The effect of a highly specific serotonin agonist on osmoregulated vasopressin secretion in healthy man

OBJECTIVE: To explore a possible interaction of the serotonin neurotransmitter system and posterior pituitary function, we have looked at the effect of fluoxetine treatment on osmoregulated vasopressin secretion in normal men in two placebo controlled studies. DESIGN: In each study subjects took in random order for 7 days one capsule daily of placebo or 40 mg fluoxetine. On the 8th day subjects underwent assessment. Study 1 A hypo-osmotic stimulus of an oral water load of 20 ml/kg. Study 2 A hyperosmotic stimulus of intravenous infusion of 5% (855 mmol/l) saline at 0.06 ml/kg/min for 120 minutes. PATIENTS: Normal, healthy male volunteers. Study 1, 9; Study 2, 10. MEASUREMENTS: In both studies regular measures of plasma osmolality, sodium and vasopressin were made. In Study 1 urine osmolality was measured together with urine volume at set time points and an accumulative measure of percentage of water load excreted. Free water clearance was calculated. In Study 2 the relationship of plasma vasopressin to change in plasma osmolality was calculated for each subject by linear regression analysis. RESULTS: Serotonin agonism had no effect on baseline measurements in either study. Study 1 After 4 hours subjects excreted 95 and 99% of the water load after placebo and fluoxetine respectively (P = 0.407). There was no effect of fluoxetine compared to placebo on the pattern or extent of change of plasma osmolality (nadir 285.9 +/- 1.4 mosm/kg placebo, 283.1 +/- 1.1 mosm/kg fluoxetine, P = 0.145) or free water clearance or maximum urine dilution after oral water loading. Plasma vasopressin suppressed to a minimum concentration after both treatments in response to hypo-osmolality 0.5 +/- 0.1 pmol/l (placebo), 0.3 +/- 0.01 pmol/l (fluoxetine), P = 0.195. Study 2 Fluoxetine had no significant effect on the sensitivity of vasopressin release to change in plasma osmolality (0.33 +/- 0.06 pmol/l per mosm/kg placebo, 0.36 +/- 0.06 pmol/l per mosm/kg fluoxetine, P = 0.347). Nor was there a significant effect on the theoretical osmotic threshold for release of vasopressin (287.0 +/- 1.21 mosm/kg placebo, 286.9 +/- 1.09 mosm/kg fluoxetine, P = 0.700). CONCLUSION: We have found no evidence of a physiologically relevant effect of serotonin agonism on osmoregulated vasopressin release, or on the ability of normal man to excrete a water load. The possible reasons for this contrast to animal work are discussed.