membrane. Inhibiting the Na-K-ATPase would likely reduce this source of metabolic stimulation, and ATP demand would fall and concentrations would increase. Thus, low thirst would correlate with low pump activity and higher energy (ATP) levels within the cell (high thirst would correlate with high rates of sodium entrance, high rates of ATP hydrolysis, lower ATP levels, higher ADP and P_i levels, and stimulated glycolysis). In this model, high thirst correlates with high pump activity and lower steady-state ATP levels. If the cellular trigger for thirst were related to lower ATP levels (energy depletion), then this might explain the analogous condition of high ADH release (Baylis and Robertson, 1980; Baylis et al., 1981) with either intracellular glucopenia or 2-deoxyglucose (2dG). If glucose were either unavailabel (glucopenia) or unable (2-DG) to fuel glycolysis, then steady-state ATP levels would fall (energy depletion), thereby stimulating ADH release and thirst. Depending upon the situation (glucose concentration, insulin, etc.), elevated glucose levels might elevate the ATP levels and inhibit thirst, but even higher levels might deplete ATP levels by producing excess hexose phosphates. This difficult concept is summarized in Table 14-2.

Table 14-2 provides logic that thirst and ADH release can both be defined or regulated in terms of energy balance rather than the more common approach using water deficits and elevated osmolalities. This concept is relatively sophisticated and useful because it unifies a number of observations that on the surface are either unrelated or difficult to interpret