SUMMARY In this study we found that, in 31 normal subjects, close to 90% of circulating arginine vasopressin (AVP), measured by radiolmmunoassay, was associated with platelets. By using routine methods of centrifugation, which do not completely separate platelets, the normal range of plasma vasopressin was higher by twofold than the normal range in platelet-free plasma prepared by differential centrifugation, which was 1.4 ± 1.0 SD pg/ml. Platelet vasopressin was 12.9 ± 5.7 pg/ml. Patients with congestive heart failure had, on average, an elevated platelet-free plasma AVP, as did two patients with thrombocytopenia and one with thrombocytosis. Patients with essential hypertension had slightly high levels of platelet-free plasma AVP and demonstrated an abnormal inverse relationship between platelet-free plasma AVP and serum osmolality. Immunoreactive platelet vasopressin was slightly low in patients with essential hypertension and was subnormal in patients with congestive heart failure. These studies demonstrate that platelets normally present in centrifuged plasma cause an overestimation of the plasma vasopressin levels. Until the physiological meaning of plasma and platelet-bound AVP is understood, studies of circulating vasopressin should probably assess both plasma and platelet AVP levels. 2 Moreover, vasopressin is reported to be involved in various forms.of experimental hypertension, 1 * 9 and lesions of the medulla oblongata have been shown to cause excessive release and acute hypertension. 10 In addition, elevations of urine and plasma vasopressin have been reported in some patients with hypertension"" 13 despite the presence of lower serum sodium.
13On the other hand, the role of vasopressin in the maintenance of hypertension is controversial, 14 " 17 and there is evidence to suggest that vasopressin cannot maintain chronic hypertension because of a compen- Nonetheless, measurements of circulating vasopressin in human hypertension may be of value by providing clues to underlying mechanisms. Vasopressin is released into the circulation in response to either osmotic stimuli or volume-dependent baroreceptor-mediated stimuli.
"23 Therefore, assessment of vasopressin levels in relation to serum osmolality should provide evidence of which mechanism is operative in patients who have abnormal levels of vasopressin.In the present study we evaluated arginine vasopressin (AVP) levels in relation to the serum osmolality, in a group of untreated patients with essential hypertension. In the course of this investigation, using a sensitive radioimmunoassay method for vasopressin, 24 we observed abnormally high levels in plasma samples that were contaminated with the buffy coat. We then found that vasopressin levels measured in a platelet fraction were close to tenfold higher than in plateletfree plasma and, that platelet-free plasmas had significantly lower values than in most previous reports. 22 In this paper we report these findings and also the results of vasopressin measurements in platelets and platelet-