It may be somewhat surprising to find that even after centuries of research, the fundamental system of control of thermal balance in humans, as well as other mammals and homeotherms, remains controversial. One of the reasons may be that thermal balance is not confined to or dominated by a specific organ, such as the kidneys with renal control and fluid balance; accordingly, its physiological control is inherently integrative and open to theorization. Currently, three major models have been advanced to help explain thermal homeostasis in humans. The first two models propose that temperature is the regulated variable and share similar ideas concerning the underlying neural architecture. The third model is fundamentally different from the first two models, in that rather than temperature per se, body heat content is the regulated variable and body temperature a by-product of that regulation.
The traditional model of temperature regulation was proposed by Hammel and has been termed the set point model (Hammel et al. 1963). In this model, which explains thermal homeostasis using a set point signal as a comparator for body temperature, a corrective response is executed when body temperature differs from the comparator signal. A relevant basic analogy for this model is a household thermostat, which initiates either cooling or heating based on comparison of house temperature with an internally adjustable set temperature. This simple analogy breaks down in that, unlike a thermostat with its on/off response of maximal heating or cooling, humans are capable of graded thermoregulatory responses (e.g., higher-intensity shivering in more muscles) with greater deviations from normal temperatures. In the set point model, thermal afferents from throughout the body core and peripheries are integrated into an overall thermal signal at one or more central sites, generally assumed to reside within the central nervous system (CNS) and likely within the hypothalamus. This is informed in part by cold and warm receptors under human skin existing at an average depth of 0.15 to 0.17 mm and 0.3 to 0.6 mm (Boulant 2006), respectively. Further neural support for the model comes from the observation that the hypothalamus contains both cold- and heat-sensitive neurons, which increase their rate of firing based on either cooling or heating. This thermal signal is then compared to an internally stored thermal signal, and heat gain or heat loss responses are activated appropriately.
One common argument against this “set point” model is the false assumption that it cannot explain deviations from a presumably fixed and constant reference temperature stored within the CNS. Such deviations, for example a higher regulated body temperature, can occur normally over the course of the menstrual cycle, with baseline temperature during the luteal phase regulated at approximately 0.5 °C higher than during the follicular phase (see figure 2.4). Also, body temperature is generally defended appropriately, though at a higher absolute core temperature, during fever. In response, proponents for the set point model argue that there is no requirement for a fixed and unchanging reference temperature hard-wired into the CNS, but rather that the set point temperature may vary over a range based on the relative activity of heat- and cold-sensitive neurons throughout the body (Cabanac 2006).