Sand surfaces are a popular alternative training venue for a range of firm ground sports (i.e., team sports, distance runners, etc), with research showing that there are distinct physiological and biomechanical differences associated with exercise on such training surfaces (5,7–9,11). These include a significant alteration in energy cost, kinematics, and muscle activation patterns when compared with firmer ground (7,9). Despite these differences, there is evidence to suggest that a transfer of training effects may be apparent between the 2 surfaces (2,8). Gortsila et al. (2) showed that sand-based agility training leads to significant improvements (p < 0.05) in agility performance (T test and Illinois test) measured on both sand and grass surfaces, despite no agility training on grass (2). Similarly, plyometric training on sand has been shown to improve jumping and sprinting abilities on firm surfaces (4). These findings suggest that training adaptations gained on a sand surface can transfer to firm ground performance. However, there is no evidence to show that the transfer of adaptations can occur in the opposite direction (i.e., from firm to sand surfaces), suggesting that surface specificity may be essential to gain any performance improvements on sand. Overall, the findings from previous research suggest that there may be a 1e-way transfer of training effects from sand to firm surfaces, with the adaptations unique to sand training also having a positive effect on firm ground performance.
With this in mind, the aim of this study was to determine the effect of a sand-based vs. grass-based training program on 20-m sprint times measured on both sand and grass surfaces. It was hypothesized that the sand-based training group would achieve significant improvements in 20-m sprint times on both sand and grass, whereas the grass-based training group would only achieve improvements on grass.
Experimental Approach to the Problem
Athletes were required to complete an 8-week training program in a sand-based (SAND) or grass-based (GRASS) training group, with 20-m sprint performance measured on both surfaces at baseline (week 0) and at the end of weeks 4 and 8. Training involved three 1-hour sessions per week, consisting of 2 sessions on the group-specific training surface (sand or grass) and an additional whole-group session on grass.
Twelve well-trained team-sport athletes were recruited from the Western Australian Institute of Sport Elite Field Hockey and Netball programs for participation in this investigation. The subjects were split equally by sport and then randomly assigned into 1 of 2 training groups (SAND and GRASS), with 5 women and 1 man in both the SAND (age: 22.2 ± 2.7 years; mass: 73.5 ± 9.9 kg; height: 176.2 ± 10.5 cm) and GRASS (age: 20.2 ± 1.9 years; mass: 68.4 ± 10.5 kg; height: 171.5 ± 6.1 cm) conditions. Participants were informed of the requirements and risks associated with their involvement in this study, before written consent acknowledging these details was obtained. Institutional Review Board (IRB) approval for the use of human subjects during this investigation was granted by the Human Research Ethics Committee of The University of Western Australia (IRB: RA/4/1/4373).
Before the commencement of the training period, all athletes completed a 4-week period of sand familiarization, comprising two 30-minute beach (sand) sessions a week that progressed from light jogging and walking up to higher intensity activity, which was ultimately similar to that performed in week 1 of the training intervention. Such an approach ensured that all athletes had an equal exposure to sand training leading into the experimental trials.
The athletes were instructed to attend all testing sessions in a well-rested and hydrated manner. Before each session, the athletes were also instructed to refrain from consuming alcohol and caffeine or from participating in intensive exercise for the previous 24 hours. Water was allowed ad libitum through each testing session, up to a maximum of 1 standard-sized drink bottle (600 mL). Furthermore, all testing sessions were conducted on the same day and time of the week to avoid any diurnal influences of circadian rhythm. All sand testing was conducted on a level area of soft dry beach sand, and all grass testing was conducted on a well-maintained sporting ground. Surface conditions were also quantified before each testing session to control for variations in surface moisture and stiffness (see Experimental Protocols). Athletes were barefoot during the sand testing and wore their normal running shoes on grass.
The 8-week training intervention was designed to replicate a generic team-sport preseason, comprising conditioning-based training sessions that progressed from interval running (15–45 seconds) to short sprint and agility efforts (approximately 5 seconds) and then on to simulated game play drills. Athletes were required to complete three 1-hour sessions per week, including 2 surface-specific sessions (SAND or GRASS) and 1 group session on grass. The training sessions were standardized between the groups, with work intervals matched for time. Additionally, training outside of these 3 sessions was controlled for, with similar training minutes (SAND: 2,441 ± 309 minutes and GRASS: 2,316 ± 197 minutes) and training loads (SAND: 11,934 ± 1,812 and GRASS: 10,978 ± 1,413 arbitrary units) reported between the 2 groups (p > 0.05). Here, training load was calculated via previously used methods (1).
20-m Sprint Performance
Sprint performance was assessed in the week leading up to the training program (week 0) and also before the group training session at the end of weeks 4 and 8 of the training intervention. Athletes completed three 20-m sprint efforts (with 2 minutes between repetitions) on each training surface, with testing conducted on the sand surface first, subsequent to a 10-minute warm-up (on sand). This consisted of a 4-minute continuous jog at moderate intensity, followed by a 3-minute period of self-selected stretching of at least 3 different muscle groups, then 3 minutes of run-throughs, with sprint pace gradually increased from moderate to maximum pace. A 3-minute recovery period followed, before the athletes completed the first 20-m sprint. Following the sand testing, the athletes had a 5-minute passive recovery period, after which they completed a second short warm-up on the grass surface (2-minute jog and 3 minutes of run-throughs). Again, a 3-minute recovery period followed the warm-up, before the athletes completed the first 20-m sprint on grass. Sprint times were recorded using electronic timing gates (±0.01 seconds; Swift Performance Equipment, Lismore, Australia) placed at the 0-m and 20-m marks of a straight line and level area of surface for both sand and grass. Athletes were instructed to start each sprint from a stable standing position, with no body movement or rocking allowed before takeoff. The fastest of the three 20-m sprint efforts was recorded for each training surface. The reliability of this protocol has previously been investigated in our laboratory with a similar group (i.e., gender, age, and performance) of team-sport athletes, yielding a typical error of measurement (TE) of 0.05 seconds.
Ground Surface Stiffness Determination
Surface stiffness for SAND and GRASS trials was determined immediately before each testing session using a 2.25 kg Clegg impact hammer (SD Instrumentation, Wiltshire, England) dropped from a height of 0.457 m. On each occasion, 10 samples (spread over the entire training area) were taken to determine the peak impact deceleration forces exerted by the surface. Additionally, sand surface samples were taken after training in areas of heavy use to quantify any compacted sand areas. The peak deceleration force (N) of each training surface was calculated in accordance with previously used methods (7–9).
All results are expressed as mean and standard deviation. Two-way repeated measures analysis of variance was used to analyze the group (SAND and GRASS) and surface differences (sand and grass) in 20-m sprint performance over the 8-week training program. Mauchly’s test of sphericity was used to assess the data variance, and a Greenhouse-Geisser correction was applied in the case of an assumption violation. Post hoc (Fisher’s Least Significant Difference) paired samples t-tests were used to determine specific differences. The alpha level was set at p ≤ 0.05. Where applicable, trends in performance were interpreted using Cohen’s d effect sizes (3). However, only effect sizes >0.40 were reported here. The thresholds for the qualitative descriptors of these effects were set at 0.40–0.79 as “moderate” and ≥0.80 as “large” (10). Finally, the smallest practically important performance change was also assessed with a 95% level of confidence that the absolute change was greater than the TE for 20-m sprint performance (3).
Ground Surface Conditions
There was no significant time effect (p = 0.869); however, a significant trial (p = 0.001) and time × trial interaction (p = 0.004) showed that the peak deceleration forces on the sand surface were significantly lower than those on grass for week 0 (SAND: 342 ± 160 N and GRASS: 1,245 ± 189 N), week 4 (SAND: 439 ± 137 N and GRASS: 1,135 ± 145 N), and week 8 (SAND: 413 ± 134 N and GRASS: 1,315 ± 113 N) (p < 0.05).
Sprint times improved on the sand surface from weeks 0 to 4 and from weeks 4 to 8 in the SAND group only (Table 1; p < 0.05), with no improvements observed in the GRASS group (p > 0.05). For sprint times on the grass surface, improvements were seen from weeks 0 to 4 and from weeks 0 to 8 in both the SAND and GRASS groups (p < 0.05). There were no significant trial, or time × trial interactions evident between the training groups for sprint times measured on either surface at weeks 0, 4, and 8 (p > 0.05). However, there was a large effect size (d = 0.80) to suggest that the 20-m sprint time on grass at week 8 was faster in the GRASS group. The mean sprint times were significantly faster on the grass vs. sand surface for all 3 tests and in both training groups (p < 0.05).
In addition to absolute sprint times, the change in sprint performance between weeks 0, 4, and 8 was also assessed for both surfaces, including an analysis of the smallest worthwhile change (Table 1). On sand, there was no significant time effect for the change in sprint performance (p = 0.99); however, a significant trial (p = 0.013) and time × trial interaction (p = 0.012) showed that the mean improvement in sprint performance was significantly greater for SAND when compared with GRASS for weeks 0 to 4 and weeks 0 to 8 (p < 0.05). Also, there was a “very likely” (97%) chance that the sand-based vs. grass-based training had a substantial benefit on sprint performance on the sand surface over the 8 weeks. For the grass surface, a significant time effect (p = 0.009) showed that the improvement in sprint performance from weeks 0 to 8 in the GRASS group was greater than the changes observed between weeks 0 to 4 and weeks 4 to 8 alone (p < 0.05). There were no trial or time × trial interactions between the training groups for changes in sprint times on the grass surface (p > 0.05). However, there was a large effect size (d = 0.82) to suggest that the improvement in 20-m sprint time on grass between week 0 and week 4 was greater in the SAND group. There was also a moderate effect size (d = 0.55) to suggest that the 20-m sprint improvement on grass between week 4 and week 8 was greater in the GRASS group.
An 8-week sand-based training program significantly improved 20-m sprint performance on both sand and grass surfaces, when compared with a grass-based training program that resulted in improvements on grass surfaces only. This training response may demonstrate that there is a 1-way transfer of training effects between sand and grass surfaces for improvements in straight-line sprint performance.
Currently, there is limited research on the effects of a sand surface when running at higher speeds, with evidence to suggest that the physiological and biomechanical differences between sand and grass surfaces are reduced as running speed increases (5,7–9). However, significantly slower 20-m sprint times on sand compared with grass were found, even when the sand sprinting was completed with a relatively “fresher” athlete, because of its completion before the grass sprint testing. This confirms that significant differences are still evident between the training surfaces at higher running speeds. Specifically, the unique low-impact nature of a sand training surface can result in a diminished energy return to the body, forcing a change in the kinematics and muscle activation strategies when compared with exercise on firm ground surfaces (9). These changes on sand include a significantly increased cadence, a greater forward trunk lean and hip range of motion, and an increased plantar flexion when compared with grass (9). Therefore, an improved 20-m sand performance within the SAND training group may only indicate that the athletes were trained to adapt their running technique, recruiting additional musculature that might have assisted with their sprint performance on sand. With this in mind, it should also be noted that the SAND training resulted in a larger improvement in sprint times on both sand and grass surfaces within the first 4 weeks of training, despite a greater focus on sprint development in the second 4-week block. This may be indicative of a rapid initial adaptation to the unique requirements of exercise on sand, before the athlete is able to adjust to the demands of the surface, and a plateau of training effects is seen. It has previously been suggested that repeated exposure to sand training can result in specific biomechanical and physiological entrainments that positively influence movement efficiency on sand (8). Future research should aim to investigate the kinematic differences between sand and grass 20-m sprint trials to better understand any biomechanical or technique adaptations that may result from habitual sand exposure. Furthermore, although sand training appears to be essential to gain 20-m sprint improvements on a sand surface, there is also evidence to suggest that sand training can positively influence firm ground performance (2,4,8).
Sand is commonly used as a training resource in firm ground sports not only for conditioning purposes but also for recovery and rehabilitation sessions. A higher energy cost training stimulus experienced on sand (5,7–9,11), as well as the benefits associated with the low-impact nature of the training surface such as a reduction in muscle damage markers (4,6), make sand an attractive alternative training venue to more traditional firm ground surfaces. However, one potential criticism of sand training is that it could compromise the training specificity needed for firm ground performance gains. It has been suggested that a sand surface is less effective than firm ground for inducing the neuromuscular adaptations needed for improvement in activities requiring the stretch-shortening cycle, such as jumping and sprinting (4). Specifically, the lower stiffness ratings on sand can reduce the stress placed on the musculoskeletal system during exercise, thereby limiting the resultant training effects on the efficiency of the muscle-tendon complex (4). However, the results of the current investigation showed similar improvements in 20-m sprint performance on the grass in both training groups, indicating that there was no limitation of a sand-based training program on firm ground sprint adaptations.
In this study, the SAND group only performed 2 hours of training per week on the sand surface and also had exposure to firm ground surfaces during the rest of the training week. This volume of training on sand was chosen in consultation with a number of team-sport coaches and was ultimately considered to be the maximum amount of training away from the primary (sport-specific) surface that team-sport programs would practically consider. However, our findings suggest that 2 hours of exposure to the sand surface per week was enough to cause significant improvements in straight-line performance on this alternate training surface. It is possible that greater changes in sprint performance would have been seen with a greater training load on sand; however, it is unclear whether this would come at a detriment to firm ground (grass) sprint performance. As such, this is an area of investigation that requires further assessment to determine the ideal integration of sand and firm ground surfaces into the daily training environment that may allow for athletes to achieve significant straight-line sprint improvements on both training surfaces.
Although novel findings to the effects of sand vs. grass training surfaces are presented here, the conclusions of this study are somewhat limited by the small subject cohort and gender imbalance. The recruitment of subjects in this study was ultimately limited by the availability of well-trained team-sport athletes over a 10-week period involving other training and competitive commitments. That said, this investigation involved an equal balance of high-quality team-sport athletes in each training group, and despite the small sample size, there were still significant differences seen within and between these groups. Clearly, such results indicate an apparent effect of the different training surfaces on straight-line sprint performance, with further research needed on a larger subject cohort to uncover any further differences that may be evident between sand and grass surfaces in this training context. With this in mind, the findings presented here should be interpreted as preliminary data interpretations. Furthermore, the training intervention used here was designed to replicate a generic team-sport preseason and therefore involved not only sprint and agility training but also aerobic interval training and game-play simulation drills. It is possible that greater changes in straight-line sprint performance would have been seen if a more specific speed-development program had been employed, and as such, this area should also be explored by further investigations.
The results of this preliminary investigation show that 2 hours of sand training per week is enough to cause significant improvements in straight-line sprint performance on a sand surface, when compared with firm ground training alone. This improvement on sand may be indicative of a surface-specific technique adaptation or as a result of physiological and biomechanical entrainments unique to habitual sand exposure. Furthermore, the SAND group also showed similar improvements to the GRASS group in firm ground performance gains, suggesting that the adaptations unique to sand training can also have a positive effect on firm ground performance gains. These results support the integration of sand and firm ground training venues when seeking improvements in straight-line sprint performance in firm ground sports. However, further research is needed in this area to determine the ideal mix of these training venues that will allow for the maximum training benefits to be gained from both surfaces.
For coaches and sport science practitioners working with firm ground sports, our results suggest that 2 hours of sand training per week is enough to cause significant improvements in 20-m sprint performance on both sand and grass surfaces. When compared with firm ground training alone, the integration of sand into the conditioning phase of a training program (i.e., 1–2 times per week) will not compromise firm ground speed gains.
The authors wish to acknowledge the assistance of Neil Hawgood, Michelle Wilkins and the athletes of the WAIS Hockey and Netball programs. They also thank Jim Crandell (Dr Baden Clegg Pty Ltd) and Peter Ruscoe (Sports Turf Technology Pty Ltd) for their assistance and expertise in dealing with the Clegg impact hammer.
1. Foster C, Florhaug JA, Franklin J, Gottschall L, Hrovatin LA, Parker S, Doleshal P, Dodge C. A new approach to monitoring exercising training. J Strength Cond Res 15: 109–115, 2001.
2. Gortsila E, Theos A, Smirnioti A, Maridaki M. The effect of sand-based training in agility of pre-pubescent volleyball players. In: Book of Abstracts. Cable T, George K eds. Liverpool, United Kingdom: XVI Annual Congress of the European College of Sport Science, 2011. pp. 643.
3. Hopkins WG. How to analyze a straightforward crossover trial (Excel spreadsheet). Available at: newstats.org/xcrossover.xls
. Accessed May 17, 2012.
4. Impellizzeri FM, Rampinini E, Castagna C, Martino F, Fiorini S, Wisloff U. Effect of plyometric training on sand versus grass on muscle soreness and jumping and sprinting ability in soccer players. Br J Sports Med 42: 42–46, 2008.
5. Lejeune TM, Willems PA, Heglund NC. Mechanics and energetics of human locomotion on sand. J Exp Biol 201: 2071–2080, 1998.
6. Miyama M, Nosaka K. Influence of surface on muscle damage and soreness induced by consecutive drop jumps. J Strength Cond Res 18: 206–211, 2004.
7. Pinnington HC, Dawson B. The energy cost of running on grass compared to soft dry beach sand. J Sci Med Sport 4: 416–430, 2001.
8. Pinnington HC, Dawson B. Running economy of elite surf iron men and male runners, on soft dry beach sand and grass. Eur J Appl Physiol 86: 62–70, 2001.
9. Pinnington HC, Lloyd DG, Besier TF, Dawson B. Kinematic and electromyography analysis of submaximal differences running on a firm surface compared with soft, dry sand. Eur J Appl Physiol 94: 242–253, 2005.
10. White J, Dawson B, Landers G, Croft K, Peeling P. Effect of supplemental oxygen on post-exercise inflammatory response and oxidative stress. Eur J Appl Physiol 113: 1059–1067, 2013.
11. Zamparo P, Perini R, Orizio C, Sacher M, Ferretti G. The energy cost of walking or running on sand. Eur J Appl Physiol 65: 183–187, 1992.