Article Text

Download PDFPDF

Impact of different climatic conditions on peak core temperature of elite athletes during exercise in the heat: a Thermo Tokyo simulation study
  1. Lennart P J Teunissen1,
  2. Kaspar M B Jansen1,
  3. Emiel Janssen1,
  4. Boris R M Kingma2,
  5. Johannus Q de Korte3,
  6. Thijs M H Eijsvogels3
  1. 1Department of Design Engineering, Delft University of Technology, Delft, The Netherlands
  2. 2Department of Training & Performance Innovations, TNO, Unit Defence, Safety & Security, Soesterberg, The Netherlands
  3. 3Radboud Institute for Health Sciences, Department of Physiology, Radboud University Medical Center, Nijmegen, The Netherlands
  1. Correspondence to Dr Thijs M H Eijsvogels; thijs.eijsvogels{at}


Objectives To evaluate how separate and combined climatic parameters affect peak core temperature during exercise in the heat using computer simulations fed with individual data.

Methods The impact of eight environmental conditions on rectal temperature (Tre) was determined for exercise under heat stress using the Fiala-thermal-Physiology-and-Comfort simulation model. Variations in ambient temperature (Ta±6°C), relative humidity (RH±15%) and solar radiation (SR+921 W/m2) were assessed in isolation and combination (worst-case/best-case scenarios) and compared with baseline (Ta32°C, RH 75%, SR 0 W/m2). The simulation model was fed with personal, anthropometric and individual exercise characteristics.

Results 54 athletes exercised for 46±10 min at baseline conditions and achieved a peak core temperature of 38.9±0.5°C. Simulations at a higher Ta (38°C) and SR (921 W/m2) resulted in a higher peak Tre compared with baseline (+0.6±0.3°C and +0.5±0.2°C, respectively), whereas a higher RH (90%) hardly affected peak Tre (+0.1±0.1°C). A lower Ta (26°C) and RH (60%) reduced peak Tre by −0.4±0.2°C and a minor −0.1±0.1°C, respectively. The worst-case simulation yielded a 1.5±0.4°C higher Tre than baseline and 2.0±0.7°C higher than the best-case condition.

Conclusion Combined unfavourable climatic conditions produce a greater increase in peak core temperature than the sum of its parts in elite athletes exercising in the heat.

  • Heat stress
  • Thermoregulation
  • Athlete
  • Olympics

This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See:

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Key messages

What is already known

  • The magnitude of exercise-induced increases in core temperature and subsequent performance loss or health risks depends on the severity of climatic conditions.

  • Elite athletes are recommended to participate in heat preparedness programs before competition in hot and humid environments, such as the Tokyo 2020 Olympic Games. Still, actual climatic conditions may vary from the simulated conditions.

What are the new findings

  • Realistic variations in air temperature, relative humidity (RH) and solar radiation (SR) have distinct effects on the estimated peak core temperature of elite athletes exercising in the heat.

  • The worst-case scenario simulation (Ta 37.5°C: RH: 79%, SR: 921 W/m2) predicted a peak Tre of 40.3±1.0°C after 46±10 min of high-intensity exercise, with 56% of the athletes demonstrating a Tre >40°C and 17% a Tre >41°C.

  • The combination of unfavourable climatic conditions led to a synergistic response with a greater effect on Tre than the sum of its parts, highlighting the added value of computer modelling beyond physiological testing in an environmental chamber.


The Tokyo Olympic and Paralympic Games were organised under challenging climatic conditions. Analyses of historical data revealed that an ambient temperature (Ta) of 31.3±3.1°C and relative humidity (RH) of 59%±10% were expected around noon,1 whereas solar radiation (SR) could exceed 900 W/m2. These environmental conditions accelerate exercise-induced increases in core temperature (Tc),2 which is known to negatively impact exercise performance3–7 and increase the risk for the development of heat-related illnesses.8 Therefore, preparatory testing for Tokyo 2020 was recommended by experts,9 and many countries implemented heat preparation programmes for their Olympic teams.10–12

The Dutch Olympic Team launched the Thermo Tokyo program, in which individual heat tolerance was determined during an exercise test in simulated Tokyo conditions.13 Although such tests provide useful insight into the responses to a hot humid climate in general, climatic conditions may vary from day to day, and changes in Ta, RH and SR are known to impact human thermoregulation.14 As it is not feasible to subject elite athletes to exercise testing across a large variety of climatic conditions, we aimed to evaluate the impact of various Tokyo summer climatic conditions on Tc responses using computer simulations of the individual exercise tests, fed with individual athlete data.


Study cohort and design

This is a substudy of the Thermo Tokyo project of which the rationale, design and measurements have been described in detail previously.13 In short, Dutch outdoor athletes (≥16 years) were recruited via TeamNL infrastructures. All athletes were active at the elite level as they competed at international tournaments and top-level competitions.15 Detailed anthropometric measurements were conducted, and all athletes completed an incremental cycling test until volitional exhaustion in simulated Tokyo conditions (Ta 32°C, RH 75%). Data from 54 participants (14 endurance-trained, 8 mixed-trained, 12 power-trained and 20 skill-trained athletes) were used to feed the thermophysiological simulation model in order to assess the impact of distinct climatic conditions on peak Tc.

Data collection

Anthropometric characteristics were derived from a DXA scan (Discovery-A, Hologic, Bedford, USA) and a full-body 3D scan (Artec 3D, Luxembourg). Height, weight, fat percentage and characteristics of 24 body segments were derived from these measurements. Gastrointestinal temperature (Tgi) and power output (W) were continuously measured during the exercise test using the myTemp system16 and a bicycle ergometer (Lode B.V., Groningen, Netherlands) or Tacx Neo Smart T2800 (Tacx B.V., Wassenaar, Netherlands), respectively. Finally, the time to exhaustion of the incremental exercise test was established.

Computer simulation

The validated Fiala-thermal-Physiology-and-Comfort (FPC) simulation model (V.5.4)17 18 was used to model the impact of eight different environmental conditions (table 1) on rectal temperature (Tre). Individual characteristics (sex, age, VO2max, acclimatisation status), anthropometrics and exercise intensities during the test (metabolic equivalent of task score (METs), calculated from power output)19 served as input for the model. First, the baseline condition was modelled to reproduce the physiological responses collected during the exercise test in the climatic chamber. Then, simulations with adjustments in Ta (±6°C), RH (±15%) and SR (+921 W/m2) were conducted. Finally, a best-case and worst-case scenario simulation was conducted based on the 95% CI boundaries for the expected Ta, RH and SR at noon during the Tokyo Olympics.1 Wind speed was kept constant at exercise test conditions in the climatic chamber (standard airflow of 0.2 m/s; convective cooling by pedalling movements was accounted for by an adjusted airflow of 1.0 m/s at the legs). An average gross efficiency factor of 0.2 was used. Standardised clothing settings were selected for shirt and shorts. Peak Tre was used as the primary outcome. The reduction in the time to reach the peak Tre attained in the baseline simulation was included as a secondary outcome parameter.

Table 1

Summary of the eight distinct environmental conditions that were entered into the Fiala-thermal-Physiology-and-Comfort simulation model


Fifty-four elite athletes (53.7% female) participated in this Thermo Tokyo substudy, and all data were collected successfully. Characteristics of the study cohort are summarised in table 2.

Table 2

Characteristics of the complete cohort as well as sex-specific groups

Model performance

A peak Tgi of 38.9±0.5°C was measured at the end of the exercise test in the environmental chamber. Computer simulations using the FPC model yielded an estimated peak Tre of 38.8±0.6°C at baseline conditions (Ta 32°C, RH 75%, no SR). Accordingly, the difference between the measured versus simulated peak core temperatures was −0.1±0.5°C.

Impact of climatic conditions

A typical example of exercise-induced changes in Tre of an individual athlete for each of the eight climatic conditions in the FPC simulations is shown in figure 1. On a group level, a higher Tre was predicted following exercise in the condition with a 6°C higher Ta (+0.6±0.3°C) and the condition with maximal SR (+0.5±0.2°C) compared with the baseline condition. Peak Tre of the baseline simulation was attained 9±7 and 9±8 min earlier in these respective conditions. A 15% increase or decrease in RH had little effect on peak Tre (+0.1±0.1°C and −0.1±0.1°C, respectively), and the time to reach baseline peak Tre was only 2±9 min shorter in the +15% condition. A 6°C lower Ta resulted in a lower peak Tre than baseline (−0.4±0.2°C). Peak Tre was 1.5±0.4°C higher in the worst-case scenario compared with the baseline condition, with an absolute predicted value of 40.3±1.0°C. As many as 56% of the athletes had a peak Tre>40°C, and 17% had a peak Tre>41°C. Furthermore, peak Tre was attained 19±6 min (41%±13%) earlier during exercise in worst-case conditions. The best-case simulation yielded a 0.5±0.3°C lower Tre than baseline and 2.0±0.7°C lower Tre than the worst-case condition. An overview of all FPC model predictions is summarised in figure 2.

Figure 1

Typical outcome of the computer simulation of a single participant, with exercise-induced increases in rectal temperature for each of the eight simulated climatic conditions. Ta, ambient temperature; RH, relative humidity; SR, solar radiation.

Figure 2

Outcomes of the computer simulation of eight distinct climatic conditions on peak rectal temperature (Tre) of elite athletes exercising in the heat. An isolated increase in air temperature and solar radiation led to a higher peak Tre, but the impact of the worst-case scenario on peak Tre was greater than the sum of its parts. FPC, Fiala-thermal-Physiology-and-Comfort.


This simulation study showed that small to moderate variations in climatic conditions could considerably impact the peak Tre of elite athletes exercising in the heat. A 6°C increase in Ta or the addition of high-intensity SR contributed to an accelerated exercise-induced Tre. In contrast, the impact of variations in RH on peak Tre was limited. Beyond the effect of isolated climatic conditions, the worst-case model highlighted that a combination of heat stressors leads to more significant heat stress, evidenced by predicted peak Tre values >40°C in 56% of the elite athletes. Therefore, the outcomes of this study provide important insight into the thermoregulatory responses of elite athletes during exercise under various challenging climatological conditions.

FPC model

We used the FPC model to simulate the impact of different climatic conditions on Tre. After feeding the model with physiological and anthropometric data for every individual athlete that participated in the present study,7 13 the average difference between the measured and predicted Tre for baseline climatic conditions was acceptably small (−0.1±0.5°C). Hence, we used these FPC settings to simulate the remaining conditions.

Climatic conditions

We found a substantial variability of peak Tre across the simulated climatic conditions. The greatest increase in peak Tre was observed at a 6°C increase in Ta. This finding aligns with a recent study in which Ta was the most important climatic parameter to affect peak performance during endurance running events.20 We also found that the impact of Ta on predicted Tre was not linear, as a 6°C decrease in Ta had less effect on peak Tre compared with a 6°C increase in Ta (−0.4±0.2 vs +0.6±0.3). These observations stress the added value of computer simulations beyond exercise tests with physiological measurements, as proper predictions can optimise heat mitigation strategies across multiple scenarios of climatic conditions. Maximal SR appeared to have a large effect on peak Tre. This is in line with previous research, indicating a significant increase in the body’s heat gain with increasing radiation intensity.21

In contrast to our hypothesis and existing literature,5 we found little impact of variations in RH on peak Tre. This may be due to the small magnitude of RH variations (±15%) and/or the low wind speed, limiting the possibility to evaporate sweat. Indeed, variations in RH between 52% and 71% have previously been shown not to result in significant changes in Tre during a 1-hour steady state run with little air flow.22 In addition, a change in skin temperature may potentially affect the impact of RH variations on heat loss. For example, increases in skin temperature at +15% RH or decreases in skin temperature with −15% RH may attenuate the heat loss effect due to the altered water vapour pressure difference between skin and environment. This could have contributed to the low impact of RH level in the current study but needs to be confirmed in future studies as skin temperature was unavailable in our simulations.

Practical application

We showed that computer modelling of exercise-induced increases in Tre provides additional insights beyond physiological testing in an environmental chamber. Hence, this approach may be adopted in preparation for future World Championships, Olympic Games and other exercise events in environmentally stressful conditions. Computer simulations provide athletes and coaches unique insight into changes in individual thermophysiological and performance indicators across different climatic scenarios without the burden of extensive exercise testing. Hence, this information can be used to select the most appropriate precooling/per-cooling strategies,23 hydration plan24 and race strategy.

Strengths and limitations

Strengths of the study include the large sample size (n=54) and a large number of simulations (8 distinct conditions) to assess the impact of variations in the Tokyo climate on the predicted Tre of Olympic athletes. This study, however, is also subject to limitations. First, we assumed a similar exertional heat production across climatic conditions, whereas athletes will generally adjust their pacing in an anticipatory way dependent on conditions,25 potentially affecting their physiological responses. Further, the wind speed was kept stable in this study, whereas it would have been interesting to explore the thermal impact of sport-specific wind speeds. Nevertheless, our results highlight the impact of different realistic climatic scenarios on the thermal strain, and the consequences if preparatory measures and anticipatory signals are not taken seriously.


Isolated climatic conditions, such as a higher Ta and high SR, significantly increase peak Tre in elite athletes exercising in the heat. Moreover, the combination of unfavourable climatic conditions in a worst-case scenario led to a synergistic response with a greater effect on Tre than the sum of its parts.

Ethics statements

Patient consent for publication

Ethics approval

This study was approved by CMO region Arnhem-Nijmegen (#2018-4640). Participants gave informed consent to participate in the study before taking part.


The authors would like to recognize the help of Bert Naagen, Toon Huysmans, Federico Altobelli, Shiannah Danen, Laura de Kubber and Jan-Willem van Dijk with data collection for the present study.



  • Contributors LT, KJ and TE designed the study. LT, BK, JdK and TE contributed to the data collection and analyses. EJ, LT and TE created the figures. LT and TE drafted the manuscript, while KJ, EJ, BK and JdK critically revised the manuscript. All authors gave their final approval of the version published.

  • Funding This project was funded by ZonMW (#546001003).

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.