The Relationship of Aerobic Endurance and Linear Speed on Repeat Sprint Ability Performance in Female International Footballers

The aim of this study was to investigate the relationship between aerobic endurance and linear speed on repeat sprint ability performance in female international footballers. Twenty-five female international football players completed 10-m, 20-m, 30-m linear speed, repeat sprint ability (RSA) and yo-yo intermittent recovery, level 1 tests. Pearson’s product correlation (r) was used to assess the relationship between 10-m, 20-m and 30-m sprint times (S10-m, S20-m, S30-m) sprint time on the first trial of the repeated sprint test (RST1) and yo – yo intermittent recovery test, level 1 (YYIR1) with total repeated sprint time (RSTT). Linear regressions were used to describe the relationship between the dependent variable (RSTT) and independent variables (S10-m, S20-m, S30-m, RST1 and YYIR1). Pearson’s correlation analysis indicated a large negative correlation between RSTT and YYIR1 performance (r = -0.58, P = .002), a very large significant relationship was established between S10-m and S20-m whilst an almost perfect relationship was present for S30-m and RST1. The results of the linear regression proved the independent variable RST1 explained 85% (adjusted R) of the variation in RSTT scores. S10-m, S20-m and S30-m performance accounted for 62 – 83% of the variation in RSTT scores whilst YYIR1 performance only accounted for 33.6%. The results of this study indicate that fast linear speed can positively impact on RSTT performance whilst reduced aerobic endurance capabilities may negatively influence RSTT performance in female international footballers.


Introduction
Elite female international football is a dynamic intermittent field sport that physiologically stresses the metabolic and neuromuscular systems [1,2]. Due to the random nature of football match play numerous contextual factors will influence the physical demands withstood by players such as playing position, formation, tactics, time in possession, team ranking and score line [3][4][5]. Players transition from long periods of aerobic activity to short bouts of high intensity anaerobic activities such as high speed running, sprinting and repeated high speed and sprinting actions [6 -8].
During match play total distances covered fall within the range of 9 -12km with the majority of distance covered walking or jogging with high speed running (> 18 km-h) been shown to account for 1.53 -1.68 km of that total distance [7,5,9,10]. Total match play sprint distance reported within the literature differs due to the variation of 148 The Relationship of Aerobic Endurance and Linear Speed on Repeat Sprint Ability Performance in Female International Footballers the sprint speed threshold implemented (19.1 km-h versus 20 km.h versus 25.1 km.h) [6,11,12]. Though players have been shown to cover mean distances in the range of . 16 -.46 km sprinting during games and complete between 180 and 240 accelerations during a senior international match [5,9,10,112,13]. Repeated high speed actions consisting of two or more sprints with less than 20 seconds recovery between efforts occur approximately 31 -33 times during international match play [6,8]. Although, it is not uncommon for players to perform up to 6 repeated high speed bouts with less than 20 second recovery between each bout [6,8,11]. However, the most frequent mean recovery time duration between repeated high speed running bouts in female international matches has been demonstrated to be 10 seconds or less [6]. Such repeated high speed and sprint efforts with minimal recovery time can result in a deterioration of high speed running, sprint and RSA performance, throughout international match-play [5,8,9,14].
RSA has been defined within the literature as short sprints typically less than ten seconds in duration with recovery periods sixty seconds or less [15]. Assessment of RSA identifies a player's capacity to maintain maximal effort and recovery during multiple successive high speed or sprint efforts a key attribute for athletes in field based team sports such as football [16]. Numerous assessment methods have been utilised to measure RSA performance in various populations of female football players such as 5 x 10-m Ajax sprint test, 7 x 30-m sprints with 30 seconds recovery, 10 x 30-m with 30 seconds recovery and 6 x 15 second cycles of 20-m maximal sprinting followed by a 10-m deceleration and a 10-m active recovery jog ( Figure 1) making direct comparisons between studies challenging [17 -21].
RSA performance is highly influenced by the testing protocol utilised (i.e. sprint duration, distance and number of repetitions as well as type of and duration of recovery) therefore any RSA assessment protocol should be reflective of the movement demands of the sport [16]. In addition, it must be noted that variables sprint duration, distance, number of repetitions and duration of recovery during match play will differ between playing positions [6,11]. Spencer et al [17] stated that mean sprint times and distances covered in field based team sports to be 2 -3 seconds in duration and between 10 -20m., findings supported by Gabbett et al [22] demonstrating mean sprint times of (2.1 + 0.7 seconds). More recently Nakamura et al [11] highlighted similar results reporting mean sprint times of (2.5 + 0.5 seconds) and mean sprint distances of (15.7 + 2.6 m) in elite female footballers. Furthermore, the recovery period between 2 repeat high speed and sprint bouts the most frequent in female international football has been shown to be (9.42 + 2.68 and 10.22 + 4.42 seconds) respectively [6,8]. In addition, RSA assessment incorporating active recovery opposed to passive recovery may potentially reduce muscle acidosis and negate a decline in RSA performances [17]. Considering 85% of the total distance covered during elite female match play is at walking and jogging speeds and approximately 95% of the activity between repeat high speed or sprint bouts is active in nature RSA assessment should include active recovery between each repeat sprint [9,22].
Similar to Lockie et al [23] this study will implement the RSA assessment protocol developed by Gabbet [21] for elite female footballers (see figure 1) as based on the presented literature the protocol appears to specifically address the RSA demands of elite female football match play. Gabbet [21] previously stated that RSA tests should be performed in conjunction with a maximal aerobic endurance and linear speed assessments. As a highly developed aerobic endurance capacity may benefit recovery between repeated high speed or sprint bouts during match play [19]. Therefore, the purpose of this study was to undertake the hypothesis stated by Gabbet [21] and examine the relationship of aerobic endurance (YYIR1) and linear speed (S10-m, S20-m, S-30m and RST1) on RSA performance in female international footballers

Experimental Approach to the Problem
Physical assessments were conducted in the following order 10-m, 20-m, 30-m linear sprints, RSA and Yo-Yo Intermittent Recovery Test, level 1 (YYIR1). Assessments took place in two separate testing sessions on the first day of each squads relevant training camp during 2019. All tests were performed in the afternoon between the hours of (12:00 and 16:00). Subjects were provided at least a 15-minute recovery period between tests to assuage any effects of fatigue. Every test was performed indoors on a rigid surface with all subjects wearing running shoes. Prior to the initiation of the testing session, participants underwent a structured 10 minute warm up protocol consisting of jogging, running, lower limb mobility and dynamic stretching movements. On completion of the warm up a 5-minute recovery period was employed to assuage any effects of fatigue. Any participants diagnosed with a lower limb injury were excluded from testing.

Participants
Twenty-five female elite level football players selected to represent the Ireland u19 or senior international team participated in the study (mean + SD; age 19.1 + 3.4 years; height 167 + 6 cm; mass 62.5 + 7.4 kg). Ethical approval was obtained by the college Research Ethics Committee. The participants were informed of the risks of the study in person and writing. An informed consent document was signed before the beginning of data collection and participants were free to withdraw from the study at any time. For those under the age of 18 years' age parental or guardian signed consent was obtained. The study was conducted in accordance to the Declaration of Helsinki.

Linear Speed
Speed was assessed over 10-m, 20-m, 30-m using photocell timing gates (Witty-gate, Microgate, Italy) equipment which has been deemed to be a reliable assessor of linear speed in female footballer players [24]. Participants started 0.5m behind the initial timing gate in a two point split stance and were instructed to set off in their own time and run maximally to a marker place 2-m beyond the 30-m timing gate. Each subject performed 2 sub maximal efforts sprints prior to 3 maximal efforts separated with a minimum of 2 minutes of rest but no longer than 3 minutes. Times were recorded to the nearest 0.01 with the fastest time of the three efforts at 10-m, 20-m, and 30-m used for analysis. ICC and CVs for 10-m ,20-m and 30-m linear speed assessments fell between the range of (ICC = .90 -.98, CV = 1 -2.4%).

Repeat Sprint Ability
As stated the RSA test protocol developed by Gabbet [21] was used for this study. The test has been shown to be a valid and reliable assessment of RSA in elite female football players (ICC = .91, TE = 1.5%) [21]. Players performed 6 x 20-m maximal effort sprints on a 15-second cycle. On the completion of each timed 20-m sprint, players performed a 10 -m deceleration followed by a 10-m active recovery jog back to the next start line before coming to a complete standstill prior to next sprint ( Figure  1) [21]. Subjects began their sprint 0.5m behind the relevant starting timing gate. Photocell timing gates (Witty-gate, Microgate, Italy) were positioned at the 10-m and 30-m start/finish lines to time each sprint effort which was recorded to the nearest 0.01 second. The researcher recorded the active recovery period between 20-m sprints using a stopwatch (TIS Pro 018, Precision, Nottingham, UK) and provided feedback on the progress of the active recovery period [23]. Each participant performed a single submaximal effort of one 15-second cycle for test protocol familiarisation. Numerous variables were calculated for the RSA test to assess the group's performance and rate of fatigue: • RSTT = the sum of all six 20-m sprint times stated in seconds [21,23]. • RST1 = the time taken to complete the first 20-m sprint also expressed in seconds [21]. • FI = the decrease in sprint performance from the first to the last sprint was calculated and presented as a percentage. The following equation was used: Fatigue index (%) = 100 X (fastest sprint -slowest sprint / fastest sprint [21,23].

Aerobic Endurance
The subjects completed the YYIR1 test as described by Krustrup et al [25] which has been deemed to be a reliable assessment of aerobic endurance [26]. The test was performed using an audio recording. Each athlete performed repeated 20m shuttle runs at increasing velocities with 10 seconds of active recovery. During the active recovery participants walked around a marker placed 5-m behind the finishing line. An individual's test was terminated when the subject failed to reach the starting line within the allotted time period on two occasions or the participant felt unable to complete another shuttle at the assigned speed. The total distance covered by a subject was used as the performance measure.

Statistical Analysis
Assumption for normality was confirmed using the Shapiro-Wilks test. Levenes test indicated equal variance to be assumed. Pearson's product moment-correlation (r) was conducted to determine the relationship between 10-m, 20-m and 30-m sprint times, sprint time on the first trial of the repeated sprint test (RST1), yo -yo intermittent recovery test, level 1 (YYIR1) and total repeated sprint time (RSTT) at significance level of (P < 0.01). Strength of the correlation was assigned by the following parameters 0.1 -0.3 (small), 0.3 -0.5 (medium) 0.5 -0.7 (large), 0.7-0.9 (very large) and 0.9 -1.0 (almost perfect) [26]. YYIR1 scores were separated in to high and low performing groups were independent samples t-test analysed differences between the groups. To determine the magnitude of difference between the high and low groups effect sizes (ES: Cohens d) were calculated and values of 0.2 ,0.5, and 0.8 were interpreted as a small, medium and large respectively [28]. Linear regressions were utilised to explain the relationship between the dependent variable (repeat sprint ability total time) and independent variables 150 The Relationship of Aerobic Endurance and Linear Speed on Repeat Sprint Ability Performance in Female International Footballers (10-m,20-m,30-m linear sprints, initial 20-m repeat sprint trial and yo-yo intermittent recovery test). The Durban Watson test was applied in the residual of regression analysis. Furthermore, the analysis of variance (ANOVA) test allowed to determine if the linear regression tests were significant.

Discussion
This study is the first of its kind to investigate the relationship between aerobic endurance (YYIR1) and linear speed (S10-m, S20-m and S30-m) on RSA (RSTT and RST1) performance in female international footballers. The RSTT for the female international football players in this study (21.42 + .97 seconds) was slower than the national level players (20.09 + .5 seconds) evaluated by Gabbett [21] but quicker than state (23.3 + .04 seconds) and collegiate level players (22.09 + .99 seconds) [21,23]. The fatigue index also termed performance decrement (PD) for the players in this study (8.34 + 2.5 %) was greater in comparison to that of the national (6.2 + 1.6 %) and state level (6.8 + 2.7 %) players examined by Gabbet [21] but less than the collegiate players (8.37 + 5.51 %) investigated by Lockie et al [23]. In addition, this study was the first to describe the Sdec (4.38 + .26 %) in female international football players using this RSA test protocol. The Sdec method has been shown to the most valid and reliable method of measuring fatigue in RSA assessment as it accounts for all sprints performed opposed to FI or PD which only considers the first and last sprint a scenario which may be considerably influenced by one good or bad repetition during a RSA assessment [29]. However, in this cohort Sdec showed no significant relationship with RSTT (r = 0.38, P =. 19). Future studies should further investigate if Sdec is a more reliable measure of fatigue during RSA assessment of elite female football players as both FI or PD have previously been shown not to be in national, state and collegiate level football players as well as the cohort in this current investigation (r = 0.29 P = .17) [21,23].
Aerobic endurance would appear to play a significant role in RSA performance as a large negative correlation highlighted players that covered greater distances in the YYIR1 achieved faster RSTT times (Fig 2). Although the test protocols differed the results of this study would concur with Shalfawi et al [19] findings which reported similar large negative correlations (r = -0.59, P = <0.01) between RSTT (10 x 30-m with 30 seconds recovery between reps) and aerobic fitness (multi stage fitness test) in elite female club level footballers. In contrast Lockie et al [23] using the same test protocols as this study reported no significant relationship between RSTT and YYIR1 performance (r = -0.03, P = 0.91). Furthermore, the presence of significant medium to large negative relationships between YYIR1 and S10-m, S20-m, S30-m and RST1 replicates a trend previously established in elite female club level players [19]. Similar to Sanders et al [20] findings on female colligate level football players 30-m sprint performance it would appear that the players in this study who possessed higher aerobic endurance capabilities on average were faster over S10-m, S20-m and S30-m distances. A growing body of research emanating from numerous populations of female football players indicates a strong relationship between linear speed performance over various distances and RSTT [19,20,23]. Contrary to Gabbett [21] (r = 0.02, P < 0.05) this study found a very large significant relationship between S10-m and RSTT performance which was greater than the significantly large relationship previously reported by Lockie et al [23] in collegiate level players (r =0.50, P = 0.03). Over the longer distances of 0-20-m and 0-30-m this current study as well asShalfawi et al [19] (r = 0.73, P < 0.01) both highlighted very large significant relationship between S20-m performance and RSTT. While Lockie et al [23] displayed a similar trend between 30-m and RSTT performance (r = 0.71, P < 0.01) which was further supported by the research findings ofthis study which highlighted an almost perfect relationship. Furthermore, akin to Gabbet (r = 0.96, P < 0.001) an almost perfect significant relationship (r = 0.92, P < 0.01) was present between RST1 and RSTT performance in this cohort of l players [21]. Linear regression analysis revealed RST1 was the most strongly correlated with RSTT accounting for 85% variance in comparison to S10-m, S20-m and S30-m which accounted for 62.1%, 79.9% and 83.8% variance in RSTT performances. YYIR1 was deemed to have a more moderate influence on RSTT in comparison to all linear speed assessments with the aerobic endurance assessment being responsible for 33.6% of the variance in RSTT.

Conclusions
There are a number of limitations to this investigation that should be addressed. The sample size of this study is a combination of two international age groups (U19 and senior) therefore the generalisation of these results must be interpreted with caution. Future RSA research should specifically examine all competitive international age categories (i.e., U17, U19 and senior). In addition, no physiological markers such as heart-rate [21] were assessed during the RSA test which may have been of benefit. However, significant relationships were observed between aerobic endurance and linear speed with RSA performance. The results from this study indicate a players RSA will be influenced by the inter-play of both their aerobic endurance and linear speed capabilities opposed to either one independently. Therefore, it may be suggested that separate training strategies for aerobic endurance and linear speed are required to improve a player's RSA.

Practical Application
In order to fully optimise RSA performance of elite female footballer's coaches, strength and conditioning coaches and sports scientist alike should implement a concurrent training strategy. For instance, small sided games are common practice in the development of female football player's aerobic capacity though they have been shown not to meet the high speed, very high speed and sprinting demands of international match play [30]. One may suggest that aerobic endurance development can be targeted via various sized small sided games whilst linear speed development can be addressed through sprint training and power development. Though it must be noted that any training program designed to optimise RSA is structured so both aerobic and linear speed development are emphasised in conjunction with the team's competition demands and the individual capabilities of the player.