first two steps, with track and field sprinters and reporting PB100m were sought for inclusion and analysis. Thirty-six full-text articles were reviewed. Several biomechanical determinants of sprinters have been identified. In the “Set” position, an anthropometry-driven block setting facilitating the hip extension and a rear leg contribution should be encouraged. At the push-off, a rapid extension of both hips and greater force production seems to be important. After block exiting, shorter flight times and greater propulsive forces are the main features of best sprinters. This systematic review emphasizes important findings and recommendations that may be relevant for researchers and coaches. Future research should focus on upper limbs behavior and on the analysis of the training drills used to improve starting performance.


Introduction
The 100 m race is perhaps the highlight of the Olympic Games, as it defines who is the fastest man and woman in the world. In this type of event, the block start performance and the subsequent first two steps can be of critical importance since they have a direct influence on the overall 100 m time [1][2][3][4][5][6][7][8]. Given the importance of the sprint start, a new body of research has emerged in the past two decades that involved advanced technologies, highprecision methods, and sprinters of a higher performance level. For this reason, several technical (kinematic) and dynamic (kinetic) aspects are currently identified as determinant factors for starting block phase and initial sprint acceleration performances [1,4,6,[9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25]. However, the concepts, outcomes, and findings between studies are sometimes inconsistent and difficult to interpret and conclude from. These inconsistencies may be accounted for by different study designs, methods, technologies of measure (e.g., external reaction forces under or on the blocks), statistical analyses, or more importantly, the ambiguity between samples of sprinters with different performance levels (e.g., elite, sub-elite, well-trained or trained) and/or between-group analyses based on the overall 100 m performance (i.e., personal best at 100 m-PB100m), and not on block performance. Although two important narrative reviews have already been published [26,27], to our knowledge, no previous review conducted a systematic search of literature exploring the inter-individual variability on block start performance across different performance levels. Thus, the main purposes of this systematic review were: (a) determine the biomechanical parameters of greatest influence on the sprint start, including the "set" position and push-off phase, and the first two steps of initial sprint acceleration and (b) identify the kinematic and kinetic biomechanical variables that best differentiate sprinters of different performance levels in each of those three phases of the sprint start. Considering the impact of the sprint in the sports field and the absence of systematic studies on the kinematics and kinetics factors that determine success in block starts and initial sprint acceleration, we hypothesized that this systematic review will have a relevant impact on researchers to better design experimental/intervention studies, as well as constituting relevant support for coaches and athletes in the definition of efficient strategies for performance in the 100 m race.

Article Search, Eligibility, Inclusion, and Exclusion Criteria
The systematic search of relevant articles was conducted based on PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analyses) guidelines [28]. PubMed, Web of Science, and SPORTDiscus databases were searched for the following mesh terms: "Block Start" OR "Track and Field" OR "Sprint Running" OR "Acceleration" AND "Kinetics and Kinematics" pairing them with all possible combinations. In addition, filters for 'English' and 'articles' have been applied. The last search took place on 30 June 2021.
The inclusion criteria were: publications in English; original observational and experimental studies published in peer-reviewed journals; studies mainly focused on the block phase and/or one or two of the subsequent stance phases concerning kinematic and kinetic variables; and studies that included track and field sprinters with the indication of their PB100m. The following types of records were excluded: conference abstracts; studies focused exclusively on the acceleration phase (beyond the first two stance phases) or mainly focused on limitations imposed by motor and neurological impairments; studies reporting data referring to samples evaluated in previously published papers; studies not mentioning the performance level of the sprinters through their PB100m; case reports; and studies without reference to biomechanical variables.
The records identified from the databases with the aforementioned mesh terms were exported to the reference manager software EndNote X8 that eliminated duplicates. All articles' eligibility was then assessed independently by two reviewers' authors (JMA and FC). The articles identified were first screened by title and abstract for relevance. Studies that raised any uncertainty in exclusion were conservatively retained for subsequent fulltext review. The full text of the articles selected as relevant or having raised uncertainty in exclusion was read and further scrutinized for meeting the inclusion criteria and their quality was evaluated. Disagreements on final inclusion or exclusion of studies were resolved by consensus, and if disagreement persisted, a third reviewer (first author, MJV) was available for adjudication. Articles that did not meet the selection criteria or presented a quality score below 50% were excluded.

Quality of the Studies
The study quality of each publication was evaluated according to the guidelines of the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Initiative [29]. This analysis was based on 22 items. Title and abstract. Introduction: background and rationale. Methods: study design, setting, participants, variables, data sources, bias, sample size, quantitative variables, and statistical methods. Results: participants, descriptive data, outcome data, main results, and other analyses. Discussion: key results, limitations, interpretation, and generalizability. Funding. These criteria were scored on a binary scale (1 = yes, 0 = no) independently by two of the authors, and a quality score was then calculated for each study by adding its binary scores and dividing the result by the maximum possible score the study could have achieved. This was then expressed as a percentage to reflect a measure of methodological quality. The quality scores were classified as follows (a) low methodological quality for scores < 50%; (b) good methodological quality for scores between 50% and 75%; and (c) excellent methodological quality for scores > 75%. The studies with a score lower than 50% [30] were excluded from the systematic review. The inter-rater reliability analysis was evaluated by the Cohen's Kappa for nominal variables (2 dimensions) [31]. Standards for strength of agreement for the kappa coefficient were: ≤0 = poor, 0.01-0.20 = slight, 0.21-0.40 = fair, 0.41-0.60 = moderate, 0.61-0.80 = substantial, and 0.81-1 = almost perfect [32].

Data Extraction
An Excel form was used for data extraction. Of each manuscript selected for review, the following information was extracted from each included study: (a) the primary focus of study, means the phase of sprint start, e.g., block phase, first stance, and study design; (b) the main purpose, e.g., associations between biomechanical variables of starting blocks and the sprint start performance, comparing athletes of different performance levels, comparing different footplate spacing and block angles; (c) type of kinematic and kinetic analyses systems used-two dimensional (2D) or three dimensional (3D) analysis and starting blocks instrumented or placed on force platforms; (d) study sample-the number per gender of participants, and per level of expertise of participants according with the authors, and their PB100m; (e) biomechanical measurement protocols-the variables used to characterize the biomechanical factors of sprint start, number and distance of repeated trials; and (f) key findings of sprint start kinematic and kinetic factors.

Search Results
The initial search identified 756 titles in the described databases. With the reference manager software, 406 duplicates were eliminated automatically. The remaining 350 articles were then screened according to title and abstract for relevance, resulting in another 289 studies being eliminated from the database. The full text of the remaining 61 articles was read and another 22 were rejected for not meeting the inclusion criteria defined for the current study and 3 studies were excluded for not meeting the quality criteria (quality index < 50%). A total of 36 studies was fully reviewed.
Studies were excluded in the screening stage due to not including track and field athletes or sprint starts using starting blocks (n = 289). In the eligibility stage, there were several reasons for exclusion, namely studies with results focused exclusively on the acceleration phase (n = 8), case studies (n = 4), studies reporting data referring to samples of previously published papers (n = 3) or mainly focused on the limitations of disability (n = 3), lack of information about the PB100m (n = 2) and studies presenting only results for electromyography and reaction time data (n = 2). Figure 1 presents the complete flow diagram.

Quality of Studies
In the evaluation of methodological quality, the inter-rater reliability analysis achieved a Kappa value of 0.91 (0.84-0.98), indicating almost perfect agreement between raters. The mean quality score of the included studies was 74.92%. None of the studies achieved the maximum score of 100% and 3 studies (excluded) scored below 50%. Sixteen studies were classified with good methodological quality (quality score between 50 and 75%), while 20 studies had excellent methodological quality (quality score > 75%). The main deficiencies in methodological quality were related to the estimation of sample size and study limitations discussion.

Quality of Studies
In the evaluation of methodological quality, the inter-rater reliability analysis achieved a Kappa value of 0.91 (0.84-0.98), indicating almost perfect agreement between raters. The mean quality score of the included studies was 74.92%. None of the studies achieved the maximum score of 100% and 3 studies (excluded) scored below 50%. Sixteen studies were classified with good methodological quality (quality score between 50 and 75%), while 20 studies had excellent methodological quality (quality score > 75%). The main deficiencies in methodological quality were related to the estimation of sample size and study limitations discussion.

Basic Characteristics of Included Studies
Fifteen studies [2,3,[10][11][12]17,20,21,23,25,[33][34][35][36][37] focused specifically on the block phase, 18 studies [1,[4][5][6][7][8][13][14][15][16]18,19,24,[38][39][40][41][42] on the block phase and, at least one of the subsequent two flight and stance phases, and 3 studies [9,22,43] on the initial acceleration (the first and/or the second step). A summary of all the individual studies reviewed is presented in Table 1. The asymmetrical nature of the block phase is most pertinent at the knee joint, and the leg extensor energy is predominantly generated at the hip joint in both the front and the rear block whereas during 1st stance, energy generation favors the ankle joint as a result of a significant reduction in relative hip work. A widened stance width at the "set" position affects the hip-joint kinematics in both legs and enhanced the hip power generation in the rear leg during the block start phase. However, when considering sprinting performance during the whole block start phase, there was no significant effect of the widened stance width on block-induced power and the subsequent sprint time. Describes for the 1st time the mechanical characteristics of the acceleration phase in elite and sub-elite sprinters: (i) while step length increases regularly during the acceleration phase, step frequency is almost instantaneously leveled at the maximal possibility of elite athletes; (ii) F-V and P-V relationships during sprints were well described by linear and quadratic models, respectively; and (iii) the effectiveness of force application greatly accounts for the differences in performance among highly trained athletes.

74.31
Milanese, Bertucco [41] Block phase and first 2 stance phases. 3D kinematic full-body model. During the 1st step, maximal power was predominately generated by the hip (54%) followed by the knee (31%) and the ankle (15%). The importance of power generation at the knee decreased at second stance since it only accounted for 9% of total power generation and the importance of the ankle increased up to 38%.

64.77
Aerenhouts, Delecluse [1] Block phase and initial acceleration (first 5 steps). Instrumented start blocks and a universal laser velocity sensor. The higher muscularity of senior athletes did not result in significantly higher forces against the starting blocks nor block velocity compared with the junior athletes. The more muscular senior athletes had a better running acceleration than the junior athletes. In female athletes, a higher body fat percentage negatively correlated with 1st step length.

79.32
Slawinski, Dumas [8] Block phase and 1st step. The pre-tensed and conventional starts produced similar performance. The increased propulsive force exerted through the legs in the early part of the block acceleration phase in the pre-tensed starts was counteracted by an increased backward force exerted through the hands during the same period.

72.95
Mero, Kuitunen [19] "Set" position (block phase and 1st step Decreasing front block obliquity induced neural and mechanical modifications that contribute to increasing the block start velocity without any increase in the duration of the push-off phase. In the 1st contact after leaving the blocks there was a significant braking phase and the force produced in the propulsion phase was associated with running velocity; Muscle strength strongly affected running velocity in sprint start.

57.27
2D-two-dimensional analysis; 3D-three-dimensional analysis; AHEP-average horizontal external power; BM-body mass; CM-center of mass; COP-center of pressure; EMG-electromyography; F-female sample; F-V-force-velocity; GRF-ground reaction forces; MTU-muscle-tendon unit; M-male sample; P-V-power-velocity; ROM-range of motion; WR-world record; (a) 100 m world record at the study time was 9.58 s; (b) 100 m U20 world record at the study time was 9.97 s; (c) all sample was divided into 3 groups according to the Cormic Index (12 brachycormic, 19 metricormic, and 11 macrocormic); (d) sample divided into two groups: 5 elite sprinters and remaining 52 sprinters; (e) all subjects included in a single experimental group; (f) sample divided into 2 experimental groups: adult/senior vs. junior sprinters; (g) sample divided into 4 experimental groups.

Data Organization and Analysis
There was a very large diversity of kinematic and kinetic variables reported among selected studies. Since it is impossible to discuss them all, we will highlight those reported as explicative of high levels of the sprint start performance and that best differentiate faster from slower sprinters. Based on the main findings highlighted in Table 1, the explanatory variables of superior performance levels were identified and systematized in a sequence of tables in Appendixes A-C, related to the "Set" position (Appendix A Table A1), block phase (Appendix B Tables A2 and A3), and first two steps of the initial acceleration (Appendix C Tables A4 and A5). With this strategy of results presentation, it is expected that readers will have access to the primary data extracted from all the studies included in the systematic review. Therefore, Appendix A Table A1 summarizes the kinematic variables in the "Set" position, showing that anthropometry-driven block setting and muscle-tendon unit (MTU) length have an important role in the block start performance. Furthermore, faster sprinters tend to move their center of mass (CM) closer to the starting line and closer to the ground. Concerning joint angles, the knee angular position seems to be a greater performance predictor than any other lower limb joint. At the push-off phase (Appendix B Tables A2 and A3, for kinematic and kinetic variables, respectively) a rear hip extension range of motion (ROM) and a rapid extension of both hips appear to be positively associated with block performance. Moreover, greater average force production during the push against the blocks, especially from the rear leg and particularly the hip, appears to be important for performance. A posterior COP location on block surfaces can also improve sprint performance. Immediately after exiting the blocks, shorter first flight durations and longer first stance durations (allowing more time to generate propulsive force) are the kinematic features of best sprinters (Table A4). During the first two steps of initial acceleration, higher levels of performance seem to be associated with shorter flight times, longer contact times, and the ability to extend the knee throughout both stance phases (Table A5).

Discussion
This paper systematically reviews the kinematic and kinetic biomechanical variables of the block start and initial sprint acceleration phase that influence performance and best differentiate sprinters of different levels. Despite the large number of variables reported in the reviewed studies it was possible to identify some that effectively best describe the influential factors of these events as they are associated with better performance outcomes or best differentiate sprinters of different performance levels. However, notice should be made to the difficulty in analyzing data between studies as there are still no standards for reporting the data, such as measurement units (e.g., m vs. cm) [12,17,18,35], joint angular measurement norms and conventions [3,4,6,12,13,36,38] and/or data normalization methodologies (e.g., for fullheight/lower limb length, body mass/body weight) [2,4,17,22,24,25]. Additionally, there is some subjectivity associated with inconsistent descriptors of performance level [26], confirmed by the variability of the sprinter's classifications used (e.g., from just sprinters to well-trained sprinters, elite sprinters, world-class sprinters, or high-level sprinters) [5,7,16,22,36,38,42]. Another critical factor that somehow may influence data variability between studies is the period of the season in which the data collection took place (e.g., prior to the competition phase of the indoor season vs. during the competitive indoor season or beginning of the summer season) [18].
To better understand the determinant factors of sprint start, the findings from the reviewed studies have been organized into three focuses: (i) the "set" position, (ii) the push-off phase, and (iii) the first two steps of initial acceleration, according to the data presented in Appendixs A-C.

The "Set" Position
The "Set" position is the first performance key factor in the block start performance because it depends on block settings and the body posture assumed by sprinters. For the question: "Is there one optimal "Set" position which should be adopted by sprinters?" the answer seems to be no. The researched studies [3,38] showed that it is not an important differentiating factor of performance, since it does not present any correlation with PB100m or normalized block power [3]. However, there are some interesting aspects that sprinters should look out for in a more effective "Set" position [5,12]. The ideal "Set" position depends on the individual anthropometric features [12], strength [38], and morphologic characteristics and motor abilities [13].

Block Settings
The "Set" position depends largely on the anteroposterior block distance, which defines the type of start used. There are three types of block starts based on inter-block spacing: bunched-less than 0.30 m; medium-0.30 to 0.50 m; and elongated-greater than 0.50 m [27,37].
Studies that reported block spacing based on the individual sprinter's preferences [5,12,13,18,35] reported distances between 23.5 ± 1.9 cm (for female sprinters; PB100: 11.97 ± 2.6 s) [13] and 32 ± 5 cm (for male sprinters; PB100m: 10.79 ± 0.21) [18]. This suggests that most sprinters adopt distances within or very close to the bunched start type, favoring CM positioning closer to the starting line [7,38]. Slawinski, Dumas [8] have demonstrated that elongated start settings increase the block velocity (i.e., horizontal CM velocity at the block clearing [7]), but linked to an increase in the pushing time on the blocks which implies a significantly worse performance at 5 and 10 m compared to the bunched start. The same authors showed that the medium start offers the best compromise between the pushing time and the force exerted on the blocks, allowing better times at 10 m [8]. Additionally, more recently, Cavedon, Sandri [12] have demonstrated that the anthropometry-driven block setting based on the sprinter's leg length has an important role in the block start performance leading to a postural adaptation that promotes several kinematic and kinetic advantages [12]. Adjusting inter-block spacing to the relative lengths of the sprinter's trunk and lower limbs (increasing 25.02% the usually bunched start inter-block spacing), allows greater force and impulse on the rear leg and greater total normalized average horizontal external power (NAHEP) [12], the latter one identified as the best descriptor of starting block performance [2].
Other blocks setting features that should be considered in the "set" position are the feet plate obliquity and the amount of pre-tension exerted on the blocks prior to the gunshot. The block inclination (relative to the track) affects the plantar flexor muscle-tendon units' (MTU) initial lengths and determines the muscle mechanics and the external force parameters during the block phase [19,25,34]. Faster sprinters presumably produce the peak torque at longer MTU lengths and adopting a more crouched position would allow them to produce a higher force on the block phase [38]. Research data shows that reductions in both footplates' inclinations (from 65 to 40 • ), meaning more muscle-tendon pre-stretch, lead to acute increases in block velocity and higher peak joint moments and powers, especially in the ankle [19]. Reductions in front block inclination alone (from 70 to 30 • ) also acutely increase block velocity without affecting push-off phase duration [34]. In another study [25], however, a greater mean rear block horizontal force was achieved by switching the rear foot to a steeper position (to 65 • ). This potential conflict between evidence might have arisen from differences in the location of the COP and the length of the footplates' surface between studies since a better sprint start performance is accomplished with a higher and more to the rear COP on the starting block surface [20,35]. Conversely, a pre-tensioned start does not seem to yield a performance advantage over a conventional start, because the increase in the propulsive force of the lower limbs is reversed by an increase in the back force exerted through the hands during the same period [17].

Sprinter Body Posture
Apart from block configuration, the choice of the sprinter's body posture also determines the effectiveness of the "Set" position on the subsequent block push-off phase. The horizontal distance between starting line and the vertical projection of the CM to the ground in the "Set" position (XCM) [7] is a factor that differentiates sprinters with different performance levels. As said before, faster sprinters tend to move their CM closer to the starting line [7,38] and closer to the ground [38]. Elite (PB100: 10.27 ± 0.14 s) and well-trained (PB100: 11.31 ± 0.28 s) male sprinters showed XCM of 22.9 and 27.8 cm, re-spectively [7]. Likewise, world-class (PB100: 11.10 ± 0.17 s) and elite (PB100: 11.95 ± 0.24 s) female sprinters presented XCM of 16.2 and 24.8 cm, respectively [38]. This more crouched position is only possible due to the high explosive strength of best sprinters, which allows them to produce higher levels of strength in the blocks [38] and reduce the horizontal travel distance of the CM. This body position is complemented by a more advanced shoulder position, putting more tension on the arms, allowing greater blocking speed during the subsequent phase [7].
Related to sprinter joint angles configuration in the "set" position, Milanese and Bertucco [41] have shown that horizontal CM velocity at the block take-off and along the first two steps increases significantly when the rear knee angle is set to 90 • instead of 135 • or 115 • . A 90 • rear knee angle allows for a better push-off of the rear leg than larger angles, showing such condition may be a strategy that allows some elite sprinters to maximize their strength capacity [41]. A more flexed front knee may facilitate the optimal joint moment production, but only in sprinters with exceptionally high levels of explosive strength [38].

The Push-Off Phase
The "block-phase" or "push-off phase" in the starting blocks initiates immediately after the gunshot and is considered a complex motor task that helps to determine sprint start performance [1]. Reaction time is the first factor in the time sequence of the block phase and it is the period from the gun signal to the first measurable change of pressure detected in the instrumented blocks [16]. While a sprinter's ability to react is undeniably important, it is related to the information-processing mechanisms that do not seem to correlate with the performance level [7,45] and, therefore, is beyond the scope of our review (for a review of factors that affect response times, see Milloz, Hayes [46]). Having reacted, the aim of the block phase is to maximize horizontal velocity in as little time as possible. The motion variables during the block phase are, therefore, the focus of this section.

Push-Off Kinematics Analysis
The efficiency of the starting action depends mainly on the compromise between horizontal start velocity (or block velocity) and the block time (referring to the time elapsing from the first movement at the "set" position to the exiting from the block [7]), resulting in the horizontal start acceleration [13]. Despite the horizontal block velocity could be considered the main parameter for an efficient sprint start [13], it cannot be used solely [2] because an increased block velocity could be due to either an increase in the net propulsion force generated or to an increased push-off duration [2,18]. Thus, best sprinters tend to present higher block velocity and greater block acceleration than slower sprinters [1,5,7,13,16,22,39,42], because they are able to produce a greater impulse in a shorter time [2,5,36] and optimize their force production on the blocks [16,19]. In fact, if sprinters increase their anteroposterior force impulse (FI = force × time) from a longer block time, they decrease their block acceleration [2,42] and the performance at 5 and 10 m [8]. Studies comparing data between sprinters of different performance levels mostly show higher block velocities ( [1,5,7,42] for faster sprinters. Furthermore, higher performance levels also appear to be slightly related to lower block vertical velocities [38] and more horizontal CM projection angles (i.e., resultant direction from the CM horizontal and vertical block exit velocities) [33,39].
Lower limbs joints pattern during the pushing phase (i.e., from movement onset until block exit) is mostly associated with extension movements, especially on the hips and knees [3,4,6,25,36]. The front leg joints typically extend through a considerable ROM in a proximal-to-distal extension pattern [3], reaching their maximum at the beginning of the flight phase (e.g., hip: 183.2 ± 6.8 • , knee: 177.4 ± 5.2 • , and ankle: 133.1 ± 6.7 • ) [6]. Contrarily, the rear leg does not exhibit the same proximal-to-distal extension strategy, with the knee reaching its peak angular velocity before the hip and the ankle [3,36]. This happens perhaps due to considerably less ROM of the rear knee compared to the front knee [3], as it starts from a more extended angle in the "set" position (e.g., rear knee: 120.7 ± 9.7 • ; front knee: 91.0 ± 9.8 • ). The movement of the ankles is more complex because it involves first a dorsiflexion and after an extension resulting in a stretch-shortening cycle of the triceps surae muscle [3,6,25,36]. The duration of the ankle's flexion is greater for the rear ankle (50% of the block phase) than for the front ankle (20% of the block phase) [36]. Experimental manipulations on footplates' inclinations [19,34] have shown an inverse association between block angles and muscle-tendon lengths of the gastrocnemius and soleus, highlighting that block angles steeper than 65 • could have disadvantageous effects on plantar flexor function [19]. Peak angular velocities at both hips are reached by a combination of flexionextension, abduction-adduction, and internal-external rotation [23,36], reinforcing the importance of a 3D analysis of the sprint start [36]. Whilst there is a consistent trend among sprinters in the joint angular velocity sequence during the block phase, the lack of comparative data between sprinters of different performance levels does not allow to highlight the technical aspects critical to success. However, a rapid hip extension should be one of the first aspects to consider on a sprinter's technique during the start, as peak angular velocities at both hips and rear hip range of extension are positively associated with block power (r = 0.49) [3].
Although upper body kinematics in the push-off phase has been the focus of a small number of studies, some important findings are noteworthy. The action of the upper limbs is more variable between sprinters than that observed for the lower limbs [36]. Despite this, it is possible to recognize a 3D movement pattern for shoulders and trunk with a combination of flexion-extension, abduction-adduction, and internal-external rotation movements, while the elbows exhibit an extension and pronation movement [36]. The velocity of the rear shoulder tends to be slightly greater than that of the other joints, but the peak resultant angular velocities at the upper limb joints are comparable to those at lower limbs during the push-off phase, particularly that of both knees and front ankle [36]. However, there is no evidence linking different upper limb kinematic patterns with any block phase performance predictor, and further research is needed to compile relevant recommendations for athletes and coaches.

Push-Off Kinetic Analysis
According to Newton's second law of motion, horizontal CM acceleration requires net propulsive forces to be applied to the athlete's body in the sprinting direction. Therefore, as said before, the horizontal force impulse, made up by the mean horizontal force and push-off time, is the determining factor of the horizontal velocity at block exit [2,5,36,42]. The relationship between these factors (i.e., horizontal force and push-off time) shows that the application of a greater amount of horizontal force is a key performance factor [42], as an increase in the time action (block time) conflicts with the criterion for 100 m performance: 'shortest time possible'. Thus, best sprinters generate greater average forces [10,22], higher rates of force development [7,25], and larger net [7] and horizontal [5] block impulses than their slower counterparts. Likewise, Graham-Smith, Colyer [39] comparing senior to junior athletes also showed that sprinters with faster PB100m (senior athletes) exhibit higher relative horizontal force during the initial block phase and higher forces during the transition from bilateral to unilateral pushing [39]. The evident importance of the force generated against the blocks for proficient execution of the starting block phase has encouraged researchers to gain a deeper understanding of the kinetic determinants of such a crucial phase of sprinting. Bezodis, Salo [2] tried to find the push-off performance measure that was more adequate, objective, and possible to quantify in the field. From their analysis, the NAHEP was identified as the most appropriate measure of performance because it objectively reflects, in a single measure, how much sprinters are able to increase their velocities and the associated length of time taken to achieve this, whilst accounting for variations in morphologies between sprinters [2]. Later, the identification of the magnitude of the force applied to both blocks and their optimal orientation as major determinants of performance encouraged researchers to gain a deeper understanding of the push-off forces applied against each block separately. Consequently, some studies support the importance of the force generated by the front leg for forwards propulsion [6,42] and show that faster sprinters are able to produce higher force impulses in the front block than slower sprinters [5,33] (for example: 221.3 ± 15.8 N·s vs. 178.3 ± 13.1 N·s for faster and slower sprinters, respectively [5]). Colyer, Graham-Smith [33] reinforce this feature highlighting that higher front block force production during the transition (when the rear foot leaves the block, 54% of the block push) and a more horizontally orientated front block force vector in the block phase (81-92%) are important performance-differentiating factors. However, other evidence ensures that the rear block force magnitudes are the most predictive external kinetic features of block power [10,33] and sprint performance [5,7,12,16]. For example, Coh, Peharec [5] found that a faster group of sprinters (PB100m = 10.66 ± 0.18 s; 913 ± 89.23 N) produced greater total forces against the rear block than a group of slower sprinters (PB100m = 11.00 ± 0.06 s; 771 ± 55.09 N). A longer relative rear leg push (i.e., as a percentage of the total push-off phase) is also positively associated (r = 0.53 [3]) with greater block power [3,10] and is present in sprinters with faster PB100m [5,7,33]. Modulations of the COP on the starting block surface showed that COP location may also be related to initial sprint performance [20,35]. Better sprint start performance appears to be achieved with a higher and more to the rear COP during the force production phase [20]. Thus, athletes and coaches should keep in mind that pushing the calcaneus onto the block (posterior location) may improve the 10 m time and/or horizontal external power for some individuals [35].
Forces under the hands have been reported in relatively few studies [10,33,42], showing somewhat contradictory results. While some point to a primary support role [42], others point out that the best athletes produced less negative horizontal impulse under hands compared with their slower counterparts [33]. Therefore, the importance of the hands' kinetics during the push-off phase remains unclear and should be the subject of future research.
In addition to external kinetic analyses, which provide valuable insight into starting block performance, the analysis of internal kinetics (i.e., joint kinetics) helps to increase the understanding of the segment motions that are responsible for CM acceleration. Recent research of joint kinetics has shown that 55% of the variance in NAHEP of a group of sprinters with a PB100m of 10.67 s was mainly accounted for by rear ankle joint moment (23%), front hip joint moment (15%), and front knee joint power (15%). The remaining 2% was shared by the remaining lower limbs joint kinetic variables [11]. In the rear block, the magnitude of the horizontal force produced is determined by the rear hip extensor moment and the rear hip extensor power coupled with large ankle joint plantarflexion moment [4,11,19], without any significant knee joint contribution [4,11]. At the front block, a proximal-distal pattern of peak joint power is evident [4], highlighting a strategy often adopted in power demanding tasks, with the main periods of positive extensor power at the front ankle and knee occurring after the rear foot has left the block [4]. In a study with 12 sprinters from the University of Tokyo team (PB100m: 10.78 ± 0.19 s), Sado, Yoshioka [23] showed that the peak lumbosacral extension moment was significantly larger than any other lumbosacral and lower-limb moment, being positively correlated with the starting performance. This peak value appeared in the double-stance phase where both hip joints exerted extension moments. The aforementioned evidence supports the findings of Slawinski, Bonnefoy [36] who showed that the lower limbs and the head-trunk segments are the two main segments that contribute to the kinetic energy of the total body. Upper limbs contribute 22% to the total body kinetic energy, demonstrating that their actions in the pushing phase on the blocks are not negligible [36].

The First Two Steps
The primary goal of the first steps is to generate a high horizontal velocity [40]. However, the transition between block start and the first steps represents a specific biomechanical paradigm: integrate temporal and spatial acyclic movements into a cyclic action [5]. The efficiency of this transition depends on the biomechanical demands of the first stances after block clearance, which are very different from the other stances during acceleration [14]. The sprinter aims to generate maximal forward acceleration during the transition from start block into sprint running [2,14,22,42] while generating sufficient upward acceleration to erect itself from a flexed position in the start blocks to a more extended position [6,14]. Specific technical (kinematic) and dynamic (kinetic) skills are therefore needed to successfully achieve this transition, and they are the focus of this section.

First Two Steps Kinematic Analysis
The primary goal of the initial steps of a sprint running is to generate a high horizontal sprint velocity, which results from the product of the length and frequency of the sprinter's steps [22,40]. Spatiotemporal parameters have shown that the sprinter's step length increases regularly during the acceleration phase, while step frequency is almost instantaneously leveled to the maximum possible [22]. Typically, the step frequency reaches the maximal values very quickly (80% at the first step and about 90% after the third step) [22], achieving around 4 Hz immediately after block exit [26,40]. The length of the first steps is more variable between sprinters, ranging from 0.82 to 1.068 m (senior females) [1,38] or 0.85 to 1.371 m (senior males) [1,7] on the first step, and from 1.06 to 1.30 m (senior females) [1,13] [38]. This may be a consequence of the lower vertical velocity of the CM at the block clearing shown by faster sprinters, allowing them to travel a longer distance despite shorter flight times [38]. Indeed, the kinematics of faster sprinters is also characterized by a tendency to assume long ground contact times in the first two steps (e.g., mean first contact duration for Diamond League sprinters is 0.210 s for males and 0.225 s for females, which is greater than those of lower-level Italian junior sprinters: 0.176 and 0.166 s, respectively), associated to short flight times (0.045 and 0.064 s, for the first flight of world-class and elite male sprinters, respectively) [38]. This strategy allows the high-level sprinters to optimize the time during which propulsive force can be generated, minimizing the time spent in flight where force cannot be generated. Combined with this, best sprinters have their CM projected further forward [7] at the first touchdown, putting the foot behind the vertical projection of the CM [3], and minimizing the braking phase. At the takeoff of the first and second steps, the CM horizontal position is also greater in elite than well-trained sprinters [7]. This means that the CM resultant and horizontal velocity in the first two steps are generally greater in high-level sprinters [7,15]. Slawinski, Bonnefoy [7], for example, reported that elite sprinters have a CM resultant velocity 5.8% higher than well-trained sprinters, at the end of the first step (4.69 ± 0.15 vs. 4.42 ± 0.11 m·s −1 for elite and welltrained sprinters, respectively). Furthermore, high-level sprinters also show slightly lower vertical velocities [7,39] and more horizontal CM projection angles at the end of the first two support phases [39].
Lower limb joints pattern during the first two steps is associated with a proximal-to-distal sequence of the hip, knee, and ankle of the stance leg [4,9,43]. During both first and second steps, the ankle joint undergoes dorsiflexion during the first half of stance (e.g., 17 ± 3 • and 18 ± 3 • for the first and second steps, respectively [43]) and subsequently a plantarflexion movement (e.g., 45 ± 6 • and 44 ± 5 • for the first and second steps, respectively [43]).
The hip performs extension for the entire stances, the knee extends until the final 5% of stances, and the ankle is dorsi-flexed during the first half of stances before the plantar flexing action [6]. After leaving the rear block, there is a small increase in ankle joint dorsiflexion during the swing phase, preceding the plantarflexion that occurs just before touchdown [6]. Although the ankle plantar-flexes slightly at the end of the flight, the ankle is in a dorsi-flexed position at initial contact (e.g., first stance: 70.6 ± 5.8 • and second stance: 72.4 ± 7.1 • [6]). During both first and second steps, the ankle joint dorsi-flexes during the first half of stance (e.g., 17 ± 3 • and 18 ± 3 • for the first and second steps, respectively [43]) and subsequently performs a plantarflexion movement (e.g., 45 ± 6 • and 44 ± 5 • for the first and second stance, respectively [43]). Note that a reduction in the range of dorsiflexion during early stance, requiring high plantar flexor moments, has already been associated with increases in first stance power [47]. Maximal plantarflexion occurs immediately following takeoff reaching, for example, 111.3 • at the first stance and 107.1 • at the second stance [6]. The extension of both knees occurs just after the block exit and reaches its maximum at the beginning of the flight phase, with larger extension in the front compared with the rear leg (e.g., rear: 134.9 ± 11.2 • ; front: 177.4 ± 5.2 • ) [6]. From a flexed position at initial contact, the knee extensors generate power to induce extension throughout stance and to attain maximal extension at takeoff, achieving peak extension angles of around 160-170 • (not full extension; e.g., first stance: 165.2 ± 20.6 • ; second stance: 163.6 ± 17.7 • [6]). This extension action of the knee during stances on its own may play a role in the rise of the CM during early acceleration [26]. The hip joints extend during block clearance to reach maximal extension during the beginning of the flight phase. During stance, the hips are in a flexed position at initial contact and continue to extend throughout stance, achieving maximal extension immediately following takeoff (e.g., first stance: 180.6 ± 20.9 • ; second stance: 181.1 ± 20.0 • [6]). There is also a considerable ROM in hip and pelvis rotation during stance as well as abduction. Although there are detailed descriptions of the lower limb angular kinematics during the first two stances and flight phases [3,6], there seems to be no clear evidence about the joint kinematic features that differentiate faster from slower sprinters. Furthermore, there is also a lack of experimental data on arm actions during early acceleration and its relationship to performance descriptors, making necessary future research in this area to help identify the most important performance features.

First Two Steps Kinetic Analysis
As said before, fast acceleration is a crucial determinant of performance in sprint running, where a high horizontal force impulse in a short time [13] is essential to reach high horizontal velocity [43]. Thus, as the highest CM acceleration during a sprint occurs during the first stances [7,9,14] (e.g., first stance: 0.36 ± 0.05 m·s −2 ; second stance: 0.23 ± 0.04 m·s −2 [14]), the ability to generate during this phase greater absolute impulse [7,18], maximal external power [39,42], and a forward-leaning force oriented in the sagittal plane [21,22,24,42] is linked to an overall higher sprint performance. Larger propulsive horizontal forces are particularly important during early acceleration, being a discriminating factor for superior levels of performance [48]. Experienced male sprinters (PB100m: 10.79 ± 0.21 s) can produce propulsive horizontal forces of around 1.1 bodyweight during the first stance [18]. However, a negative horizontal force has also been reported during the first contact after the block exit, even if the foot is properly placed behind the vertical projection of the CM [18]. During the first stance, for example, the braking phase represents about 13% of the total stance phase and the magnitude of the braking forces can reach up to 40% of the respective propulsive forces [18].
Furthermore, 3D analysis studies also highlight a lower body motion outside the sagittal plane during the first few ground contact phases [6,21,22,24,36,42]. In fact, during the first steps of a sprinter, a stance medial deviation is often observed that results from an impulse in the transverse plane. Although the medial impulse is the smallest of the three orthogonal stance impulses [21,22,42], the fact that it is non-zero can have an effect on the motion of the CM and on step width. However, it has been shown that well-trained sprinters present similar step widths in the early acceleration to those of the trained and non-trained sprinters [42]. Moreover, manipulations of both "set" position [21] and first step [24] widths have shown no effect on block-induced power nor braking force or net anteroposterior impulse, showing that smaller step width is not a discriminator factor of superior performance levels. Therefore, the perception that the adoption of a widened stance during initial acceleration (referred to as "skating style") is detrimental to performance is not at all proven, and further research is needed to clarify the joint and muscular factors that contribute to the sprinters' lateral motion in the initial phase of acceleration.
At joint level, the hip, knee, and ankle joints generate energy during stance leg extension [6], although it appears that the ankle joint is the main contributor to CM acceleration [14]. However, experimental and simulation studies highlight that the knee plays an important role during the first stance, being decisive for forward and upward CM acceleration [4,6,14,15]. The importance of power generation at the knee seems to be specific for the first stance when the knee is in a more flexed position and the sprinter is leaning forward. From the second stance onwards, the knee becomes less and the ankle more dominant since the plantar flexors are in a better position to contribute to forward progression [6]. As the knee is in a flexed position during the first step, the sprinter favors the immediate power generation of the knee extensors rather than preserving a stretch-shortening cycle [6]. In contrast, a stretch-shortening mechanism can be confirmed at the hip and ankle [4,6,14,15]. Hip extensors maximal power generation occurs near touchdown [4,6] where the hip extensors actively pull the body over the touchdown point [6]. The hip can effectively generate large joint moments and power [14], but only contributes minimally to propulsion and body lift during the first two stances [14]. Ankle plantar flexors act throughout both the first and second stances under a stretch-shortening cycle. There is therefore an initial phase of power absorption preceding the forceful power generation at take-off [4,14]. As a major contributor to CM acceleration, the ankle joint can generate up to four times more power than it absorbs during the first two stances [43]. Nevertheless, the importance of ankle stiffness during the first two stances remains unclear. While Charalambous, Irwin [49], in a case report, found a correlation between greater ankle stiffness and greater horizontal CM velocity at take-off (r = 0.74), Aeles, Jonkers [9] did not, still highlighting the lack of differences between faster (senior) and slower (junior) sprinters. Future work is therefore needed to further clarify this issue. Furthermore, it remains unclear whether ankle stiffness is influenced by foot structure and function (e.g., planus, rectus cavus, clubfoot) as well as other important performance variables such as greater maximal power, a forward-leaning force oriented in the sagittal plane, or COP location during push-off.
Concerning kinetic factors differentiating senior and junior athletes, Graham-Smith, Colyer [39] reported that, contrarily to the block phase where there are marked differences between groups, the force and power waveforms relating to the first two steps did not differ considerably across groups. Still, senior sprinters are able to produce greater horizontal power during the initial part (10-19% of the stance phase) of the first and second ground contact (first step: 25.1 ± 3.6 W·kg −1 vs. 23.1 ± 6 W·kg −1 and second step: 26.7 ± 3.6 W·kg −1 vs. 24.9 ± 4.5 W·kg −1 , forsenior and junior sprinters, respectively), and also exhibit a higher proportion of forces immediately after braking forces are reversed (from 9% to 15% and 25% to 29% of stance phase) [39]. Furthermore, Debaere, Vanwanseele [15] also highlight that adult sprinters are able to generate more joint power at the knee during the first step compared to young sprinters, inducing longer step length and therefore higher velocity [15]. Younger sprinters tend to prioritize a different technique: the hip contributes more to total power generation, while the knee contributes far less [15]. This indicates that younger sprinters lack the specific technical skills observed in adult sprinters, likely due to less musculature than adults [1,9,15]. However, there is no evidence of differences in ankle joint stiffness, range of dorsiflexion, or plantar flexor moment between young and adult sprinters [9]. This indicates that the technical performance-related parameters of the first stances are not likely to explain the better 100 m sprint times in adult compared to young sprinters [9].

Strengths, Limitations, and Recommendations
A strength of this review was that it allowed us to identify a body of knowledge that provides fundamental information for athletes and coaches as relevant data that can contribute to improving the training and/or preparation strategies for better performance, supported by scientific evidence.
A possible limitation of this systematic review is that it only includes studies written in English, thereby potentially overlooking other relevant publications in other languages. Additionally, the present article reviewed only studies with mention to sprinters' PB100m, eventually precluding publications with relevant samples that could also add knowledge. Furthermore, extending the biomechanical analysis to muscular features beyond the simple kinematic and kinetic approach might have allowed a further understanding of the discriminating factors of superior performance levels. Another obvious limitation is the limited amount of research with female sprinters. Indeed, in the reviewed studies, there is a clear imbalance between the amount of female and male sprinters included (179 females vs. 587 males), questioning whether the biomechanical characteristics of the sprint start previously associated to female sprinters are attributable to sex-related aspects, or, rather, to aspects related to the 100 m time. Moreover, some of the studies included in this review were based on a relatively small sample size, especially when elite or world-class sprinters were included. This problem reflects the difficult access to high-level athletes, preventing the clear identification of discriminatory factors of superior performance levels. Finally, the conflicting classifications of sprinters level and the scarcity of information on effectively high-level or world-class sprinters, makes it difficult to compare sprinters of different performance levels. Considering entry standards for 100 m sprint event at the 2022 European Athletics Championships (10.16 s for men and 11.24 s for women), it can be said that a very small percentage of elite and/or world-class sprinters [50] was included in the reviewed studies.
Research on the biomechanics of the block and/or first stance phases has been the subject of growing interest in the past few years. Nonetheless, there are some unclear features in the studies published so far, which should be investigated in future studies for a better understanding of: (i) the association between different upper limb patterns and the main block start performance predictors; (ii) the influence of foot type (e.g., planus, rectus cavus, clubfoot on sprint start performance; (iii) the association between ankle stiffness during dorsiflexion and the horizontal CM velocity at take-off; (iv) the specificity characteristics of training drills, utilizing temporal organization and intra-limb joint coordination analyses, to help the process of exercise selection to enhance block starting performance; (v) how technical and/or physical training can improve ankle and knee function during first steps and increase horizontal velocity in the early acceleration; (vi) the influence of sex (such as physical or muscle structures and/or anthropometric characteristics) on sprint start performance descriptors. A major challenge for researchers is to align these research lines with the need for greater information on world-class sprinters during competition. Whenever possible, research based on a marker-less methodology and obtained during official top-level sprint competitions, during which the sprinters are supposedly more motivated to produce their best performance, should be encouraged.
It is worth mentioning two new studies [51,52] published after the date of this systematic review, which, meeting the defined inclusion criteria, could have added important knowledge on some of the issues mentioned above.

Conclusions
Based on this review, some important conclusions and recommendations to help athletes and coaches can be made, namely: (i) the choice of an anteroposterior block distance relative to the sprinter's leg length may be beneficial for some individuals, promoting greater block start performance (greater normalized average horizontal external power); (ii) the use of footplate inclinations that individually facilitate initial dorsiflexion should be encouraged-footplate angles around the 40 • are recommended and block angles steeper than 65 • should be avoided; (iii) pushing the calcaneus onto the block (posterior location) may be beneficial for some individuals, improving the 10 m time and/or horizontal external power; (iv) short block exit flight times and optimized first stance contact times should be encouraged, as they maximize the time during which propulsive force can be generated; (v) focus attention on the magnitude of force applied on the rear block, as it is considered to be a primary determinant of block clearance; (vi) rapid hip extension during the push-off phase should be a priority in sprinter focus and coach feedback; (vii) the large role played by the hips on the push-off phase and by both the knee and ankle at the early stance must be acknowledged within physical and technical training to ensure strength and power are developed effectively for the nature of the sprint start.

Conflicts of Interest:
The authors declare no conflict of interest.
Appendix A Table A1. Summary of the kinematic variables in the "Set" position. Data are the magnitude of the mean ± SD presented in the reviewed studies. Groups are male, female, and mixed (when authors joined data without discriminating by sex) sprinters. Studies are listed, in each variable, in reverse chronological order. Data, terms, conditions, and sprinters' performance levels are presented according to the original authors. Statistical differences between groups are marked with asterisks (* p < 0.05; *** p < 0.001). Mero [18] Trained sprinters 39 ± 7 (a) Slovene national sprinters 103.38 ± 6.97 (c) * Mero [18] Trained sprinters 96 ± 12 (c) Bezodis, Salo [3] World-class to university sprinters 77 ± 9 (a) Debaere, Delecluse [6] Elite sprinters 107.1 ± 9 (b) Mero [18] Trained sprinters 126 ± 16 (c) CM-center of mass; (a) internal angle between the thigh and trunk in flexion/extension plane; (b) relative angle between the pelvis and the thigh according to the Biomechanical Convention [53]; (c) relative angle between the thigh and the shank according to the Medical Convention [53]; (d) relative angle between the shank and the foot according to the Biomechanical Convention [53]; (e) rear leg hip angle measured as front-rear leg angle; (f) relative angle between the vector from hip to shoulder and the horizontal plane.   Maulder, Bradshaw [40] National and regional level sprinters 22 ± 7 CM-center of mass; ROM-range of motion; (a) block time calculated from the difference between the average data of total block time and reaction time data; (b) probably an incorrect data from the original paper; (c) presumably the negative signal is a gap in the data reported in the original paper; (d) the take-off or push-off angle is the angle between the horizontal and the line passing through the most front part of the contact foot and the center of mass at block clearance; (e) center of mass projection angle is calculated as the resultant direction from the horizontal and vertical block exit velocities of the center of mass; (f) angular displacement during rear block contact only; (g) higher magnitude of dorsiflexion was correlated to a faster stretch velocity, which was related to increased force generation (maximal rate of force development, maximal resultant and horizontal push force, and also normalized average horizontal block power); (h) the angle, measured relative to the horizontal, between the line passing through the hip and shoulder (trunk segment) of the side of the body in which the athlete's front foot at the block take off instant; (i) relative angle between the pelvis and the thigh according to the Biomechanical Convention [53]; (j) relative angle between the thigh and the shank according to the Medical Convention [53]; (k) relative angle between the shank and the foot according to the Biomechanical Convention [53]. Table A3. Summary of the kinetic variables in the "Block Phase". Data are the magnitude of the mean ± SD presented in the reviewed studies. Groups are male and mixed (when authors joined data without discriminating by sex) sprinters. Studies are listed, in each variable, in reverse-chronological order, followed by alphabetically for studies published in the same year. Data, terms, conditions, and sprinters' performance levels are presented according to the original authors. Statistical differences between groups are marked with asterisks (* p < 0.05; ** p < 0.01; *** p < 0.001; # Cohen's d-large effect size (>0.8); § § § large effect size [1.     [2]; (g) normalized average horizontal external power is the average horizontal external power normalized to the mass and the leg length of the sprinter [2]; (h) for normalization, the body height was used instead of the sprinter's leg length [25]; (i) joint data normalized to the mass and the leg length of the sprinter; (j) significantly larger (p < 0.05, Cohen's d = 2.02-11.09) than any other lower-limb and lumbosacral torques, although quantitative data for the remaining joint torques are not available. Table A4. Summary of the kinematic variables in the "first two steps". Data are the magnitude of the mean ± SD presented in the reviewed studies. Groups are male, female, and mixed (when authors joined data without discriminating by sex) sprinters. Studies are listed, in each variable, in reverse-chronological order, followed by alphabetically for studies published in the same year. Data, terms, conditions, and sprinters' performance levels are presented according to the original authors. Statistical differences between groups are marked with asterisks (* p < 0.05; ** p < 0.01; *** p < 0.001; # significant different from adults; § small effect size [0.   Chen, Wu [37] Bunched start 33.4 ± 7.0 Medium start 34.9 ± 5.9

Appendix B
Elongated start 36.7 ± 5.9 * Maulder, Bradshaw [40] National and regional level sprinters CM-center of mass; (a) horizontal distance relative to stat line: (b) represents the horizontal distance (divided by leg length) between the CM and the stance leg metatarsal-phalangeal joint (negative value means that foot is behind the CM; (c) center of mass projection angle is calculated as the resultant direction from the horizontal and vertical block exit velocities of the center of mass; (d) the angle, measured relative to the horizontal, between the line passing through the most front part of the contact foot and the CG during takeoff; (e) the angle, measured relative to the horizontal, between the line passing through the hip and shoulder (trunk segment) of the support leg; (f) internal angle between the thigh and trunk in flexion/extension plane; (g) relative angle between the pelvis and the thigh according the Biomechanical Convention [53]; (h) relative angle between the thigh and the shank according the Medical Convention [53]; (i) relative angle between the shank and the foot according the Biomechanical Convention [53]; (j) data referring to the values recorded in the entire acceleration phase (0-40 m) excluding the block phase. Table A5. Summary of the kinetic variables in the "first two steps". Data are the magnitude of the mean ± SD presented in the reviewed studies. Groups are male and mixed (when authors joined data without discriminating by sex) sprinters. Studies are listed, in each variable, in reverse chronological order, followed by alphabetically for studies published in the same year. Data, terms, conditions, and sprinters' performance levels are presented according to the original authors. Statistical differences between groups are marked with asterisks (* p < 0.05; *** p < 0.001; # significant different from adults;