International Journal of Sports Physiology and Performance, 2014, 9, 621-626 http://dx.doi.org/10.1123/IJSPP.2013-0263 © 2014 Human Kinetics, Inc.
www.IJSPP-Journal.com ORIGINAL INVESTIGATION
Impact of Maximum Speed on Sprint Performance During High-Level Youth Female Field Hockey Matches: Female Athletes in Motion (FAiM) Study Jason D. Vescovi The aim of this study was to examine the impact of maximum sprint speed on peak and mean sprint speed during youth female field hockey matches. Two high-level female field hockey teams (U-17, n = 24, and U-21, n = 20) were monitored during a 4-game international test series using global position system technology and tested for maximum sprint speed. Dependent variables were compared using a 3-factor ANOVA (age group, position, and speed classification); effect sizes (Cohen d) and confidence limits were also calculated. Maximum sprint speed was similar between age groups and positions, with faster players having greater speed than slower players (29.3 ± 0.4 vs 27.2 ± 1.1 km/h). Overall, peak match speed in youth female field hockey players reaches approximately 90% of maximum sprint speed. Absolute peak match speed and mean sprint speed during matches were similar among the age groups (except match 1) and positions (except match 2); however, peak match speed was greater for faster players in matches 3 and 4. No differences were observed in the relative proportion for mean sprint speeds for age groups or positions, but slower players consistently displayed similar relative mean sprint speeds by using a greater proportion of their maximum sprint speed. Keywords: adolescents, nondifferential GPS, motion analysis Sprinting is believed to be an important component of field sports, and recent research has described the influence of maximum sprint speed on match sprinting ability in male soccer players.1 Mendez-Villanueva et al1 reported that players with greater maximal sprint speed reached higher velocities in matches than their slower counterparts and that position influenced the ability to achieve greater speeds. Similar to soccer, the expression of maximum sprint speed in field hockey is likely influenced by tactical and possibly position demands; however, it is currently unknown if there are any limitations due to the mechanical disadvantage of carrying a stick, as well as the biomechanical position (ie, continued hip flexion to play the ball) of the body inherent to the sport. Activity profiles of elite female field hockey players have been reported,2,3 but the relationship between maximum sprint speed and speed achieved while sprinting in matches is unknown. It would be of interest to sport scientists and strength and conditioning professionals to understand the proportion of maximum sprint speed achieved during matches, as it could help direct training programs. Thus, the aim of this study was to examine the expression of maximum sprint speed during youth female field hockey matches.
Methods This was an observational study designed to evaluate how maximum sprint speed affects sprint speeds achieved during youth female field hockey matches. All athletes were verbally informed of the experimental procedures, and written informed consent was obtained from a parent and assent provided by the participants (if
The author is with the Faculty of Kinesiology & Physical Education, University of Toronto, Toronto, ON, Canada. Address correspondence to
[email protected].
younger than 18 y) before the study began. The study was approved by the office of research ethics and conducted in accordance with the Declaration of Helsinki.
Participants Female field hockey players from the U-21 (n = 20) and U-17 (n = 24) national teams volunteered for this study and were monitored during 4 matches during an international test series. Player position was generically described as forward, midfielder, or defender. Each player was categorized into 1 of 2 groups (faster or slower) based on maximum sprint speed, which was operationally defined by the 90th percentile (≥ 29.0 km/h) of female field-sport athletes.4 Sample-size distribution among the groups is presented in Table 1. Participants had a minimum of 4 years of playing experience and typically performed 2 to 4 hours of training per week with their local club teams while coming together for specific national-team commitments 3 or 4 times a year.
Match Configuration and Analysis All matches had 35-minute halves and were played on the same standard-size, outdoor, water-based artificial surface with 11 players a side. Matches were played in accordance with standard rules from the International Hockey Federation and refereed by qualified officials. The match schedule was customary for field hockey, with the first 2 matches played on consecutive days, followed by 1 day off and then the final 2 matches played on consecutive days. A total of 5 practice sessions were conducted during the 3 days preceding the first match. The matches were played at the same time each day, with the U-17 at 11 AM and the U-21 at 1 PM. Before each match participants were outfitted with a nondifferential global positioning system (GPS) unit (SPI Pro, GPSports, Canberra, Australia), which operated at a sampling frequency of 621
622 Vescovi
Table 1 Sample Size (n) for Testing and Matches Match Testing
1
2
3
4
U-21
20
16
14
17
16
U-17
24
15
15
16
14
forward
14
10
9
9
9
midfielder
19
12
12
16
14
defender
11
9
8
8
7
slower
33
22
20
26
21
faster
11
9
9
7
9
Age group
Position
Speed classification
5 Hz. Units were worn between the shoulder blades in specially made vest. Other investigators have reported on the reliability and validity of nondifferential GPS,5,6 with improved reliability and validity demonstrated at greater collection frequency.7 The GPS brand and model used in this study has also been shown to be valid and reliable for measuring sprint distance and speed.8,9 During all matches, 7 to 12 satellites were available for signal transmission (GPSports, Team AMS R1 2012.9), which is optimal for assessment of human movement.7 The horizontal dilution of precision is not available from the Team AMS software; however, any value in excess of 4 results in an automatic rejection of the data (personal communication). This threshold is well below the maximum value (50) reported to result in inaccurate outcomes.10 A digital watch that received satellite time identified the start and end of each half, as signaled by the referee’s whistle. These times were used, in addition to substitution rotations, to delineate the raw GPS data file so that only time on the pitch during the matches was included for analysis. After each match the data were extracted using proprietary software (GPSports, Team AMS R1 2012.9) for analysis. Any locomotor activity lasting more than 1 second and exceeding 20.0 km/h was considered a sprint and included in the analysis. The sprint threshold is based on previously published reports describing female field-sport athletes.4,11,12 Peak speed for all sprints during every match was recorded for each player and a relative proportion of maximum sprint speed calculated for subsequent analysis.
Maximum-Sprint-Speed Testing Four days before the initial match the participants completed a comprehensive battery of tests that included speed, power, agility, and aerobic fitness. Testing was performed on an indoor tennis court, and linear sprint speed was the first test in the protocol. All athletes performed a standardized warm-up of approximately 15 minutes that included general exercises such as jogging, shuffling, sprinting, multidirectional movements, and dynamic stretching exercises. Participants wore shorts, T-shirt, and running shoes during testing. Linear sprint speed was evaluated over 35 m using infrared timing gates (Brower Timing Systems, Draper, UT, USA) positioned at the start line and at 5, 10, 20, 30, and 35 m at a height of approximately 1.0 m. Participants stood upright, with their lead foot positioned approximately 5 cm behind the initial infrared beam (ie, start line) and began when ready. The athletes were instructed to run
at maximal speed through the final pair of sensors. Timing started when the laser of the starting gate was broken (ie, first movement). Two or three trials were completed with at least 3 minutes between trials, and the best score recorded for subsequent analysis. The split (eg, 20–30 m) that provided maximum speed was used for analysis. The assessment of linear sprints using infrared timing gates is highly reliable and does not require familiarization.13–15
Statistical Analysis Dependent variables included maximum sprint speed (testing), peak match speed (highest recorded speed achieved during the match), mean sprint speed (average of all peak sprint speeds performed during the match), and relative peak and mean sprint speed in relation to maximum (testing) sprint speed. No 2 matches consisted of the exact same players, so separate 3-factor ANOVAs compared the dependent variables for each match using position (forward, midfield, defender), age group (U-21, U-17), and speed classification (>90th percentile, 19.0 km/h vs >20.0 km/h). Based on previous research11 we can approximate there would be a 0.75% reduction in the proportion of total sprint distance if we were to increase the threshold by 1.0 km/h, which would result in about 191 m of sprinting (3.45% of 5541 m). Nonetheless, this distance still exceeds a majority of the outcomes from the current study (except U-21 match 1) and highlights a performance gap between standards that must be overcome when advancing from youth teams to senior squads. Sprinting ability in female athletes tends to peak by approximately 15 to 16 years,20 which is supported by the lack of difference between age groups in the current study. Moreover, 8 of the 11
624
Table 3 Number of Sprints and Mean and Total Sprint Distance (m) During Matches Match 1
Match 2
Match 3
Match 4
Sprints
Mean distance
Total distance
Sprints
Mean distance
Total distance
Sprints
Mean distance
Total distance
Sprints
Mean distance
Total distance
U-21
10 ± 5*
15.6 ± 2.9
188 ± 66*
8±5
13.0 ± 2.3
118 ± 66
8±4
12.8 ± 3.8
83 ± 52
9±4
14.0 ± 5.2
124 ± 48
U-17
5±3
15.1 ± 4.7
117 ± 49
6±3
12.6 ± 2.3
80 ± 44
7±4
13.6 ± 3.5
93 ± 54
7±5
12.7 ± 4.4
88 ± 66
forward
8±5
15.1 ± 3.9
140 ± 70
7±4
13.4 ± 3.1
96 ± 54
7±3
14.1 ± 3.7
89 ± 43
7±5
13.1 ± 3.8
99 ± 78
midfielder
9±4
13.9 ± 2.6
176 ± 77
8±5
12.4 ± 2.4
116 ± 65
8±4
12.5 ± 3.3
86 ± 49
8±4
13.9 ± 5.4
119 ± 52
defender
5±3
17.6 ± 4.3
138 ± 44
5±3
12.8 ± 1.6
76 ± 48
7±4
13.7 ± 4.4
91 ± 70
8±4
12.9 ± 5.3
93 ± 49
slower
8±5
15.5 ± 3.5
159 ± 70
6±5
12.9 ± 2.6
94 ± 53
7±4
13.3 ± 3.9
80 ± 53
7±4
13.1 ± 5.5
94 ± 49
faster
7±4
15.1 ± 4.7
141 ± 66
9±4
12.6 ± 2.0
109 ± 70
10 ± 2
12.8 ± 2.7
116 ± 40
10 ± 6
14.1 ± 2.7
138 ± 72
Age group
Position
Speed classification
*Different from U-17 (P ≤ .010).
Figure 1 — Proportion of peak match speed to maximum sprint speed for positions (top), and speed classifications (bottom). *Different from midfielders and defenders. **Different from faster players.
Figure 2 — Proportion of mean sprint speed to maximum sprint speed for positions (top), and speed classifications (bottom). *Different from faster players. 625
626 Vescovi
players identified in the faster group were U-17 players. Of practical importance is that maximal sprint speed was observed in the final 5 m of the test for 26 (60%) of the participants and between 20 and 30 m for the remaining 18 (40%) athletes. Similarly, the highest values for maximal sprint speed reported for a group of high-level female soccer players occurred in the final 15 m of a 35-m sprint.4 Thus, to ensure that maximal speed is captured in female field-sport athletes, distances of 30 to 35 m should be used with measured splits during the final 5 to 15 m. The use of GPS technology in field sports is not without limitations—namely, the accuracy and reliability of measuring high-speed movements. This issue has been addressed by other researchers,5,6 with improved validity and reliability for units collecting at 5 Hz compared with 1 Hz.7 However, different manufacturers might also contribute to measurement error; in fact, Petersen et al8 demonstrated lower bias for sprinting 20 to 30 m using the same brand and model as in the current study (10–15%) than a commercially available competitor (20–37%). Furthermore, Waldron et al9 reported acceptable levels of accuracy (
