Background It is still not certain whether it is worth using theophylline in

The following four intravenous treatments were administered in a balanced, randomized Latin square design to eight healthy volunteers: (‐)‐adrenaline (0.06 microgram kg‐1 min‐1 for 90 min) + vehicle control (+)‐glucose infusion (60 min), salbutamol (120 ng kg‐1 min‐1 for 30 min) + vehicle control (+)‐glucose infusion (90 min), (‐)‐adrenaline (0.06 microgram kg‐1 min‐1 for 90 min) + salbutamol (120 ng kg‐1 min‐1 for 30 min) and two vehicle control infusions of (+)‐glucose. All active solutions were preceded by a 1 h control infusion and the control infusion was continued for 1 h following the active solutions. Both the active solutions, (‐)‐adrenaline and salbutamol were increased stepwise to the above doses. Heart rate and blood pressure were recorded at frequent intervals throughout and venous blood was taken for the estimation of potassium, insulin, glucose, catecholamine and salbutamol levels. Adrenaline levels similar to those seen in acute illness were achieved using this infusion protocol. Salbutamol levels rose throughout the period of the salbutamol infusions and steady‐state was not achieved. Potassium levels were unchanged on the control + control study day and fell on all active treatments (0.45 mmol l‐1 following (‐)‐adrenaline + control; 0.48 mmol l‐1 following salbutamol + control; 0.93 mmol l‐1 following (‐)‐adrenaline + salbutamol). Insulin levels rose insignificantly after salbutamol alone and fell slightly on all other treatments.(ABSTRACT TRUNCATED AT 250 WORDS)

1 We have previously shown that salbutamol induced hypokalaemia, like adrenaline induced hypokalaemia, is the result of stimulation of a membrane bound 132-adrenoreceptor linked to Na+/K+ ATPase. We have also demonstrated that adrenaline induced hypokalaemia is potentiated by therapeutic concentrations of theophylline. 2 In a single-blind study of 14 normal volunteers, we infused salbutamol in doses used in clinical practice and examined the effects of the addition of theophylline alone or combined with (-)-adrenaline on plasma potassium levels, heart rate and blood pressure. The combinations studied were (i) salbutamol + vehicle control adrenaline infusion + placebo theophylline; (ii) salbutamol + vehicle control adrenaline infusion + theophylline; (iii) salbutamol + adrenaline + theophylline. 3 In a randomised, balanced placebo controlled design oral slow release theophylline or placebo was given for 9 days. Subjects were studied twice on the active limb (days 7 and 9) and once on the placebo limb (day 9) and the procedure was identical on each of the 3 study days except for the solutions administered. 4 Theophylline increased salbutamol induced hypokalaemia and in some individuals profound hypokalaemia (< 2.5 mmol 1-1) was observed with these relatively low doses of salbutamol and theophylline. Adrenaline did not further increase the magnitude of the fall in potassium observed. Combining theophylline with salbutamol increased the tachycardia resulting from the salbutamol infusion. Salbutamol infusion caused a fall in diastolic and rise in systolic blood pressure on all 3 study days and this was not altered by either theophylline or adrenaline alone or together. 5 We conclude that theophylline significantly increases salbutamol induced hypokalaemia and tachycardia and that the addition of adrenaline does not further increase hypokalaemia. Intensive bronchodilator therapy with these two agents in acutely ill, hypoxic patients with asthma or chronic obstructive lung disease may increase the risk of serious cardiac arrhythmias secondary to hypokalaemia.

Regulation of magnesium balance is poorly understood. However, hypomagnesaemia has been reported in patients in clinical situations where circulating catecholamines are raised including myocardial infarction, cardiac surgery and insulin-induced hypoglycaemia stress tests. The effects of L-adrenaline infusions, sufficient to achieve pathophysiological levels of adrenaline, and of therapeutic intravenous infusions of salbutamol, a beta 2-agonist, on plasma magnesium, plasma potassium, plasma glucose and plasma insulin levels were studied in a placebo-controlled design in eight normal subjects. Plasma magnesium levels fell significantly during the adrenaline infusion and also during the salbutamol infusion, though more slowly. In a 1 h period of observation after cessation of the infusions no recovery of plasma magnesium levels was seen. Significant falls in plasma potassium levels were also observed during both infusions with spontaneous recovery within 30 min after the infusions. No significant changes in plasma insulin levels occurred with either salbutamol or L-adrenaline compared with control. Plasma glucose levels rose significantly during the adrenaline infusion. The study suggests that both L-adrenaline and salbutamol cause shifts in plasma magnesium which are not mediated by insulin. We propose that intracellular shifts of magnesium occur as a result of beta-adrenergic stimulation.