Study objective: To determine whether antibodies to vasopressin play a role in the development of diabetes insipidus or interfere with diagnosis and treatment.
Design: Random plasma or serum samples for determining of antibodies to vasopressin were collected from patients and controls.
Setting: Referral university hospital with most patients studied in the clinical research center.
Patients: Twenty-nine healthy controls and 113 patients with polyuria (15 with primary polydipsia; 86 with neurogenic diabetes insipidus [60 studied before, 28 during, and 10 after antidiuretic hormone treatment]; and 12 with nephrogenic diabetes insipidus).
Interventions: Antibodies were detected by incubating samples with radiolabeled 125I-arginine vasopressin. The effect of antibodies on diagnosis was studied by examining the relation of plasma vasopressin to osmolality measured during dehydration or infusion of hypertonic saline and the relation of urine osmolality to plasma vasopressin measured during dehydration.
Measurements and main results: Antibodies to vasopressin were not detected in patients with primary polydipsia, nephrogenic diabetes insipidus, or neurogenic diabetes insipidus studied before therapy with antidiuretic hormone. Antibodies were detected in 6 of 28 patients studied during such treatments. All 6 patients reported decreased antidiuretic response to previously effective therapy with arginine or lysine vasopressin but had normal response to desmopressin or chlorpropamide.
Conclusion: Diabetes insipidus does not result from spontaneously occurring antibodies to vasopressin. The antibodies occasionally develop during treatment with antidiuretic hormone and, when they do, almost always result in secondary resistance to its antidiuretic effect. Antibodies usually do not impair the response to other forms of therapy; they only rarely interfere with the diagnosis of diabetes insipidus, by falsely suggesting the presence of the partial nephrogenic form.