1. The effects of metoprolol and propranolol on heat production and body temperature have been studied in six male subjects during insulin-induced hypoglycaemia in a thermoneutral environment. Hypoglycaemia was induced by insulin infusion on three occasions in each subject, accompanied by the infusion of sodium chloride solution (154 mmol/l) (control), metoprolol (β 1 -selective antagonist) or propranolol (non-selective antagonist). 2. During the period of hypoglycaemia in the control experiments mean heat production (calculated from respiratory gas exchange) increased by 1·07 ± sem 0·13 kJ/min and remained elevated for 30−40 min. This heat production response was reduced by metoprolol and abolished by propranolol. During the recovery period, heat production was significantly reduced in the presence of propranolol. 3. Skin and core temperatures fell during the period of hypoglycaemia in all three experiments. The fall in skin temperature was significantly greater in the presence of propranolol (−2·51 ± 0·47°C). The reductions in core temperature recorded during the three experiments were similar (control −0·73 ± 0·17, metoprolol −0·99 ± 0·21, propranolol −0·88 ± 0·22°C), but core temperature was still falling at the end of the propranolol experiment. 4. The cardiovascular responses to hypoglycaemia were similar in the control and metoprolol experiments but were substantially modified by propranolol. During the period of hypoglycaemia in the control experiments, plasma adrenaline levels rose to 7·78 ± 1·79 nmol/l; significantly higher levels were measured in the metoprolol (10·11 ± 1·64) and propranolol (22·76 ± 7·02) experiments. The very high adrenaline levels may have been responsible for the modified cardiovascular responses to hypoglycaemia observed in the propranolol experiment.

1. The present experiments were designed to elucidate the reasons for the fall in central body temperature during hypoglycaemia. 2. The first experiment was carried out at a room temperature of 25 °C on 11 male subjects. Hypoglycaemia was induced by infusion of insulin. Heat production (calculated from respiratory gas exchange) rose from a baseline of 5.10 ± 0.13 kJ/min (mean ± sem ) to a peak of 6.25 ± 0.21 kJ/min ( P < 0.001), but core temperature fell concurrently by 0.51 ± 0.08°C and skin temperature fell by 1.1 ± 0.2°C. The net heat loss was due to peripheral vasodilatation and sweating. 3. To determine the effect of insulin-induced hypoglycaemia on thermoregulation in a cool environment, the experiment was repeated at a room temperature of 18–19°C on five of the subjects who had air blown over them until shivering was sustained. During this time heat production rose to 10.13 ± 1.67 kJ/min, but core temperature remained constant. Shivering stopped as plasma glucose fell below 2.5 mmol/l during insulin infusion and the subjects said they no longer felt cold. 4. During hypoglycaemia in the cold peripheral vasodilatation and sweating occurred, skin temperature fell by up to 0.8°C and core temperature fell below 35°C, so subjects had to be rewarmed. 5. Recovery of plasma glucose after hypoglycaemia in the cold was impaired at low body temperatures, but shivering was restored within seconds when glucose was given intravenously.