Statistical Modelling of Physiological Heat Stress Response by Means of a Nonlinear Two-Stage Model

Abstract

In the context of medical or biological studies, very often parameters of interest are measured repeatedly over time under a given set of conditions. This results in a set of (often similarly shaped) time series. Then, the objective is the determination of the functional relationship between the parameter of interest and time on the one hand, and the analysis of the variation of this functional relationship between experiments, on the other hand. This may be done by means of a two-stage model. The present work describes the theory of the two-stage model and its application to the increase of human core temperature for a set of 678 experiments where the subjects were exposed to warm and hot environments. The data originating from 6 European research institutes, have been pooled into one database for the Heat Stress Project within the scope of the BIOMED 2 programme of the European Union. A nonlinear two-stage model was applied, with a logistic function modelling the nonlinear time course of the core temperature, and with its parameters depending on air temperature, mean radiant temperature, air velocity, partial vapour pressure, clothing insulation, metabolic rate, gender, acclimatisation status and body surface area. We conclude that acclimatisation, clothing insulation, body surface area, air temperature, air velocity, partial vapour pressure, metabolic rate, and the difference between mean radiant temperature and air temperature play an important role for work in warm and hot environments. We show how our results can be used for the estimation of allowable exposure times for work in hot environments.