Pathophysiology of transient cranial diabetes insipidus during pregnancy.

OBJECTIVE: We investigated the possible mechanisms underlying transient cranial diabetes insipidus during pregnancy.

DESIGN AND PATIENTS: A woman who developed clinical diabetes insipidus during the third trimester of pregnancy was studied through a total of three pregnancies and postpartum.

MEASUREMENTS: Plasma AVP, urine and plasma osmolality, urine volume and specific gravity were measured during water deprivation tests and hypertonic saline infusion. Plasma and urine osmolality were measured after subcutaneous injection of AVP. The water deprivation and AVP test were repeated after proven inhibition of urinary PGE2 with aspirin. Serum vasopressinase activity was measured during one of the pregnancies affected with diabetes insipidus and compared with that obtained between 26 and 38 weeks from 13 normal pregnancies.

RESULTS: The patient was found to have cranial diabetes insipidus which responded to low dose intranasal 1-desamino-8-D-arginine vasopressin. Inhibition of PGE2 with aspirin did not enhance urine concentrating ability or the response to a test dose of subcutaneous AVP. Plasma levels of vasopressinase remained within the physiological range for normal pregnancy.

CONCLUSIONS: These studies indicate that subclinical cranial diabetes insipidus may be unmasked in late pregnancy. This effect is not related to AVP resistance resulting from PGE2 production or excessive vasopressinase activity, but may be due to a combination of physiological vasopressinase secretion with reduced AVP secretory capacity and reduction in the thirst threshold that accompanies normal pregnancy. We relate these findings to a previously described group of women with transient diabetes insipidus during pregnancy who had impaired liver function.