hyperphagia would probably lead to hyperthermia. (At 40°C, rats will stop feeding altogether, and if force fed by intubation, they suffer heat stress and occasionally die [Hamilton, 1967].) Either way, reduced intake in the heat would seem to be adaptive. The only issue concerns activity, which, if intensified, ought to place extra demands on energy stores. The prudent recommendation for heat exposure would seem to be to allow for reduced intake but to avoid, as much as possible, strenuous activity, which not only requires more energy but also generates more undesirable heat, and which also puts fluid balance in jeopardy. Reduced activity is a natural response to heat exposure. If bursts of activity are unavoidable, care should be taken to allow, as much as possible, for longer than normal metabolic recovery periods.
Although eating causes thermogenesis, it does not follow automatically that all thermogenesis will feed back as a regulator of eating. Glick et al. (1989) abolished the thermogenic response of brown adipose tissue during and after feeding in rats but found no indication that meal size was augmented, as would be expected under, say, Brobeck's (1948) theory, if heat provided a satiety signal. It remains true that "increases in body and brain temperature do not coincide exactly with the cessation of feeding" (Balagura, 1973). Of course, the apparent unimportance of BAT thermogenesis in the control of appetite does not mean that endogenous heat in general is irrelevant to the regulatory control of appetite. Rampone and Reynolds (1991) have recently outlined a proposal—a "fine-tuning" of Brobeck's (1948) proposal—in which diet-induced thermogenesis feeds back to activate temperature-sensitive neurons in the rostral hypothalamus, which in turn activate the ventromedial hypothalamus to induce satiety. They explain hyperphagia and weight gain as the result of inadequate diet-induced thermogenesis and consequent inadequate satiety, with the result that the animal takes in more energy than it expends. Consistent with this notion is the finding that animals with rostral lesions both overeat and become hyperthermic (Hamilton and Brobeck, 1964).
It should be noted that even if the abolition of all thermogenesis failed to affect satiety or the duration and/or size of a meal, it would not follow logically that thermogenesis is unimportant in the control of appetite. Conceivably, feeding might be responsive to the lack of energy/heat in the "system." Meal-induced thermogenesis might not act as a satiety signal, but still serve to delay the onset of a drop in heat below some threshold that serves as a hunger signal. In other words, the focus in this chapter on heat as a satiety signal fails to address the initiation of eating. Perhaps energy depletions as hunger signals ought to be considered, in which case heat might well remain an important determinant of feeding but more at the onset end than at the offset end. The unwarranted but prevalent assumption that the same types of signals control both meal termination and meal initiation—as in Rampone and Reynolds' (1991) hypothesis that heat induces