Potassium-Aldosterone-Renin Interrelationships THEP HIMATHONGKAM, ROBERT G. DLUHY, AND GORDON H. WILLIAMS Endocrine-Metabolic Unit, Peter Bent Brigham Hospital and the Department Harvard Medical School, Boston, Massachusetts
In six normal subjects the potassium-lowering effect of glucose ingestion (0.25 g/kg/15 min over a 2-h period) was assessed. The mean maximal potassium decrement below control 0.3 meq/liter (8%) at 90 min was coincident with the mean plasma aldosterone decrement below control of 11 ng/100 ml (46%). Plasma aldosterone then rose to peak levels at 180 min (mean increment 22 ng/ 100 ml above nadir) while potassium levels remained below control. The rise in plasma aldosterone was associated with a parallel but more rapid rise in plasma renin activity, peaking at a level 108% above control. Ninety minutes after the termination of the glucose ingestion, plasma aldosterone returned to control levels but now in the setting of reduced levels of plasma potassium and elevated levels of plasma renin activity. The data support an important role for physiologic changes in extracellular potassium concentration in the control of aldosterone secretion and indicate that interpretation of studies assessing acute changes in plasma aldosterone must carefully consider minor simultaneous changes in plasma potassium levels. The data also document that acute changes in extracellular potassium concentration play a role in the regulation of renin secretion. (J Clin Endocrinol Metah 41: 153, 1975)
ABSTRACT. The present study was performed to assess the sensitivity of the renin-angiotensinaldosterone axis to small changes in plasma potassium concentration within the physiologic range. Small increments in potassium levels were accomplished by graded constant infusions of potassium chloride over 2 h (0.17 meq/min; 0.33 meq/min; 0.5 meq/min) in 8 normal subjects on a 10 meq sodium-100 meq potassium intake. Plasma levels of aldosterone, renin activity, angiotensin II, cortisol, potassium and sodium were measured at frequent intervals. There were no significant changes observed in plasma sodium, renin activity or angiotensin II levels while cortisol levels declined in the expected diurnal pattern. During the 0.17 meq/min (10 meq/h) infusion potassium levels did not increase significantly until 120 min while plasma aldosterone levels rose significantly at 30-60 min. The mean increment above control during the lowest infusion rate was 0.2 meq/liter (5%) for plasma potassium and 13 ng/100 ml (46%) for plasma aldosterone. Although there was a stepwise increase in the increments above control of both potassium and aldosterone levels as the rate of the infusion was increased, the most sensitive area of the dose response curve appears to be 0.1-0.5 meq/liter.
P
REVIOUS studies in man have demonstrated that changes in potassium balance can produce large changes in aldosterone secretion (1-4). Since only minor changes in blood potassium levels were recorded with changes in potassium balance it has been suggested by some investigators that intracellular potassium content was regulating aldosterone secretion (5). However, recent studies have demonstrated that acute changes in serum potassium can alter plasma aldosterone levels while potassium balance was held constant (6-9). These studies suggest a role for extracellular potassium in the control of aldosterone secretion. However, regulaReceived November 5, 1974. Reprint requests to: Gordon H. Williams, M.D., Peter Bent Brigham Hospital, 721 Huntington Avenue, Boston, Massachusetts 02115.
of Medicine,
tion of aldosterone secretion by physiologic fluctuations in blood potassium levels is speculative since the serum potassium changes induced in these studies were large (1-2 meq/liter). Therefore the present study was designed to investigate the sensitivity of the adrenal glomerulosa to small increases and decreases in serum potassium within the physiologic range. Elevations of serum potassium were produced by graded infusions of potassium chloride while small decrements were produced by oral glucose ingestion. Materials and Methods Fourteen normal volunteers, 8 males and 6 females, ages 22 to 35, were admitted to the Clinical Center of the Peter Bent Brigham hospital. Complete history, physical exam and routine laboratory tests were within normal limits. All subjects denied use of drugs of any type
153
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 September 2016. at 12:28 For personal use only. No other uses without permission. . All rights reserved.
154
HIMATHONGKAM, DLUHY AND WILLIAMS
in the weeks immediately preceding the admission and informed written consent was obtained in all cases. All subjects received a constant dietary intake of 10 meq sodium/100 meq potassium, 2500 ml isocaloric diet during the course of their hospitalization. When the subjects had achieved metabolic balance as judged by their urinary sodium-potassium excretion, one or more of several protocols were performed. At least one week separated two studies performed on any one individual. All studies were performed with patients supine for 12 h after an overnight fast beginning at 8 AM. Protocol Intravenous potassium infusion. Samples were obtained before and then 10, 20, 30, 60, 90, and 120 min after the start of an intravenous potassium infusion. Isosmotic potassium chloride in water was infused at a constant rate of 0.17 meq/min for 2 h (10 meq/h) using a Harvard infusion pump in 8 subjects and 0.33 meq/min (20 meq/h) and 0.5 meq/min (30 meq/h) for 2 h in 5 of these 8 subjects during separate studies. In those subjects receiving more than one infusion, the sequence was randomized, and at least one week was allowed between studies. Plasma samples for aldosterone, cortisol, potassium, sodium, angiotensin II and renin activity were obtained at each of the time periods.
J C E & M 1975 Vol 41 < No 1
sponses were analyzed statistically by computing the t value for the response at each time interval compared with its appropriate control with pooled variances derived from a two-way analysis of variance. The computations were done on the log transform of the data. In all cases, the variances were assessed as being homogenous on the log transform data by Bartlett's test (13). T values were then found in Dunnett's tables for comparing multiple tests with a single control (14). The results are expressed as mean plus or minus standard error of the mean and significance was P < 0.05 unless otherwise stated. Results Potassium infusions
Regardless of the rate of potassium infusion, plasma renin activity, angiotensin II and sodium levels never changed significantly (Table 1). Plasma cortisol levels declined in an expected diurnal pattern during all infusions. In all studies, aldosterone and potassium levels significantly correlated with each other. While the basal potassium levels were comparable in the three groups, the basal aldosterone levels varied. This variation can probably be accounted for by the variation in the basal plasma renin activity in the three studies. FollowGlucose ingestion ing the 0.17 meq/min (10 meq/h) potassium In 6 subjects, glucose (0.25 g/kg/15 min) was infusion, venous potassium levels did not ingested for a 2-h period with plasma samples change for 60 min and did not rise signifiobtained before and then at 10, 20, 30, 60, 90, cantly above basal (4.0 ± 0 . 1 meq/liter) until 120, 150, 180 and in some cases 210 min after the end of the infusion (4.2 ± 0 . 1 meq/ the start of the ingestion. Plasma samples were liter). However, aldosterone levels began to assayed for renin activity, cortisol, potassium, sodium, aldosterone, glucose, and serum insulin rise by 30 min (33 ± 5 ng/100 ml) and rose significantly by 60 min (38 ± 7 ng/100 ml). levels. The peak potassium increment was 0.2 meq/ liter (5% above control) while the aldoLaboratory methods sterone increment was 13 ng/100 ml (46% Plasma aldosterone, cortisol, renin activity, above control). Thus, during this infusion angiotensin II and serum insulin were measured study, aldosterone levels rose at a time when by radioimmunoassays as previously described cortisol levels were falling, renin activity (10-12). Blood samples were spun immediately and the plasma separated for determination of was unchanged and before changes in potassium and sodium levels by flame photom- plasma potassium levels were detectable. etry and glucose by an autoanalyzer technique. During the 0.33 meq/min (20 meq/h) poThe level of sensitivity of the plasma potas- tassium infusion, both plasma aldosterone sium determination was ±0.05 meq/liter. Re- and potassium levels rose significantly by
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 September 2016. at 12:28 For personal use only. No other uses without permission. . All rights reserved.
POTASSIUM-ALDOSTERONE-RENIN INTERRELATIONSHIPS
155
TABLE 1. Response of plasma renin activity, angiotensin II, aldosterone, cortisol and potassium to graded constant potassium chloride infusions over 2 h Time min
Potassium meq/liter
Cortisol ^.g/100 ml
Aldosterone ng/100 ml
Renin activity ng/ml/h
Angiotensin II pg/ml
0.17 meq KCl/min n = 8 0 10 20 30 60 90 120
4.0 4.0 4.0 4.0 4.1 4.2 4.2
± 0.1 ±0.1 ± 0.1 ± 0.1 ± 0.1 ± 0.1 ± 0.1*
28 28 28 33 38 41
d- 7 i: 7 i: 6 d: o dh 7*** dt 6***
11 ± 1 12 ± 2 10 ± 1
8± 1 6± 1** 7± 1** 5 ± 1***
35 ± 6**
5.9 5.4 5.4 5.4 6.8 6.3 5.0
± ± ± ± ± ± ±
1.0 0.9 0.8 0.8 1.3 1.0 0.4
45 48 46 42 47 47 54
dE d: d: d: dt dt dt
5.6 5.7 5.3 6.0 6.1 7.0 6.7
± ± ± ± ± ± ±
1.1 1.3 1.4 1.2 1.6 1.6 1.8
43 dt 9 41 dt 8 43 dt 7 50:t 9 46 :t 11 5 5 : t 13 4 8 : t 12
4.7 ± 5.3 dt 5.0 dt 6.3 dt 6.2 dt 5.0: t 6.0 dt
1.5 1.7 1.3 2.2 1.5 0.9 2.0
42:t 55 :t 47:t 48:t 48:t 4 5 :t 57:t
7
5 5
7 8 7 8
0.33 meq KCl/min n = 5 0 10 20 30 60 90 120
4.3 4.3 4.4 4.7 4.9 5.0 5.0
± 0.2 d: 0.2 dtO.2 dt 0.3* db 0.3** dt 0.2*** ± 0.4***
27 ±4 32 ± 4 34 48 50 48 49
±6 ± 7*** ± 6*** ± 8*** ± 9*** 0.5
0 10 20 30 60 90 120
4.2 ± 0.1 4.4 :t 0.1* 4.5 dt 0.1** 4.6:t 0.1*** 4.7:t 0.2*** 4.9:t 0.1*** 5.1 :t 0.2***
12 ± 2 13 ± 1 12 ±2 9± 2* 10 ± 1* 7 -H 2*** 6 ± 1*** meq KCl/min n = 5
16 ± 4 3 21 :t 2 2 3 : t 1* 31 :h 4*** 3 3 : ,. 4*** 3 3 : t 6*** 18:t
14 ± 2 12 ± 1 13 ± 2 11 ± 2 9 ± 1** 7 ± 2*** 8 ± 2***
14 17 17 9 14 11 13
Mean ± SEM of normal subjects on a 10 meq Na+/100 meq K+ diet. * P < 0.05 significantly different from control. ** P < 0.02 significantly different from control. *** P < 0.01 significantly different from control.
30 min with potassium peaking at 60-90 min and aldosterone at 30-60 min. The mean peak increment in plasma potassium was 0.7 meq/liter (16% above control) while the increment of plasma aldosterone was 23 ng/100 ml (85% above control) (Table 1). The 0.5 meq/min (30 meq/h) potassium infusion produced a significant rise in potassium at 10 min and a peak level at 90-120 min (mean peak increment 0.9 meq/liter, 21% above control). In contrast, plasma aldosterone levels did not significantly rise until 30 min with peak levels at 60-90 min (mean peak increment 17 ng/100 ml, 106% above control). Thus, there was a stepwise increase in the peak potassium and aldosterone levels as the rate of infusion was increased. However, the relationship between the increments in
potassium and aldosterone levels was not constant. During the three potassium infusions, the initial increments in potassium were accompanied by a 3 ng/100 ml mean rise in aldosterone per 0.1 meq/liter. Yet, the final potassium increments produced a mean increment in aldosterone of only 0.7 ng/100 ml per 0.1 meq/liter. Glucose ingestion study Plasma glucose levels rose significantly by 20 min and peaked between 60 and 90 min (mean increment 65 mg/100 ml, 81% above control) (Fig. 1). Mean insulin levels also rose significantly above control (15 ± 1 /u,U/ml) to mean peak levels at 90-120 min of 70 ± 5 /uU/ml. Sodium levels did not change and cortisol levels declined
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 September 2016. at 12:28 For personal use only. No other uses without permission. . All rights reserved.
HIMATHONGKAM, DLUHY AND WILLIAMS
156
JCE & M • 1975 Vol 41 • No 1
l40r
Z
FIG. 1. Responses of plasma potassium, renin activity, cortisol and aldosterone following glucose ingestion (0.25 g/kg/15 min from 0-120 min). (Mean
30
± SEM; n = 6).
£8 Q c
20
10 20 >-
{-
15
10
UJ 0
20
60
90
120
150
180
210
MINUTES
in an expected diurnal pattern. Potassium levels began to decline at 30 min reaching a nadir at 120 min, then rising slightly in the recovery period (mean decrement 0.3 meq/liter, 8% below control). Aldosterone levels exhibited a biphasic response. Plasma aldosterone began to decline at 10 min with a significant decrement at 90 min (mean decrement 11 ng/100 ml, 46% below control) (Fig. 1). Plasma aldosterone levels then began to rise, peaking at 180 min, a highly significant (P < 0.01) increment
above the nadir (mean increment 22 ng/100 ml) as well as a significant (P < 0.05) increment above control. However, by 210 min (90 min after the last glucose ingestion) aldosterone levels had returned to control. Plasma renin activity did not change until 30 min when levels began to rise, peaking at 90 min (mean increment 8.5 ng/ml/h, 108% above control). Plasma renin activity then returned toward control levels over the next 90 min. The peak renin level coin-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 September 2016. at 12:28 For personal use only. No other uses without permission. . All rights reserved.
POTASSIUM-ALDOSTERONE-RENIN INTERRELATIONSHIPS cided both with the plasma potassium and aldosterone nadirs with the subsequent rise in aldosterone levels following persistently elevated plasma renin levels. Sixty to 90 min after the termination of the glucose ingestion, plasma aldosterone levels returned to control levels; however, basal levels were achieved in the setting of reduced plasma potassium and elevated plasma renin activity levels. Discussion The present data indicate that small increments and decrements of plasma potassium within the normal physiologic range can alter the secretion of aldosterone in the sodium depleted state. Studies in experimental animals (15,16) and man have documented that potassium can alter the secretion of aldosterone. However, previous potassium infusion studies in normal subjects induced large increases in potassium concentration and, therefore, the sensitivity of the adrenal glomerulosa could not be adequately evaluated (6-8). Likewise, in anephric subjects, glucose and insulin administration induced large decrements in potassium levels (mean 1.7 meq/liter) and associated declines in plasma aldosterone concentration (9). In the present study, glucose ingestion and 10 meq/h (0.17 meq/ min) potassium chloride infusions produced changes of 0.2-0.3 meq/liter. In both studies, minimal 5-8% increments and decrements in potassium concentration were associated with significant (40-50%) changes in plasma aldosterone levels. Since changes of this magnitude have been reported in subjects on varying dietary potassium intakes (1-4,17,18), alterations in aldosterone secretion associated with changes in potassium balance could be related to extracellular potassium concentration. Furthermore, interpretation of studies assessing the acute plasma aldosterone response to saline loading or ACTH infusion must carefully consider the simultaneous associated changes in plasma potassium levels.
157
The precise relationship of the potassium level and the aldosterone response is complex since in individual subjects the sensitivity of the adrenal cortex to acute changes in extracellular potassium varied. In some subjects, a significant rise in aldosterone was recorded without a significant elevation in venous potassium levels. In other subjects, increments in potassium concentration greater than 0.3 meq/liter were required to increase aldosterone secretion. Significant increments in plasma aldosterone levels recorded in subjects without elevations in venous potassium concentration is also consistent with the hypothesis that aldosterone secretion is regulated by changes in adrenal intracellular potassium concentration. However, in these subjects, small increments in arterial potassium concentration might not be apparent since only venous levels were sampled. Similarly, the glucose induced declines in aldosterone levels may be explained either by a change in extracellular potassium concentration or some alteration in adrenal intracellular potassium content. The adrenal intracellular regulation theory postulates that a decrease in intracellular potassium produces a decline in aldosterone levels. Since the change in extracellular potassium concentration is secondary to the insulin mediated net movement of potassium into cells, the glomerulosa cell would presumably be insensitive to insulin. As in the study of Cooke et al. (9), changes in intracellular potassium content cannot be inferred from our data and thus the problem remains unresolved. The time course of the relationship between changes in potassium concentration and aldosterone secretion is best seen in the 0.33 meq/min and 0.5 meq/min infusion studies. A significant increment and a plateau in aldosterone levels were observed 20-30 min after a change in potassium concentration. This relationship is also supported by previous in vitro studies where potassium concentrations in the incubation media were altered, as well as a recent study in normal man reporting plasma aldosterone levels following the infusion of
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 September 2016. at 12:28 For personal use only. No other uses without permission. . All rights reserved.
158
HIMATHONGKAM, DLUHY AND WILLIAMS
angiotensin II (19-22). From the present study, it is also possible to define the upper limit of maximal adrenal responsiveness following increments in serum potassium levels. Increments above control less than 0.4-0.5 meq/liter were associated with greater increments in plasma aldosterone levels than increments between 0.5-1 meq/ liter. Although the adrenal can still respond to greater rises in potassium concentration (6), the most sensitive area of the dose response curve appears to be 0.1-0.5 meq/ liter. It is unlikely that the recorded changes in aldosterone during the potassium infusion studies were related to alterations in ACTH or renin secretion. Plasma cortisol levels declined in the expected diurnal pattern while plasma renin levels remained relatively constant. It is also unlikely that the sodium restricted diet altered the responses since it has been shown that the adrenal response to potassium chloride infusion is unchanged by the level of sodium intake (6). Finally, infusing sodium chloride, dextran or 5% glucose in water over the time course utilized in the present study produces either a rapid decline or no significant change in the plasma concentrations of renin, angiotensin or aldosterone. Thus, it is improbable that the infusion technique per se is responsible for these alterations (23,24). It is well established that plasma renin levels are influenced by changes in dietary potassium. Miiller (25) and later other investigators (17,18) have demonstrated that potassium restriction increased peripheral renin levels while potassium loading suppressed the secretion of renin. Acute infusion studies in the dog suggest that increased blood levels of potassium can suppress the secretion of renin (26) although such studies in man have not been reported. The present study suggests that acute changes in extracellular potassium have a variable effect on renin secretion in sodiumrestricted normal man. Increments in plasma potassium produced by potassium chloride infusion had no significant effect on plasma renin activity. However, acute declines in
JCE & M • 1975 Vol 41 • No 1
potassium levels associated with glucose ingestion consistently produced rises in levels of plasma renin activity. Since these subjects were on a severely sodium-restricted diet, it is unlikely that alterations in sodium balance can explain the large changes observed. Other unknown factors related to glucose ingestion, e.g., a shift in gastrointestinal fluids or a change in hepatic blood flow, or a change in juxtaglomerular or macula densa intracellular potassium content, remain possibilities but a conclusion cannot be inferred from our data. It is of further interest in the glucose ingestion studies to note that the net effect of the decline of potassium and elevation in plasma renin levels was to return aldosterone concentration to control values. This is in contrast to the study of Cooke et al. in anephric subjects where the lowering of potassium levels following glucose and insulin administration produced a sustained fall in aldosterone levels (9). The difference between the studies appears to be related to the effect of glucose ingestion (and presumably decreased potassium levels) on renin secretion in the normal subjects. Thus, in normal man where aldosterone secretion is involved in both extracellular fluid volume and potassium homeostasis, it appears that the increase in renin secretion following a decline in potassium levels prevents a rapid decline in aldosterone secretion and may thereby offset major changes in sodium homeostasis. Acknowledgments These studies were supported in part by a grant from the John A. Hartford Foundation. The clinical studies were carried out on The Clinical Research Center of the Peter Bent Brigham Hospital, supported by Grant 8-M01-FR-31-14.
References 1. Laragh, J. H., and H. C. Stoerk, J Clin Invest 36: 383, 1957. 2. Bartter, F. C , I. H. Mills, E. G. Biglieri, and C. S. Delea, Recent Progr Horm Res 15: 311, 1959. 3. Gann, D. S., C. S. Delea, J. R. Gill, Jr., J. P. Thomas, and F. C. Bartter, Am J Physiol 207: 104, 1964.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 September 2016. at 12:28 For personal use only. No other uses without permission. . All rights reserved.
POTASSIUM-ALDOSTERONE-RENIN INTERRELATIONSHIPS 4. Cannon, P. J., R. P. Ames, and J. H. Laragh, 7 Clin Invest 45: 865, 1966. 5. Banmber, J. S., J. O. Davis, J. A. Johnson, and R. T. Witty, Amy Physiol 220: 1094, 1971. 6. Dliihy, R. G., L. Axelrod, R. H. Underwood, and G. H. Williams, J Clin Invest 51: 1950, 1972. 7. Birkhauser, M., R. Gaillard, A. M. Riondel, D. Scholer, M. B. Vallotton, and A. F. Miiller, EurJ Clin Invest 3: 307, 1973. 8. Scholer, D., M. Birkhauser, A. Peytremann, A. M. Riondel, M. B. Vallotton, and A. F. Miiller, Acta Endocrinol 72: 293, 1973. 9. Cooke, C. R., J. S. Horvath, M. A. Moore, T. Bledsoe, and W. G. Walker, J Clin Invest 52: 3028, 1973. 10. Underwood, R. H., and G. H. Williams, J Lab Clin Med 79: 848, 1972. 11. Emanuel, R. L., J. P. Cain, and G. H. Williams, J Lab Clin Med 81: 632, 1973. 12. Dluhy, R. G., L. Axelrod, and G. H. Williams, y Appl Physiol 33: 22, 1972. 13. Snedecor, G. H., and W. G. Cochran, Statistical Methods, ed. 6, Iowa State University Press, Ames, Iowa, p. 296, 1967. 14. Dunnett, C. W., Biometrics 20: 482, 1964.
159
15. Boyd, J. E., W. P. Palmore, and P. J. Mulrow, Endocrinology 88: 556, 1971. 16. Funder, J. W., J. R. Blair-West, J. P. Coghlan, D. A. Denton, B. A. Scoggins, and R. D. Wright, Endocrinology 88: 281, 1969. 17. Dluhy, R. G., R. H. Underwood, and G. H. Williams, J Appl Physiol 28: 299, 1970. 18. Brunner, H. R., L. Baer, J. E. Sealey, J. G. G. Ledingham, and J. H. Laragh, J Clin Invest 49: 2128, 1970. 19. Kaplan, N. M.J Clin Invest 44: 2029, 1965. 20. Miiller, J., Acta Endocrinol 50: 301, 1965. 21. Williams, G. H., L. M. McDonnell, S. A. S. Tait, and J. F. Tait, Endocrinology 91: 948, 1972. 22. Hollenberg, N. K., W. R. Chenitz, D. F. Adams, and G. H. Williams, J Clin Invest 54: 34, 1974. 23. Tuck, M. L., R. G. Dluhy, and G. H. Williams, y Clin Invest 53: 988, 1974. 24. Wong, P. Y., R. C. Talamo, G. H. Williams, and R. W. Colman,y Clin Invest 1975 (In press). 25. Veyrat, R., H. R. Brunner, E. L. Mannine, and A. F. Mueller, Urol Nephrol 73: 271, 1967. 26. Dluhy, R. G., G. L. Wolf, A. R. Christlieb, R. A. Hickler, and D. P. Lauler, Circulation 38: Suppl. 6, VI-66, 1968.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 September 2016. at 12:28 For personal use only. No other uses without permission. . All rights reserved.
In six normal subjects the potassium-lowering effect of glucose ingestion (0.25 g/kg/15 min over a 2-h period) was assessed. The mean maximal potassium decrement below control 0.3 meq/liter (8%) at 90 min was coincident with the mean plasma aldosterone decrement below control of 11 ng/100 ml (46%). Plasma aldosterone then rose to peak levels at 180 min (mean increment 22 ng/ 100 ml above nadir) while potassium levels remained below control. The rise in plasma aldosterone was associated with a parallel but more rapid rise in plasma renin activity, peaking at a level 108% above control. Ninety minutes after the termination of the glucose ingestion, plasma aldosterone returned to control levels but now in the setting of reduced levels of plasma potassium and elevated levels of plasma renin activity. The data support an important role for physiologic changes in extracellular potassium concentration in the control of aldosterone secretion and indicate that interpretation of studies assessing acute changes in plasma aldosterone must carefully consider minor simultaneous changes in plasma potassium levels. The data also document that acute changes in extracellular potassium concentration play a role in the regulation of renin secretion. (J Clin Endocrinol Metah 41: 153, 1975)
ABSTRACT. The present study was performed to assess the sensitivity of the renin-angiotensinaldosterone axis to small changes in plasma potassium concentration within the physiologic range. Small increments in potassium levels were accomplished by graded constant infusions of potassium chloride over 2 h (0.17 meq/min; 0.33 meq/min; 0.5 meq/min) in 8 normal subjects on a 10 meq sodium-100 meq potassium intake. Plasma levels of aldosterone, renin activity, angiotensin II, cortisol, potassium and sodium were measured at frequent intervals. There were no significant changes observed in plasma sodium, renin activity or angiotensin II levels while cortisol levels declined in the expected diurnal pattern. During the 0.17 meq/min (10 meq/h) infusion potassium levels did not increase significantly until 120 min while plasma aldosterone levels rose significantly at 30-60 min. The mean increment above control during the lowest infusion rate was 0.2 meq/liter (5%) for plasma potassium and 13 ng/100 ml (46%) for plasma aldosterone. Although there was a stepwise increase in the increments above control of both potassium and aldosterone levels as the rate of the infusion was increased, the most sensitive area of the dose response curve appears to be 0.1-0.5 meq/liter.
P
REVIOUS studies in man have demonstrated that changes in potassium balance can produce large changes in aldosterone secretion (1-4). Since only minor changes in blood potassium levels were recorded with changes in potassium balance it has been suggested by some investigators that intracellular potassium content was regulating aldosterone secretion (5). However, recent studies have demonstrated that acute changes in serum potassium can alter plasma aldosterone levels while potassium balance was held constant (6-9). These studies suggest a role for extracellular potassium in the control of aldosterone secretion. However, regulaReceived November 5, 1974. Reprint requests to: Gordon H. Williams, M.D., Peter Bent Brigham Hospital, 721 Huntington Avenue, Boston, Massachusetts 02115.
of Medicine,
tion of aldosterone secretion by physiologic fluctuations in blood potassium levels is speculative since the serum potassium changes induced in these studies were large (1-2 meq/liter). Therefore the present study was designed to investigate the sensitivity of the adrenal glomerulosa to small increases and decreases in serum potassium within the physiologic range. Elevations of serum potassium were produced by graded infusions of potassium chloride while small decrements were produced by oral glucose ingestion. Materials and Methods Fourteen normal volunteers, 8 males and 6 females, ages 22 to 35, were admitted to the Clinical Center of the Peter Bent Brigham hospital. Complete history, physical exam and routine laboratory tests were within normal limits. All subjects denied use of drugs of any type
153
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 September 2016. at 12:28 For personal use only. No other uses without permission. . All rights reserved.
154
HIMATHONGKAM, DLUHY AND WILLIAMS
in the weeks immediately preceding the admission and informed written consent was obtained in all cases. All subjects received a constant dietary intake of 10 meq sodium/100 meq potassium, 2500 ml isocaloric diet during the course of their hospitalization. When the subjects had achieved metabolic balance as judged by their urinary sodium-potassium excretion, one or more of several protocols were performed. At least one week separated two studies performed on any one individual. All studies were performed with patients supine for 12 h after an overnight fast beginning at 8 AM. Protocol Intravenous potassium infusion. Samples were obtained before and then 10, 20, 30, 60, 90, and 120 min after the start of an intravenous potassium infusion. Isosmotic potassium chloride in water was infused at a constant rate of 0.17 meq/min for 2 h (10 meq/h) using a Harvard infusion pump in 8 subjects and 0.33 meq/min (20 meq/h) and 0.5 meq/min (30 meq/h) for 2 h in 5 of these 8 subjects during separate studies. In those subjects receiving more than one infusion, the sequence was randomized, and at least one week was allowed between studies. Plasma samples for aldosterone, cortisol, potassium, sodium, angiotensin II and renin activity were obtained at each of the time periods.
J C E & M 1975 Vol 41 < No 1
sponses were analyzed statistically by computing the t value for the response at each time interval compared with its appropriate control with pooled variances derived from a two-way analysis of variance. The computations were done on the log transform of the data. In all cases, the variances were assessed as being homogenous on the log transform data by Bartlett's test (13). T values were then found in Dunnett's tables for comparing multiple tests with a single control (14). The results are expressed as mean plus or minus standard error of the mean and significance was P < 0.05 unless otherwise stated. Results Potassium infusions
Regardless of the rate of potassium infusion, plasma renin activity, angiotensin II and sodium levels never changed significantly (Table 1). Plasma cortisol levels declined in an expected diurnal pattern during all infusions. In all studies, aldosterone and potassium levels significantly correlated with each other. While the basal potassium levels were comparable in the three groups, the basal aldosterone levels varied. This variation can probably be accounted for by the variation in the basal plasma renin activity in the three studies. FollowGlucose ingestion ing the 0.17 meq/min (10 meq/h) potassium In 6 subjects, glucose (0.25 g/kg/15 min) was infusion, venous potassium levels did not ingested for a 2-h period with plasma samples change for 60 min and did not rise signifiobtained before and then at 10, 20, 30, 60, 90, cantly above basal (4.0 ± 0 . 1 meq/liter) until 120, 150, 180 and in some cases 210 min after the end of the infusion (4.2 ± 0 . 1 meq/ the start of the ingestion. Plasma samples were liter). However, aldosterone levels began to assayed for renin activity, cortisol, potassium, sodium, aldosterone, glucose, and serum insulin rise by 30 min (33 ± 5 ng/100 ml) and rose significantly by 60 min (38 ± 7 ng/100 ml). levels. The peak potassium increment was 0.2 meq/ liter (5% above control) while the aldoLaboratory methods sterone increment was 13 ng/100 ml (46% Plasma aldosterone, cortisol, renin activity, above control). Thus, during this infusion angiotensin II and serum insulin were measured study, aldosterone levels rose at a time when by radioimmunoassays as previously described cortisol levels were falling, renin activity (10-12). Blood samples were spun immediately and the plasma separated for determination of was unchanged and before changes in potassium and sodium levels by flame photom- plasma potassium levels were detectable. etry and glucose by an autoanalyzer technique. During the 0.33 meq/min (20 meq/h) poThe level of sensitivity of the plasma potas- tassium infusion, both plasma aldosterone sium determination was ±0.05 meq/liter. Re- and potassium levels rose significantly by
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 September 2016. at 12:28 For personal use only. No other uses without permission. . All rights reserved.
POTASSIUM-ALDOSTERONE-RENIN INTERRELATIONSHIPS
155
TABLE 1. Response of plasma renin activity, angiotensin II, aldosterone, cortisol and potassium to graded constant potassium chloride infusions over 2 h Time min
Potassium meq/liter
Cortisol ^.g/100 ml
Aldosterone ng/100 ml
Renin activity ng/ml/h
Angiotensin II pg/ml
0.17 meq KCl/min n = 8 0 10 20 30 60 90 120
4.0 4.0 4.0 4.0 4.1 4.2 4.2
± 0.1 ±0.1 ± 0.1 ± 0.1 ± 0.1 ± 0.1 ± 0.1*
28 28 28 33 38 41
d- 7 i: 7 i: 6 d: o dh 7*** dt 6***
11 ± 1 12 ± 2 10 ± 1
8± 1 6± 1** 7± 1** 5 ± 1***
35 ± 6**
5.9 5.4 5.4 5.4 6.8 6.3 5.0
± ± ± ± ± ± ±
1.0 0.9 0.8 0.8 1.3 1.0 0.4
45 48 46 42 47 47 54
dE d: d: d: dt dt dt
5.6 5.7 5.3 6.0 6.1 7.0 6.7
± ± ± ± ± ± ±
1.1 1.3 1.4 1.2 1.6 1.6 1.8
43 dt 9 41 dt 8 43 dt 7 50:t 9 46 :t 11 5 5 : t 13 4 8 : t 12
4.7 ± 5.3 dt 5.0 dt 6.3 dt 6.2 dt 5.0: t 6.0 dt
1.5 1.7 1.3 2.2 1.5 0.9 2.0
42:t 55 :t 47:t 48:t 48:t 4 5 :t 57:t
7
5 5
7 8 7 8
0.33 meq KCl/min n = 5 0 10 20 30 60 90 120
4.3 4.3 4.4 4.7 4.9 5.0 5.0
± 0.2 d: 0.2 dtO.2 dt 0.3* db 0.3** dt 0.2*** ± 0.4***
27 ±4 32 ± 4 34 48 50 48 49
±6 ± 7*** ± 6*** ± 8*** ± 9*** 0.5
0 10 20 30 60 90 120
4.2 ± 0.1 4.4 :t 0.1* 4.5 dt 0.1** 4.6:t 0.1*** 4.7:t 0.2*** 4.9:t 0.1*** 5.1 :t 0.2***
12 ± 2 13 ± 1 12 ±2 9± 2* 10 ± 1* 7 -H 2*** 6 ± 1*** meq KCl/min n = 5
16 ± 4 3 21 :t 2 2 3 : t 1* 31 :h 4*** 3 3 : ,. 4*** 3 3 : t 6*** 18:t
14 ± 2 12 ± 1 13 ± 2 11 ± 2 9 ± 1** 7 ± 2*** 8 ± 2***
14 17 17 9 14 11 13
Mean ± SEM of normal subjects on a 10 meq Na+/100 meq K+ diet. * P < 0.05 significantly different from control. ** P < 0.02 significantly different from control. *** P < 0.01 significantly different from control.
30 min with potassium peaking at 60-90 min and aldosterone at 30-60 min. The mean peak increment in plasma potassium was 0.7 meq/liter (16% above control) while the increment of plasma aldosterone was 23 ng/100 ml (85% above control) (Table 1). The 0.5 meq/min (30 meq/h) potassium infusion produced a significant rise in potassium at 10 min and a peak level at 90-120 min (mean peak increment 0.9 meq/liter, 21% above control). In contrast, plasma aldosterone levels did not significantly rise until 30 min with peak levels at 60-90 min (mean peak increment 17 ng/100 ml, 106% above control). Thus, there was a stepwise increase in the peak potassium and aldosterone levels as the rate of infusion was increased. However, the relationship between the increments in
potassium and aldosterone levels was not constant. During the three potassium infusions, the initial increments in potassium were accompanied by a 3 ng/100 ml mean rise in aldosterone per 0.1 meq/liter. Yet, the final potassium increments produced a mean increment in aldosterone of only 0.7 ng/100 ml per 0.1 meq/liter. Glucose ingestion study Plasma glucose levels rose significantly by 20 min and peaked between 60 and 90 min (mean increment 65 mg/100 ml, 81% above control) (Fig. 1). Mean insulin levels also rose significantly above control (15 ± 1 /u,U/ml) to mean peak levels at 90-120 min of 70 ± 5 /uU/ml. Sodium levels did not change and cortisol levels declined
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 September 2016. at 12:28 For personal use only. No other uses without permission. . All rights reserved.
HIMATHONGKAM, DLUHY AND WILLIAMS
156
JCE & M • 1975 Vol 41 • No 1
l40r
Z
FIG. 1. Responses of plasma potassium, renin activity, cortisol and aldosterone following glucose ingestion (0.25 g/kg/15 min from 0-120 min). (Mean
30
± SEM; n = 6).
£8 Q c
20
10 20 >-
{-
15
10
UJ 0
20
60
90
120
150
180
210
MINUTES
in an expected diurnal pattern. Potassium levels began to decline at 30 min reaching a nadir at 120 min, then rising slightly in the recovery period (mean decrement 0.3 meq/liter, 8% below control). Aldosterone levels exhibited a biphasic response. Plasma aldosterone began to decline at 10 min with a significant decrement at 90 min (mean decrement 11 ng/100 ml, 46% below control) (Fig. 1). Plasma aldosterone levels then began to rise, peaking at 180 min, a highly significant (P < 0.01) increment
above the nadir (mean increment 22 ng/100 ml) as well as a significant (P < 0.05) increment above control. However, by 210 min (90 min after the last glucose ingestion) aldosterone levels had returned to control. Plasma renin activity did not change until 30 min when levels began to rise, peaking at 90 min (mean increment 8.5 ng/ml/h, 108% above control). Plasma renin activity then returned toward control levels over the next 90 min. The peak renin level coin-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 September 2016. at 12:28 For personal use only. No other uses without permission. . All rights reserved.
POTASSIUM-ALDOSTERONE-RENIN INTERRELATIONSHIPS cided both with the plasma potassium and aldosterone nadirs with the subsequent rise in aldosterone levels following persistently elevated plasma renin levels. Sixty to 90 min after the termination of the glucose ingestion, plasma aldosterone levels returned to control levels; however, basal levels were achieved in the setting of reduced plasma potassium and elevated plasma renin activity levels. Discussion The present data indicate that small increments and decrements of plasma potassium within the normal physiologic range can alter the secretion of aldosterone in the sodium depleted state. Studies in experimental animals (15,16) and man have documented that potassium can alter the secretion of aldosterone. However, previous potassium infusion studies in normal subjects induced large increases in potassium concentration and, therefore, the sensitivity of the adrenal glomerulosa could not be adequately evaluated (6-8). Likewise, in anephric subjects, glucose and insulin administration induced large decrements in potassium levels (mean 1.7 meq/liter) and associated declines in plasma aldosterone concentration (9). In the present study, glucose ingestion and 10 meq/h (0.17 meq/ min) potassium chloride infusions produced changes of 0.2-0.3 meq/liter. In both studies, minimal 5-8% increments and decrements in potassium concentration were associated with significant (40-50%) changes in plasma aldosterone levels. Since changes of this magnitude have been reported in subjects on varying dietary potassium intakes (1-4,17,18), alterations in aldosterone secretion associated with changes in potassium balance could be related to extracellular potassium concentration. Furthermore, interpretation of studies assessing the acute plasma aldosterone response to saline loading or ACTH infusion must carefully consider the simultaneous associated changes in plasma potassium levels.
157
The precise relationship of the potassium level and the aldosterone response is complex since in individual subjects the sensitivity of the adrenal cortex to acute changes in extracellular potassium varied. In some subjects, a significant rise in aldosterone was recorded without a significant elevation in venous potassium levels. In other subjects, increments in potassium concentration greater than 0.3 meq/liter were required to increase aldosterone secretion. Significant increments in plasma aldosterone levels recorded in subjects without elevations in venous potassium concentration is also consistent with the hypothesis that aldosterone secretion is regulated by changes in adrenal intracellular potassium concentration. However, in these subjects, small increments in arterial potassium concentration might not be apparent since only venous levels were sampled. Similarly, the glucose induced declines in aldosterone levels may be explained either by a change in extracellular potassium concentration or some alteration in adrenal intracellular potassium content. The adrenal intracellular regulation theory postulates that a decrease in intracellular potassium produces a decline in aldosterone levels. Since the change in extracellular potassium concentration is secondary to the insulin mediated net movement of potassium into cells, the glomerulosa cell would presumably be insensitive to insulin. As in the study of Cooke et al. (9), changes in intracellular potassium content cannot be inferred from our data and thus the problem remains unresolved. The time course of the relationship between changes in potassium concentration and aldosterone secretion is best seen in the 0.33 meq/min and 0.5 meq/min infusion studies. A significant increment and a plateau in aldosterone levels were observed 20-30 min after a change in potassium concentration. This relationship is also supported by previous in vitro studies where potassium concentrations in the incubation media were altered, as well as a recent study in normal man reporting plasma aldosterone levels following the infusion of
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 September 2016. at 12:28 For personal use only. No other uses without permission. . All rights reserved.
158
HIMATHONGKAM, DLUHY AND WILLIAMS
angiotensin II (19-22). From the present study, it is also possible to define the upper limit of maximal adrenal responsiveness following increments in serum potassium levels. Increments above control less than 0.4-0.5 meq/liter were associated with greater increments in plasma aldosterone levels than increments between 0.5-1 meq/ liter. Although the adrenal can still respond to greater rises in potassium concentration (6), the most sensitive area of the dose response curve appears to be 0.1-0.5 meq/ liter. It is unlikely that the recorded changes in aldosterone during the potassium infusion studies were related to alterations in ACTH or renin secretion. Plasma cortisol levels declined in the expected diurnal pattern while plasma renin levels remained relatively constant. It is also unlikely that the sodium restricted diet altered the responses since it has been shown that the adrenal response to potassium chloride infusion is unchanged by the level of sodium intake (6). Finally, infusing sodium chloride, dextran or 5% glucose in water over the time course utilized in the present study produces either a rapid decline or no significant change in the plasma concentrations of renin, angiotensin or aldosterone. Thus, it is improbable that the infusion technique per se is responsible for these alterations (23,24). It is well established that plasma renin levels are influenced by changes in dietary potassium. Miiller (25) and later other investigators (17,18) have demonstrated that potassium restriction increased peripheral renin levels while potassium loading suppressed the secretion of renin. Acute infusion studies in the dog suggest that increased blood levels of potassium can suppress the secretion of renin (26) although such studies in man have not been reported. The present study suggests that acute changes in extracellular potassium have a variable effect on renin secretion in sodiumrestricted normal man. Increments in plasma potassium produced by potassium chloride infusion had no significant effect on plasma renin activity. However, acute declines in
JCE & M • 1975 Vol 41 • No 1
potassium levels associated with glucose ingestion consistently produced rises in levels of plasma renin activity. Since these subjects were on a severely sodium-restricted diet, it is unlikely that alterations in sodium balance can explain the large changes observed. Other unknown factors related to glucose ingestion, e.g., a shift in gastrointestinal fluids or a change in hepatic blood flow, or a change in juxtaglomerular or macula densa intracellular potassium content, remain possibilities but a conclusion cannot be inferred from our data. It is of further interest in the glucose ingestion studies to note that the net effect of the decline of potassium and elevation in plasma renin levels was to return aldosterone concentration to control values. This is in contrast to the study of Cooke et al. in anephric subjects where the lowering of potassium levels following glucose and insulin administration produced a sustained fall in aldosterone levels (9). The difference between the studies appears to be related to the effect of glucose ingestion (and presumably decreased potassium levels) on renin secretion in the normal subjects. Thus, in normal man where aldosterone secretion is involved in both extracellular fluid volume and potassium homeostasis, it appears that the increase in renin secretion following a decline in potassium levels prevents a rapid decline in aldosterone secretion and may thereby offset major changes in sodium homeostasis. Acknowledgments These studies were supported in part by a grant from the John A. Hartford Foundation. The clinical studies were carried out on The Clinical Research Center of the Peter Bent Brigham Hospital, supported by Grant 8-M01-FR-31-14.
References 1. Laragh, J. H., and H. C. Stoerk, J Clin Invest 36: 383, 1957. 2. Bartter, F. C , I. H. Mills, E. G. Biglieri, and C. S. Delea, Recent Progr Horm Res 15: 311, 1959. 3. Gann, D. S., C. S. Delea, J. R. Gill, Jr., J. P. Thomas, and F. C. Bartter, Am J Physiol 207: 104, 1964.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 September 2016. at 12:28 For personal use only. No other uses without permission. . All rights reserved.
POTASSIUM-ALDOSTERONE-RENIN INTERRELATIONSHIPS 4. Cannon, P. J., R. P. Ames, and J. H. Laragh, 7 Clin Invest 45: 865, 1966. 5. Banmber, J. S., J. O. Davis, J. A. Johnson, and R. T. Witty, Amy Physiol 220: 1094, 1971. 6. Dliihy, R. G., L. Axelrod, R. H. Underwood, and G. H. Williams, J Clin Invest 51: 1950, 1972. 7. Birkhauser, M., R. Gaillard, A. M. Riondel, D. Scholer, M. B. Vallotton, and A. F. Miiller, EurJ Clin Invest 3: 307, 1973. 8. Scholer, D., M. Birkhauser, A. Peytremann, A. M. Riondel, M. B. Vallotton, and A. F. Miiller, Acta Endocrinol 72: 293, 1973. 9. Cooke, C. R., J. S. Horvath, M. A. Moore, T. Bledsoe, and W. G. Walker, J Clin Invest 52: 3028, 1973. 10. Underwood, R. H., and G. H. Williams, J Lab Clin Med 79: 848, 1972. 11. Emanuel, R. L., J. P. Cain, and G. H. Williams, J Lab Clin Med 81: 632, 1973. 12. Dluhy, R. G., L. Axelrod, and G. H. Williams, y Appl Physiol 33: 22, 1972. 13. Snedecor, G. H., and W. G. Cochran, Statistical Methods, ed. 6, Iowa State University Press, Ames, Iowa, p. 296, 1967. 14. Dunnett, C. W., Biometrics 20: 482, 1964.
159
15. Boyd, J. E., W. P. Palmore, and P. J. Mulrow, Endocrinology 88: 556, 1971. 16. Funder, J. W., J. R. Blair-West, J. P. Coghlan, D. A. Denton, B. A. Scoggins, and R. D. Wright, Endocrinology 88: 281, 1969. 17. Dluhy, R. G., R. H. Underwood, and G. H. Williams, J Appl Physiol 28: 299, 1970. 18. Brunner, H. R., L. Baer, J. E. Sealey, J. G. G. Ledingham, and J. H. Laragh, J Clin Invest 49: 2128, 1970. 19. Kaplan, N. M.J Clin Invest 44: 2029, 1965. 20. Miiller, J., Acta Endocrinol 50: 301, 1965. 21. Williams, G. H., L. M. McDonnell, S. A. S. Tait, and J. F. Tait, Endocrinology 91: 948, 1972. 22. Hollenberg, N. K., W. R. Chenitz, D. F. Adams, and G. H. Williams, J Clin Invest 54: 34, 1974. 23. Tuck, M. L., R. G. Dluhy, and G. H. Williams, y Clin Invest 53: 988, 1974. 24. Wong, P. Y., R. C. Talamo, G. H. Williams, and R. W. Colman,y Clin Invest 1975 (In press). 25. Veyrat, R., H. R. Brunner, E. L. Mannine, and A. F. Mueller, Urol Nephrol 73: 271, 1967. 26. Dluhy, R. G., G. L. Wolf, A. R. Christlieb, R. A. Hickler, and D. P. Lauler, Circulation 38: Suppl. 6, VI-66, 1968.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 September 2016. at 12:28 For personal use only. No other uses without permission. . All rights reserved.
