Hormones Other Than Aldosterone May Contribute to Hypertension in 3 Different Subtypes of Primary Aldosteronism

Abstract

Blood pressure (BP) level is similar in patients with 3 subtypes of primary aldosteronism (PA), even though aldosterone levels may vary. Glucocorticoids and adrenomedullary hormones may be influenced and may contribute to hypertension in PA. The authors' objective was to investigate the influence of PA on adrenal gland secretion and the roles of these hormones in hypertension. Patients diagnosed with PA (229 cases) were enrolled and classified into 3 subgroups: aldosterone‐producing adenoma (APA), unilateral nodular adrenal hyperplasia (UNAH), and idiopathic hyperaldosteronism (IHA). Patients with essential hypertension served as the control group (100 cases). Concentration of the above hormones was measured and compared between groups. Level of plasma adrenocorticotrophic hormone (ACTH) in patients with APA was significantly lower than that in patients with IHA (P<.001) and UNAH (P<0.5). The 24‐hour urinary free cortisol and adrenomedullary hormone levels were highest in patients with IHA, lower in patients with APA, and lowest in patients with UNAH. Systolic BP level was positively correlated with 8 am plasma cortisol level (r=0.142, P=.039) and plasma ACTH level (r=0.383, P=.016). Cortisol and adrenomedullary hormones were different between PA subtypes and they might involve regulation of BP in those patients.

Primary aldosteronism (PA) is the most common form of secondary hypertension, accounting for up to 5% to 13% of all hypertensive cases.1, 2 The main subtypes of PA are aldosterone‐producing adenoma (APA), unilateral nodular adrenal hyperplasia (UNAH), and idiopathic hyperaldosteronism, also known as idiopathic aldosteronism (IHA).1, 3 Research shows that PA is associated with more cardiovascular,4 renal,5 and metabolic6 complications than essential hypertension (EH). In spite of significantly different aldosterone and serum potassium levels7, 8 among different PA subtypes, there was typically no difference in blood pressure (BP) levels, the degree of kidney function,5 or the glucose and lipid metabolism across PA subtypes.9 Besides mineralcorticoids, we know that the adrenal gland also produces cortisol and adrenomedullary hormones that influence the regulation of BP. The changes of adrenal gland hormone secretion as a result of excess aldosterone secretion and the differences across 3 PA subtypes have not been studied. The extent of those hormones contributing to the clinical appearance of patients with PA is still unknown.

In this study, clinical data from 229 patients with PA are analyzed, including 3 PA subtype groups and a control group with EH. The aims of this study were to (1) elucidate the patterns of adrenal gland hormone secretion change when PA occurred and their roles in regulation of BP and (2) clarify the potential mechanisms for the manifestation.

Materials and Methods

Patients

Patients diagnosed as having PA in the endocrine and metabolism department of Ruijin Hospital, Shanghai Jiaotong University School of medicine, from January 2005 to December 2011 were enrolled. A total of 229 patients (aged 18 to 70 years) were registered. Secondary causes of hypertension other than hyperaldosteronism, such as renovascular hypertension, pheochromocytoma, or Cushing's syndrome, as well as other adrenal gland diseases, were excluded by laboratory test and/or radiological imaging. All patients were classified into 3 PA subgroups: 103 patients had APA, 52 had UNAH, and 74 had IHA.

A total of 100 patients with EH served as control patients. These patients were recruited at our hypertension department and were selected by frequency matching according to the inclusion criteria to exclude confounding variables such as age and sex. Their possibilities of secondary causes of hypertension were also excluded.

The clinical database of all patients with PA and EH was established and followed up for 1 year, including medical history, biochemical parameters, and clinical manifestations.

Diagnosing PA and Its Subtypes

All patients diagnosed with PA were screened by aldosterone‐to‐renin ratio (ARR) and confirmed by saline‐loading test. Patients with EH were screened by ARR to exclude diagnosis of PA.

Prior to evaluation, patients with hypokalemia received oral potassium supplements, and treatment with mineralocorticoid receptor antagonists had been withdrawn for at least 6 weeks. Other agents, such as diuretics, β‐blockers, angiotensin‐converting enzyme (ACE) inhibitors, and angiotensin type 1 receptor blockers (ARBs), were stopped for at least 2 weeks. A long‐acting calcium channel blocker (CCB) Isoptin SR (verapamil sustained‐release tablets, Abbott GmbH & Co. KG, Abbott Park, IL) was used to minimize the risk of uncontrolled hypertension.10 BP was measured in the sitting position with a period of Isoptin SR use. A trained nurse determined the BP level with a mercurial sphygmomanometer. The BP was measured 3 times at 5‐minute intervals. Office BP was determined by calculating the average BP of the second and third measurements.11 Before examination, patients with EH in this study were suspected as having PA. For diagnosis, their antihypertension medication was also changed to Isoptin SR.

All patients underwent plasma aldosterone and plasma renin activity testing after 2 hours in the upright position as described previously.12 A positive screening test for PA was based on an elevated aldosterone‐to‐renin ratio ≥30 ((ng/dL)/(ng/mL/h)) together with elevated plasma aldosterone (≥20 ng/dL).13 Diagnosis was confirmed by the absence of plasma aldosterone suppression after a saline‐loading test with plasma potassium to normal (plasma aldosterone ≥9 ng/dL).14

All PA patients firstly underwent an adrenal computerized tomographic (CT) scan for subtype differentiation to exclude large masses such as adrenocortical carcinoma. PA patients with no typical adrenal gland adenoma on CT scans were diagnosed as having BAH (n=52).

On CT scan, clear unilateral adrenal masses were found in 177 PA patients. Adrenal venous sampling (AVS) was employed to determine the distinction between unilateral and bilateral adrenal disease before surgery10 (patients who refused to undergo AVS were not enrolled in this study). After AVS, patients with unilateral aldosterone oversecretion received unilateral adrenalectomy (n=155). The diagnosis was further confirmed by histological examination of the surgical specimens and by outcomes after diverse treatments.

Patients presenting with bilateral oversecretion of aldosterone after AVS were considered to have BAH (n=22). Those patients were treated with spironolactone and other antihypertensive agents and followed‐up.

Secretion Status of the Adrenal Cortex and Medulla

Hormone secretion status of the adrenal cortex was evaluated by plasma cortisol circadian, 24‐hour urinary excretion of free cortisol, and plasma adrenocorticotrophic hormone (ACTH) level. Secretion status of adrenal medulla was accessed via measurements of epinephrine, norepinephrine, and dopamine. Sex hormones were not analyzed in this study because the hormones secreted by adrenal glands are not the main components in blood and not directly related to hypertension.

Measurement of Biochemical Parameters

Blood samples were obtained from patients at 8 am, 4 pm, and 12 am on the same day to determine patients' cortisol circadian rhythm. The blood sample taken at 8 am was also measured for plasma ACTH level. A 24‐hour urinary test was obtained for measurement of cortisol, epinephrine, norepinephrine, and dopamine.

Plasma and urinary aldosterone were measured using a radioimmunoassay (RIA) kit (Beckman Coulter; Diagnostic Systems Laboratories, Brea, CA). Plasma renin activity was measured as the generation of angiotensin I in vitro with a commercially available RIA kit (Diasorin, Stillwater, MN). Plasma ACTH level, plasma cortisol, and urinary free cortisol levels were assessed regularly in our clinical center with radioimmunoassay (RIA) kits (Diagnostic Systems Laboratories). Testing for 24‐hour urinary epinephrine, norepinephrine, and dopamine excretion were regularly performed in the Shanghai Hypertension Institute using high‐pressure liquid chromatography.

Blood biochemistry (including serum potassium and 24‐hour urinary potassium excretion) was evaluated in our hospital with Beckman automatic biochemical instruments.

Statistical Analysis

Data were expressed as mean±standard error (continuous variables) or number and percentage (categorical variables). Differences between groups were examined with analysis of variance or chi‐square test. Correlations were evaluated using Pearson's correlation coefficients. Multiple linear regression analysis was used to identify potential risk factors. SPSS for Windows, version 15.1 (SPSS Inc, Chicago, IL) was used for all analyses. A P value <.05 was considered statistically significant.

Results

Baseline Characteristics of the Study Population

Patients with PA and EH were comparable in regards to age and sex distribution (P=.211 and .156, respectively). Patients with PA had significantly higher BP than patients with EH (P=.025 and .021 for systolic BP and diastolic BP, respectively). The plasma (P=.000, both supine and upright) and urinary (P=.000) aldosterone levels were higher and serum potassium (P=.000) and plasma renin activity (P=.000, both supine and upright) were lower in the PA group compared with the EH group.

Among the 3 PA subgroups, we found that BP level did not differ among the 3 groups. Plasma (P=.000 for supine and P=.002 for upright) and urinary (P=.000) aldosterone levels were significantly higher in the APA group than in the IHA group, while serum potassium concentration and plasma renin activity were significantly lower in the APA group compared with in the IHA group. The UNAH group's values were between those of the APA and IHA group (Table 1).

Table 1.

Baseline Characteristics of the Study Population

Parameter	Group					P Valuea
	APA (n=103)	UNAH (n=52)	IHA (n=74)	PA (n=229)	EH (n=100)
Age, y	40.3±8.6	41.4±7.6	43.5±9.0	42.0±8.7	43.5±14.0	.211
Sex distribution, female/male	44/59	27/25	46/28	110/112	45/55	.156
SBP, mm Hg	153.1±24.2	154.0±19.2	154.0±19.6	153.7±21.2	146.3±17.2	.025
DBP, mm Hg	96.5±13.6	93.2±13.5	93.2±12.8	94.3±13.2	89.5±11.9	.021
Supine aldosterone, ng/L	594.9±365.2c,e	392.6±180.4	292.8±324.5	373.8±339.0	144.4±88.2	.000
Upright aldosterone, ng/L	531.8±384.6c	521.4±442.4c	323.3±232.8	401.9±318.8	193.1±100.8	.000
Supine PRA, ng/mL/h	0.07±0.12c	0.10±0.19	0.15±0.25	0.11±0.21	1.47±1.13	.000
Upright PRA, ng/mL/h	0.25±0.32	0.18±0.22d	0.31±0.39	0.26±0.35	2.90±1.39	.000
Urinary aldosterone excretion, µg/24 h	26.65±17.67c	25.16±19.34c	15.21±10.72	21.23±16.05	8.15±6.57	.000
Plasma K+, mEq/L	2.41±0.55c	2.57±0.65c	2.95±0.72	3.06±0.62	3.68±0.39	.000
Urinary potassium excretion, mmol/24 h	48.07±30.24	52.65±28.66	49.61±30.90	49.63±30.27	41.32±13.11	.071

Abbreviations: APA, aldosterone‐producing adenoma; DBP, diastolic blood pressure; EH, essential hypertension; IHA, idiopathic hyperaldosteronism; PA, primary aldosteronism; PRA, plasma renin activity; SBP, systolic blood pressure; UNAH, unilateral nodular adrenal hyperplasia. Data are expressed as mean±standard deviation or distributions. P ^a=unpaired t test between the PA and EH groups. Compared with the IHA group. ^b P<.001. ^c P<.05. Compared with the UNAH group. ^e P<.05.

Plasma Cortisol and ACTH Secretion in the PA and EH Groups

There was no significant difference in plasma cortisol levels at all 3 time points (8 am, 4 pm, and 12 am) between the PA group and the EH group or among the 3 PA subtypes (Table 2).

Table 2.

Changing Patterns of the Pituitary Adrenal Axis and Adrenal Medulla in the APA, UNAH, IHA, and EH Groups

	Group					P Valuea
	APA (n=103)	UNAH (n=52)	IHA (74)	PA (n=299)	EH (n=100)
Plasma cortisol (8 am), ug/dL	11.8±4.7	11.9±4.1	12.2±3.9	12.0±4.2	12.6±5.2	.597
Plsma cortisol (4 pm), ug/dL	7.1±3.2	6.8±3.7	6.7±2.4	6.8±2.9	8.7±3.3	.050
Plasma cortosol (12 pm), µg/dL	3.8±2.3	4.2±2.6	4.6±2.8	4.3±2.6	5.9±5.7	.086
Urinary free cortisol, µg/24 h	78.2±34.2c	74.7±33.3c	91.0±46.2	84.0±41.2	96.0±34.2	.008
Urinary epinephrine, µg/24 h	5.1±2.4b,d	3.1±1.4b	8.1±4.2	6.2±3.8	8.7±4.9	.043
Urinary norepinehrine, µg/24 h	44.2±21.6b	41.1±17.0b	55.2±19.3	48.9±20.7	52.2±19.4	.885
Urinary dopamine, µg/24 h	157.5±63.2b	149.3±84.8b	191.8±59.7	172.4±68.3	176.1±64.3	.295
ACTH, pg/mL	33.1±22.4b,e	40.4±19.5	43.7±19.5	39.4±21.1	–	–

Abbreviations: ACTH, adrenocorticotrophic hormone; APA, aldosterone‐producing adenoma; EH, essential hypertension; IHA, idiopathic hyperaldosteronism; PA, primary aldosteronism; UNAH, unilateral nodular adrenal hyperplasia. Data are expressed as mean±standard deviation. P ^a=unpaired t test between the PA and EH groups. Compared with the IHA group: ^b P<.001. ^c P<.05. Compared with group UNAH. ^d P<.001. ^e P<.05.

Urinary free cortisol levels were higher in the EH group compared with the PA group (P=.008; Figure 1a), and urinary free cortisol levels varied significantly in all 4 studied groups. The urinary free cortisol level in the IHA group was significantly higher than that in the APA group (P=.045) and the UNAH group (P=.027; Figure 1b); it was comparable between the APA group and the UNAH group.

Figure 1.

Evaluation and comparison of the status of the pituitary adrenal axis in the primary aldosteronism (PA) and essential hypertension (EH) groups. The status of the pituitary adrenal axis was determined by plasma and urinary free cortisol levels, as well as plasma adrenocorticotrophic hormone (ACTH) concentration. Data are presented as means±standard errors. (a) Comparison of urinary free cortisol level in patients with PA and EH. ^a P<.05 PA vs EH. (b) Comparison of urinary free cortisol levels in patients in the three PA subgroups. ^b1 P<.05. ^b2 P<.001 vs the IHA group. (c) Plasma ACTH concentrations in the three PA subgroups. ^b2 P<.001 vs the IHA group. ^c P<.05 vs the UNAH group.

The ACTH secretion was different from cortisol secretion in the APA, UNAH, and IHA groups. The plasma ACTH level of patients with APA was significantly lower than that of patients with IHA (P=.009) and UNAH (P=.046), while plasma ACTH level was comparable between the IHA and UNAH groups (Figure 1c). There were no data available on ACTH secretion in the EH group.

Adrenal Medulla Secretion Functions in the 4 Groups

Urinary epinephrine level was significantly higher in patients with EH than that in patients with PA (P=.000). Urinary excretion of norepinephrine and dopamine was comparable in EH and PA groups (Table 2; Figure 2a).

Figure 2.

Assessment of adrenal medulla secretion function in the 4 groups. Adrenal medulla secretion functions were determined by 24‐hour urinary epinephrine, norepinephrine, and dopamine. (a) Comparison of 24‐hour urinary epinephrine secretion in the primary aldosteronism (PA) and essential hypertension (EH) groups. ^a P<.05 PA vs EH. Twenty‐four–hour urinary norepinephrine and dopamine were comparable in the two groups. (b) Comparison of 24‐hour urinary epinephrine and dopamine excretion in patients in the 3 PA subgroups. ^b2 P<.001 vs the IHAgroup. ^c2 P<.001 vs the UNAH group. (c) Comparison of 24‐hour urinary norepinephrine excretion in patients in the 3 PA subgroups. ^b2 P<.001 vs the IHA group. (d) Comparison of 24‐hour urinary dopamine excretion in patients in the 3 PA subgroups. ^b2 P<.001 vs the IHA group.

Among the 3 PA subgroups, urinary epinephrine level was highest in the IHA group and lowest in the UNAH group, with the APA group's values being in the middle. The differences between the 3 groups were statistically significant (P=.000).

The features of norepinephrine and dopamine secretion were similar to epinephrine in the 3 subgroups. There was a significant difference between the APA and IHA groups (P=.000) and between the UNAH and IHA groups (P=.000). There was no significant difference between the APA and UNAH groups on urinary excretion of norepinephrine and dopamine (P>.05; Figure 2b).

Correlation Analysis of Biochemical Parameters and Clinical Features

Based on Pearson correlation analysis, there was a significant negative correlation between serum potassium levels and aldosterone levels (r=−0.374, P=.000). ACTH concentration was positively correlated with 8 am plasma cortisol concentration (r=0.269, P=.006). No correlation was found between plasma or urinary aldosterone levels or plasma ACTH concentration. However, there was a positive correlation between systolic BP and 8 am plasma cortisol concentration (r=0.142, P=.039) in all patients. There was no statistically significant correlation between BP and plasma aldosterone concentration, urinary aldosterone level, and plasma potassium concentration (all P>.05). Multiple linear regression analysis indicated that systolic BP was positively correlated with plasma ACTH levels (r=0.383, P=.016) and negatively correlated with urinary free cortisol levels (r=−0.305, P=.034), but did not have any significant relationship with plasma or urinary aldosterone levels in all patients. Serum potassium concentration had no correlation with those hormones including cortisol, ACTH, epinephrine, norepinephrine, and dopamine (all P>.05). No significant correlation was found between urinary free cortisol and systolic BP in each group.

Discussion

We have observed that patients with different PA subtypes had almost the same level of BP even with significant different concentration of aldosterone and plasma potassium. Our results showing similar BP in different subtypes of PA coincide with previous reports15, 16, 17, 18, 19, 20 but partly differ from other studies.5, 6, 9

Most of the patients with PA received long‐acting CCBs, which might diminish the difference in level of BP, because it may have a stronger influence in patients with higher BP than in those with lower BP.21 This defect could not be avoided because drugs have to be used to minimize the risk of too high blood pressure.10 For patients with diverse subtypes of PA, hypertension was mainly caused by oversecreted aldosterone. Antihypertensive agents such as Isoptin SR had only limited effect on BP in PA unlike mineral receptor antagonists.22, 23 The influence of this medication in this study could not be avoided but should be small.

Compared with EH, secretion of cortisol and adrenal medulla hormones was inhibited in the PA group. All these hormones are associated with BP and thus these changes in PA might influence the regulation of BP.

From the perspective of anatomy, cortisol, aldosterone, and sex hormones are generated in different zona of adrenal gland cortex. Epinephrine, norepinephrine, and dopamine are generated in the adrenal medulla. Adrenal adenoma or hyperplasia in zona glomerulosa may lead to inhibited secretion of other parts of the adrenal gland. This hypothesis is also supported by data showing significant lower cortisol and adrenal medullary hormones in APA and UNAH compared with IHA.

Cortisol and adrenal medullary hormones are all related to chronic stress. It is known that PA is a series of diseases in the adrenal gland (IHA can also occur in other parts of the human body24), which may cause stress of the adrenal gland. However, stress cannot explain the condition because it seems that stress is more severe in APA than in IHA. For further exploration, stress levels in different subtypes of PA should be assessed.

Feedback inhibited by oversecretion of aldosterone, plasma renin, and angiotensin was less in the APA and UNAH groups compared with the IHA group with a higher aldosterone level in patients with APA and UNAH than IHA. It is well‐known that angiotensin II is a stimulus for catecholamine secretion. The mechanism could be one of the reasons for lower catecholamine levels in patients with APA and UNAH.

We observed that systolic BP level was positively correlated with 8 am plasma cortisol level and plasma ACTH level without significant correlation with aldosterone. Firstly, it is known that all hormones, including aldosterone, cortisol, and catecholamines, can increase BP. The secretion of those hormones in PA and the composition of those hormones in different subtypes of PA are complicated. Our data imply an important role of those hormones. But we cannot say that there is no correlation between BP and aldosterone and catecholamines because the correlation is obvious.

Secondly, in different PA subtypes, there are higher aldosterone levels with lower cortisol and adrenal medullary hormones in APA and UNAH, which is reversed in IHA. All of these hormones can increase BP, which was determined to be similar. This may also explain why BP was similar in the 3 groups.

No correlation between aldosterone and BP in PA patients can be explained by aldosterone escape. Aldosterone escape can occur after treatment with medications that block the renin‐angiotensin‐aldosterone system or in undiagnosed PA.25 Before the study, some of the patients were given ACE inhibitor or ARB treatment, which can induce aldosterone escape in some patients, but ACE inhibitors or ARBs had been withdrawn for at least 4 weeks, which could diminish the influence. Aldosterone escape by PA itself could not be avoided.

Study Strengths and Limitations

There are some advances in this study. First, the number of patients studied was large. In addition, AVS had been practiced in our hospital since 2005, which allowed for an exact subtype diagnosis.

Some limitations of our study need to be highlighted. The BP was taken while the patients were taking verapamil SR, which may influence hormone levels. This could not be avoided for the purpose of minimizing the risk of uncontrolled hypertension in this study, however, and it may have different effects on different kinds of hypertension. We did not analyze whether CCBs affect aldosterone levels differently in patients with different forms of PA.

Before diagnosing PA and its subtypes, most patients had hypokalemia, which may influence secretion of aldosterone, cortisol, and adrenal medullary hormones.26 It is difficult to normalize serum potassium level in PA patients, however, especially in APA patients without mineralocorticoid receptor antagonists.27

Conclusions

APA‐related BP was almost the same as the other subtypes of PA, although more aldosterone is secreted in APA than the other types. Assessments of hormones indicated that higher secretion of aldosterone was accompanied by lower plasma potassium concentration, low function of adrenal zona fasciculate and medulla, and vice versa. This suggests that the similar BP between PA subgroups might be partially due to the changes of other adrenal hormones.

Disclosures

The authors confirm that this article is an original work, has not been previously published, and is not currently under consideration by another journal. The article is accurate, free of plagiarism, and does not infringe on anyone's copyright or other rights.

J Clin Hypertens (Greenwich). 2013;00:00–00. © 2013 Wiley Periodicals, Inc.