Background: Animal studies suggest that vasopressin has cognitive-enha
ncing properties and oxytocin may have amnestic effects. A clinical re
port suggests that the acute increase in oxytocin-associated neurophys
in predicts clinical response to electroconvulsive therapy (ECT) in de
pressed patients. Methods: Medication-free patients with major depress
ion were randomized to receive right unilateral or bilateral ECT admin
istered with electrical stimulus intensity at either just above seizur
e threshold or at 150% above seizure threshold. The associations betwe
en plasma vasopressin, oxytocin, ECT treatment parameters, clinical ou
tcome, and cognitive effects were assessed. Results: The sample compri
sed 55 patients. At the second ECT, patients receiving ECT at 150% abo
ve initial seizure threshold had significantly greater increases in pl
asma vasopressin than patients receiving low-dose ECT (ps < .01-.04),
with no effects of electrode placement. At the second and ninth ECT tr
eatments, the vasopressin or oxytocin surges were not associated with
clinical improvement, seizure duration, time to orientation, or memory
test performance. There were inverse trend-level associations between
the acute surge in oxytocin levels at the ninth ECT and clinical resp
onse, contradicting a report in the literature. Conclusions: Overall,
these findings do nor support the hypothesis that diencephalic seizure
propagation is central to the mechanism of action of ECT. (C) 1998 So
ciety of Biological Psychiatry.