Increased thirst and plasma arginine vasopressin levels during 2-deoxy-D-glucose-induced glucoprivation in humans.

Insulin-induced hypoglycemia by unknown mechanism(s) increases plasma arginine vasopressin (AVP) levels in humans. Mechanisms for increased AVP levels during central nervous system glucoprivation were investigated by administering 20-min i.v. infusions of 2-deoxy-d-glucose (50 mg/kg), a competitive inhibitor of glucose utilization, or normal saline (sham), to 24 normal volunteers. Some of the infusions were administered in combination with neuropharmacological blocking agents (placebo). The behavioral, physiological, metabolic, and hormonal correlates of 2-deoxy-d-glucose (2DG)-induced gluco-privation and AVP secretion were studied in a group (n = 5) pretreated for 1 wk with either mazindol (1 mg per os three times per day), a potent norepinephrine and dopamine-reuptake blocker, or placebo. A second group (n = 5) received either propranolol (3 mg/3 min followed by 80 mug/min) or normal saline infusion before and during 2DG administration. With 2DG alone, plasma AVP levels increased from 1.3+/-0.3 pg/ml at base line to a peak of 4.5+/-1.4 pg/ml at 60 min and remained elevated for 150 min. From 30 to 180 min after 2DG administration, the 2DG-infused volunteers increased their water intake in comparison with sham-infused volunteers. Marked increases in epinephrine and slight increases in norepinephrine were associated with increases in plasma glucose and renin activity and decreases in plasma potassium. Plasma sodium and osmolality increased transiently and mean arterial pressure (MAP) fell. These changes, however, were small and inconstant and could not account for the observed increases in thirst and AVP levels. Pretreatment with mazindol prevented the decrease in MAP and the increase in plasma renin activity (PRA) following 2DG infusions without modifying increased thirst, water intake, or AVP responses to glucoprivation. Pretreatment with propranolol effectively blocked beta-adrenoreceptors as evidenced by increased MAP and plasma epinephrine, and abolition of the RPA increases during 2DG-induced glycoprivation, but did not suppress AVP and thirst responses. A cervical cord-sectioned patient lacking descending sympathetic out-flow had a potentiated thirst response to 2DG-induced glucoprivation in the absence of increases in sodium, catecholamines, and PRA. Thus 2DG administration activates mechanisms for increased thirst and AVP which are unrelated to changes in peripheral catecholamines, MAP, PRA, and osmolality.