To examine the impact of hot ambient conditions on physical performance and physiological responses during football match-play.
Two experimental games were completed in temperate (∼21°C; CON) and hot ambient conditions (∼43°C; HOT). Physical performance was assessed by match analysis in 17 male elite players during the games and a repeated sprint test was conducted after the two game trials. Core and muscle temperature were measured and blood samples were obtained, before and after the games.
Muscle and core temperatures were ∼1°C higher (P<0.05) in HOT (40.3±0.1 and 39.5±0.1°C, respectively) compared to CON (39.2±0.1 and 38.3±0.1°C). Average heart rate, plasma lactate concentration, body weight loss as well as post-game sprint performance were similar between the two conditions. Total game distance declined (P<0.05) by 7% and high intensity running (>14 km⋅h−1) by 26% in HOT compared to CON), but peak sprint speed was 4% higher (P<0.05) in HOT than in CON, while there were no differences in the quantity or length of sprints (>24 km⋅h−1) between CON and HOT. In HOT, success rates for passes and crosses were 8 and 9% higher (P<0.05), respectively, compared to CON. Delta increase in core temperature and absolute core temperature in HOT were correlated to total game distance in the heat (r = 0.85 and r = 0.53, respectively; P<0.05), whereas, total and high intensity distance deficit between CON and HOT were not correlated to absolute or delta changes in muscle or core temperature.
Total game distance and especially high intensity running were lower during a football game in the heat, but these changes were not directly related to the absolute or relative changes in core or muscle temperature. However, peak sprinting speed and execution of successful passes and crosses were improved in the HOT condition.
Citation: Mohr M, Nybo L, Grantham J, Racinais S (2012) Physiological Responses and Physical Performance during Football in the Heat. PLoS ONE7(6): e39202. https://doi.org/10.1371/journal.pone.0039202
Editor: Maria Moran, Instituto de Investigación Hospital 12 de Octubre, Spain
Received: November 25, 2011; Accepted: May 21, 2012; Published: June 19, 2012
Copyright: © 2012 Mohr et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: The study was funded and supported by Aspetar, Qatar Orthopaedic and Sport Medicine Hospital. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Playing football in the heat has been shown to result in high body temperatures –, which may have a negative impact on performance and accelerate the development of fatigue during a game. While the total distance covered in a football match may be lower in relation to temperate conditions even in moderate heat , the total game-distance alone has been shown to be a poor football match-performance indicator , . With regards to sprint performance, maximal sprint performance may be improved when the muscle temperature is elevated prior to competition , whereas the ability to perform repeated high intensity exercise has been reported to deteriorate with hyperthermia even though higher muscle temperatures are reached , . A possible explanation may be that hyperthermia limits whole-body exercise even before peripheral muscle factors became limiting , , . Thus, hyperthermia appears to affect various physiological systems and types of physical performance components in a complex manner, making its net-effect on a multi-faceted activity like football difficult to resolve.
Only two studies have compared the effect of elevated ambient temperature on match performance in football. Ekblom  compared a game at moderate heat (30°C) to a game at a normal temperature (20°C), while Ozgunen et al.  compared two games in the heat with only two degrees difference. Neither of these studies could elucidate the physiological and performance responses of athletes normally competing in temperate conditions when exposed to severe heat stress (>40°C) which often occurs during international tournaments and will be expected at the upcoming football World Cup in Qatar in 2022. Mohr et al.  have reported that in the last 15 minutes of a football match played in moderately high environmental temperatures (∼30°C), high intensity running decreased markedly and the muscle temperature in some players was in excess of 41°C at the end of the match. Therefore, it appears that the ability to perform high intensity running towards the end of a match may be more affected in the heat compared to observations from temperate conditions. However, the mechanisms for this response have not been elucidated or directly compared to a matched control situation.
In an integrated physiological system, fatigue/altering the pacing strategy in hyperthermic settings appears to be mediated by a complex interplay between peripheral, central and perceptual mechanisms –, as well as anticipatory responses . However, hyperthermia-induced performance decrement has been suggested to be directly affected by an elevated cerebral temperature that may provoke central fatigue , . Fatigue during exercise has been associated with numerous physiological impairments and is highly dependent on the type of exercise performed . In football, where prolonged intermittent exercise is combined with multiple sprints, high intensity actions and challenging co-ordination tasks, fatigue is a complicated phenomenon involving interplay between several physiological mechanisms , . Therefore we hypothesized that the different components of football performance would respond differently to heat stress. Thus, the present study investigated the effect of environmental heat stress on body temperatures and activity and performance patterns during a football game. This was achieved by having elite Scandinavian football players, unfamiliar with exercising in severe heat stress, play a control game (CON) in a temperate environment (∼21°C) and a game in hot ambient conditions (HOT; ∼43°C).
Materials and Methods
All subjects were informed of potential risks and discomforts associated with the experiment before giving their written consent to participate. The study conformed to the code of Ethics of World Medical Association (Declaration of Helsinki) and was approved by the Research Committee from Aspetar, Qatar Orthopaedic and Sports Medicine Hospital in Doha, Qatar.
Seventeen of the 20 outfield players from two teams played the full 90 minutes in the two game trials and were included in this study. They were elite male football players (age: 26.6±1.2 yrs; height: 1.84±0.01 cm; body mass: 80.1±1.6 kg) from two Scandinavian countries (Faroe Islands and Denmark). Collectively, the players encompassed all outfield playing positions and a Yo-Yo intermittent recovery level 1 (Yo-Yo IR1) performance of 2236±57 m, which was completed as part of the experimental testing, and indicates that the participants were well-trained , .
Two experimental football game trials separated by six days were completed towards the end of the competitive season. The first experimental game (control; CON) took place at normal environmental temperature (21°C) in an indoor hall (Aspire Dome, Doha, Qatar) while the second game was played outside in a hot environment (43°C; HOT). Relative humidity during the two games was 55 and 12%, respectively. The training and preparation routines were identical prior to the two game trials. Both games were played on artificial grass on pitches with similar dimensions at the same time of the day. The pre-game and half-time procedures, as well as the coaching during the game, were similar to competitive game scenarios. The same match officials controlled both games. Twenty outfield players were part of the study initially, however, due to injuries, only 17 players completed both game trials. Substitute players that were not participating as subjects replaced the injured players. During the entirety of both games the players’ activity profile was determined by a multiple-camera tracking system. Baseline repeated sprint tests were completed prior to the experimental games in resting condition as well as after both games (all repeated sprint tests were performed on the same indoor track, in a thermo-neutral environment). Finally, physiological measurements were performed prior to and after the CON and HOT games.
Match Analysis Procedures
The players’ work profile was determined by the Amisco® multiple-camera semi-automatic passive tracking system , . Data for total distance covered, high-intensity running (>14 km⋅h−1) and sprinting (>24 km⋅h−1) was computed in 5 and 15-minute intervals, as well as for each half and the entire game. Two permanent game fatigue indexes (P-FI; ) were calculated as the percentage difference in high intensity running between the first (0–15 min) and the last (75–90 min) 15-minute interval of the game (P-FI1; see also ) and the percentage difference between the average high intensity running distance in the first five 15-minute periods (from 0–75 min) and the final 15-minutes (75–90 min) of the game (P-FI2). In addition, a temporary game-fatigue index (T-FI) was calculated based on the percentage difference between amounts of high-intensity running during the most intense 5-minute period and the 5-minute period following the most intense period (see also ). Finally, data on sprint characteristics during the game such as average and peak sprinting speed and sprint length were recorded.
Amisco® Multiple-camera System
The Amisco® system is a multiple-camera match analysis system (Amisco Pro®, version 1.0.2, Nice, France). The movements of all outfield players were recorded during the game by three stable synchronized cameras positioned at a height of approximately 10 m above the pitch and sampled data at a frequency of 25 Hz. Signals and angles obtained by the encoders were sequentially converted into digital data and recorded on a computer for post-match analysis. From the stored data, the distance covered, time spent in the different movement categories and the frequency of occurrence for each activity was determined by Athletic Mode Amisco Pro® (Nice, France) , , . The following locomotor categories were used: Total distance covered, high intensity running (>14 km⋅h−1) and sprinting (>24 km⋅h−1), which are similar to movement categories described by others , , , . The multiple-camera match analysis approach has been shown to be a precise method of assessing game activities in football  and is programmed to detect work rate (distance covered within different speed thresholds) changes throughout a game . Correlation analyses were performed on match activities and the physiological response during the games. In addition to the movement pattern assessment, technical parameters were determined using the multiple-camera match analysis approach as previously described , .
Heart rate was recorded throughout the games (Polar Team system 2, Polar Electro, Kempele, Finland). The maximal heart rate of the participants recorded during the Yo-Yo IR1 (see ) was 189±2 bts⋅min−1.
Quadriceps muscle temperature (Tm) was determined in the medial part of vastus lateralis at a depth of 3 cm adjusting for the thickness of the skin by a needle thermistor (MKA08050-A, Ellab A/S, Rødovre, Denmark), with a precision of 0.1°C as previously reported by Mohr et al. . All thermistors were calibrated against a mercury thermometer. Prior to insertion of the sterile needle thermistors, the skin was disinfected by ethanol (70%). Tm was determined in 12 players representing all outfield playing positions immediately after each of the two halves. Core temperature (Tc) was monitored via a wireless ingestible thermometer pill (VitalSense, precision 0.01°C, Mini Mitter, Respironics, Herrsching, Germany) swallowed early in the morning before breakfast, to maximize the transit time before the experimental games. However, due to technical failures, this was substituted by rectal measures using clinical electronic rectal thermometers (Philips, HF, 365, China – precision 0.01°C) at a depth of ∼2 cm following the first and second half of the CON game in 10 of the 17 participants. Ingestible pills have been proven to be a viable alternative to rectal temperature  and validated as displaying similar values and circadian variation as rectal temperature .
Blood was drawn according to standard procedures from an antecubital vein into heparinized Plasma-Vacutainers (Becton Dickinson, USA). Blood was taken directly before and immediately after the experimental matches. The samples were then stored on ice and transported within 15 minutes to the laboratory. After centrifugation the plasma was analyzed directly for lactate, glucose and sodium on a Cobas b221 Blood Gas Analyzer (Roche, Mannheim, Germany). The analytical variation, expressed as Percentage of Coefficient of Variation (CV%) was 1.5% for lactate, 3.5% for glucose and 0.4% for sodium.
Fluid loss in each player was also determined after voiding nude body mass was measured before the warm-up and after the game (TANITA HD-316, Tokyo, Japan) and the individual fluid intake during the game was accounted for. The players consumed a standardized diet prepared by a nutritionist during the days preceding both games, but only water was provided during the game. The sweat rate was calculated based on the total fluid loss, correcting for urinal loss and fluid ingested.
Repeated Sprint Testing
Within 2 to 5 minutes after the end of the game, players carried out a repeated sprint test composed of 3×30 m straight-line sprints interspersed with 25 seconds of active recovery (easy jog back to the start line) , . The test was performed indoors in a thermo-neutral environment (∼20°C) on an indoor track. Sprinting time was measured by photocell gates placed 1.0 m above ground level (Polifemo Radio Light, Micro Gate, Bolzano, Italy). Each sprint was initiated from a standing position with the arms raised to chest height, 20 cm behind the photocell gate which started a digital timer as described by Mohr et al , . The time of each sprint was recorded and reported. The players were familiarized with the test prior to commencement of the study.
Values are presented as mean ± SEM. Differences in activity parameters, heart rate response, blood lactate, fluid loss and sprint performance between CON and HOT games were determined using the paired Student’s t-test including Bonferroni correction. Differences between the two halves, as well as 5 and 15-minute periods within the match in movement activities and body temperatures, were determined using an analysis of variance test with two-way repeated measurements (time x condition). In the case of a significant difference between time periods, a Tukey post hoc test was used to identify the points of difference. The relationship between selected variables was evaluated using Pearson’s product-moment correlation coefficient. Significance level was set as P<0.05.
Tc was higher in HOT than CON (Figure 1, P<0.05), both after the 1st half (39.6±0.1 vs. 38.7±0.2°C) and 2nd half (39.6±0.1 vs. 38.3±0.1°C) of the game. Peak Tc was 39.7±0.1°C in HOT and higher (P<0.05) than in CON (38.8±0.2°C). Tc was higher (P<0.05) after the 1st half than after the 2nd half in CON, while Tc tended (P = 0.058) to be higher after the 2nd than after the 1st half in HOT (Figure 1). Tm was not significantly different between trials after the 1st half (40.2±0.1°C in HOT and 39.7±0.2 in CON°C), but was higher in HOT than CON after the 2nd half (40.4±0.0°C vs. 39.2±0.1°C, P<0.05) (Figure 1). Peak Tm was higher in HOT than CON (40.4±0.1°C vs. 39.8±0.2°C, P<0.05). Tm was higher (P<0.05) after the 1st half than after the 2nd half in CON, while there was no difference between halves in HOT (Figure 1).
Total distance covered was 7% shorter (P<0.05) in HOT compared to CON (Figure 2A). A greater distance was covered (P<0.05) in the 1st than in the 2nd half in both HOT (2.7%) and CON (4.2%). In both halves less distance was covered in HOT than in CON (Figure 2A). High intensity running distance was 26% shorter (P<0.05) in HOT than in CON (Figure 2B). There was no difference between halves in HOT while in CON the players ran less (23.5%; P<0.05) at high intensities in the 2nd than in the 1st half (Figure 2B). Players performed 41% less (P<0.05) high intensity running in the first half in HOT compared to CON, with no differences between game trials in the second half (HOT: 883±45 m and CON: 978±97) (Figure 2B). No difference was observed between HOT and CON in total sprint distance (Figure 2C). Moreover, no difference in sprint distance between game trials in either the 1st or the 2nd half of the game was evident (Figure 2C).
N = 17. *Significant different between conditions; #significant different from the first half (P<0.05).
In CON, more (P<0.05) high intensity running was performed during the first four 15-minute intervals (range: 361–422 m) than in the final 15-minutes (300±33 m), while no difference was evident between 15-minute periods in HOT (range: 249–350 m). Thus, the P-FI1 was 28.0±5.6% in CON, which was higher (P<0.05) than in HOT (−7.2±12.3%). The PFI2 was also higher (P<0.05) in CON than in HOT (36.9±8.2% vs. −0.9±9.7%). Sprint distance was not significantly different between the 15-minute periods of the game in either HOT (range: 29–47 m) or CON (range 30–45 m). No significant difference was evident between the two games in sprinting distance in 15-minute intervals of the game.
Less (16%; P<0.05) high intensity running was performed during the most intense 5-minute periods in HOT compared with CON (207±9 vs. 241±15 m). However, in the 5-minute period immediately after the most intense 5-minute interval, 74% more (P<0.05) high intensity distance was covered in HOT compared to CON (99±11 vs. 57±12 m, respectively) (Figure 3). In CON, the amount of high intensity distance in the 5-minute period following the most intense periods was 114% lower (P<0.05) than the average distance covered at high intensities in all other 5-minute intervals, with no significant difference in HOT (Figure 3). The T-FI was lower (P<0.05) in HOT (50.6±5.8%) than in CON (78.1±3.7%).
Average for all other 5-min periods are shown as the solid and dashed horizontal line for CON and HOT, respectively. N = 17. *Significant different between conditions (P<0.05).
Peak sprinting speed reached in the game was 4% faster (P<0.05) in HOT (33.3±0.4 km⋅h−1) than in CON (32.1±0.4 km⋅h−1). There was no difference in sprint frequency (11±1 and 13±1) and length (19.6±0.7 and 19.9±0.8 m) between HOT and CON respectively. Peak speed reached in a sprint was higher (P<0.05) in HOT than CON in the first, third and fourth 15-minute interval (Table 1). Average sprinting speed was also higher (P<0.05) in HOT in comparison to CON in the first 15-minutes of each half (Table 1). No significant difference in average length of sprint in 15-minute intervals between HOT and CON (Table 1) was evident.
In HOT there was a higher success rate for all passes (73.8±1.2 vs. 66.1±2.1%) and forward passes (67.7±2.1 vs. 59.7±2.7%, respectively) (Figure 4) compared to CON (both P<0.05). Additionally the success rate for crosses in HOT was higher (P<0.05) than in CON (39.3±3.2 vs. 29.9±3.0%, respectively) (Figure 4). There was no significant difference in the total number of passes, forward passes, crosses or average length of passes and crosses performed in HOT and CON. The higher success rate of passes and crosses implies increased ball possession in HOT compared to CON (26.5 vs. 21.2 min, respectively). Conversely, both gain and loss of ball possession were higher (P<0.05) in CON (14.0±1.3 and 21.1±1.5, respectively) in comparison to HOT (6.4±0.8 and 9.8±1.1, respectively). Additionally, the average number of challenges was higher (P<0.05) in CON than in HOT (5.8±0.3 vs. 4.2±0.3). The average sum of ground and air duels was also higher (P<0.05) in CON (2.2±0.2 and 1.8±0.1, respectively) than HOT (1.4±0.2 and 0.8±0.1, respectively).
Heart Rate Response
Average heart rate during the HOT and CON games was 158±2 and 160±2 bts⋅min−1, respectively, corresponding to 83.7±0.8 and 84.6±0.8% of the maximal heart rate. The peak heart rate reached during the games was 185±3 and 183±2 bts⋅min−1 in HOT and CON, respectively, which is 97.7±0.7 and 97.0±0.8% of the maximal heart rate. There was no significant difference in average or peak heart rate between the HOT and CON trials.
Plasma Lactate, Glucose and Sodium
The plasma lactate concentration was 1.7±0.1 and 1.6±0.1 mmol⋅l−1 prior to HOT and CON and increased (P<0.05) to 4.9±0.5 and 3.3±0.4 mmol⋅l−1 immediately after the games, respectively, with no significant difference between trials. Plasma glucose was 5.0±0.2 and 5.3±0.1 mmol⋅l−1 prior to HOT and CON, respectively and increased (P<0.05) to 6.8±0.2 and 6.4±0.2 mmol⋅l−1 after the games with no significant difference between HOT and CON. The plasma Na+ concentration increased (P<0.05) from 138±0 to 142±1 mmol⋅l−1 during the HOT game trial and from 138±3 to 140±1 mmol⋅l−1 in CON, but was not different across conditions.
Fluid Loss and Intake
Total sweat volume was 4.14±0.16 L in HOT, which was higher (P<0.05) than CON (2.59±0.14 L) and corresponded to average sweat rates of 2.48±0.09 and 1.55±0.09 L⋅h−1, respectively. The water intake was greater (P<0.05) during HOT than CON (2.6±0.2 L vs. 1.1±0.1 L, respectively) making the game-induced decrease in body mass similar in HOT and CON (1.9±0.3 and 1.8±0.1%, respectively). The percentage decrease in total game distance from CON to HOT was not correlated (r = 0.34, P = 0.16) to the game-induced loss in body mass.
Repeated Sprint Performance
Compared to baseline, sprint performance after both games was reduced (P<0.05), but there were no significant differences in post-game repeated sprint performance between HOT and CON (Figure 5). Total sprint test time was 13.36±0.11 seconds in a rested state before the experimental games and decreased (P<0.05) by 1.7 and 2.1% respectively after HOT and CON. The 1st and 3rd sprints were slower (P<0.05) after HOT compared to baseline, while the 2nd and 3rd sprints were slower (P<0.05) in CON (Figure 5).
Delta increase in core temperature in HOT was correlated (r = 0.85; P<0.05) to the total distance covered (Figure 6), but not in CON (r = 0.43; NS). Absolute core temperature at the end of HOT was also correlated (r = 0.53; P<0.05) to total distance covered and distance covered in the last 15 minutes of the HOT game (r = 0.48; P<0.05). The relative distance-deficit between CON and HOT was not correlated to peak core temperature, while the delta increase in core temperature from CON to HOT did not correlate to percentage reduction in total distance covered from CON to HOT. Also, plasma lactate concentration at the end of HOT was not correlated to percentage reduction in total distance covered from CON to HOT. Core temperature was not correlated to neither T-FI nor P-FI1 and PFI2 in the HOT game. Total distance covered and high intensity running during CON was correlated (r = 0.79 and 0.76, respectively; P<0.05) with Yo-Yo IR1 performance, whereas in HOT, high intensity running correlated to Yo-Yo IR1 performance (r = 0.65, P<0.05), but not total distance covered (r = 0.37, NS). The P-FI1 and P-FI2 in CON correlated inversely with Yo-Yo IR1 performance (r = −0.69 and r = −0.50, respectively; P<0.05), but not in HOT. Total sprint time in the baseline repeated sprint test correlated inversely (r = −0.53; P<0.05) with high intensity running in CON, but not in HOT.
Our data showed that well-trained football players have a moderate reduction in total distance (7%) and a marked decline (26%) in the amount of high intensity running completed when a game is played in hot ambient conditions compared to a temperate environment. In contrast, peak sprint velocity was higher in the heat and the technical quality, as indicated by percentage of successful passes and ball possession, was enhanced rather than impaired. The reductions in total and high intensity running were not correlated to absolute or delta changes in core or muscle temperature. Neither changes in plasma lactate, heart rate nor the net-loss of body mass could predict the hyperthermia-induced decline in work performance. In HOT the players who achieved the highest core temperatures were those with the highest work capacity, whereas there was no relationship between core temperature changes and work performance in the control game.
The altered pacing strategy in the HOT game is in accordance with responses observed when games are completed under moderate heat stress . In the present study the movement patterns throughout the game were studied in detail and showed several discrepancies in different match performance variables between the two game trials. For example, in the CON game there was a greater distance deficit in high intensity running in the second half compared to the first half which did not occur in the HOT game. The observation of a higher exercise intensity in the first compared to the second half in games played in thermo-neutral conditions agrees with numerous studies , , , . A possible explanation for the different responses in CON and HOT is that the high work rate in the first half of the CON game will deteriorate physical performance in the second half, which is supported by Rampinini et al. . On the other hand, the pacing strategy was markedly different in HOT with a relatively low work rate during the first half (41% less high intensity running than in CON) potentially leading to less accumulative fatigue. Consequently, players were able to achieve similar absolute exercise intensity as the second half in the CON game, despite a 23°C warmer ambient temperature and markedly higher body temperatures. Another suggestion is that the players change their behavior tactically in the heat analogously to the changes in the duration of duels and rest between duels during tennis in environmentally challenging situations  and by anticipatory altered pacing during time-trial cycling in heat , .
In addition to a more constant but lower running intensity in HOT than CON, the technical data substantiated large differences between the two games. Indeed, time in ball possession was ∼25% higher and the success rate of passes, forward passes and crosses was significantly greater in HOT compared to CON. In contrast, gain and loss of ball possession as well as the number of duels was higher in the CON trial, indicating a more transition-based type of play in CON compared to the more possession-orientated play in HOT. Others have shown that technical variables are affected by playing standard , surface , field dimensions , playing formation  and physiological fatigue . In the present study the same players, similar turf and field dimension as well as the same type of playing formation were applied in both game-trials. This suggests that the higher running intensity in CON is likely to have provoked a higher number of technical mistakes. In part this might be consequent to an increase in pressure from the opposition when in ball possession and due to a greater degree of transient muscle fatigue throughout the CON game. Hyperthermia may affect cognitive performance during complex tasks , however it appears that any negative effect of heat stress on the players technical skills was compensated for by the lower amount of high intensity running and hence less pressure on the player in ball possession.
In normal temperate conditions fatigue has been shown to develop in the final 15-minute interval of competitive high level games ,  as well as in a simulated football protocol . The P-FI1 and P-FI2 represent the ability to exercise at a high intensity during the final 15-minutes of a game, compared to earlier in the game . These indicators of accumulative game-induced fatigue or altered pacing were noticeable higher in CON than HOT. Thus the exercise intensity during the CON game resulted in a much more pronounced decline in running intensity in the final and critical stage of the CON game compared to HOT. However, despite the markedly different pacing strategy in HOT than in CON, repeated sprint test performance was equally compromised after the games irrespective of the ambient temperature. This indicates that the overall fatigue response after the two game trials was the same. In addition, during the most intense 5-minute period of the game, where temporary fatigue occurs that mandates a short recovery , , , , , 16% more high intensity running was performed in CON than in HOT. This created a greater decline in high intensity running in the following 5-minute interval and produced a concomitant higher T-FI in CON compared to HOT (Figure 3). Therefore, the higher exercise intensity in CON also appears to increase the degree of the more transient type of performance decrement  in CON compared to the HOT game. This notion is further supported by the fact that the total sprint distance was the same in CON and HOT, but the peak and average sprint speed was higher in HOT than in CON. Hence the pacing patterns appear to be greatly affected by hot ambient temperatures, resulting in markedly lower but more constant exercise intensity with improved sprinting quality in comparison to a temperate environment.
In further support of hyperthermia-induced effects on the pacing strategy in HOT is the fact that Yo-Yo IR1 performance was correlated to the total distance covered and both P-FIs in CON, but not in HOT. Thus the pacing in football can partly be predicted from training status at normal temperatures but less in severe heat where other physiological (e.g. thermal) and behavioral (i.e. pacing) mechanisms , , , , , ,  might dictate the limitations for physical performance.
In the present study, core temperature was more than 1°C higher in HOT than in CON. The core temperature levels were above 39°C during most of the HOT game and peak levels exceeded 39.5°C which is similar to temperatures reported previously in a football game in hyperthermic conditions . The increase in core temperature was correlated to the total game distance in the HOT game, but the distance deficit in high intensity running from CON to HOT was not directly correlated to the player’s peak core temperature or the delta increase in core temperature from CON to HOT. Therefore, physical activity during a hot game is not simply a function of absolute core temperature, but may potentially be dictated by individual differences in the response to and the perception of heat, leading to different pacing strategies.
While body core hyperthermia may hamper prolonged exercise performance or central activation during repeated sprints –, , , a potential benefit of the elevated muscle temperature may be gained. Specifically, increased muscle temperature has been shown to enhance sprint performance during a football game . Our data showed that hot ambient conditions have a variable impact on sprint performance since the total sprint distance was the same in HOT and CON, whereas the peak and average speed was higher in HOT than CON. Several studies have reported improved muscle contractile properties  leading to higher power production and, in turn, better performance  in hot compared to neutral ambient conditions. This seems to be specifically the case in the morning, before the diurnal increase in body temperatures (due to circadian rhythms) acts as a warm-up in the late afternoon . In the current study, the kickoff of both games was at 1100 hours confirming the potential positive effect of hot ambient temperature on sprint capacity at this time of day. In addition the average speed in the sprints was higher in HOT than in CON during the first 15 minutes of each half where the core temperature was relatively low, indicating that elevated muscle temperature improves sprint ability when core temperature is maintained below ∼39°C.
It is also likely that the lower average intensity in the most intense periods of the game in HOT stressed peripheral fatiguing mechanisms such as muscle acidosis and, to a lesser extent, extracellular potassium accumulations , . Therefore, the decreased work rate in the hot environment probably lowered the level of peripheral fatigue which in turn might be beneficial for sprint performance during the game. However in the present study, blood lactate concentrations after the HOT game were not lower than after CON, suggesting that glycolytic activity in the heat was at least as high as the control game despite the reduced amount of high intensity running. Additionally, the muscle glycogen levels measured in muscle biopsies obtained immediately after both games were similar in HOT and CON i.e. higher than 250 mmol⋅kg−1 d.w. (data not shown). Therefore it is unlikely that the degree of depleted muscle glycogen stores contributed to the differences in match performance between HOT and CON.
High fluid-loss may affect both the capacity to perform prolonged submaximal and intense exercise . However in the present study, the net-fluid loss was only moderate (<2% of body mass) and was similar in HOT and CON due to a markedly higher fluid intake in the hot game. Interestingly, when ad libitum water was provided to the players, they voluntarily drank a sufficient amount to maintain their body weight. Additionally, the percentage loss of body mass in HOT did not correlate to the distance-deficit from CON to HOT. Taken together, these observations suggest that the lower exercise intensity was not mediated by dehydration, which also is supported by others . The sweat rate was very high in HOT and is in line with recent reports from high level competitive football games played at high environmental temperatures .
In conclusion, the present study shows, for the first time with detailed analysis of the players activity patterns, how competitive football in a hot environment affects the various component of physical performance. Heat stress lowers the average work rate in the game, as well as high intensity efforts, but reduces intensity fluctuations throughout the game and increases peak sprinting speed and certain technical parameters of the game. The core and muscle temperature changes were not strong predictors of the performance changes in hot ambient condition.
The authors would like to thank the football players involved as well as their coaches and clubs. We thank the nurses, physiotherapists and sports medicine physicians from Aspetar for taking care of the players. Sarah Jackman, Joanna Bowtell and Stephen Bailey are acknowledged for grammatical advice. The study was performed within Aspetar, Qatar Orthopaedic and Sports Medicine Hospital and Aspire, Academy for Sports Excellence in Doha, Qatar.
Conceived and designed the experiments: MM LN JG SR. Performed the experiments: MM LN JG SR. Analyzed the data: MM LN JG SR. Contributed reagents/materials/analysis tools: MM LN JG SR. Wrote the paper: MM LN JG SR.
- 1. Ekblom B (1986) Applied physiology in soccer. Sports Med 3: 50–60.B. Ekblom1986Applied physiology in soccer.Sports Med35060
- 2. Mohr M, Mujika I, Santisteban J, Randers MB, Bischoff R, et al. (2010) Examination of fatigue development in elite soccer in a hot environment: a multi-experimental approach. Scand J Med Sci Sports 20: 125–132.M. MohrI. MujikaJ. SantistebanMB RandersR. Bischoff2010Examination of fatigue development in elite soccer in a hot environment: a multi-experimental approach.Scand J Med Sci Sports20125132
- 3. Ozgunen KT, Kurdak SS, Maughan RJ, Zeran C, Korkmaz S, et al. (2010) Effect of hot environmental conditions on physical activity patterns and temperature response of football players. Scand J Med Sci Sports 20: 140–147.KT OzgunenSS KurdakRJ MaughanC. ZeranS. Korkmaz2010Effect of hot environmental conditions on physical activity patterns and temperature response of football players.Scand J Med Sci Sports20140147
- 4. Mohr M, Krustrup P, Bangsbo J (2003) Match performance of high-standard soccer players with special reference to development of fatigue. J Sport Sci 21: 439–449.M. MohrP. KrustrupJ. Bangsbo2003Match performance of high-standard soccer players with special reference to development of fatigue.J Sport Sci21439449
- 5. Bradley PS, Carling C, Archer D, Roberts J, Dodds A, et al. (2011) The effects of playing formation on high-intensity running and technical profiles in English Premier League soccer matches. J Sports Sci 29: 821–30.PS BradleyC. CarlingD. ArcherJ. RobertsA. Dodds2011The effects of playing formation on high-intensity running and technical profiles in English Premier League soccer matches.J Sports Sci2982130
- 6. Mohr M, Krustrup P, Nybo L, Nielsen JJ, Bangsbo J (2004) Muscle temperature and sprint performance during soccer matches – beneficial effects of re-warm up at half-time. Scand J Med Sci Sports 14: 156–162.M. MohrP. KrustrupL. NyboJJ NielsenJ. Bangsbo2004Muscle temperature and sprint performance during soccer matches – beneficial effects of re-warm up at half-time.Scand J Med Sci Sports14156162
- 7. Drust B, Rasmussen P, Mohr M, Nielsen B, Nybo L (2005) Elevations in core and muscle temperature impairs repeated sprint performance. Acta Physiol Scand 183: 181–190.B. DrustP. RasmussenM. MohrB. NielsenL. Nybo2005Elevations in core and muscle temperature impairs repeated sprint performance.Acta Physiol Scand183181190
- 8. Mohr M, Rasmussen P, Drust B, Nielsen B, Nybo L (2006) Environmental heat stress, hyperammonemia and nucleotide metabolism during intermittent exercise. Eur J Appl Physiol 97: 89–95.M. MohrP. RasmussenB. DrustB. NielsenL. Nybo2006Environmental heat stress, hyperammonemia and nucleotide metabolism during intermittent exercise.Eur J Appl Physiol978995
- 9. Nybo L (2008) Hyperthermia and fatigue. J Appl Physiol 104: 871–878.L. Nybo2008Hyperthermia and fatigue.J Appl Physiol104871878
- 10. Nybo L, Secher NH (2004) Cerebral perturbations provoked by prolonged exercise. Prog Neurobiol 72: 223–261.L. NyboNH Secher2004Cerebral perturbations provoked by prolonged exercise.Prog Neurobiol72223261
- 11. Racinais S, Girard O (2012) Neuromuscular failure is unlikely to explain the early exercise cessation in hot ambient conditions. Psychophysiology 2012 Mar 14 [Epub ahead of print]. S. RacinaisO. Girard2012Neuromuscular failure is unlikely to explain the early exercise cessation in hot ambient conditions.Psychophysiology 2012 Mar 14 [Epub ahead of print]
- 12. Tatterson AJ, Hahn AG, Martin DT, Febbraio MA (2000) Effects of heat stress on physiological responses and exercise performance in elite cyclists. J Sci Med Sport 3: 186–193.AJ TattersonAG HahnDT MartinMA Febbraio2000Effects of heat stress on physiological responses and exercise performance in elite cyclists.J Sci Med Sport3186193
- 13. Nybo L, Nielsen B (2001) Hyperthermia and central fatigue during prolonged exercise in humans. J Appl Physiol 91: 1055–1060.L. NyboB. Nielsen2001Hyperthermia and central fatigue during prolonged exercise in humans.J Appl Physiol9110551060
- 14. Nybo L (2010) Cycling in the heat: performance perspectives and cerebral challenges. Scand J Med Sci Sports 20: 71–79.L. Nybo2010Cycling in the heat: performance perspectives and cerebral challenges.Scand J Med Sci Sports207179
- 15. McKenna MJ, Bangsbo J, Renaud JM (2008) Muscle K+, Na+ and Cl- disturbances and Na+-K+ pump inactivation: implication for fatigue. J Appl Physiol 104: 288–95.MJ McKennaJ. BangsboJM Renaud2008Muscle K+, Na+ and Cl- disturbances and Na+-K+ pump inactivation: implication for fatigue.J Appl Physiol10428895
- 16. Mohr M, Krustrup P, Bangsbo J (2005) Fatigue in soccer: A brief review. J Sport Sci 23: 593–599.M. MohrP. KrustrupJ. Bangsbo2005Fatigue in soccer: A brief review.J Sport Sci23593599
- 17. Bangsbo J, Iaia FM, Krustrup P (2008) The Yo-Yo Intermittent Recovery Test: A useful tool for evaluation of physical performance in intermittent sports. Sports Med 38: 37–51.J. BangsboFM IaiaP. Krustrup2008The Yo-Yo Intermittent Recovery Test: A useful tool for evaluation of physical performance in intermittent sports.Sports Med383751
- 18. DiSalvo V, Gregson W, Atkinson G, Tordoff P, Drust B (2009) Analysis of high intensity activity in Premier League soccer. Int J Sports Med 30: 205–212.V. DiSalvoW. GregsonG. AtkinsonP. TordoffB. Drust2009Analysis of high intensity activity in Premier League soccer.Int J Sports Med30205212
- 19. Randers MB, Mujika I, Hewitt A, Santisteban J, Bischoff R, et al. (2010) Application of four different football match analysis systems: a comparative study. J Sports Sci 28: 171–82.MB RandersI. MujikaA. HewittJ. SantistebanR. Bischoff2010Application of four different football match analysis systems: a comparative study.J Sports Sci2817182
- 20. Rampinini E, Impellizeri FM, Castagna C, Coutts AJ, Wisloff U (2009) Technical performance during soccer matches of the Italian Seria A: effect of fatigue and competitive level. J Sci Med Sport 12: 227–33.E. RampininiFM ImpellizeriC. CastagnaAJ CouttsU. Wisloff2009Technical performance during soccer matches of the Italian Seria A: effect of fatigue and competitive level.J Sci Med Sport1222733
- 21. McKenzie JE, Osgood DW (2004) Validation of a new telemetric core temperature monitor. J Thermal Biol 29: 605–611.JE McKenzieDW Osgood2004Validation of a new telemetric core temperature monitor.J Thermal Biol29605611
- 22. Edwards B, Waterhouse J, Reilly T, Atkinson G (2002) A comparison of the suitabilities of rectal, gut, and insulated axilla temperatures for measurement of the circadian rhythm of core temperature in field studies. Chronobiol Int 19: 579–597.B. EdwardsJ. WaterhouseT. ReillyG. Atkinson2002A comparison of the suitabilities of rectal, gut, and insulated axilla temperatures for measurement of the circadian rhythm of core temperature in field studies.Chronobiol Int19579597
- 23. Andersson H, Ekblom B, Krustrup P (2008) Elite football on artificial turf versus natural grass: movement pattern, technical standard and player impressions. J Sport Sci 26: 113–122.H. AnderssonB. EkblomP. Krustrup2008Elite football on artificial turf versus natural grass: movement pattern, technical standard and player impressions.J Sport Sci26113122
- 24. Rampinini E, Coutts AJ, Castagna C, Impellizeri FM, Sassi A (2007) Variation in top level soccer match performance. Int J Sports Med 28: 1018–24.E. RampininiAJ CouttsC. CastagnaFM ImpellizeriA. Sassi2007Variation in top level soccer match performance.Int J Sports Med28101824
- 25. Morante SM, Brotherhood JR (2008) Autonomic and behavioral thermoregulation in tennis. Br J Sports Med 42: 679–685.SM MoranteJR Brotherhood2008Autonomic and behavioral thermoregulation in tennis.Br J Sports Med42679685
- 26. Tucker R, Rauch L, Harley YX, Noakes TD (2004) Impaired exercise performance in the heat is associated with an anticipatory reduction in skeletal muscle recruitment. Pflugers Arch 448: 422–30.R. TuckerL. RauchYX HarleyTD Noakes2004Impaired exercise performance in the heat is associated with an anticipatory reduction in skeletal muscle recruitment.Pflugers Arch44842230
- 27. Dellal A, Hill-Haas S, Lago-Penas C, Chamari K (2011) Small-sided games in soccer: Amateur vs. professional players’ physiological responses, physical and technical activities. J Strength Cond Res 25: 2371–81.A. DellalS. Hill-HaasC. Lago-PenasK. Chamari2011Small-sided games in soccer: Amateur vs. professional players’ physiological responses, physical and technical activities.J Strength Cond Res25237181
- 28. Owen AL, Wong Del P, McKenna M, Dellal A (2011) Heart rate responses and technical comparison between small- vs. large-sided games in elite professional soccer. J Strength Cond Res 25: 2104–10.AL OwenP. Wong DelM. McKennaA. Dellal2011Heart rate responses and technical comparison between small- vs. large-sided games in elite professional soccer.J Strength Cond Res25210410
- 29. Gaoua N, Racinais S, Grantham J, El Massioui F (2011) Alterations in cognitive performance during passive hyperthermia are task dependent. Int J Hyperthermia 27: 1–9.N. GaouaS. RacinaisJ. GranthamF. El Massioui2011Alterations in cognitive performance during passive hyperthermia are task dependent.Int J Hyperthermia2719
- 30. Bendiksen M, Bischoff R, Randers MB, Mohr M, Rollo I, et al. (2012) The Copenhagen Soccer Test: Physiological response and fatigue development. Med Sci Sports Exerc 2012 Feb 10 [epub ahead of print]. M. BendiksenR. BischoffMB RandersM. MohrI. Rollo2012The Copenhagen Soccer Test: Physiological response and fatigue development.Med Sci Sports Exerc 2012 Feb 10 [epub ahead of print]
- 31. Krustrup P, Mohr M, Steensberg A, Bencke J, Kjaer M, et al. (2006) Muscle and blood metabolites during a soccer game: implications for sprint performance. Med Sci Sports Exerc 38: 1165–1174.P. KrustrupM. MohrA. SteensbergJ. BenckeM. Kjaer2006Muscle and blood metabolites during a soccer game: implications for sprint performance.Med Sci Sports Exerc3811651174
- 32. Nielsen B, Nybo L (2001) Cerebral changes during exercise in the heat. Sports Med 33: 1–11.B. NielsenL. Nybo2001Cerebral changes during exercise in the heat.Sports Med33111
- 33. Racinais S, Blonc S, Jonville S, Hue O (2005) Time-of-day influences the environmental effects on muscle force and contractility. Med Sci Sports Exerc 37: 256–261.S. RacinaisS. BloncS. JonvilleO. Hue2005Time-of-day influences the environmental effects on muscle force and contractility.Med Sci Sports Exerc37256261
- 34. Racinais S, Hue O, Blonc S (2004) Time-of-day effects on anaerobic muscular power in a moderately warm environment. Chronobiol Int 21: 485–495.S. RacinaisO. HueS. Blonc2004Time-of-day effects on anaerobic muscular power in a moderately warm environment.Chronobiol Int21485495
- 35. Racinais S (2010) Time-of-day effect when exercising in hot environments. Scand J Med Sci Sports 20: 80–89.S. Racinais2010Time-of-day effect when exercising in hot environments.Scand J Med Sci Sports208089
- 36. Judelson DA, Maresh CM, Anderson JM, Armstrong LE, Casa DJ, et al. (2007) Hydration and muscular performance: does fluid balance affect strength, power and high-intensity endurance? Sports Med 37: 907–921.DA JudelsonCM MareshJM AndersonLE ArmstrongDJ Casa2007Hydration and muscular performance: does fluid balance affect strength, power and high-intensity endurance?Sports Med37907921
- 37. Goulet ED (2011) Effect of exercise-induced dehydration on time-trial performance: a meta-analysis. Br J Sports Med 45: 1149–1156.ED Goulet2011Effect of exercise-induced dehydration on time-trial performance: a meta-analysis.Br J Sports Med4511491156
- 38. Mohr M, Krustrup P (2012) Heat stress impairs repeated jump ability after competitive elite soccer games. J Strength Cond Res [in press]. M. MohrP. Krustrup2012Heat stress impairs repeated jump ability after competitive elite soccer games.J Strength Cond Res [in press]