Introduction

Training and competition for many team-sports often requires sustained exercise performance in high ambient temperatures. Under these conditions elevated core (T core) and mean skin (T skin) temperatures are noted, resulting in increased cardio-vascular and metabolic loads (Kozlowski et al. 1985; Morris et al. 2000), in addition to hastening neuromuscular fatigue (Kay et al. 2001) and reducing endurance (Gonzalez-Alonso et al. 1999) and intermittent-sprint (Drust et al. 2005) performance compared with normo-thermic conditions. To counter the physiological strain associated with higher environmental temperatures, pre-cooling procedures which reduce pre-exercise T skin and T core have been used in laboratory and field settings (Marino 2002).

Increased thermal stress has shown to excessively elevate T core and subsequently reduce intermittent-sprint performance (Maxwell et al. 1999; Drust et al. 2005; Morris et al. 2005). As such, methods such as cold baths, cold ambient air and ice vests have been utilised to varying effect to delay the rise in T core during prolonged duration repeated sprint efforts (Marino 2002). These procedures have improved endurance performance in exercise protocols consisting of constant-load exercise to fatigue (Lee and Haymes 1995; Hasegawa et al. 2005), total maximal work performed in a set time (Booth et al. 1997) and time to complete variable-paced efforts of set distance (Arngrímsson et al. 2004). However, minimal performance benefits have been reported by the small number of studies investigating pre-cooling and intermittent-sprint exercise (simulating team-sport activity) (Drust et al. 2000; Duffield et al. 2003; Cheung and Robinson 2004). Previous research has not shown an improvement in peak or mean power, speed or distance covered in singular (Marsh and Sleivert 1999) or repeated (Drust et al. 2000; Duffield et al. 2003; Cheung and Robinson 2004) maximal sprint efforts of 5- to 30-s in duration following pre-cooling. However, recently Castle et al. (2006) have shown some ergogenic benefits for intermittent-sprint performance following lower body pre-cooling with a 4% increase in peak power. While pre-cooling procedures are generally effective in reducing T core, T skin and the perceived thermal strain of intermittent-sprint exercise protocols (Mitchell et al. 2001; Sleivert et al. 2001), evidence for improved intermittent-sprint performance in warm–hot conditions following pre-cooling is limited.

Team-sport exercise patterns involve intermittent-sprint activity incorporating maximal sprints separated by activity ranging from passive stationary recovery to high-intensity work (Spencer et al. 2005). Traditionally, exercise protocols simulating team-sport exercise involve repeated maximal (2–60) non-specific cycle ergometer sprint efforts (5–30 s) interspersed with sub-maximal recovery bouts (passive–50% \( \ifmmode\expandafter\dot\else\expandafter\.\fi{V}{\text{O}}_{{2\max }} \)) over prolonged durations (30–80 min) (Duffield et al. 2003; Cheung and Robinson 2004). Exercise performance is assessed via measures of peak and mean power, while the sub-maximal exercise separating maximal efforts is normally standardised and ignored as a performance measure. Morris et al. (2005) have previously reported significantly reduced sub-maximal distances covered in 33°C compared to 17°C ambient conditions. Hence, given the improved endurance performance in completing maximal aerobic events, reported for pre-cooling procedures (Arngrímsson et al. 2004), it is surprising more research attention has not incorporated the measurement of sub-maximal exercise in the assessment of performance.

Therefore, given the often ambiguous and non-specific nature of the reported effects of pre-cooling on intermittent-sprint exercise performance, the purpose of the present study was to investigate the effect of two common pre-cooling procedures (ice-bath and ice-vest) on maximal sprint and sub-maximal work during intermittent-sprint exercise in warm conditions.

Methods

Participants

Participants were nine male, moderate to well trained, club-level rugby players with a mean ± SD age 21.4 ± 1.3 year, height 184.1 ± 5.1 cm and body mass 85.16 ± 5.56 kg. All participants gave verbal and written consent prior to engaging in testing procedures and Human Ethics clearance was granted by the Institutional Ethics Committee.

Overview

Participants performed an initial familiarisation session, followed by three testing sessions at the same time of day, separated by 5–7 days and were required to abstain from the ingestion of alcohol, caffeine and food substances 4 h prior to testing. All testing was conducted on a 20-m synthetic running track in an enclosed biomechanics laboratory. Testing was performed in warm/hot environmental conditions (32 ± 1°C and 30 ± 3% relative humidity), with heating provided by a customised gas heating system, supplemented by four electronic 2,000 W room heaters (Kambrook, Australia) placed at 5-m intervals alongside (within 1-m) the running track. Following familiarisation with all measures and procedures, participants performed three identical sessions in a randomised, counter-balanced order, with only the type of cooling intervention varying between sessions. The three sessions consisted of a control session (no cooling intervention), an ice-vest session (cooling intervention with an ice-vest) and an ice-bath session (ice bath and ice-vest). All participants were required to document their physical activity and dietary and fluid consumption in the 24 h prior to the first testing session and replicate these patterns for the following sessions and particularly to standardise fluid consumption.

Exercise protocol

Participants performed an intermittent-sprint protocol consisting of 2 × 30-min identical halves, separated by a 10-min recovery period and performed on an enclosed 20-m synthetic running track. Initially a warm-up consisting of 4 min of running along the 20-m running track with increments in running speed each minute and 5-min of passive stretching was performed. The intermittent-sprint protocol consisted of a maximal 15-m sprint every minute, separated by sub-maximal exercise of varying intensities. The maximal 15-m sprint was performed on the running track, with a 5-m space for deceleration before impact with a large high jump mat to simulate body contact and collisions that occur in many team sports. Further, during the first and last minute of each half, an extra sprint was performed, reducing the recovery to 30 s. Between each maximal sprint effort were bouts of self-paced, sub-maximal activity, consisting of hard running, jogging and walking, respectively. These sub-maximal bouts were performed up and back along the 15-m sprint track in a shuttle-run fashion. A sub-maximal exercise bout was started as soon as the participant had finished the maximal sprint and was ceased 10 s prior to the commencement of the next sprint to allow time for preparation for the ensuing sprint. Only one sub-maximal exercise mode was employed each minute and these were rotated through in the previously mentioned order to ensure correct replication of procedures. Participants were instructed prior to and given verbal support during each session to attempt to cover as much distance as possible during the hard running bouts, and jog or at a self-selected comfortable pace during the respective jogging and walking bouts.

Pre-cooling intervention

A cooling intervention was performed for 15-min prior to exercise, during the warm-up and during the 10-min half-time recovery period. Participants in the control condition sat in the 32°C temperature and within 5-m of a radiant heat source during both the initial 15-min and then 10-min half-time recovery periods with no cooling intervention. During the ice-vest session, participants wore an ice-vest (Artic Heat, Australia) while sitting in the 32°C temperature within 5-m of a radiant heat source for the 15-min pre-cooling and 10-min half-time cooling and also wore the vest during the warm-up and stretching period. The ice-vest was stored in crushed ice prior to and following use. During the ice-bath session, participants were immersed up to the suprasternal notch in a cold water bath (Custom design, CSU, Australia) at 14 ± 1°C for the initial 15-min before donning the ice-vest and performing the warm-up and stretching in the vest, and then donning the vest during the 10-min half-time recovery period while seated as per other conditions. Finally, in all conditions, participants were required to consume a measured 350 ml of water during the half-time break.

Measures

Performance

Exercise performance was assessed by measures of 15-m sprint time and distance covered during sub-maximal exercise bouts. Sprint time was measured by an infra-red timing system (Speed Light, Swift, Australia) while % decrement was calculated according to Dawson et al. (1993). Markings in 1-m increments alongside the 15-m running track allowed for an accurate measure of the distance covered on each individual sub-maximal exercise bout performed in the respective hard running, jogging and walking bouts.

Physiological measures

Before and after each testing session, nude mass was measured on a set of calibrated scales (HW 150 K, A & D, Australia) to estimate changes in body mass due to sweat loss. Heart rate was measured (Vantage NV, Polar, Finland) pre-intervention, post-intervention, pre-exercise and every 5-min throughout the exercise protocol. T core was measured by a telemetric pill (VitalSense, Mini Mitter, USA) that was ingested 4–5 h prior to exercise to ensure it had passed into the Gastro-Intestinal tract. Skin temperature was measured at the sternum, chest, mid-forearm, mid-quadriceps and medial calf via telemetric patches (VitalSense, Mini Mitter, USA). T core and T skin were recorded pre-intervention, post-intervention, pre-exercise and every 10-min throughout the protocol from a hand-held monitor that telemetrically received core pill and skin patch measurements (VitalSense, Mini Mitter, USA). T skin was calculated based on the equation of Ramanathan (1964), while heat storage was calculated based on the equation of Havenith et al. (1995). Ratings of perceived exertion (RPE) and Thermal comfort were obtained pre-intervention, post-intervention, pre-exercise, either side of the half-time break and post-exercise. A 100 μl of capillary blood was sampled from a hyperaemic ear lobe pre-intervention, post-intervention and pre- and post-halves for analysis of capillary blood lactate [La−], pH, Potassium (K+) and Sodium (Na+) (ABL825 Radiometer, Copenhagen, Denmark). Finally, a 60 μl sample of capillary blood was collected pre- and post-exercise and centrifuged to separate blood plasma and analysed for hematocrit (Hct).

Statistical analysis

A repeated measures (condition × time) ANOVA was used to determine significant differences between the respective conditions (Control, Ice-vest, Ice-bath). Tukey’s post-hoc HSD tests were used to determine the source of significance, which was set a priori at P = 0.05. Effect sizes (ES) (Cohen’s d) were calculated to analyse potential trends in the data comparing respective cooling conditions to the control condition. An ES of <0.2 is classified as a ‘trivial’, 0.2–0.4 as a ‘small’, 0.4–0.7 as a ‘moderate’ and >0.8 as a ‘large’ effect.

Results

The results for mean ± SD 15-m sprint time, total sprint time and % decline in sprint time for the first and second half, respectively, for all conditions are presented in Table 1. No significant differences (P>0.05) and small ES were present between conditions for mean and total sprint time and % decline. The mean ± SD and total distance covered during hard running, jogging and walking exercise bouts for the first and second half and overall, respectively, for all conditions is presented in Table 2. The overall total distance covered during all sub-maximal exercise bouts were not significantly different (P = 0.09) for the ice-bath, ice-vest and control conditions (4865 ± 546, 4551 ± 418 and 4493 ± 403 m, respectively). No significant differences (P > 0.05) were noted between conditions for the mean or total distance covered in hard running (P = 0.09), jogging (P = 0.50) or walking (P = 0.61) either in the first or second half or overall. However, large ES data indicated a greater mean and total distance covered in the ice-bath session for hard running in both halves (ES = 0.84 and 0.86) and overall (ES = 0.88). Mean ± SD individual hard running efforts are presented in Fig. 1

Table 1 Mean ± SD 15-m sprint time, total sprint time for all sprints and % decline in sprint time for the first and second halves for the Control, Ice-vest and Ice-bath conditions
Full size table
Table 2 Mean and total ± SD distance covered for hard running, jogging and walking in the first and second halves respectively for the Control, Ice-vest and Ice-bath conditions
Full size table
Fig. 1
figure 1

Mean ± SD distance covered during individual free-paced hard running efforts through out the exercise protocol for Ice-bath, Ice-vest and Control Conditions. * Significant difference between Ice-bath and Control conditions (P < 0.05)

Full size image

Results for mean ± SD T core, T skin and chest temperature (T chest) for all conditions are presented in Fig. 2. A significantly lower T core (P < 0.05) was evident following the intervention, prior to the warm-up, in the ice-bath compared to the ice-vest and control condition and remained lower until the fortieth minute. Further, large ES data (ES = 1.0–1.8) indicated a lower T core in the ice-bath condition following the intervention for the entire exercise protocol. No significant differences (P > 0.05) and moderate ES data (ES = 0.3–0.5) were evident between the ice-vest and control conditions. A significantly lower T skin (P < 0.05) was evident following the intervention in the ice-bath compared to the ice-vest and control conditions throughout the warm-up and until the twentieth minute, with large ES data (ES = 1.0–12.0) indicating lower T skin in the Ice-bath condition following the intervention for the entire exercise protocol. No significant differences (P > 0.05), but large ES data (ES = 0.8–1.1) were evident between T skin in the ice-vest and control conditions until the twentieth minute. A significant difference (P < 0.05) and large ES (0.8–3.1) were evident in lower T chest values in the ice-bath and ice-vest conditions compared to the control condition following cooling until the tenth minute and again after the half-time cooling procedure. No differences (P > 0.05) were present between the respective cooling conditions. Results for change in heat storage over the testing protocol are presented in Fig. 3. In the ice-bath condition an initial loss of body heat was noted following cooling, with a reduction in heat storage until the tenth minute (P < 0.05).

Fig. 2
figure 2

a Mean ± SD Core temperature b Mean ± SD skin temperature and c mean ± SD chest temperature for Ice-bath, Ice-vest and Control conditions over the duration of the exercise protocol. * Significant difference of Ice-bath to Ice-vest and Control conditions respectively (P < 0.05). # Significant difference of both Ice-bath and Ice-vest to Control condition (P < 0.05)

Full size image
Fig. 3
figure 3

Mean ± SD heat storage for Ice-bath, Ice-vest and Control conditions over the duration of the exercise protocol. * Significant difference between Ice-bath and Ice-vest and Control conditions respectively (P < 0.05)

Full size image

Mean ± SD heart rate data are presented in Fig. 4 and show a significantly lower heart rate in the ice-bath condition following the cooling intervention until the tenth minute of the exercise protocol (including the warm-up). Large ES data also indicate reduced heart rates until 15-min of the exercise protocol (ES = 0.8–1.8) and again after half-time (ES = 0.90) in the ice-bath condition. A significant difference (P < 0.05) in the amount of sweat loss (as measured by pre-post difference in nude body mass) was evident, with a lower change in mass in the ice-bath than control condition (1.73 ± 0.17, 1.92 ± 0.24 and 2.13 ± 0.19 kg for ice-bath, ice-vest and control conditions, respectively). The large ES data indicated both ice-bath and ice-vest conditions had a reduced sweat loss compared to the control condition.

Fig. 4
figure 4

Mean ± SD heart rate for Ice-bath, Ice-vest and Control conditions over the duration of the exercise protocol. * Significant difference between Ice-bath and Control conditions (P < 0.05)

Full size image

The mean ± SD capillary blood [La−], pH, Na+, K+ and Hct data is presented in Fig. 5. No significant differences (P < 0.05) were evident between conditions for [La−], pH, Na+ or K+ throughout the exercise protocol. Large ES data (ES = 0.70) indicated lower pH values at the end of the first-half in the ice-bath condition; however, all other ES data indicated trivial to small differences. Significant differences (P < 0.05) were noted between the ice-bath and ice-vest respectively and control condition for the change in Hct. Attenuated changes in Hct were evident for the ice-bath (1.2 ± 0.4%) and ice-vest (1.6 ± 0.5%) compared to the control condition (2.3 ± 0.5 %).

Fig. 5
figure 5

a Mean ± SD capillary blood lactate b mean ± SD capillary blood pH, c mean ± SD capillary blood Sodium (Na+) and d mean ± SD capillary blood Potassium (K+) for the Ice-bath, Ice-vest and Control conditions

Full size image

RPE and Thermal comfort ratings throughout the protocol are presented in Table 3. No significant differences (P > 0.05) and small ES data (<0.4) were present between all conditions for RPE values. Significantly (P < 0.05) reduced Thermal comfort ratings were present in the ice-bath condition before the start of each half and in the ice-vest before the first-half compared to the control condition. Large ES (d = 1.0–3.5) were present for attenuated ratings in the ice-bath and ice-vest conditions at the start of each half.

Table 3 Mean ± SD Rate of perceived exertion (RPE) and rating of Thermal comfort (Therm) pre-intervention (Pre-int), pre-warm up (pre-wup), post warm-up (post-wup), pre-exercise (0 min), end of first-half (30-min), end of half-time (40-min) and end of exercise (70-min) for the Control, Ice-vest and Ice-bath conditions
Full size table

Discussion

The aim of the current study was to determine the effect of two different pre-cooling procedures on maximal sprint and sub-maximal work during intermittent-sprint exercise in warm conditions. Results for exercise performance indicated no significant differences between conditions for the time or % decline in 15-m sprint efforts or the distance covered during sub-maximal work bouts; however, large ES data indicate a greater distance covered during hard running following ice-bath cooling procedures. Physiological measures indicated lowered T core, T skin, heart rate, sweat loss and thermal comfort following ice-bath cooling compared with ice-vest or control conditions, with no differences present in capillary blood measures for [La−], pH, K+ or Na+.

Similar to most previous studies investigating the effects of pre-cooling on intermittent-sprint performance in the heat, limited ergogenic effects were apparent (Drust et al. 2000; Duffield et al. 2003; Cheung and Robinson 2004). Castle et al. (2006) recently reported a 4% increase in peak power following leg cooling, but no difference in mean or total work done for any cooling procedure (bath, vest or pack) compared to a control condition. Previous prolonged intermittent-sprint protocols that have shown limited declines in sprint performance in the heat, have reported limited beneficial effects of pre-cooling (Duffield et al. 2003; Cheung and Robinson 2004). In contrast, Castle et al. (2006) demonstrated a significant decline in sprint performance in the control condition and as such reported an ergogenic influence of a cooling intervention to maintain peak power without any significant change in work done.

Exercise protocols that allow sufficient recovery (>60 s) and repletion of PCr between sprints (Bogdanis et al. 1998), with low resistance to sprint efforts may result in limited declines in performance and hence limited pre-cooling ergogenic benefits (Duffield et al. 2003). It is possible that greater resistance and durations tease out declines in sprint performance and allow for a potential ergogenic pre-cooling influence. In the present study, a moderate decline in sprint performance was evident, however, little amelioration of the decline was evident following ice-bath cooling procedures. As such, it is possible that the effects of pre-cooling on intermittent-sprint performance are minor and only manifest when the thermal and exercise stress is sufficient to induce heat strain.

Although limited effects of pre-cooling on sprint performance were evident, a unique aspect of the current study, which has not been previously reported in pre-cooling studies on intermittent-sprint exercise, was the measurement of free-paced distance covered between sprints. The results indicated an increased distance covered during hard-running following ice-bath pre-cooling procedures (Fig. 1). Precooling has been reported to significantly increase the distance covered in 30-min of continuous running by ∼300-m (Booth et al. 1997) and decrease time to complete 5-km treadmill efforts by 13-s (Arngrímsson et al. 2004). In the current study, ice-bath cooling increased the total distance covered in the 18 respective 45-s hard-running bouts (∼14-min in total) by ∼200-m, which approach the distances reported by Booth et al. (1997).

The strategy of combining whole-body pre-cooling prior to warm-up to reduce T core and T skin, and then maintenance of cooling procedures during the warm-up (ice-vest) was effective in improving the distance covered during hard-running, compared to the less effective cooling procedures such as torso cooling alone (ice-vest). Moreover, while minimal differences were evident between conditions for distance covered in jogging and walking, respectively, the total distance covered for all sub-maximal bouts shows that the combined ice-bath and vest cooling procedure increased the distance covered by ∼300–400-m compared to control or ice-bath only. Thus, while pre-cooling procedures may have limited influence on intermittent-sprint performance, pre-cooling is potentially effective in improving the sub-maximal work of team-sport athletes in hot conditions, similar to data previously reported for endurance exercise (Booth et al. 1997; Arngrímsson et al. 2004; Hasegawa et al. 2005).

The proposition that exercise performance is reduced once T core reaches either a critical level or an increment of 2°C is well documented (Gonzalez-Alonso et al. 1999). As a result, physiological mechanisms likely to assist performance improvements following pre-cooling relate to the reduction of T core, T skin and efficiency in heat storage and removal. In the current study, the 15-min ice-bath procedure was effective at blunting the rise in T core throughout the exercise protocol. The effectiveness of cooling strategies varies with the extent and duration of cooling applied (Olshewski and Bruck 1988; Lee and Haymes 1995) and as such it is not unexpected that whole-body immersion maintained a lower T core and T skin compared to the torso-only cooling (only T chest reduced), however, this has not been as evident in all cases (Castle et al. 2006). Increases in heat storage and T core commenced from initial exposure to the warm environment in ice-vest and control conditions, yet were absent in the ice-bath condition. Several studies have indicated that there is a duration of 30–40-min before the thermal and cardio-vascular effects of pre-cooling wane (Hessemer et al. 1984; Wilson et al. 2002). Hence, the importance of maintaining the physiological advantages provided by whole-body pre-cooling until as close as possible to the start of exercise are evident, and therefore including a pre-cooling procedure during the warm-up is likely to be of benefit to performance (Arngrímsson et al. 2004; Webborn et al. 2005).

The reduced T skin and reduction in stored heat in the ice-bath condition blunted the extent of the exercise-induced rise in T core. In turn, a lower T core is likely to be a key contributor to a delayed onset of neurally-mediated peripheral vasodilation and sweating mechanisms, respectively (Kruk et al. 1990). The delay in redistribution of cardiac output to supply cutaneous requirements for the transfer of metabolically generated heat allows the maintenance of a greater central blood volume (Gonzalez-Alonso and Calbet 2003). Further, the maintenance of a lower T skin after pre-cooling results in a more efficient heat transfer gradient and reduces the requirement for evaporative sweat loss (Lee and Haymes 1995). Accordingly, the early presence of a reduced heart rate following ice-bath pre-cooling combined with an ameliorated decline in the respective post-exercise nude mass and Hct indirectly support the maintenance of a greater blood volume and more efficient thermo-regulatory control following pre-cooling; which may be of possible benefit to exercise performance.

The maintenance of a greater central blood volume will potentially reduce cardio-vascular load and provide an increased skeletal muscle blood supply (Gonzalez-Alonso and Calbet 2003) and theoretically assist muscle performance. However, given that differences in nude mass equate to a greater blood volume retention of ∼ 400 ml in the ice-bath condition and differences in heart rate were only present up to the tenth min, it is apparent that these factors alone may not sufficiently explain differences in self-selected work patterns. Centrally-mediated mechanisms resulting in a ‘feed-forward’ control of pacing in endurance exercise have been suggested to result in the selection of lower exercise intensities in warmer conditions or with higher T core (Kay et al. 2001). Recent intermittent-sprint data (Drust et al. 2005; Castle et al. 2006) has further indicated the role of central processes in adjusting motor-unit recruitment based on both the rate of rise and absolute T core per se. Previous research (Morris et al. 2000; Drust et al. 2005; Castle et al. 2006), as with the current study, have reported minimal differences in blood metabolite accumulation for intermittent-sprint exercise in hot and moderate conditions with cooling, further implicating the role of a central fatigue mechanism for the reduction in intermittent-sprint performance. In the current study, exercise was not terminated due to hyperthermia and therefore the adjustment to reduce exercise intensity was self-selected. This is indicated in the present data (Figs. 1, 2) where an increased distance covered during hard running and delayed heat storage following ice-bath cooling, suggests that subjects could perform more work for the same heat strain, especially when the rise in T core was blunted. As such, it is feasible to speculate that the role of a centrally-mediated increase in work may be a result of a greater central drive to skeletal musculature in the pre-cooling condition (Kay et al. 2001; Tucker et al. 2004).

In conclusion, pre-cooling methods did not significantly improve intermittent-sprint performance; however, a combined pre-cooling strategy of an ice-bath followed by ice-vest during warm-up did indicate an effect of increasing distance covered during sub-maximal hard-running bouts. Further, the combined ice-bath and vest procedure resulted in significantly lower T core, T skin, heart rate, sweat loss and thermal comfort than the vest or control conditions respectively. As such, the ergogenic benefits of effective pre-cooling procedures in warm conditions for team-sports may be predominantly evident during sub-maximal bouts of exercise. Further, for maximal benefit, pre-cooling interventions should be continued until as close as possible to exercise (game) commencement.