Predictive validity of a computer model of body temperature during exercise

Area of Science:

• Physiology
• Environmental Health
• Computational Biology

Background:

• Accurate prediction of human thermoregulation is crucial for understanding physiological responses to environmental conditions.
• Computer models offer a potential tool for simulating and predicting thermal strain in various scenarios.

Purpose of the Study:

• To assess the predictive validity of a computer model simulating human temperature regulation.
• To compare model predictions with experimental data across a range of exercise intensities and thermal environments.

Main Methods:

• Three male subjects performed exercise on a cycle ergometer at varying work rates (Basal Metabolic Rate to 250 W).
• Experiments were conducted in a controlled environmental facility with effective temperatures ranging from 13°C to 29°C.
• Core (rectal, tympanic) and skin temperatures were monitored, and compared with computer model simulations.

Main Results:

• A strong negative correlation (r = -0.87) was observed between effective temperature and the mean absolute difference in mean body temperature between experimental data and model simulations.
• The computer model demonstrated increased predictive validity under conditions of higher heat stress (Effective Temperature > 25°C), with a mean absolute difference < 0.3°C.
• Predictive accuracy decreased in colder environments (Effective Temperature < 16°C), showing a mean absolute difference > 0.8°C.

Conclusions:

• The computer model for human temperature regulation shows promising predictive validity, particularly under heat stress.
• Model performance is influenced by environmental temperature, with greater accuracy at higher effective temperatures.
• Further refinement may be needed to improve predictions in colder environmental conditions.