(T₂) was removed immediately after the fourth work bout (approximately 11:30 a.m.). The final blood sample (T₃) was withdrawn after the final walk (approximately 3:30 p.m.). Collecting tubes were immediately placed into ice and transported to the laboratory for centrifugation (4°C, 10,000 rpm); EDTA plasma or serum was removed and stored (-20°C) for subsequent analysis.

Aldosterone was quantitated in serum using commercially available kits purchased from the Diagnostic Products Corporation (Los Angeles, California) and following techniques described in their technical bulletin (Aldosterone, No Extraction, Coat-A-Count^R). This technique provides an approximate detection limit of 16 picograms (pg) per ml and is extremely specific for aldosterone; the range for this hormone is usually 4 to 31 nanograms (ng) per deciliter in salt-replete, standing adults (Aldosterone, No Extraction, Coat-A-Count^R). Plasma renin activity was estimated by the quantitation of angiotensin I in EDTA plasma. Commercially available test kits (RIANEN Angiotensin I [¹²⁵I] RIA Kit) were purchased from DuPont NEN Products (Boston, Massachusetts), and the assay was performed according to techniques outlined in their technical manual (RIANEN Assay System, Angiotensin I, Instruction Manual). When endogenous converting enzyme and angiotensinases of human plasma are appropriately inhibited, then angiotensin I formation quantitatively reflects PRA. Control values in adult men ordinarily range from 1 to 4.1 ng angiotensin I formed per ml per hour (Young, 1987).

Arginine-vasopressin was quantitated in EDTA plasma according to the techniques outlined by LaRose et al. (1985). One ml of EDTA plasma was treated with 10 µl per ml of 50 percent trifluoroacetic acid to acidify the sample to a pH of 4.0 to 4.5. Rabbit antibody to arg⁸-vasopressin was purchased from the Calbiochem Corporation (San Diego, California), and ¹²⁵I-AVP was purchased from the DuPont NEN Corporation. Prepared standards were purchased from the Incstar Corporation (Stillwater, Minnesota). The range of circulating AVP in healthy adult males has been reported to be non-detectable to 4.7 pg per ml (Incstar, Vasopressin ¹²⁵I RIA Kit).

Repeated measures analysis of variance was performed using statistical package BMDP4V (BMDP Statistical Software, Los Angeles, California). Tukey's mean critical difference test was applied post hoc to determine significant differences of appropriate mean values. The null hypothesis was rejected at p < .05.