0021-972X/90/7106-1536$02.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1990 by The Endocrine Society
Vol. 71, No. 6 Printed in U.S.A.
Nonosmotic Stimuli Alter Osmoregulation in Patients with Spinal Cord Injury* HUGH H. WILLIAMS, BARRY M. WALL, JEANNE M. HORAN, DANIEL N. PRESLEY, JOAN T. CROFTON, LEONARD SHARE, AND C. ROBERT COOKE Veterans Affairs Medical Center and the University of Tennessee, Memphis, Tennessee 38104
ABSTRACT. Studies on two quadriplegic patients who developed severe hyponatremia during episodes of acute respiratory distress were performed to determine whether differences in osmoregulation of vasopressin release could be identified in these patients compared to other quadriplegic subjects previously studied in a similar manner. Both patients were clinically stable and normonatremic, with no signs or symptoms of respiratory distress, when the studies were performed. However, both exhibited evidence of hemodynamic instability in the sitting posture. Linear regression analysis of the plasma vasopressin/plasma osmolality (Pavp:Posm) relationship during infusions of 0.85 M sodium chloride showed no significant differences in either the slope (sensitivity) or abscissal intercept (osmotic threshold) of this relationship compared to that of other quadriplegic subjects when the patients were supine. In contrast, when the patients were studied in the sitting posture there was a marked shift in the relationship of Pavp:Posm indicative of increased sensitivity and reduced osmotic threshold for vasopressin release. The
U
NDER most physiological conditions vasopressin release is primarily responsive to osmotic stimulation (1, 2). However, nonosmotic stimuli, especially those associated with hypotension and/or hypovolemia, are also capable of eliciting a vasopressin response and can modify the relationship between plasma vasopressin (Pavp) and plasma osmolality (Posm) without disrupting their characteristically close association (3-7). In recently reported studies on subjects with cervical spinal cord injury, changes in the Pavp:Posm relationship consistent with increased sensitivity of vasopressin release to osmotic stimulation were observed in association with the assumption of erect (sitting) posture (7). The circumscribed neurological deficit in such individuals provides a unique opportunity to study the interaction between osmotic and nonosmotic control of vasopressin release in human subjects. The afferent neural pathways inReceived April 25, 1990. Address all correspondence and requests for reprints to: Hugh H. Williams, M.D., Nephrology Section, Veterans Administration Medical Center (111B), 1030 Jefferson Avenue, Memphis, Tennessee 38104. * This work was supported by V.A. Merit Review funding and in part by USPHS Research Grant HL-19209 from the NHLBI.
slopes of the Pavp:Posm relationships were 0.249 and 0.178 for
the two patients, respectively, compared to 0.092 ± 0.03 (±SD) for previously studied quadriplegic subjects. Oral water-loading studies performed on one patient revealed marked impairment of urine-diluting ability and free water clearance in the sitting posture compared with observations in similar studies performed when the patient was supine. Impairment of renal water excretion could not be attributed to an effect of vasopressin, which was reduced to unquantifiable levels by water loading. These studies have shown that hemodynamic stress related to autonomic dysfunction in quadriplegic patients may result in marked alteration of osmoregulation of vasopressin release in more severely affected individuals. Such altered osmoregulation, which may also be associated with vasopressin-independent impairment of renal water excretion in the sitting posture, may be a predisposing factor in the development of hyponatremia, especially in the presence of other potent nonosmotic stimuli. (J Clin Endocrinol Metab 7 1 : 1536-1543,1990)
volved in the modulation of vasopressin release are intact in quadriplegic subjects, while the efferent pathways from the central nervous system that are normally involved in the maintenance of cardiovascular homeostasis in erect posture are defective (8-10). This defect in the efferent limb of the baroreflex arc is responsible for the occurrence of orthostatic hemodynamic instability in quadriplegia. Such instability may be a source of nonosmotic stimulation of vasopressin release via the aforementioned afferent neural pathways, but the extent to which these nonosmotic influences may modify the relationship between Pavp and Posm has not been fully delineated. We have observed two quadriplegic patients whose ability to remain out of bed in a wheelchair was markedly affected by hemodynamic instability in the sitting posture, and who became severely hyponatremic during episodes of acute respiratory distress. To determine whether abnormalities in osmoregulation and water handling could be identified as possible predisposing factors in the development of hyponatremia, the interaction between osmotic and nonosmotic control of va-
1536
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 August 2016. at 22:21 For personal use only. No other uses without permission. . All rights reserved.
NONOSMOTIC STIMULI ALTER OSMOREGULATION
sopressin release was investigated when the patients were clinically stable, normonatremic, and showing no signs or symptoms of respiratory distress. Our observations in these studies suggest that hemodynamic instability, as shown by orthostatic hypotension, results in potent nonosmotic stimulation of vasopressin release and marked alteration of vasopressin responsiveness to osmotic stimulation.
Materials and Methods Clinical observations durifig severe hyponatremia Patient 1, a 60-yr-old white male with quadriplegia due to spinal cord injury at the C-5 level in 1949, was admitted to the V.A. Medical Center with a history of lethargy, weakness, and altered mental state of 2-day duration. Findings on x-ray examination of the chest Were consistent with right lower lobe pneumonia. Admission laboratory studies revealed the following results: serum sodiunj concentration, 113 mmol/L; potasssium, 3.1 mmol/L; chloride, 83 mmol/L; bicarbonate, 24 mmol/ L; blood urea nitrogen, 1.3 mmol/L urea; and glucose, 6.9 mmol/ L. Arterial blood pH wa$ 7.45, pCO2 was 28 mm Hg, and pO2 was 53 mm Hg. He received respiratory therapy and antibiotics as treatment for pneumonia. The following morning his serum sodium concentration was 107 mmol/L, potassium Was 2.3 mmol/L, chloride was 81 mmol/ L, and bicarbonate was 23 mmol/L. Blood urea nitrogen concentration was 1.1 mmol/L urea. Serum osmolality was 211 mosmol/kg H2O, and urine osmolality (Uosm) was 568 mosmol/ kg H2O. Urinary sodiufrn concentration was 112 mmol/L. Plasma cortisol concentration and thyroid function test results were normal. The patient was treated with iv infusion of 0.51 M sodium chloride and furosemide, followed by continuous infusion of 0.153 M sodium chloride and supplemental potassium chloride. Subsequently, his serum electrolytes were: sodium, 124 mmol/L; potassium, 2.9 mmol/L; chloride, 90 mmol/ L; and bicarbonate, 26 mmol/L. There was some improvement in his mental status, but respiratory distress persisted. Arterial blood pH 28 h after admission was 7.13, pCO2 was 84 mm Hg, and pO2 was 81 mm Hg. Endotracheal intubation was performed, and mechanical ventilation was required over the next 3 days. There was gradual improvement in his pulmonary status, and at the time of extubation his serum sodium concentration was 139 mmol/L. Patient 2, a 69-yr-old white male with quadriplegia due to spinal cord injury at the C-4-5 level 11 yr previously, was admitted to the V.A. Medical Center for investigation of nephrolithiasis. Intravenous pyelography revealed a staghorn calculus and multiple smaller renal stones within the left kidney in addition to a calculus in the bladder. The bladder calculus was removed cystoscopically, and a left percutaneous nephrolithotomy was performed under general anesthesia. He tolerated the procedure well,, but in the postoperative period he developed pulmonary congestion, a nonproductive cough, and dyspnea. Chest x-ray examination showed left lower lobe atelectasis. Four days postoperatively, the serum sodium concentration was 135 mmol/L, chloride was 92 mmol/L, and bicarbonate
1537
was 33 mmol/L. The serum glucose concentration was 5.2 mmol/L, blood urea nitrogen was 2.5 mmol/L urea, and serum creatinine was 62 /rnnol/L. Dyspnea persisted despite intermittent postive pressure breathing, chest percussion, postural drainage, and therapy with bronchodilators. Three days later respiratory stridor was noted. Laboratory studies at this time showed a marked reduction in the serum sodium concentration to 112 mmol/L. The serum potassium concentration was 3.5 mmol/L, chloride was 83 mmol/L, bicarbonate was 25 mmol/ L, glucose was 6.9 mmol/L, blood urea nitrogen was 5.3 mmol/ L urea, and serum creatinine was 53 ^mol/L. Arterial blood pH was 7.52, pCO2 was 36 mm Hg, and pO2 was 136 mm Hg. The urinary sodium concentration was 60 mmol/L, Uosm was 340 mosmol/kg H2O, and serum osmolality was 236 mosmol/kg H2O. Water intake was restricted, and the patient was observed closely. The following day, he was treated with an iv infusion of 0.85 M sodium chloride, receiving a total of 270 mL over 90 min. The serum sodium concentration and serum osmolality increased from 109 mmol/L and 236 mosmol/kg H2O, respectively, to 127 mmol/L and 271 mosmol/kg H2O. The plasma arginine vasopressin concentration before hypertonic sodium chloride infusion was 6.7 ^U/mL and increased further during the infusion to 15.5 jiU/mL. Respiratory stridor persisted, endotracheal intubation and mechanical ventilation were required, and tracheostomy was subsequently performed. With water restriction and improvement in the patient's respiratory status, his serum sodium concentration and plasma osmolality became normal and remained so for the rest of his hospital
Study protocol Serum sodium concentrations were normal in both patients when these studies were performed, and neither patient exhibited signs of respiratory distress, fluid retention, or dehydration. They had been eating a normal diet without restriction of fluid intake and had been receiving no medications known to affect water handling. Both refrained from the use of tobacco or ethanol for at least 16 h before the studies. The studies were approved by the Human Studies Subcommittee and the Research and Development Committee of the V.A. Medical Center, and informed consent was obtained in writing before the studies were performed. Hypertonic sodium chloride infusion studies The patients were studied on two separate occasions, once while supine and once while sitting quietly in a wheelchair. All studies began at approximately 0900 h, after an overnight fast. On the morning of each study, catheters were inserted into antecubital or forearm veins of both arms, and blood pressure and heart rate (HR) were measured at 3-min intervals over the next 30 min using a thigh cuff and an automated recording device (Dinamap, Critikon, Inc., Tampa, FL). At the end of this period of equilibration, blood samples were obtained for determinations of Posm, plasma sodium (Pna), plasma aldosterone (Pa), plasma cortisol (Pc), Pavp, total protein concentrations, and PRA. Intravenous infusions of 0.85 M sodium chlo-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 August 2016. at 22:21 For personal use only. No other uses without permission. . All rights reserved.
JCE & M • 1990 Vol 71 • No 6
WILLIAMS ET AL.
1538
ride (0.05 mL/kg-min) were then begun and continued for 90 min. Blood pressure and heart rate were measured every 3 min during the infusions, and blood samples were obtained every 15 min for determinations of Posm, Pna, Pavp, and total protein. Final blood samples were obtained at the end of the infusions for determinations comparable to those performed on preinfusion blood samples. The total volume of all blood samples in each study was approximately 120 mL. The protocol followed in these studies was identical to that employed in studies on 15 other quadriplegic subjects who had not previously been hyponatremic. Data from the studies on these subjects were previously reported (7) and are presented only for comparison with the observations in the studies on the 2 patients who provide the basis for this report. The 15 normonatremic quadriplegic subjects were all males, ranging in age from 28-72 yr.
ing on which extractions were performed. Arginine vasopressin was extracted with octadecylsilane (Sep-Pak C18 cartridges, Waters Associates, Milford, MA). The limit of detection of the assay, defined as the smallest quantity of hormone statistically separable from zero, is 0.18 /xU/mL. Pavp concentrations are at or below this level during sustained water diuresis in normal subjects (our unpublished data). USP posterior pituitary reference standard (U.S. Pharmacopeial Convention, Inc., Rockville, MD) was used as the vasopressin standard. The average recovery of arginine vasopressin was 86.8 ± 1.1% (n = 57 samples from 11 assays). The interassay coefficient of variation was 9.3%; the within-assay coefficient of variation was 5.4 ± 1.2%. Plasma values were not corrected for incomplete recovery.
Water load studies performed on patient 1
The relationship of Pavp:Posm in each patient, supine and sitting, was evaluated by linear regression analysis (13, 14). Slope values were compared to the mean regression slopes of data derived from our previously reported studies on quadriplegic subjects (7). Values for all data, including the Pavp:Posm slope values, from the hypertonic sodium chloride infusion studies were considered to be significantly different from the previously reported group data if they were outside 2 SD of the group means. Data from water load studies on patient 1 were compared to observations obtained in similar studies on four quadriplegic subjects with no prior history of hyponatremia and three normal control subjects.
The renal response to oral water loading was assessed in studies on patient 1 in both the supine and the sitting posture. On both occasions, a standard water load (20 mL/kg BW) was administered orally over a period of 30 min, followed by additional hourly water intake equal to the previous hour's urine output over a total period of 3 h. Blood samples for determinations of Posm, Pna, Pavp, and plasma creatinine concentration were drawn at the beginning of water loading, at 15-min intervals during the first and second hours, and at the end of the third hour of the studies. Urine samples were collected from an indwelling bladder catheter at 15-min intervals during the first and second hours and at the end of the third hour for measurements of urine volume, osmolality, sodium, chloride, and creatinine concentrations. Blood pressure and HR were measured every 15 min using an arm cuff and an automated recording device. Comparable studies were performed on four quadriplegic subjects with no prior history of hyponatremia and three normal male control subjects. Analytical methods Blood for determinations of PRA and Pavp was drawn into prechilled plastic syringes containing EDTA (500 nL EDTA/ 10 mL blood). Blood for other determinations was collected into heparinized prechilled plastic syringes. Samples were kept in ice until centrifuged at 4 C. The separated plasma was stored at -20 C until the analyses could be performed. Pna was measured by ion-selective electrode (Beckman Instruments, Inc., Brea, CA), Posm by freezing point depression (Precision Systems, Inc., Sudbury, MA), and plasma creatinine concentration by the Jaffe reaction. PRA was measured by a modification of the RIA method of Haber et al. (11), using antibody and reagents provided by New England Nuclear (Billerica, MA), Pa by solid phase RIA performed on unextracted plasma using antibody-coated tubes supplied by Diagnostic Products Corp. (Los Angeles, CA), and Pc by RIA performed on unextracted plasma using antibody-coated assay tubes provided by Immuchem Corp. (Carson, CA). Details of our extraction technique and RIA procedure for determination of Pavp have been previously reported (12). Briefly, plasma was thawed in the morn-
Statistical methods
Results Hypertonic sodium chloride infusion studies Posm and Pavp before and at the end of hypertonic sodium chloride infusions in the studies on both patients, supine and erect, are shown in Table 1. In the studies performed while the patients were supine, Pavp increased in patient 1 to a postinfusion level that exceeded 2 SD of the mean postinfusion Pavp (group data) in other quadriplegic subjects. However, as shown in Fig. 1, Pavp during the infusions in both patients were otherwise well TABLE 1. Patient data compared to group data before (pre) and after (post) hypertonic sodium chloride infusion, supine and erect (sitting) Posm (mosmol/kg H2O) Supine Patient 1 Patient 2 Group Erect Patient 1 Patient 2 Group
Pavp (MU/mL)
Pre
Post
Pre
Post
279 274 284 ± 5.0
298 307 302 ± 5.4
0.20 0.06 0.23 ± 0.12
2.08" 1.36 1.30 ± 0.35
281 280 284 ± 4.3
299 297 302 ± 5.8
1.82° 2.27° 0.34 ± 0.19
6.41" 5.79° 2.03 ± 0.74
All values for group data are the mean ± SD. a More than 2 SD ± mean of group data.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 August 2016. at 22:21 For personal use only. No other uses without permission. . All rights reserved.
NONOSMOTIC STIMULI ALTER OSMOREGULATION
1539
SUPINE 65FIG. 1. The relationship of Pavp:Posm in both patients compared to that in previously studied quadriplegic subjects (Group) in the supine positure. , The lower limit of sensitivity of the RIA for Pavp.
E 3
X Patient 1 • Patient 2 O Group
4P1
QL
32-
260
270
280
290
300
310
320
POSM (mosm/kg H2O) within the range of Pavp observed in the group studies. Linear regression analysis of the data showed highly significant correlation between Pavp and Posm in both patients in the supine posture (r = 0.941; P < 0.002 for patient 1 and r = 0.950; P < 0.002 for patient 2). Slope values were 0.089 and 0.033 for patients 1 and 2, respectively. When compared to the collective slope values for group subjects, the slopes of the Pavp:Posm relationships in both patients were within 2 SD of the group mean (0.061 ± 0.02). The osmotic threshold was virtually the same in the two patients as in the group subjects, especially if the osmotic threshold is assumed to be at the level at which Pavp is detectable by the RIA (Fig. 1). In the studies performed while the patients were erect, Pavp both before and at the end of the infusions exceeded 2 SD of the mean pr$- and postinfusion Pavp in the comparable group. As shown in Fig. 2, there was a marked shift in the distribution of Pavp in both patients compared to the Pavp in the other subjects. Linear regression analysis of the data showed significant correlations between Pavp and Posm (r = 0.983; P < 0.001 for patient 1; r = 0.816; P < 0.05 for patient 2) in the erect posture. However, slope values (0.249 for patient 1 and 0.178 for patient 2) were higher than those determined by analysis of the Pavp:Posm relationships when the patients were supine. The slopes of these relationships in the erect posture also exceeded 2 SD of the mean slope of comparable data from the group studies (0.092 ± 0.03). The osmotic thresholds, as shown by extrapolation of the regression lines to the abscissa, were reduced in both patients compared to the abscissal intercept of the group data (Fig. 2).
Pa, Pc, PRA, mean arterial pressure (MAP), and HR before and at the end of the infusions, supine and erect, are shown in Table 2. Data from studies on patients 1 and 2 are compared to those derived from the studies on previously reported quadriplegic subjects. Pa, Pc, and PRA were higher in the erect than in the supine posture in all subjects. In both the supine and the erect posture, PRA was suppressed by hypertonic sodium chloride infusion in our previously reported normonatremic quadriplegic subjects. In contrast, patients 1 and 2 in the erect posture exhibited less suppression of PRA; postinfusion PRA was 2 SD higher than the group mean. In patient 2, pre- and postinfusion HR in the erect posture as well as postinfusion HR in the supine posture were also more than 2 SD higher than the mean values for the previously reported group. Water load studies performed on patient 1 Posm in patient 1 decreased during water loading from 286 to 270 mosmol/kg H2O in the supine posture and from 284 to 269 mosmol/kg H2O in the erect posture. Minimum Uosm and maximum free water clearance were achieved after 105 min of water loading in the supine posture and during the third hour of water loading in the erect posture. Data from these studies are shown in Table 3 together with observations from comparable studies on four quadriplegic subjects with no prior history of hyponatremia and three normal male control subjects. All of these data shown are derived from analysis of urine samples collected when Uosm was at its lowest level and free water clearance was maximal. These values in the
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 August 2016. at 22:21 For personal use only. No other uses without permission. . All rights reserved.
WILLIAMS ET AL.
1540
JCE & M • 1990 Vol71«No6
ERECT
7c.
/
X Patient 1 • Patient 2 O Group
ml)
5-
/
FIG. 2. The relationship of Pavp:Posm in both patients compared to that in previously studied quadriplegic subjects (Group) in the erect posture. - - -, The lower limit of sensitivity of the RIA for Pavp.
4-
•/A*
3 Q. >
3-
y
°o
Y*
V2 A 2-
/
PI
O
1-
cc
& ^
0-
260
270
280
290
300
310
320
POSM (mosm/kg H2O) TABLE 2. Patient data compared to group data before (pre) and after (post) hypertonic sodium chloride infusion, supine and erect (sitting) Pa (ng/dL)
Supine Patient 1 Patient 2 Group Erect Patient 1 Patient 2 Group
Pc (Mg/dL)
PRA (ng/mL-h)
MAP (mm Hg)
HR (beats/min)
Pre
Post
Pre
Post
Pre
Post
Pre
Post
Pre
Post
1.6 5.3
0.2 1.0
0.2 1.4
0.3 0.7
12.6
14.1
8.4
8.5
70 82
70 97
73 82
78 86°
6.9 ± 4.3
3.3 ± 1.6
0.8 ± 1.04
0.3 ± 0.31
10.3 ± 3.1
11.5 ± 3.5
87 ± 16.6
90 ± 15.5
62 ± 11
8.9
4.0 8.5
2.4 2.1
27.5 22.6 17.6 ± 7.0
112 117
1.7 ± 1.78
30.3 19.7 19.2 ± 7.0
77 106
5.9 ± 3.1
1.5° 1.7° 0.5 ± 0.46
92 ± 15.5
110 ± 10.1
10.2 19.4 ± 10.5
62 ± 11
75
76
100° 73 ± 11
100" 69 ± 11
All values for group data are the mean ± SD. " More than 2 SD ± mean of group data. TABLE 3. Data from water load studies performed on patient 1, four quadriplegic subjects with no prior history of hyponatremia, and three normal male control subjects Ccr
(mL/min) Supine Patient 1 Quad control Normal control Erect Patient 1 Quad control Normal control
Uosm (mosmol/kg H2O)
CH 2 O
(mL/min)
126
56
9.1
143 ± 18.8 143 ± 29.6
66 ± 32.8 78 ± 11.8
6.2 ± 2.3 10.8 ± 2.3
153
213
0.4
126 ± 18.3 152 ± 24.4
65 ± 14.5 66 ± 7.0
4.2 ± 1.1 10.0 ± 3.8
Cosm (mL/min)
UnaV (mmol/min)
FEna (%)
V/GFR (%)
MAP (mm Hg)
2.36 1.71 ± 0.36 3.78 ± 0.44
0.14 0.09 ± 0.02 0.25 ± 0.04
0.84 0.50 ± 0.12 1.30 ± 0.04
8.0
5.2 ± 1.9 9.5 ± 1.6
83 84 ± 6 98 ± 1 1
1.40 1.24 ± 0.17 2.86 ± 0.77
0.02 0.03 ± 0.01 0.14 ± 0.03
0.08 0.20 ± 0.07 0.69 ± 0.07
0.4
70
3.8 ± 1.2 7.6 ± 1.5
76 ±20 98 ± 17
Control values are the mean ± SD. Ccr, Creatinine clearance; CH2O, free water clearance; Cosm, osmolar clearance; UnaV, urinary sodium excretion. FEna, fractional excretion of sodium. All data are from periods of lowest Uosm and maximal CH,o-
supine posture were markedly different from those achieved when the patient was erect (sitting). Although creatinine clearance was not reduced in the erect posture, osmolar clearance, sodium excretion, fractional excretion
of sodium, and distal delivery of filtrate [vol/glomerular filtration rate (V/GFR) X 100] were all much lower than in the studies performed when the patient was supine. All of these parameters were also markedly reduced in
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 August 2016. at 22:21 For personal use only. No other uses without permission. . All rights reserved.
NONOSMOTIC STIMULI ALTER OSMOREGULATION
the erect posture compared to the levels demonstrated in the studies on erect frormal control subjects. Minimum Uosm was higher, and maximum free water clearance was lower in patient 1 in the erect posture than in both quadriplegic and normal control subjects. The other difference in patient 1 compared to quadriplegic as well as normal control subjects was a very low V/GFR. Pavp concentration before water loading in patient 1 were 0.23 /ilJ/mL in the supine posture and 0.47 juU/mL in the erect posture. P&vp before water loading in control subjects was 0.22 ± 0f06 /uU/mL in the supine posture and 0.29 ± 0.12 /LtU/rJnL in the erect posture. In both supine and erect postures (patient 1 and controls), Pavp was below the lower liifrit of sensitivity of our assay (0.18 fdJ/mL) when urinary dilution and free water clearance were maximal. MAP in patient 1 ranged from 70-95 mm Hg (median MAP, 80 mm Hg) in the supine posture and from 51-71 mm Hg (median MAP, 64 mm Hg) in the erect posture. Arm cuff measurements of blood pressure were used in the calculation of MA£ in these studies.
Discussion These studies on two quadriplegic patients who became severely hyponatremic during episodes of acute respiratory distress have sho^vn marked nonosmotic effects on vasopressin release related to erect (sitting) posture at times when both patients were clinically stable, without compromised respiratory function, and when plasma sodium concentrations and Posm were normal. Hypertonic sodium chloride infusion studies performed while the patients were supine showed that, with the exception of a higher postinfusion Pavp in one patient (patient 1), the Pavp:Posm relationship was similar to that shown in previously reported studies on normonatremic quadriplegic subjects (7). Linear regression analysis of these data did not show differences in either the threshold or the sensitivity of osmotically stimulated vasopressin release in either patient compared to these parameters for the PavpiPosm relationship in the previously reported group studies. In contrast, studies on both patients in the sitting posture revealed that Pavp was markedly increased at all levels Of Posm compared to Pavp over a comparable range of Posm in the other quadriplegic subjects. The slope of the Pavp:Posm relationship in both patients, as shown by linear regression, was also increased compared to the mean of collective slope values for the group subjects in the eirect posture. However, Pavp concentrations in both patients in erect posture were increased before hypertoftic sodium chloride infusion, and
1541
in the absence of observations at lower levels of Posm, the threshold and response characteristics of the Pavp:Posm relationship cannot be fully delineated. Studies on rats have shown previously that when Posm is increased in the presence of hypovolemia induced by sc injections of polyethylene glycol the effects of simultaneous osmotic and nonosmotic stimuli on vasopressin release are synergistic (6). In contrast, the increase in Pavp associated with polyethylene glycol-induced volume contraction was shown to be markedly attenuated by hypoosmolality induced by ad libitum water intake or intragastric water loading. Thus, although increased sensitivity of vasopressin release to osmotic stimulation in the erect posture is suggested by linear regression analysis of the data from the studies on both patients, the slope of these lines might be substantially altered by the inclusion of Pavp:Posm data pairs closer to a theoretical osmotic threshold for vasopressin release. Despite the lack of observations at lower levels of Posm, the Pavp:Posm relationship depicted by linear regression analysis is indicative of a reduced osmotic threshold for vasopressin release in both patients in the erect posture. Our observations on suppression of Pavp during the water load studies performed on patient 1 are also consistent with this assumption. In the studies performed when the patient was supine, suppression of Pavp to an unquantifiable level occurred when Posm was reduced from 286 to 281 mosmol/kg H2O after 15 min of water drinking. In the studies performed when the patient was erect, Pavp remained above this level until Posm was reduced from 284 to 272 mosmol/kg H2O after 45 min of water drinking. Although inhibition of vasopressin release by the pharyngeal reflex associated with water drinking cannot be excluded as a factor in the suppression of Pavp (15-17), the timing of water drinking and the volume of water ingested in the first 30 min of water loading were identical in the two studies. The most likely explanation for the observed shift in the relationship of Pavp:Posm in these two quadriplegic patients, as shown by low MAP in patient 1 and increased HR in the erect posture in patient 2, is relative hemodynamic instability. Thigh cuff measurements of arterial blood pressure, which were used during the hypertonic sodium chloride infusion studies to preserve forearm venous access, often do not reflect the magnitude of orthostatic hypotension in quadriplegic subjects in the sitting posture (7). Both patients, however, exhibited signs and symptoms of hemodynamic stress, i.e. diaphoresis and complaints of headache and/or impending syncope without nausea, before termination of the infusion studies. The ability of these patients to be out of bed in a wheelchair was also severely limited by the development of such symptoms. Orthostatic hypotension was more clearly demonstrated by arm cuff measurements of
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 August 2016. at 22:21 For personal use only. No other uses without permission. . All rights reserved.
1542
WILLIAMS ET AL.
JCE & M • 1990 Vol 71 • No 6
arterial blood pressure during the water load studies performed on patient 1. Median MAP was 64 mm Hg in the erect posture compared to 80 mm Hg in the supine posture. High Pc concentrations in patient 1 and less suppression of PRA by hypertonic sodium chloride infusions in both patients in the erect posture also suggest that hemodynamic stress induced by orthostasis was greater in these two patients than in our previously studied subjects. The effect of altered hemodynamic status on cortisol secretion is mediated by cardiac and arterial baroreflex mechanisms that modulate the release of CRF and are also involved in the nonosmotic control of vasopressin release (18,19). The neural afferent component of these volume- and pressure-sensitive mechanisms is
Reductions in blood volume of approximately 6% were associated with increased osmoreceptor sensitivity, whereas larger reductions in blood volume, i.e. approximately 14%, resulted in reductions in the osmotic threshold for vasopressin release as well as further increases in osmoreceptor sensitivity. Thus, in human as well as animal subjects the effect of nonosmotic stimuli on osmoregulation of vasopressin release seems to vary in direct proportion to the potency of the nonosmotic stimuli. Studies of the renal response to water loading in patient 1 revealed marked impairment of urine-diluting ability and reduced free water clearance in the erect (sitting) posture which were not evident in the study performed while the patient was supine. Although
intact in quadriplegic subjects and is capable of eliciting
suppression of Pavp by water loading was delayed in the
erect posture compared to the prompt reduction in Pavp cortisol and vasopressin responses to orthostatic changes to unquantifiable levels in the supine posture, this does in arterial blood pressure and/or blood volume distribunot account for the impairment of urine-diluting ability tion. It has also been reported that CRF stimulates shown by the study of this patient in the erect posture. vasopressin release and that it may act synergistically to Minimum Uosm was 213 mosmol/kg H2O in the third augment the vasopressin response to osmotic stimulation hour of water loading, whereas Pavp was suppressed to (19). Similar observations regarding an effect of angioan unquantifiable level after 45 min of water drinking. tensin-II on vasopressin release have been reported (20, Osmolar clearance was reduced, and both absolute and 21). Thus, several mechanisms may contribute to nonfractional sodium excretion were quite low in the erect osmotic stimulation of vasopressin release during periods of hemodynamic stress, all of which may be implicated posture. These observations are consistent with the posin the increase in Pavp and the shift in its relationship sibility that low free water clearance in the erect posture to Posm shown by our studies on both patients in the was due to increased proximal sodium chloride reabsorperect posture. tion and decreased distal delivery of filtrate to diluting Few studies of the interaction between osmotic and segments of the renal tubules, as shown by low V/GFR nonosmotic control of vasopressin release in human sub(23). Although GFR was not reduced, it seems likely that jects have been reported. Robertson and Athar (3) re- such marked alterations in sodium and water handling ported that upright posture in hydropenic normal subby the kidneys were a reflection of reduced renal blood jects resulted in a small but significant lowering of the flow due to orthostatic hypotension in erect posture. osmotic threshold for vasopressin release, but did not Similar observations on the effect of posture on renal alter osmoreceptor sensitivity, as defined by the slope of function in quadriplegic subjects were reported by the linear relationship between Pavp and Posm. EnKooner et al. (24). Although Uosm was not reduced by hancement of osmotic stimulation of vasopressin release water loading to levels observed in studies on normal by reductions in central venous pressure induced by lower control subjects in either the sitting or supine posture in body negative pressure without changes in MAP was these studies, urine flow rate, free water clearance, and reported by Goldsmith et al. (22). In our recently reported MAP increased, as in our studies on patient 1, when the studies on stable quadriplegic subjects, modest reducquadriplegic subjects were supine. This ability to increase tions in MAP in the sitting posture were associated with free water clearance in the supine posture by improveincreased sensitivity of vasopressin release to concomiment in renal water excretion due to intrinsic renal tant osmotic stimulation without a change in the threshmechanisms and elimination of posture-related nonosold for stimulation (7). The more severe hemodynamic motic effects on vasopressin release could explain the stress, as shown by orthostatic reductions in blood presabsence of hyponatremia when the studies on our two sure, experienced by the the two patients described in patients were performed. Both patients were out of bed this report was associated with marked alterations in in a wheelchair for only brief periods of time during the both the threshold and sensitivity of vasopressin release day because of symptoms related to orthostatic hypotenin response to osmotic stimulation. These observations sion. The relatively long periods of recumbency may have are similar to those of Dunn et al. (4) in studies on rats, been sufficient to allow water retained in the sitting in which blood volume was reduced by ip injections of posture to be excreted, thus preventing the reduction in polyethylene glycol in the presence of increased Posm. serum sodium concentration that might otherwise occur.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 August 2016. at 22:21 For personal use only. No other uses without permission. . All rights reserved.
NONOSMOTIC STIMULI ALTER OSMOREGULATION
The development of severe hyponatremia during episodes of acute respiratory distress can be attributed to the superimposition of additional potent nonosmotic stimulation of vasopre$sin release, as shown by the absence of signs of volume depletion in both patients and elevated Pavp (6.7 jiU/mL) in patient 2. Hypoxemia and hypercapnic acidosis, both of which were present during the episode of acute respiratory distress in patient 1, have been shown to fesult in diminished free water clearance, marked elevation of Pavp, and reduced renal blood flow and GFR in studies on dogs (25-27) and humans (28). Elevation of Pavp was also shown in studies on patients with acute bronchial asthma reported by Baker et al. (29). In these studies increased Pavp was correlated with reductions in forced expiratory volume. Patient 2 was treated with intermittent positive pressure breathing and subsequently developed respiratory stridor before the demonstration of severe hyponatremia in association with markedly increased Pavp. Thus, conditions known to be associated with nonosmotic stimulation of vasopressin release existed during the episodes of acute respiratory distress in both patients. Whether hemodynamic instability, which seems to be the most likely explanation for enhancement of vasopressin release and marked impairment of urine-diluting ability in erect posture when th$ patients were clinically stable, may also have contributed to the development of severe hyponatremia during episodes of acute respiratory distress is uncertain. However, it may be postulated that recurrent nonosmotic stimulation of vasopressin release by the hemodynamic effects of orthostasis may increase the susceptibility of some quadriplegic individuals to the effects of other potent nonosmotic stimuli that affect vasopressin release.
7. 8. 9. 10.
11.
12. 13. 14. 15. 16. 17. 18. 19. 20. 21.
Acknowledgments
22.
We wish to thank Benjamin A. Moeller, M.D., and the staff of the Spinal Cord Injury Unit, V.A, Medical Center (Memphis, TN), for their continued support and cooperation in the performance of these studies. We thank Mary Alice Bobal for excellent technical assistance.
23. 24.
References 1. Robertson GL, Shelton RL, Athar S. The osmoregulation of vasopressin. Kidney Int. 197$;10:25-37. 2. Robertson GL. The regulation of vasopressin function in health and disease. Recent Prog Horm Res. 1977;23:333-85. 3. Robertson GL, Athar S. The interaction of blood osmolality and blood volume in regulating plasma vasopressin in man. J Clin Endocrinol Metab. 1976;42:613-20. 4. Dunn FL, Brennan TJ, Nelson AE, Robertson GL. The role of blood osmolality and volume in regulating vasopressin secretion in the rat. J Clin Invest. 1973;52:3212-9. 5. Quillen Jr EW, Cowley Jr AW. Influence of volume changes on osmolality-vasopressin relationships in conscious dogs. Am J Physiol. 1983;244:H73-9. 6. Strieker EM, Verbalis JG. Interaction of osmotic and volume
25. 26. 27.
28. 29.
1543
stimuli in regulation of neurohypophyseal secretion in rats. Am J Physiol. 1986;250:R267-75. Wall BM, Williams HH, Presley DN, et al. Altered sensitivity of osmotically stimulated vasopressin release in quadriplegic subjects. Am J Physiol. 1990;258:R827-35. Mathias CJ, Frankel HL. Autonomic failure in tetraplegia. In: Bannister R, ed. Autonomic failure. Oxford: Oxford University Press; 1984;454-88. Debarge O, Christensen NJ, Corbett JL, Eidelman BH, Frankel HL, Mathias CJ. Plasma catecholamines in tetraplegics. Paraplegia. 1974;12:44-9. Mathias CJ, Christensen NJ, Corbett JL, Frankel HL, Goodwin TJ, Peart WS. Plasma catecholamines, plasma renin activity and plasma aldosterone in tetraplegic man, horizontal and tilted. Clin Sci Mol Med. 1975;49:291-9. Haber E, Koerner T, Page LB, Kliman B, Purnode A. Application of a radioimmunoassay for angiotensin I to the physiologic measurements of plasma renin activity in normal human subjects. J Clin Endocrinol Metab. 1969;29:1349-55. Crofton JT, Dustan H, Share L, Brooks DP. Vasopressin secretion in normotensive black and white men and women on normal and low sodium diets. J Endocrinol. 1986;108:191-9. Armitage P, Berry G. Statistical methods in medical research, 2nd ed. Oxford: Blackwell; 1987;273-82. Baylis PH. Osmoregulation and control of vasopressin secretion in healthy humans. Am J Physiol. 1987;253:R671-8. Geelen G, Keil LC, Kravik SE, et al. Inhibition of plasma vasopressin after drinking in dehydrated humans. Am J Physiol. 1984;247:R968-71. Salata RA, Verbalis JG, Robinson AG. Cold water stimulation of oropharyngeal receptors in man inhibits release of vasopressin. J Clin Endocrinol Metab. 1987;65:561-7. Thompson CJ, Burd JM, Baylis PH. Acute suppression of plasma vasopressin and thirst after drinking in hypernatremic humans. Am J Physiol. 1987;252:R1138-42. Raff H, Merrill D, Skelton M, Cowley Jr AW. Control of ACTH and vasopressin in neurohypophysectomized conscious dogs. Am J Physiol. 1985;249:R281-4. Raff H, Skelton MM, Merrill DC, Cowley Jr AW. Vasopressin responses to corticotropin releasing factor and hyperosmolality in conscious dogs. Am J Physiol. 1986;251:R1235-9. Ramsay DJ, Keil LC, Sharpe MC, Shinsako J. Angiotensin II infusion increases vasopressin, ACTH, and 11-hydroxycorticosteroid secretion. Am J Physiol. 1978;234:R66-71. Klingbeil CK, Keil LC, Chang D, Reid IA. Effects of CRF and ANG II on ACTH and vasopressin release in conscious dogs. Am J Physiol. 1988;255:E46-53. Goldsmith SR, Dodge D, Cowley AW. Nonosmotic influences on osmotic stimulation of vasopressin in humans. Am J Physiol. 1987;252:H85-8. Berliner RW, Davidson DG. Production of hypertonic urine in the absence of pituitary antidiuretic hormone. J Clin Invest. 1957;36:1416-27. Kooner JS, Frankel HL, Mirando N, Peart WS, Mathias CJ. Haemodynamic, hormonal and urinary responses to postural change in tetraplegic and paraplegic man. Paraplegia. 1988;26:2337. Rose Jr CE, Anderson RJ, Carey RM. Antidiuresis and vasopressin release with hypoxemia and hypercapnia in conscious dogs. Am J Physiol. 1984;247:R127-34. Rose Jr CE, Kimmel DP, Godine RL, Kaiser DL, Carey RM. Synergistic effects of acute hypoxemia and hypercapnic acidosis in conscious dogs. Circ Res. 1983;53:202-13. Rose Jr CE, Dixon BS, Anderson RJ. Effects of hypoxemia and hypercapnic acidosis on renal water excretion and vasopressin secretion. In: Schrier RW, ed. Vasopressin. New York: Raven Press; 1985;517-23. Kilburn KH, Dowell AR. Renal function in respiratory failure. Arch Intern Med. 1971;127:754-62. Baker JW, Yerger S, Segar WE. Elevated plasma antidiuretic hormone levels in status asthmaticus. Mayo Clinic Proc. l976;51:31-4.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 August 2016. at 22:21 For personal use only. No other uses without permission. . All rights reserved.
Vol. 71, No. 6 Printed in U.S.A.
Nonosmotic Stimuli Alter Osmoregulation in Patients with Spinal Cord Injury* HUGH H. WILLIAMS, BARRY M. WALL, JEANNE M. HORAN, DANIEL N. PRESLEY, JOAN T. CROFTON, LEONARD SHARE, AND C. ROBERT COOKE Veterans Affairs Medical Center and the University of Tennessee, Memphis, Tennessee 38104
ABSTRACT. Studies on two quadriplegic patients who developed severe hyponatremia during episodes of acute respiratory distress were performed to determine whether differences in osmoregulation of vasopressin release could be identified in these patients compared to other quadriplegic subjects previously studied in a similar manner. Both patients were clinically stable and normonatremic, with no signs or symptoms of respiratory distress, when the studies were performed. However, both exhibited evidence of hemodynamic instability in the sitting posture. Linear regression analysis of the plasma vasopressin/plasma osmolality (Pavp:Posm) relationship during infusions of 0.85 M sodium chloride showed no significant differences in either the slope (sensitivity) or abscissal intercept (osmotic threshold) of this relationship compared to that of other quadriplegic subjects when the patients were supine. In contrast, when the patients were studied in the sitting posture there was a marked shift in the relationship of Pavp:Posm indicative of increased sensitivity and reduced osmotic threshold for vasopressin release. The
U
NDER most physiological conditions vasopressin release is primarily responsive to osmotic stimulation (1, 2). However, nonosmotic stimuli, especially those associated with hypotension and/or hypovolemia, are also capable of eliciting a vasopressin response and can modify the relationship between plasma vasopressin (Pavp) and plasma osmolality (Posm) without disrupting their characteristically close association (3-7). In recently reported studies on subjects with cervical spinal cord injury, changes in the Pavp:Posm relationship consistent with increased sensitivity of vasopressin release to osmotic stimulation were observed in association with the assumption of erect (sitting) posture (7). The circumscribed neurological deficit in such individuals provides a unique opportunity to study the interaction between osmotic and nonosmotic control of vasopressin release in human subjects. The afferent neural pathways inReceived April 25, 1990. Address all correspondence and requests for reprints to: Hugh H. Williams, M.D., Nephrology Section, Veterans Administration Medical Center (111B), 1030 Jefferson Avenue, Memphis, Tennessee 38104. * This work was supported by V.A. Merit Review funding and in part by USPHS Research Grant HL-19209 from the NHLBI.
slopes of the Pavp:Posm relationships were 0.249 and 0.178 for
the two patients, respectively, compared to 0.092 ± 0.03 (±SD) for previously studied quadriplegic subjects. Oral water-loading studies performed on one patient revealed marked impairment of urine-diluting ability and free water clearance in the sitting posture compared with observations in similar studies performed when the patient was supine. Impairment of renal water excretion could not be attributed to an effect of vasopressin, which was reduced to unquantifiable levels by water loading. These studies have shown that hemodynamic stress related to autonomic dysfunction in quadriplegic patients may result in marked alteration of osmoregulation of vasopressin release in more severely affected individuals. Such altered osmoregulation, which may also be associated with vasopressin-independent impairment of renal water excretion in the sitting posture, may be a predisposing factor in the development of hyponatremia, especially in the presence of other potent nonosmotic stimuli. (J Clin Endocrinol Metab 7 1 : 1536-1543,1990)
volved in the modulation of vasopressin release are intact in quadriplegic subjects, while the efferent pathways from the central nervous system that are normally involved in the maintenance of cardiovascular homeostasis in erect posture are defective (8-10). This defect in the efferent limb of the baroreflex arc is responsible for the occurrence of orthostatic hemodynamic instability in quadriplegia. Such instability may be a source of nonosmotic stimulation of vasopressin release via the aforementioned afferent neural pathways, but the extent to which these nonosmotic influences may modify the relationship between Pavp and Posm has not been fully delineated. We have observed two quadriplegic patients whose ability to remain out of bed in a wheelchair was markedly affected by hemodynamic instability in the sitting posture, and who became severely hyponatremic during episodes of acute respiratory distress. To determine whether abnormalities in osmoregulation and water handling could be identified as possible predisposing factors in the development of hyponatremia, the interaction between osmotic and nonosmotic control of va-
1536
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 August 2016. at 22:21 For personal use only. No other uses without permission. . All rights reserved.
NONOSMOTIC STIMULI ALTER OSMOREGULATION
sopressin release was investigated when the patients were clinically stable, normonatremic, and showing no signs or symptoms of respiratory distress. Our observations in these studies suggest that hemodynamic instability, as shown by orthostatic hypotension, results in potent nonosmotic stimulation of vasopressin release and marked alteration of vasopressin responsiveness to osmotic stimulation.
Materials and Methods Clinical observations durifig severe hyponatremia Patient 1, a 60-yr-old white male with quadriplegia due to spinal cord injury at the C-5 level in 1949, was admitted to the V.A. Medical Center with a history of lethargy, weakness, and altered mental state of 2-day duration. Findings on x-ray examination of the chest Were consistent with right lower lobe pneumonia. Admission laboratory studies revealed the following results: serum sodiunj concentration, 113 mmol/L; potasssium, 3.1 mmol/L; chloride, 83 mmol/L; bicarbonate, 24 mmol/ L; blood urea nitrogen, 1.3 mmol/L urea; and glucose, 6.9 mmol/ L. Arterial blood pH wa$ 7.45, pCO2 was 28 mm Hg, and pO2 was 53 mm Hg. He received respiratory therapy and antibiotics as treatment for pneumonia. The following morning his serum sodium concentration was 107 mmol/L, potassium Was 2.3 mmol/L, chloride was 81 mmol/ L, and bicarbonate was 23 mmol/L. Blood urea nitrogen concentration was 1.1 mmol/L urea. Serum osmolality was 211 mosmol/kg H2O, and urine osmolality (Uosm) was 568 mosmol/ kg H2O. Urinary sodiufrn concentration was 112 mmol/L. Plasma cortisol concentration and thyroid function test results were normal. The patient was treated with iv infusion of 0.51 M sodium chloride and furosemide, followed by continuous infusion of 0.153 M sodium chloride and supplemental potassium chloride. Subsequently, his serum electrolytes were: sodium, 124 mmol/L; potassium, 2.9 mmol/L; chloride, 90 mmol/ L; and bicarbonate, 26 mmol/L. There was some improvement in his mental status, but respiratory distress persisted. Arterial blood pH 28 h after admission was 7.13, pCO2 was 84 mm Hg, and pO2 was 81 mm Hg. Endotracheal intubation was performed, and mechanical ventilation was required over the next 3 days. There was gradual improvement in his pulmonary status, and at the time of extubation his serum sodium concentration was 139 mmol/L. Patient 2, a 69-yr-old white male with quadriplegia due to spinal cord injury at the C-4-5 level 11 yr previously, was admitted to the V.A. Medical Center for investigation of nephrolithiasis. Intravenous pyelography revealed a staghorn calculus and multiple smaller renal stones within the left kidney in addition to a calculus in the bladder. The bladder calculus was removed cystoscopically, and a left percutaneous nephrolithotomy was performed under general anesthesia. He tolerated the procedure well,, but in the postoperative period he developed pulmonary congestion, a nonproductive cough, and dyspnea. Chest x-ray examination showed left lower lobe atelectasis. Four days postoperatively, the serum sodium concentration was 135 mmol/L, chloride was 92 mmol/L, and bicarbonate
1537
was 33 mmol/L. The serum glucose concentration was 5.2 mmol/L, blood urea nitrogen was 2.5 mmol/L urea, and serum creatinine was 62 /rnnol/L. Dyspnea persisted despite intermittent postive pressure breathing, chest percussion, postural drainage, and therapy with bronchodilators. Three days later respiratory stridor was noted. Laboratory studies at this time showed a marked reduction in the serum sodium concentration to 112 mmol/L. The serum potassium concentration was 3.5 mmol/L, chloride was 83 mmol/L, bicarbonate was 25 mmol/ L, glucose was 6.9 mmol/L, blood urea nitrogen was 5.3 mmol/ L urea, and serum creatinine was 53 ^mol/L. Arterial blood pH was 7.52, pCO2 was 36 mm Hg, and pO2 was 136 mm Hg. The urinary sodium concentration was 60 mmol/L, Uosm was 340 mosmol/kg H2O, and serum osmolality was 236 mosmol/kg H2O. Water intake was restricted, and the patient was observed closely. The following day, he was treated with an iv infusion of 0.85 M sodium chloride, receiving a total of 270 mL over 90 min. The serum sodium concentration and serum osmolality increased from 109 mmol/L and 236 mosmol/kg H2O, respectively, to 127 mmol/L and 271 mosmol/kg H2O. The plasma arginine vasopressin concentration before hypertonic sodium chloride infusion was 6.7 ^U/mL and increased further during the infusion to 15.5 jiU/mL. Respiratory stridor persisted, endotracheal intubation and mechanical ventilation were required, and tracheostomy was subsequently performed. With water restriction and improvement in the patient's respiratory status, his serum sodium concentration and plasma osmolality became normal and remained so for the rest of his hospital
Study protocol Serum sodium concentrations were normal in both patients when these studies were performed, and neither patient exhibited signs of respiratory distress, fluid retention, or dehydration. They had been eating a normal diet without restriction of fluid intake and had been receiving no medications known to affect water handling. Both refrained from the use of tobacco or ethanol for at least 16 h before the studies. The studies were approved by the Human Studies Subcommittee and the Research and Development Committee of the V.A. Medical Center, and informed consent was obtained in writing before the studies were performed. Hypertonic sodium chloride infusion studies The patients were studied on two separate occasions, once while supine and once while sitting quietly in a wheelchair. All studies began at approximately 0900 h, after an overnight fast. On the morning of each study, catheters were inserted into antecubital or forearm veins of both arms, and blood pressure and heart rate (HR) were measured at 3-min intervals over the next 30 min using a thigh cuff and an automated recording device (Dinamap, Critikon, Inc., Tampa, FL). At the end of this period of equilibration, blood samples were obtained for determinations of Posm, plasma sodium (Pna), plasma aldosterone (Pa), plasma cortisol (Pc), Pavp, total protein concentrations, and PRA. Intravenous infusions of 0.85 M sodium chlo-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 August 2016. at 22:21 For personal use only. No other uses without permission. . All rights reserved.
JCE & M • 1990 Vol 71 • No 6
WILLIAMS ET AL.
1538
ride (0.05 mL/kg-min) were then begun and continued for 90 min. Blood pressure and heart rate were measured every 3 min during the infusions, and blood samples were obtained every 15 min for determinations of Posm, Pna, Pavp, and total protein. Final blood samples were obtained at the end of the infusions for determinations comparable to those performed on preinfusion blood samples. The total volume of all blood samples in each study was approximately 120 mL. The protocol followed in these studies was identical to that employed in studies on 15 other quadriplegic subjects who had not previously been hyponatremic. Data from the studies on these subjects were previously reported (7) and are presented only for comparison with the observations in the studies on the 2 patients who provide the basis for this report. The 15 normonatremic quadriplegic subjects were all males, ranging in age from 28-72 yr.
ing on which extractions were performed. Arginine vasopressin was extracted with octadecylsilane (Sep-Pak C18 cartridges, Waters Associates, Milford, MA). The limit of detection of the assay, defined as the smallest quantity of hormone statistically separable from zero, is 0.18 /xU/mL. Pavp concentrations are at or below this level during sustained water diuresis in normal subjects (our unpublished data). USP posterior pituitary reference standard (U.S. Pharmacopeial Convention, Inc., Rockville, MD) was used as the vasopressin standard. The average recovery of arginine vasopressin was 86.8 ± 1.1% (n = 57 samples from 11 assays). The interassay coefficient of variation was 9.3%; the within-assay coefficient of variation was 5.4 ± 1.2%. Plasma values were not corrected for incomplete recovery.
Water load studies performed on patient 1
The relationship of Pavp:Posm in each patient, supine and sitting, was evaluated by linear regression analysis (13, 14). Slope values were compared to the mean regression slopes of data derived from our previously reported studies on quadriplegic subjects (7). Values for all data, including the Pavp:Posm slope values, from the hypertonic sodium chloride infusion studies were considered to be significantly different from the previously reported group data if they were outside 2 SD of the group means. Data from water load studies on patient 1 were compared to observations obtained in similar studies on four quadriplegic subjects with no prior history of hyponatremia and three normal control subjects.
The renal response to oral water loading was assessed in studies on patient 1 in both the supine and the sitting posture. On both occasions, a standard water load (20 mL/kg BW) was administered orally over a period of 30 min, followed by additional hourly water intake equal to the previous hour's urine output over a total period of 3 h. Blood samples for determinations of Posm, Pna, Pavp, and plasma creatinine concentration were drawn at the beginning of water loading, at 15-min intervals during the first and second hours, and at the end of the third hour of the studies. Urine samples were collected from an indwelling bladder catheter at 15-min intervals during the first and second hours and at the end of the third hour for measurements of urine volume, osmolality, sodium, chloride, and creatinine concentrations. Blood pressure and HR were measured every 15 min using an arm cuff and an automated recording device. Comparable studies were performed on four quadriplegic subjects with no prior history of hyponatremia and three normal male control subjects. Analytical methods Blood for determinations of PRA and Pavp was drawn into prechilled plastic syringes containing EDTA (500 nL EDTA/ 10 mL blood). Blood for other determinations was collected into heparinized prechilled plastic syringes. Samples were kept in ice until centrifuged at 4 C. The separated plasma was stored at -20 C until the analyses could be performed. Pna was measured by ion-selective electrode (Beckman Instruments, Inc., Brea, CA), Posm by freezing point depression (Precision Systems, Inc., Sudbury, MA), and plasma creatinine concentration by the Jaffe reaction. PRA was measured by a modification of the RIA method of Haber et al. (11), using antibody and reagents provided by New England Nuclear (Billerica, MA), Pa by solid phase RIA performed on unextracted plasma using antibody-coated tubes supplied by Diagnostic Products Corp. (Los Angeles, CA), and Pc by RIA performed on unextracted plasma using antibody-coated assay tubes provided by Immuchem Corp. (Carson, CA). Details of our extraction technique and RIA procedure for determination of Pavp have been previously reported (12). Briefly, plasma was thawed in the morn-
Statistical methods
Results Hypertonic sodium chloride infusion studies Posm and Pavp before and at the end of hypertonic sodium chloride infusions in the studies on both patients, supine and erect, are shown in Table 1. In the studies performed while the patients were supine, Pavp increased in patient 1 to a postinfusion level that exceeded 2 SD of the mean postinfusion Pavp (group data) in other quadriplegic subjects. However, as shown in Fig. 1, Pavp during the infusions in both patients were otherwise well TABLE 1. Patient data compared to group data before (pre) and after (post) hypertonic sodium chloride infusion, supine and erect (sitting) Posm (mosmol/kg H2O) Supine Patient 1 Patient 2 Group Erect Patient 1 Patient 2 Group
Pavp (MU/mL)
Pre
Post
Pre
Post
279 274 284 ± 5.0
298 307 302 ± 5.4
0.20 0.06 0.23 ± 0.12
2.08" 1.36 1.30 ± 0.35
281 280 284 ± 4.3
299 297 302 ± 5.8
1.82° 2.27° 0.34 ± 0.19
6.41" 5.79° 2.03 ± 0.74
All values for group data are the mean ± SD. a More than 2 SD ± mean of group data.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 August 2016. at 22:21 For personal use only. No other uses without permission. . All rights reserved.
NONOSMOTIC STIMULI ALTER OSMOREGULATION
1539
SUPINE 65FIG. 1. The relationship of Pavp:Posm in both patients compared to that in previously studied quadriplegic subjects (Group) in the supine positure. , The lower limit of sensitivity of the RIA for Pavp.
E 3
X Patient 1 • Patient 2 O Group
4P1
QL
32-
260
270
280
290
300
310
320
POSM (mosm/kg H2O) within the range of Pavp observed in the group studies. Linear regression analysis of the data showed highly significant correlation between Pavp and Posm in both patients in the supine posture (r = 0.941; P < 0.002 for patient 1 and r = 0.950; P < 0.002 for patient 2). Slope values were 0.089 and 0.033 for patients 1 and 2, respectively. When compared to the collective slope values for group subjects, the slopes of the Pavp:Posm relationships in both patients were within 2 SD of the group mean (0.061 ± 0.02). The osmotic threshold was virtually the same in the two patients as in the group subjects, especially if the osmotic threshold is assumed to be at the level at which Pavp is detectable by the RIA (Fig. 1). In the studies performed while the patients were erect, Pavp both before and at the end of the infusions exceeded 2 SD of the mean pr$- and postinfusion Pavp in the comparable group. As shown in Fig. 2, there was a marked shift in the distribution of Pavp in both patients compared to the Pavp in the other subjects. Linear regression analysis of the data showed significant correlations between Pavp and Posm (r = 0.983; P < 0.001 for patient 1; r = 0.816; P < 0.05 for patient 2) in the erect posture. However, slope values (0.249 for patient 1 and 0.178 for patient 2) were higher than those determined by analysis of the Pavp:Posm relationships when the patients were supine. The slopes of these relationships in the erect posture also exceeded 2 SD of the mean slope of comparable data from the group studies (0.092 ± 0.03). The osmotic thresholds, as shown by extrapolation of the regression lines to the abscissa, were reduced in both patients compared to the abscissal intercept of the group data (Fig. 2).
Pa, Pc, PRA, mean arterial pressure (MAP), and HR before and at the end of the infusions, supine and erect, are shown in Table 2. Data from studies on patients 1 and 2 are compared to those derived from the studies on previously reported quadriplegic subjects. Pa, Pc, and PRA were higher in the erect than in the supine posture in all subjects. In both the supine and the erect posture, PRA was suppressed by hypertonic sodium chloride infusion in our previously reported normonatremic quadriplegic subjects. In contrast, patients 1 and 2 in the erect posture exhibited less suppression of PRA; postinfusion PRA was 2 SD higher than the group mean. In patient 2, pre- and postinfusion HR in the erect posture as well as postinfusion HR in the supine posture were also more than 2 SD higher than the mean values for the previously reported group. Water load studies performed on patient 1 Posm in patient 1 decreased during water loading from 286 to 270 mosmol/kg H2O in the supine posture and from 284 to 269 mosmol/kg H2O in the erect posture. Minimum Uosm and maximum free water clearance were achieved after 105 min of water loading in the supine posture and during the third hour of water loading in the erect posture. Data from these studies are shown in Table 3 together with observations from comparable studies on four quadriplegic subjects with no prior history of hyponatremia and three normal male control subjects. All of these data shown are derived from analysis of urine samples collected when Uosm was at its lowest level and free water clearance was maximal. These values in the
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 August 2016. at 22:21 For personal use only. No other uses without permission. . All rights reserved.
WILLIAMS ET AL.
1540
JCE & M • 1990 Vol71«No6
ERECT
7c.
/
X Patient 1 • Patient 2 O Group
ml)
5-
/
FIG. 2. The relationship of Pavp:Posm in both patients compared to that in previously studied quadriplegic subjects (Group) in the erect posture. - - -, The lower limit of sensitivity of the RIA for Pavp.
4-
•/A*
3 Q. >
3-
y
°o
Y*
V2 A 2-
/
PI
O
1-
cc
& ^
0-
260
270
280
290
300
310
320
POSM (mosm/kg H2O) TABLE 2. Patient data compared to group data before (pre) and after (post) hypertonic sodium chloride infusion, supine and erect (sitting) Pa (ng/dL)
Supine Patient 1 Patient 2 Group Erect Patient 1 Patient 2 Group
Pc (Mg/dL)
PRA (ng/mL-h)
MAP (mm Hg)
HR (beats/min)
Pre
Post
Pre
Post
Pre
Post
Pre
Post
Pre
Post
1.6 5.3
0.2 1.0
0.2 1.4
0.3 0.7
12.6
14.1
8.4
8.5
70 82
70 97
73 82
78 86°
6.9 ± 4.3
3.3 ± 1.6
0.8 ± 1.04
0.3 ± 0.31
10.3 ± 3.1
11.5 ± 3.5
87 ± 16.6
90 ± 15.5
62 ± 11
8.9
4.0 8.5
2.4 2.1
27.5 22.6 17.6 ± 7.0
112 117
1.7 ± 1.78
30.3 19.7 19.2 ± 7.0
77 106
5.9 ± 3.1
1.5° 1.7° 0.5 ± 0.46
92 ± 15.5
110 ± 10.1
10.2 19.4 ± 10.5
62 ± 11
75
76
100° 73 ± 11
100" 69 ± 11
All values for group data are the mean ± SD. " More than 2 SD ± mean of group data. TABLE 3. Data from water load studies performed on patient 1, four quadriplegic subjects with no prior history of hyponatremia, and three normal male control subjects Ccr
(mL/min) Supine Patient 1 Quad control Normal control Erect Patient 1 Quad control Normal control
Uosm (mosmol/kg H2O)
CH 2 O
(mL/min)
126
56
9.1
143 ± 18.8 143 ± 29.6
66 ± 32.8 78 ± 11.8
6.2 ± 2.3 10.8 ± 2.3
153
213
0.4
126 ± 18.3 152 ± 24.4
65 ± 14.5 66 ± 7.0
4.2 ± 1.1 10.0 ± 3.8
Cosm (mL/min)
UnaV (mmol/min)
FEna (%)
V/GFR (%)
MAP (mm Hg)
2.36 1.71 ± 0.36 3.78 ± 0.44
0.14 0.09 ± 0.02 0.25 ± 0.04
0.84 0.50 ± 0.12 1.30 ± 0.04
8.0
5.2 ± 1.9 9.5 ± 1.6
83 84 ± 6 98 ± 1 1
1.40 1.24 ± 0.17 2.86 ± 0.77
0.02 0.03 ± 0.01 0.14 ± 0.03
0.08 0.20 ± 0.07 0.69 ± 0.07
0.4
70
3.8 ± 1.2 7.6 ± 1.5
76 ±20 98 ± 17
Control values are the mean ± SD. Ccr, Creatinine clearance; CH2O, free water clearance; Cosm, osmolar clearance; UnaV, urinary sodium excretion. FEna, fractional excretion of sodium. All data are from periods of lowest Uosm and maximal CH,o-
supine posture were markedly different from those achieved when the patient was erect (sitting). Although creatinine clearance was not reduced in the erect posture, osmolar clearance, sodium excretion, fractional excretion
of sodium, and distal delivery of filtrate [vol/glomerular filtration rate (V/GFR) X 100] were all much lower than in the studies performed when the patient was supine. All of these parameters were also markedly reduced in
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 August 2016. at 22:21 For personal use only. No other uses without permission. . All rights reserved.
NONOSMOTIC STIMULI ALTER OSMOREGULATION
the erect posture compared to the levels demonstrated in the studies on erect frormal control subjects. Minimum Uosm was higher, and maximum free water clearance was lower in patient 1 in the erect posture than in both quadriplegic and normal control subjects. The other difference in patient 1 compared to quadriplegic as well as normal control subjects was a very low V/GFR. Pavp concentration before water loading in patient 1 were 0.23 /ilJ/mL in the supine posture and 0.47 juU/mL in the erect posture. P&vp before water loading in control subjects was 0.22 ± 0f06 /uU/mL in the supine posture and 0.29 ± 0.12 /LtU/rJnL in the erect posture. In both supine and erect postures (patient 1 and controls), Pavp was below the lower liifrit of sensitivity of our assay (0.18 fdJ/mL) when urinary dilution and free water clearance were maximal. MAP in patient 1 ranged from 70-95 mm Hg (median MAP, 80 mm Hg) in the supine posture and from 51-71 mm Hg (median MAP, 64 mm Hg) in the erect posture. Arm cuff measurements of blood pressure were used in the calculation of MA£ in these studies.
Discussion These studies on two quadriplegic patients who became severely hyponatremic during episodes of acute respiratory distress have sho^vn marked nonosmotic effects on vasopressin release related to erect (sitting) posture at times when both patients were clinically stable, without compromised respiratory function, and when plasma sodium concentrations and Posm were normal. Hypertonic sodium chloride infusion studies performed while the patients were supine showed that, with the exception of a higher postinfusion Pavp in one patient (patient 1), the Pavp:Posm relationship was similar to that shown in previously reported studies on normonatremic quadriplegic subjects (7). Linear regression analysis of these data did not show differences in either the threshold or the sensitivity of osmotically stimulated vasopressin release in either patient compared to these parameters for the PavpiPosm relationship in the previously reported group studies. In contrast, studies on both patients in the sitting posture revealed that Pavp was markedly increased at all levels Of Posm compared to Pavp over a comparable range of Posm in the other quadriplegic subjects. The slope of the Pavp:Posm relationship in both patients, as shown by linear regression, was also increased compared to the mean of collective slope values for the group subjects in the eirect posture. However, Pavp concentrations in both patients in erect posture were increased before hypertoftic sodium chloride infusion, and
1541
in the absence of observations at lower levels of Posm, the threshold and response characteristics of the Pavp:Posm relationship cannot be fully delineated. Studies on rats have shown previously that when Posm is increased in the presence of hypovolemia induced by sc injections of polyethylene glycol the effects of simultaneous osmotic and nonosmotic stimuli on vasopressin release are synergistic (6). In contrast, the increase in Pavp associated with polyethylene glycol-induced volume contraction was shown to be markedly attenuated by hypoosmolality induced by ad libitum water intake or intragastric water loading. Thus, although increased sensitivity of vasopressin release to osmotic stimulation in the erect posture is suggested by linear regression analysis of the data from the studies on both patients, the slope of these lines might be substantially altered by the inclusion of Pavp:Posm data pairs closer to a theoretical osmotic threshold for vasopressin release. Despite the lack of observations at lower levels of Posm, the Pavp:Posm relationship depicted by linear regression analysis is indicative of a reduced osmotic threshold for vasopressin release in both patients in the erect posture. Our observations on suppression of Pavp during the water load studies performed on patient 1 are also consistent with this assumption. In the studies performed when the patient was supine, suppression of Pavp to an unquantifiable level occurred when Posm was reduced from 286 to 281 mosmol/kg H2O after 15 min of water drinking. In the studies performed when the patient was erect, Pavp remained above this level until Posm was reduced from 284 to 272 mosmol/kg H2O after 45 min of water drinking. Although inhibition of vasopressin release by the pharyngeal reflex associated with water drinking cannot be excluded as a factor in the suppression of Pavp (15-17), the timing of water drinking and the volume of water ingested in the first 30 min of water loading were identical in the two studies. The most likely explanation for the observed shift in the relationship of Pavp:Posm in these two quadriplegic patients, as shown by low MAP in patient 1 and increased HR in the erect posture in patient 2, is relative hemodynamic instability. Thigh cuff measurements of arterial blood pressure, which were used during the hypertonic sodium chloride infusion studies to preserve forearm venous access, often do not reflect the magnitude of orthostatic hypotension in quadriplegic subjects in the sitting posture (7). Both patients, however, exhibited signs and symptoms of hemodynamic stress, i.e. diaphoresis and complaints of headache and/or impending syncope without nausea, before termination of the infusion studies. The ability of these patients to be out of bed in a wheelchair was also severely limited by the development of such symptoms. Orthostatic hypotension was more clearly demonstrated by arm cuff measurements of
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 August 2016. at 22:21 For personal use only. No other uses without permission. . All rights reserved.
1542
WILLIAMS ET AL.
JCE & M • 1990 Vol 71 • No 6
arterial blood pressure during the water load studies performed on patient 1. Median MAP was 64 mm Hg in the erect posture compared to 80 mm Hg in the supine posture. High Pc concentrations in patient 1 and less suppression of PRA by hypertonic sodium chloride infusions in both patients in the erect posture also suggest that hemodynamic stress induced by orthostasis was greater in these two patients than in our previously studied subjects. The effect of altered hemodynamic status on cortisol secretion is mediated by cardiac and arterial baroreflex mechanisms that modulate the release of CRF and are also involved in the nonosmotic control of vasopressin release (18,19). The neural afferent component of these volume- and pressure-sensitive mechanisms is
Reductions in blood volume of approximately 6% were associated with increased osmoreceptor sensitivity, whereas larger reductions in blood volume, i.e. approximately 14%, resulted in reductions in the osmotic threshold for vasopressin release as well as further increases in osmoreceptor sensitivity. Thus, in human as well as animal subjects the effect of nonosmotic stimuli on osmoregulation of vasopressin release seems to vary in direct proportion to the potency of the nonosmotic stimuli. Studies of the renal response to water loading in patient 1 revealed marked impairment of urine-diluting ability and reduced free water clearance in the erect (sitting) posture which were not evident in the study performed while the patient was supine. Although
intact in quadriplegic subjects and is capable of eliciting
suppression of Pavp by water loading was delayed in the
erect posture compared to the prompt reduction in Pavp cortisol and vasopressin responses to orthostatic changes to unquantifiable levels in the supine posture, this does in arterial blood pressure and/or blood volume distribunot account for the impairment of urine-diluting ability tion. It has also been reported that CRF stimulates shown by the study of this patient in the erect posture. vasopressin release and that it may act synergistically to Minimum Uosm was 213 mosmol/kg H2O in the third augment the vasopressin response to osmotic stimulation hour of water loading, whereas Pavp was suppressed to (19). Similar observations regarding an effect of angioan unquantifiable level after 45 min of water drinking. tensin-II on vasopressin release have been reported (20, Osmolar clearance was reduced, and both absolute and 21). Thus, several mechanisms may contribute to nonfractional sodium excretion were quite low in the erect osmotic stimulation of vasopressin release during periods of hemodynamic stress, all of which may be implicated posture. These observations are consistent with the posin the increase in Pavp and the shift in its relationship sibility that low free water clearance in the erect posture to Posm shown by our studies on both patients in the was due to increased proximal sodium chloride reabsorperect posture. tion and decreased distal delivery of filtrate to diluting Few studies of the interaction between osmotic and segments of the renal tubules, as shown by low V/GFR nonosmotic control of vasopressin release in human sub(23). Although GFR was not reduced, it seems likely that jects have been reported. Robertson and Athar (3) re- such marked alterations in sodium and water handling ported that upright posture in hydropenic normal subby the kidneys were a reflection of reduced renal blood jects resulted in a small but significant lowering of the flow due to orthostatic hypotension in erect posture. osmotic threshold for vasopressin release, but did not Similar observations on the effect of posture on renal alter osmoreceptor sensitivity, as defined by the slope of function in quadriplegic subjects were reported by the linear relationship between Pavp and Posm. EnKooner et al. (24). Although Uosm was not reduced by hancement of osmotic stimulation of vasopressin release water loading to levels observed in studies on normal by reductions in central venous pressure induced by lower control subjects in either the sitting or supine posture in body negative pressure without changes in MAP was these studies, urine flow rate, free water clearance, and reported by Goldsmith et al. (22). In our recently reported MAP increased, as in our studies on patient 1, when the studies on stable quadriplegic subjects, modest reducquadriplegic subjects were supine. This ability to increase tions in MAP in the sitting posture were associated with free water clearance in the supine posture by improveincreased sensitivity of vasopressin release to concomiment in renal water excretion due to intrinsic renal tant osmotic stimulation without a change in the threshmechanisms and elimination of posture-related nonosold for stimulation (7). The more severe hemodynamic motic effects on vasopressin release could explain the stress, as shown by orthostatic reductions in blood presabsence of hyponatremia when the studies on our two sure, experienced by the the two patients described in patients were performed. Both patients were out of bed this report was associated with marked alterations in in a wheelchair for only brief periods of time during the both the threshold and sensitivity of vasopressin release day because of symptoms related to orthostatic hypotenin response to osmotic stimulation. These observations sion. The relatively long periods of recumbency may have are similar to those of Dunn et al. (4) in studies on rats, been sufficient to allow water retained in the sitting in which blood volume was reduced by ip injections of posture to be excreted, thus preventing the reduction in polyethylene glycol in the presence of increased Posm. serum sodium concentration that might otherwise occur.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 August 2016. at 22:21 For personal use only. No other uses without permission. . All rights reserved.
NONOSMOTIC STIMULI ALTER OSMOREGULATION
The development of severe hyponatremia during episodes of acute respiratory distress can be attributed to the superimposition of additional potent nonosmotic stimulation of vasopre$sin release, as shown by the absence of signs of volume depletion in both patients and elevated Pavp (6.7 jiU/mL) in patient 2. Hypoxemia and hypercapnic acidosis, both of which were present during the episode of acute respiratory distress in patient 1, have been shown to fesult in diminished free water clearance, marked elevation of Pavp, and reduced renal blood flow and GFR in studies on dogs (25-27) and humans (28). Elevation of Pavp was also shown in studies on patients with acute bronchial asthma reported by Baker et al. (29). In these studies increased Pavp was correlated with reductions in forced expiratory volume. Patient 2 was treated with intermittent positive pressure breathing and subsequently developed respiratory stridor before the demonstration of severe hyponatremia in association with markedly increased Pavp. Thus, conditions known to be associated with nonosmotic stimulation of vasopressin release existed during the episodes of acute respiratory distress in both patients. Whether hemodynamic instability, which seems to be the most likely explanation for enhancement of vasopressin release and marked impairment of urine-diluting ability in erect posture when th$ patients were clinically stable, may also have contributed to the development of severe hyponatremia during episodes of acute respiratory distress is uncertain. However, it may be postulated that recurrent nonosmotic stimulation of vasopressin release by the hemodynamic effects of orthostasis may increase the susceptibility of some quadriplegic individuals to the effects of other potent nonosmotic stimuli that affect vasopressin release.
7. 8. 9. 10.
11.
12. 13. 14. 15. 16. 17. 18. 19. 20. 21.
Acknowledgments
22.
We wish to thank Benjamin A. Moeller, M.D., and the staff of the Spinal Cord Injury Unit, V.A, Medical Center (Memphis, TN), for their continued support and cooperation in the performance of these studies. We thank Mary Alice Bobal for excellent technical assistance.
23. 24.
References 1. Robertson GL, Shelton RL, Athar S. The osmoregulation of vasopressin. Kidney Int. 197$;10:25-37. 2. Robertson GL. The regulation of vasopressin function in health and disease. Recent Prog Horm Res. 1977;23:333-85. 3. Robertson GL, Athar S. The interaction of blood osmolality and blood volume in regulating plasma vasopressin in man. J Clin Endocrinol Metab. 1976;42:613-20. 4. Dunn FL, Brennan TJ, Nelson AE, Robertson GL. The role of blood osmolality and volume in regulating vasopressin secretion in the rat. J Clin Invest. 1973;52:3212-9. 5. Quillen Jr EW, Cowley Jr AW. Influence of volume changes on osmolality-vasopressin relationships in conscious dogs. Am J Physiol. 1983;244:H73-9. 6. Strieker EM, Verbalis JG. Interaction of osmotic and volume
25. 26. 27.
28. 29.
1543
stimuli in regulation of neurohypophyseal secretion in rats. Am J Physiol. 1986;250:R267-75. Wall BM, Williams HH, Presley DN, et al. Altered sensitivity of osmotically stimulated vasopressin release in quadriplegic subjects. Am J Physiol. 1990;258:R827-35. Mathias CJ, Frankel HL. Autonomic failure in tetraplegia. In: Bannister R, ed. Autonomic failure. Oxford: Oxford University Press; 1984;454-88. Debarge O, Christensen NJ, Corbett JL, Eidelman BH, Frankel HL, Mathias CJ. Plasma catecholamines in tetraplegics. Paraplegia. 1974;12:44-9. Mathias CJ, Christensen NJ, Corbett JL, Frankel HL, Goodwin TJ, Peart WS. Plasma catecholamines, plasma renin activity and plasma aldosterone in tetraplegic man, horizontal and tilted. Clin Sci Mol Med. 1975;49:291-9. Haber E, Koerner T, Page LB, Kliman B, Purnode A. Application of a radioimmunoassay for angiotensin I to the physiologic measurements of plasma renin activity in normal human subjects. J Clin Endocrinol Metab. 1969;29:1349-55. Crofton JT, Dustan H, Share L, Brooks DP. Vasopressin secretion in normotensive black and white men and women on normal and low sodium diets. J Endocrinol. 1986;108:191-9. Armitage P, Berry G. Statistical methods in medical research, 2nd ed. Oxford: Blackwell; 1987;273-82. Baylis PH. Osmoregulation and control of vasopressin secretion in healthy humans. Am J Physiol. 1987;253:R671-8. Geelen G, Keil LC, Kravik SE, et al. Inhibition of plasma vasopressin after drinking in dehydrated humans. Am J Physiol. 1984;247:R968-71. Salata RA, Verbalis JG, Robinson AG. Cold water stimulation of oropharyngeal receptors in man inhibits release of vasopressin. J Clin Endocrinol Metab. 1987;65:561-7. Thompson CJ, Burd JM, Baylis PH. Acute suppression of plasma vasopressin and thirst after drinking in hypernatremic humans. Am J Physiol. 1987;252:R1138-42. Raff H, Merrill D, Skelton M, Cowley Jr AW. Control of ACTH and vasopressin in neurohypophysectomized conscious dogs. Am J Physiol. 1985;249:R281-4. Raff H, Skelton MM, Merrill DC, Cowley Jr AW. Vasopressin responses to corticotropin releasing factor and hyperosmolality in conscious dogs. Am J Physiol. 1986;251:R1235-9. Ramsay DJ, Keil LC, Sharpe MC, Shinsako J. Angiotensin II infusion increases vasopressin, ACTH, and 11-hydroxycorticosteroid secretion. Am J Physiol. 1978;234:R66-71. Klingbeil CK, Keil LC, Chang D, Reid IA. Effects of CRF and ANG II on ACTH and vasopressin release in conscious dogs. Am J Physiol. 1988;255:E46-53. Goldsmith SR, Dodge D, Cowley AW. Nonosmotic influences on osmotic stimulation of vasopressin in humans. Am J Physiol. 1987;252:H85-8. Berliner RW, Davidson DG. Production of hypertonic urine in the absence of pituitary antidiuretic hormone. J Clin Invest. 1957;36:1416-27. Kooner JS, Frankel HL, Mirando N, Peart WS, Mathias CJ. Haemodynamic, hormonal and urinary responses to postural change in tetraplegic and paraplegic man. Paraplegia. 1988;26:2337. Rose Jr CE, Anderson RJ, Carey RM. Antidiuresis and vasopressin release with hypoxemia and hypercapnia in conscious dogs. Am J Physiol. 1984;247:R127-34. Rose Jr CE, Kimmel DP, Godine RL, Kaiser DL, Carey RM. Synergistic effects of acute hypoxemia and hypercapnic acidosis in conscious dogs. Circ Res. 1983;53:202-13. Rose Jr CE, Dixon BS, Anderson RJ. Effects of hypoxemia and hypercapnic acidosis on renal water excretion and vasopressin secretion. In: Schrier RW, ed. Vasopressin. New York: Raven Press; 1985;517-23. Kilburn KH, Dowell AR. Renal function in respiratory failure. Arch Intern Med. 1971;127:754-62. Baker JW, Yerger S, Segar WE. Elevated plasma antidiuretic hormone levels in status asthmaticus. Mayo Clinic Proc. l976;51:31-4.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 August 2016. at 22:21 For personal use only. No other uses without permission. . All rights reserved.
