Initiation of GH treatment in adults is frequently complicated by the development of symptomatic fluid retention. To investigate the mechanism and extent of fluid retention that occurs with dosages of GH used in the treatment of GH-deficient adults, we conducted a double blind study in which seven GH-deficient patients (aged 24-74 yr) each received in random order daily sc injections of placebo, a physiological dose of GH (0.04 U/kg, low dose), and a supraphysiological dose of GH (0.08 U/kg, high dose) for 7 days, separated by 21-day washout periods. On the seventh day, measurements were made of serum insulin-like growth factor I, body weight, exchangeable sodium, plasma volume, angiotensinogen, PRA, aldosterone, atrial natriuretic peptide (ANP), and mean 24-h ambulatory heart rate and blood pressure. GH significantly increased mean insulin-like growth factor I levels from 105 +/- 11 to 304 +/- 45 micrograms/L during low dose treatment (P = 0.006) and 400 +/- 76 micrograms/L during high dose treatment (P = 0.004). High dose GH caused a 1.2 +/- 0.3 kg increase in body weight (P = 0.01) and a 193 +/- 65 mmol increase in exchangeable sodium (P = 0.008). Low dose GH had a lesser effect, with no significant increase in body weight, but an increase in exchangeable sodium of 113 +/- 37 mmol (P = 0.02). Plasma volume was not significantly affected by GH treatment. Mean supine angiotensinogen levels were significantly higher during both GH treatments compared to placebo (low dose, P = 0.017; high dose, P = 0.028) as were mean supine PRA levels (low dose, P = 0.0002; high dose, P = 0.0025). Supine angiotensin II, aldosterone, and ANP levels were not significantly affected by GH treatment. There was no significant change from placebo in any of the sodium-regulating hormones in the erect posture. The mean 24-h heart rate was significantly higher during low dose (82 +/- 2 beats/min; P = 0.0001) and high dose (88 +/- 3 beats/min; P = 0.0001) GH treatment than during placebo (67 +/- 3 beats/min). However, no significant change in mean 24-h systolic or diastolic blood pressure was observed. In summary, acute GH administration using doses currently employed in treating adults causes a dose-related increase in body weight and body sodium, but no associated increase in blood pressure. We conclude that 1) sodium retention is a physiological effect of GH, but does not cause an acute rise in blood pressure; and 2) the mechanism of sodium and fluid retention is not primarily due to enhanced aldosterone secretion or inhibition of ANP release, but more likely to a direct renal tubular effect.