Thermoregulatory responses to sports and exercise have been widely investigated. However, investigation of the response characteristics has been conducted mainly for continuous sports such as running and cycling.1^–3 Several studies have examined thermoregulatory responses during tennis;4^–14 however, there were some limitations. The studies of Dawson et al6 and Elliott et al7 were the only ones that measured the important thermal strains of core body temperature, skin temperature, sweat rate and heart rate. Neither of these studies, however, continuously measured rectal and skin temperatures or heart rate throughout the match, instead making observations during each change of ends, approximately 2 minutes apart. This method prevents detailed plotting of these variables over the course of the match and thus reduces precision in the subsequent description of the response characteristics. Furthermore, observations were made on four different occasions in matches that lasted 1 hour, thus reflecting only four different thermal environments. Therminarias et al13 14 also measured rectal temperature, sweat loss and heart rate during competitive singles tennis. This study improved on the measurement of heart rate during tennis by using continuous logging every 15 seconds during the matches. However, rectal temperature was measured only at the beginning and end of matches, thereby failing to describe the characteristics of this response, particularly if and when a steady state is achieved. All other published studies investigating physiological responses during tennis measured only heart rate and/or oxygen consumption, which do not provide a complete description of thermoregulatory responses and their characteristics during tennis.

We undertook a more comprehensive assessment of thermoregulation in tennis in our previous study15 by making a total of 86 observations throughout each of the seasons (air temperature ranged from 14.5 to 38.4°C). The physiological observations included all of the major thermal strains: core body (rectal) temperature, skin temperature, sweat rate and heart rate. Furthermore, the study benefited from the body temperature being logged every minute and heart rate recorded every 15 seconds throughout the match. This enables the subsequent investigation of the response characteristics for each of these variables.

As no known study has examined the characteristics of thermoregulatory responses during tennis, little is known about how the body responds to environmental and exercise stresses in tennis. The intermittent nature of competitive tennis would be expected to produce different responses to previously published studies for continuous, steady-state exercise. Information on tennis is important because the thermal environment, and consequently thermal transfers, are more complex in tennis compared with laboratory studies, as tennis is generally played outdoors.

This study aimed to expand on the current knowledge base of thermoregulation during exercise and physiological responses during tennis. Case studies of body core (rectal) temperature, skin temperature and heart rate were compared during competitive tennis with continuous, steady-state running in the laboratory.

METHODS

Each participant gave their written informed consent and the project was approved by the University of Sydney Human Research Ethics Committee.

Participants

Tennis

Examples of thermoregulatory responses were selected from a subset of six different participants from our previous study. The individual characteristics for these six players were not significantly different from the values for the group of 25 subjects reported in our previous study.15 The mean (SD) age was 23.9 (3.5) years, height 181 (13) cm, body mass 76.8 (8.5) kg, maximum aerobic power (VO_2max) 56.7 (5.4) mL/kg/min, maximum heart rate (HR_max) 196 (9) beats/min, sum of nine skinfolds 89.7 (29.8) mm, and predicted body fat 11.8 (6.0)%.

Running

Examples of thermoregulatory responses were for one runner selected from a group of eight participants: age 18 years; height 183 cm; body mass 68.2 kg; VO_2max 67.0 mL/kg/min; HR_max 191 beats/min; sum 9 skinfolds 58.5 mm; predicted body fat 6.5%.

Procedures

Observations for tennis were made during “best of three”, tie-break set, singles matches played on a hard-court surface between subjects of similar playing standard. All matches adhered to the rules set by the International Tennis Federation.16 Three new tennis balls were used for each match, with players retrieving balls between points. Conventional tennis attire was worn for all matches.

Observations during running were made in a climate chamber on a motorised treadmill. A 5-minute warm-up at 5 km/h was followed by the 60-minute continuous running trial at 12 km/h. The treadmill grade remained at 0% throughout the entire trial.

Thermoregulatory responses

Subjects inserted a YSI rectal thermistor (Measurement Specialities Inc., Hampton, Virginia, USA) 10 cm beyond the anal sphincter. Four YSI thermistors were attached to the skin using Opsite adhesive tape at the chest, arm, thigh and leg17 for calculation of weighted mean skin temperature. In the tennis trials, the five thermistors were attached to a 180 g multi-channel data logger worn on a belt around the waist for the recording of core and skin temperatures every minute throughout the match. During the running trial, rectal and skin temperatures were logged by a computer every 10 seconds. A heart rate monitor (Polar S610i; Polar, Vantaa, Finland) was worn by each subject, which recorded heart rate every 15 seconds during exercise. Subjects consumed fluids during the change of ends on an ad libitium basis throughout the trials. Measurement of subjects’ maximum oxygen consumption in the laboratory and rectal temperature during tennis matches enables the prediction of on-court exercise intensity (%VO_2max) using the regression equation presented by Davies et al.18

Environmental observations

Dry bulb (air) temperature was measured on court using a whirling psychrometer shielded from direct sunlight every 20 minutes throughout the match. A 15-cm blackened copper bulb was used to measure globe temperature. This was placed on the court 1 hour before the match to be heated by the sun in order to provide an accurate observation at the start of the match. A short arm anemometer set 1.5 m from the ground was used to measure wind speed. A customised Davis Perception II weather station and WeatherLink software (both Davis Instruments Corp., Hayward, California, USA) were used to record globe temperature (solar radiation) and wind speed during each minute of play. The conditions of the climate chamber for the running trial were controlled at an air temperature of 20°C, relative humidity of 50% and an air movement of 1.9 m/s.

The examples of thermoregulatory responses (rectal temperature, skin temperature and heart rate) presented were derived from our previous study,15 and are reflective of the entire group of 25 subjects and 86 observations.

Selection of example data

Data were selected to show the characteristics of thermoregulatory responses in tennis. The three alternative actions for core body temperature response during exercise were selected from the dataset (fig 1A). The association between skin temperature and environmental temperature was explored by contrasting skin temperatures observed in cool and hot air temperatures (fig 2A). The similarity in heart-rate responses between the two players involved in a match is shown in fig 3A and is characteristic of the total of 86 observations.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 (A) Examples of body core (rectal) temperature. Rectal temperature measured at an air temperature of (a) 25.8°C, (b) 26.3°C and (c) 30.3°C. (B) Examples of body core (rectal) temperature during tennis and continuous steady-state running on a treadmill in a laboratory; in both the tennis and running assessments, the subjects paused exercising at 30 minutes to be weighed on nearby body mass scales. (C) Rectal temperature measured on one subject over a range of environmental air temperatures. Ta, air temperature.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2 (A) Examples of weighted mean skin temperature. Skin temperature measured at an air temperature of (a) 38.4°C and (b) 20.0°C. (B) Examples of skin temperature during tennis and continuous steady-state running on a treadmill in a laboratory. (C) Skin temperature measured for one subject over a range of environmental air temperatures. Ta, air temperature.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3 (A) Examples of heart-rate responses for opponents during a tennis match. Player 1: mean (range) 156.8 (85 to 187) beats/min; player 2: mean (range) 145.3 (82 to 191) beats/min.(B) Examples of heart rate during tennis and continuous steady-state running on a treadmill in the laboratory. (C) Skin temperature measured on one subject over a range of environmental air temperatures, during four different tennis matches.

Data were selected to compare thermoregulatory responses during tennis and running (figs 1B, 2B, 3B). The tennis example used to compare rectal temperature response characteristics between tennis and running was selected for its representation of the 86 observed rectal temperatures (no significant difference between the example and the total dataset: 38.31°C versus 38.40°C, respectively). Furthermore, there was no significant difference between the average of the rectal temperature observed from the entire match and the average rectal temperature from the running example (38.04°C for tennis versus 37.95°C for running). The difference in skin temperature characteristics between exercise modes was explored by selecting a skin temperature example from tennis that was measured at the same air temperature as in the laboratory for running (20.0°C air temperature for both tennis and running). The same rationale was used for the selection of a tennis example to compare heart-rate responses between the exercise modes. The average heart rate measured during running was 116.4 beats/min, therefore a heart-rate response during tennis that was not significantly different to this was selected (heart rate during tennis of 115.9 beats/min).

Differences in thermoregulatory responses for a single player during five different matches (and thermal stresses) were explored (figs 1C, 2C, 3C). These responses were characteristic of the group of 25 different players and 86 different sets of data.

Statistical analysis

Statistical analyses were preformed using SPSS V.15.0 (SPSS, Chicago, Illinois, USA). Data are expressed as mean (SD). Significance was set at p<0.05.

Results

Body core (rectal) temperature

The three contrasting responses of rectal temperature during competitive singles tennis (fig 1A), responses of rectal temperature during tennis and continuous, steady-state running in the laboratory (fig 1B) and rectal temperature observations made on five different occasions for a single subject (fig 1C) are shown. Based on the subject’s measured core temperature and maximum oxygen consumption, the exercise intensity during the tennis example was 57.3% of VO_2max. The exercise intensity for the running example was 53.8% of VO_2max, which was based on measured maximum oxygen consumption for the individual subject.

Skin temperature

Two contrasting examples of skin temperature during tennis (fig 2A) and traces of skin temperature for tennis and running at an air temperature of 20°C (fig 2B) are shown. The sharp increase for tennis and decrease for running at 30 minutes indicates the point at which subjects paused from exercise to be weighed on nearby body-mass scales. This involved the tennis player moving from the outdoor court into the warmer laboratory, and the runner moving from the climate chamber into the cool laboratory. The skin temperature responses for one subject on five different occasions are shown (fig 2C), and support the data shown in fig 2A, showing the relationship between skin temperature and air temperature.

Heart rate

The wide variation in heart rate during competitive tennis is highlighted (fig 3A) The two examples are of opponents during a tennis match, which illustrates the similarity between work and rest intervals (mean (range) HR for player 1 was 56.8 (85 to 187) beats/min; and for player 2 was 145.3 (82 to 191) beats/min). Figure 3 shows the response of heart rate for one subject over four different tennis matches. The contrasting response of heart rate during tennis and running is shown in fig 3B. Despite both examples producing a similar average heart rate (115.9 (19.1) beats/min for tennis versus 116.4 (8.0) beats/min for running), the range and maximum of these responses was very different (72 to 179 and 65 to 124 beats/min, respectively). Figure 3C illustrates the response of heart rate for one subject over four different tennis matches.

Discussion

Body core (rectal) temperature

There is a range of environmental conditions in which body core temperature is maintained at safe levels, independent of the environment.19 This is called the prescriptive zone, and varies depending on metabolic rate. The upper limit of the prescriptive zone decreases with increasing metabolic rate—that is, the higher the exercise intensity, the cooler and less humid the conditions required to allow control of body temperature to be achieved.19 In conditions that exceed the upper limit of the prescriptive zone, body core temperature rises progressively during exercise.19 The lack of association between rectal temperature and air temperature is evident in fig 1A, which shows three different rectal temperature responses during tennis over relatively similar environmental air temperatures. As expected, rectal temperature was maintained independently of environmental air temperature, which was 25.8°C, 26.3°C and 30.3°C for examples a, b and c, respectively. Each of the examples in this figure are reflective of differences in the intensity of play (metabolic rate). The first example (a) shows a continual rise in rectal temperature throughout the match, achieving thermal equilibrium (indicated by the plateau in rectal temperature) after around 80 minutes of play. This must be the result of either a continual increase in the intensity throughout the match or an increasing heat load being imposed by the environment. Further analysis using bivariate regression found that heart rate increased with match time (p<0.0001), indicating that there was a progressive increase in the intensity of play throughout the match. This may be attributed to improvement in play and/or increased motivation to win during the final stages of the match. Regression analysis also discovered a continual increase in mean radiant temperature (including air temperature, globe temperature (solar radiation) and wind speed) over the course of the match. Therefore, the difficulty in the achievement of thermal balance in this first example is due to the combination of increasing metabolic heat production and convective and radiant heat gain from the environment. The second example (b) illustrates a more typical rectal temperature response during exercise, in which thermal equilibrium is achieved after an initial increase in rectal temperature. The steep rise in rectal temperature during the first 40 minutes of play reflects the onset of exercise and increase in energy expenditure. During the initial 40 minutes, the rate of metabolic heat production and heat gain from the environment exceeds the rate of heat loss via convection and evaporation of sweat. This imbalance causes heat to be stored within the body, illustrated by a progressive rise in rectal temperature. After approximately 30 minutes, when a steady sweating rate has been established, evaporative heat loss is able to effectively dissipate the heat produced and gained, thus thermoregulation is successful and rectal temperature plateaus. The third example (c) illustrates the rectal temperature response in a tennis match in which the intensity of play is little more than resting levels. Consequently, the excess heat can be dissipated effectively from the onset of play, as seen by no or minimal initial rise in rectal temperature.

The average response of rectal temperature throughout a tennis match is different from a typical response during continuous, steady-state running. Figure 1B illustrates the longer time taken to achieve thermal equilibrium during tennis (approximately 40 minutes) compared with running (around 10 minutes). This may reflect a continual increase in exercise intensity during the tennis match as seen in fig 1A (example a), which is impossible during steady-state running. Alternatively, the heat load imposed by the environment may alter the response during tennis, as solar radiation and varied wind speed are additional factors not present in the laboratory. Achieving thermal equilibrium during exercise or sport is essential to prevent body core temperature reaching a critical level of approximately 42°C, which poses a serious risk to an individual’s health and safety.20

Skin temperature

Skin temperature, unlike body core temperature, is related to environmental air temperature.21 22 Figure 2A illustrates this by comparing the skin temperature illustrated in the second example (b; 26.93°C) obtained on a cool, winter day (20.0°C) with the skin temperature of the first (a; 35.86°C), measured on a warm, summer day during which air temperature was 38.4°C. Generally, as in example b, skin temperature is greater than air temperature. This causes heat to be transferred via convection down a thermal gradient from the skin to the surrounding environment. Therefore, the player in example b is losing metabolic heat produced during exercise via convection, which assists in the maintenance of body core temperature. In contrast, if air temperature is greater than skin temperature, as in example a, the convective heat exchange is reversed, with heat being gained from the environment. This is a concern as it adds to the metabolic heat load imposed by exercise, meaning the evaporation of sweat must occur at a greater rate in order to balance the rate of heat production and gain.

Both examples of skin temperature in fig 2B were measured at an air temperature of 20.0°C. As skin temperature and air temperature are related21 22 and both examples were measured at 20.0°C, similar skin temperatures would have been expected. However, this figure illustrates the different results between the two modes of exercise, with mean (SD) skin temperature for tennis being 26.88 (1.36)°C compared with 28.91 (1.56)°C for running (p<0.0001). A major factor contributing to this variation is the difference in wind speed between the tennis and running conditions (tennis 2.6 m/s; running 1.9 m/s). Air movement is an integral component of convective and evaporative heat exchange, with both being increased with higher wind speeds.22 Therefore, the rate of convective and evaporative heat loss was greater and the wind chill factor higher during tennis, due to the faster wind speed. Combined, this explains the lower skin temperature observed during tennis compared with running at the same air temperature. In contrast, the opposite would occur if environmental temperature was greater than skin temperature, with the rate of heat gain via convection being increased when wind speed is higher. A tennis player is likely to experience thermal discomfort if skin temperature is too low or too high, as skin temperature has been linked with an individual’s rating of thermal comfort.21^–27 Thermal comfort is an important consideration for tennis players as it is essential for enjoyment of the sport.

Heart rate

The exercise intensity of singles tennis is the same for both players regardless of their fitness level. This is shown by fig 3A, in which the mean, range and characteristics of the heart-rate response over time were similar for each opponent. This suggests that it would be advantageous for players to improve their fitness in order for them to work at a lower relative percentage of the workload.

The intermittent nature of tennis produces a different heart-rate response from that of continuous, steady-state exercise. This is evident in the comparison of tennis and running heart rates over time (fig 3B). The average heart rate for both examples was similar (115.9 beats/min for tennis and 116.4 beats/min for running). The trace for running showed little variation throughout the exercise compared with the large and rapid fluctuations between points and rest periods in tennis. Given the wide variation in heart rate during tennis, the overall average does not completely describe the intensity of play. Instead, the maximum heart rate may also be considered when examining the physiological strain imposed by tennis. This reveals much greater variation between the two exercise modes, with heart rate in the tennis example peaking at 179 beats/min compared with 124 beats/min for the running example. Thus, tennis players must be physically able to work at intensities exceeding that of continuous, steady-state exercise, and should plan training accordingly.

Conclusion

This study preovides a description of thermoregulatory responses measured during competitive singles tennis. Comparison with thermoregulatory responses measured during continuous, steady-state running reveal differences in the response of rectal temperature, skin temperature and heart rate during tennis.