Aldosterone excess and resistance to 24-h blood pressure control

Recent studies have reported that primary aldosteronism (PA) has a prevalence of approximately 20% in patients with resistant hypertension [1-4]. These studies have been based largely on office blood pressure (BP) measurements. Thus, the degree to which hyperaldosteronism contributes to resistance of 24-h blood pressure control is not known.

Multiple studies indicate that ambulatory blood pressure monitoring (ABPM) is a stronger predictor of cardiovascular morbidity and mortality than office BP [5-12]. In treated hypertensive patients, higher ABPM levels are associated with an increased number of cardiovascular events [6] and in patients with primary hypertension or resistant hypertension, APBM better predicts cardiovascular risk than office BP measurements [12].

Observational studies suggest that patients with PA are at increased risk of cardiovascular complications compared to primary hypertensive patients with similar office BP levels [13-18]. It is not known whether this higher cardiovascular risk is in part attributable to differences in ABPM levels. The present study prospectively compared the 24-h ABPM levels in resistant hypertensive patients with or without hyperaldosteronism to test the hypothesis that aldosterone excess is associated with sustained resistance to antihypertensive treatment.

Consecutive patients referred to the University of Alabama at Birmingham (UAB) Hypertension Clinic for resistant hypertension were enrolled from August 2002 to August 2006. The protocol was approved by the UAB Institutional Review Board for Human Use and all participants provided their written informed consent before study participation. Resistant hypertension was defined as uncontrolled hypertension (> 140/90 mmHg) determined at two or more clinic visits in spite of the use of three or more antihypertensive medications at optimal doses. All patients were on a stable antihypertensive regimen for at least 4 weeks before biochemical evaluation. No medications were discontinued before evaluation except for spironolactone, triamterene, or amiloride, which were discontinued for at least 6 weeks prior to evaluation. In this case, thiazide diuretics were substituted. The clinic BP was measured with a mercury sphygmomanometer after the patient had rested by sitting for at least 5 min according to AHA guidelines [19] with the mean of two readings used for analysis. Secondary causes of hypertension other than hyperaldosteronism, such as renovascular hypertension, pheochromocytoma or Cushing's syndrome, were excluded by laboratory analysis and/or radiological imaging as clinically indicated.

All patients underwent non-invasive ABPM (Spacelabs model 90207; Spacelabs Inc., Richmond, Washington, USA). The monitor recorded systolic and diastolic BP every 20 min during the daytime (0600 to 2200 h) and every 30 min at night (2200 to 0600 h). The ambulatory data were included in the analysis if the monitoring period was ≥ 20 h and there were no periods of ≥ 2 h without measurements. Nocturnal decline was defined as the percentage difference in ambulatory day versus night BP levels. Dippers were defined as patients with ≥ 10% decline in both systolic and diastolic BP.

Patients with a history of congestive heart failure, recent myocardial infarction or stroke (< 6 months), diabetes on insulin treatment, or chronic kidney disease (creatinine clearance < 60 ml/min) were excluded from study participation. Use of oral contraceptives or hormone replacement therapy did not preclude enrollment [20].

Biochemical evaluation was done in all patients on an outpatient basis and included an early-morning (0800 h) ambulant PRA. A 24-h urine collection for aldosterone (Ualdo), sodium and creatinine was obtained during ingestion of the patient's routine diet.

PRA and Ualdo were measured by a commercial laboratory (Mayo Clinic Laboratories, Rochester, Minnesota, USA) using standard techniques. PRA levels were measured by radioimmunoassay. The reference range is 1.31-3.95 ng/ml per h for upright PRA and 2-16 μg/24 h for Ualdo.

High aldosterone status (H-Aldo) was defined as suppressed PRA (< 1.0 ng/ml per h) and elevated 24-h urinary aldosterone excretion (≥ 12 μg/24 h). Normal aldosterone status (N-Aldo) was defined as urinary aldosterone excretion < 12 μg/24 h and/or the absence of suppressed renin activity (≥ 1.0 ng/ml per h). The cut-off value of urinary aldosterone ≥ 12 μg/24 h was chosen to separate the high- and normal-aldosterone groups consistent with current recommendations for diagnosing PA [21].

Values are expressed as mean ± SD, unless otherwise noted. Values between groups were compared by a two-tailed t-test or Wilcoxon's test, where applicable, and categorical values were compared using a chi-squared test or Fisher's exact test. Ualdo, PRA, body mass index, gender and age were evaluated as independent predictors of ABPM values by multivariate analysis. P < 0.05 was considered statistically significant. The current analysis is observational and so the statistical analyses were not adjusted for multiple assessments.

A total of 251 patients were evaluated; 59 were classified as H-Aldo and 192 as N-Aldo. Patients were on an average of 4.2 ± 1.1 medications, which usually included a diuretic (85%), an angiotensin-converting enzyme (ACE) inhibitor or an angiotensin receptor blocker (ARB) (85%), a calcium channel blocker (76%), and a β-antagonist (85%). The prevalence of H-Aldo was similar among African-American and white participants (27.9 versus 25.0%, respectively). There were no differences in number or type of prescribed antihypertensive mediations between the H-Aldo and N-Aldo groups. There was a higher proportion of males in the H-Aldo group than in the N-Aldo group. In spite of similar office BP levels, daytime, night-time and 24-h systolic BP and diastolic levels were significantly higher in the H-Aldo compared to N-Aldo patients.

Due to the disproportionate number of males in the H-Aldo group, statistical assessments of demographic, biochemical and ABPM values in relation to aldosterone status were determined separately for males and females. H-Aldo male patients had lower serum potassium levels and lower mean PRA than N-Aldo male patients (Table 1). There was no difference in creatinine clearance between the H-Aldo and N-Aldo groups for either gender. The H-Aldo females were significantly younger than the N-Aldo females. There was no difference in age between the H-Aldo and N-Aldo patients in the entire cohort.

Office BP levels were not significantly different in the H-Aldo compared to the N-Aldo group for either gender (Table 2). By contrast, daytime, night-time, and 24-h systolic and diastolic BP values were all higher in H-Aldo compared to N-Aldo males (Table 2). Daytime, night-time and 24-h systolic BP were higher in H-Aldo compared to N-Aldo females. Daytime, night-time and 24-h diastolic BP tended to be higher in H-Aldo than in N-Aldo females, but these differences were not statistically significant.

The prevalence of patients with 'white-coat resistant hypertension', comprising high office BP (> 140/90 mmHg) and normal daytime ABPM values (< 135/85 mmHg) in spite of use of three or more antihypertensive medications, was significantly lower in H-Aldo males (6.5 versus 33.3%) and females (7.7 versus 40%) compared to N-Aldo patients of the same gender. Systolic nocturnal decline in males (5.0 ± 8.3% versus 3.0 ± 6.1%) and females (3.0 ± 9.0% versus 0.0 ± 9.7%) and diastolic nocturnal decline in males (8.0 ± 10.0% versus 5.5 ± 8.4%) and females (8.3 ± 9.8% versus 5.1 ± 7.9%) tended to be lower in H-Aldo compared to N-Aldo patients, but the differences did not reach statistical significance (P < 0.10 for males; P > 0.10 for females). There were more dippers in N-Aldo males (29.9 versus 10.9%) and females (22.8 versus 7.7%) compared to H-Aldo patients of the same gender, but, again, this difference was not statistically significant (P < 0.10 for males; P > 0.10 for females).

Multivariate analysis indicated that overall age, gender, race and aldosterone status (H-Aldo versus N-Aldo) were the best predictors of ambulatory blood pressure levels. White race was associated with lower systolic BP levels than black race (differences ranged from 5-8 mmHg depending on time of measurement; P < 0.03). For diastolic BP, gender, race, age and aldosterone status were important predictors. White patients had lower diastolic BP values (4-10 mmHg depending on time of measurement) than black patients (P < 0.03). Women had lower diastolic BP values compared to men (4-10 mmHg, P < 0.03). Night-time diastolic BP was lowest in white women (75 mmHg), whereas all other race/gender combinations were the same (85 mmHg).

Overall, the results indicate a significant aldosterone-age interaction, such that, as age increased, the effects of aldosterone on ABPM levels were more pronounced. In H-Aldo patients, as age increased, systolic BP increased by as much as 20-30 mmHg whereas systolic BP levels for N-Aldo patients remained relatively unchanged (P < 0.01)  . Diastolic BP decreased dramatically for N-Aldo patients (20-30 mmHg), whereas it stayed relatively constant in H-Aldo patients (P < 0.01)

Recent studies have reported that PA has a prevalence of approximately 20% in patients with resistant hypertension [1-4]. The present study indicates that treatment resistance in hypertensive patients with aldosterone excess reflects sustained increases in 24-h BP, including nocturnal BP. These results have important clinical implications in confirming hyperaldosteronism as a cause of true treatment resistance and suggest that aldosterone excess imparts an increased cardiovascular risk not reflected in office BP measurements.

Studies of hypertensive patients have consistently shown that ABPM levels predict cardiovascular risk better than office BP assessments. In a prospective assessment of 86 patients with resistant hypertension, Redon et al. [12] found that patients in the highest tertile of diastolic daytime BP had a 6.4-fold greater risk of suffering a cardiovascular event during 49 months of follow-up than patients in the lowest tertile even though there were no differences in office BP values between the two groups. Studies also suggest that among patients with primary hypertension elevated night-time ambulatory BP values are more strongly related to cardiovascular morbidity and mortality than elevated daytime ambulatory BP values [6].

As in the primary hypertensive population, patients with office-resistant hypertension but normal ABPM levels are at reduced cardiovascular risk compared to patients with sustained elevation in ABPM levels [12,22,23]. ABPM studies also indicate that target-organ dysfunction is directly related to sustained hypertension. In an evaluation of patients with resistant hypertension, Mezzetti et al. [24] observed a significantly higher mean left ventricular mass index in patients with high ambulatory daytime BP levels compared to patients with normal ambulatory levels. All patients had high office BP in spite of use of a three-drug antihypertensive regimen. In a separate study, Muxfeldt et al. [22] related ABPM levels to target organ damage in 286 patients with resistant hypertension. Patients classified as having true resistant hypertension (high office and high ABPM values) had more nephropathy and left ventricular hypertrophy than patients with white-coat resistant hypertension (high office but normal daytime ABPM values).

Observational studies suggest that hypertensive patients with PA, independent of office BP measurements, are at greater risk of cardiovascular complications than primary hypertensive patients [13-18]. Based on the above reports relating higher sustained 24-h BP values to increased cardiovascular risk, the present study suggests that the higher cardiovascular morbidity and mortality associated with hyperaldosteronism is in part BP related. That is, greater cardiovascular risk in patients with hyperaldosteronism would be anticipated simply based on the higher sustained ambulatory BP levels demonstrated in the current evaluation.

Although not statistically significant, high aldosterone patients in the present study tended to have less nocturnal decline in both systolic and diastolic BP compared to normal aldosterone patients. Less dipping of nocturnal blood pressure would be expected to further contribute to an increased risk of cardiovascular complications. Reduced nocturnal dipping of BP is associated with increased target-organ damage and cardiovascular risk [25-28]. In a substudy of the Syst-Eur Trial, an inverse relationship between cardiovascular risk and nocturnal BP decline was observed, with the overall cardiovascular risk increasing by 41% for each 10% increase in the systolic night-to-day ratio [28].

Evaluations of circadian BP in patients with PA have yielded conflicting results and none has specifically evaluated patients with resistant hypertension [29-35]. Mansoor et al. [29] described no differences in the ABPM pattern between hypertensive patients with PA and control hypertensive patients without PA. In their study, ambulatory daytime, night-time and nocturnal decline were similar in PA patients and in the control group. By contrast, Zelinka et al. [30] described significantly less nocturnal decline in systolic and diastolic BP in patients with PA compared to hypertensive controls. In their study, patients with idiopathic hyperaldosteronism presented with significantly higher 24-h systolic and diastolic BP than primary hypertensive patients, but daytime and night-time BP levels were not specifically commented on and the study participants did not have resistant hypertension. In a recent study, Ceruti et al. [35] described the ABPM patterns in patients with secondary hypertension due to adrenal diseases. The prevalence of nondippers was 51.5% in patients with PA, 43.2% in patients with pheochromocytoma, 34.2% in patients with primary hypertension and 15% in normotensive individuals. The results of the present study are novel in demonstrating that aldosterone excess is associated with a resistance to 24-h BP control in spite of intensive pharmacologic treatment. This resistance to 24-h BP control would be expected to increase cardiovascular risk compared to patients with normal aldosterone levels who have similar resistance to control of office BP but are more likely to have overall lower 24-h BP levels.

A large body of experimental data indicates that aldosterone excess, in combination with a high dietary salt intake, has pro-inflammatory and profibrotic effects that contribute to target organ deterioration [36-38]. In most of these studies, these effects are independent of BP. The present study neither confirms nor excludes the direct effects of aldosterone on the progression of cardiovascular disease, but it does suggest an increased cardiovascular risk secondary to overall higher ABPM levels.

The mechanism of the high sustained ABPM levels in our patients with aldosterone excess cannot be inferred from the current study design. Aldosteronism induces inappropriate sodium and fluid retention, resulting in intravascular volume expansion. The higher sustained BP levels and blunted nocturnal dipping may be a consequence of this excess fluid retention. If so, the sustained ABPM levels may respond preferentially to the use of mineralocorticoid receptor antagonists.

In addition to increases in sodium and fluid retention, aldosterone may contribute to the development of resistant hypertension through direct effects on the vasculature. Aldosterone excess impairs endothelial function [39]. In addition, both animal and human studies indicate that aldosterone promotes hypertrophy and fibrosis of arterial vessels [40,41]. Combined, these effects may contribute to chronic elevations in blood pressure and resistance to pharmacologic treatment.

Mineralocorticoid receptor antagonists are an effective therapeutic option for the treatment of resistant hypertension [42-44]. We have previously reported that the use of low-dose spironolactone (12.5-50 mg/day) in patients with uncontrolled BP on an average of four medications provides significant additional antihypertensive benefit [42]. After 6 months of follow-up, office systolic BP was reduced on average by 25 mmHg and diastolic BP by 12 mmHg. Recently, investigators from the Anglo-Scandinavian Cardiac Outcomes Trial-Blood Pressure Lowering Arm (ASCOT) reported a similar benefit when spironolactone was used as fourth-line therapy, reducing office systolic BP and diastolic BP by 21.9 and 9.5 mmHg, respectively, in 1411 participants [44]. Whether spironolactone provides similar reductions in ambulatory blood pressure levels is not known. Prior studies of treatment of PA have demonstrated increased nocturnal dipping with spironolactone use or dietary salt restriction, but the evaluated patients did not have resistant hypertension so it is not known whether a similar benefit would occur in patients with more difficult-to-control hypertension [45,46].

The current analysis is strengthened by its prospective design, the inclusion of a large number of patients with resistant hypertension, and the measurement of 24-h urinary aldosterone excretion. Study limitations include having performed the biochemical evaluations during ongoing antihypertensive therapy and only including patients with resistant hypertension. Diuretics, ACE inhibitors and ARBs can increase PRA, which may have resulted in falsely excluding some patients from being H-Aldo. Conversely, β-blockers can have the opposite effect by suppressing PRA. Avoiding these drug effects would have required withdrawing these classes of agents, which would have been potentially hazardous in the setting of already poorly controlled hypertension. Any drug effects, however, should have been balanced between the two groups because the number and types of medications were the same in the high- and normal-aldosterone groups. Furthermore, although the above antihypertensive medications may affect PRA, they tend to have minimal effects on aldosterone secretion [47,48]. Finally, having only studied patients with resistant hypertension, we do not know to what degree the results obtained apply to patients with more easily controlled hypertension.

At least one prior study has suggested that male patients with resistant hypertension have higher ABPM levels than female patients [49]. Our multivariate analysis of the entire cohort confirms such a gender difference, which we suggest may be attributable in part to higher aldosterone levels in male patients. Interestingly, in this earlier study, both male gender and lower potassium levels were associated with higher ABPM levels, suggestive of a link to aldosterone excess.

The results of the present study indicate a novel interaction between aldosterone and age, with the effects of aldosterone on ambulatory BP levels becoming more pronounced with increasing age. To the best of our knowledge, such an interaction has not been previously described. It is interesting and perhaps related that, in ASCOT, the best predictor to response to spironolactone (at least for diastolic BP) was older age (> 60 years) [44]. Although speculative, the sodium-retaining effects of aldosterone may be enhanced in aging patients as renal function declines. Such an effect could contribute to the increased salt-sensitivity and lower PRA levels that characterize hypertension in the elderly [50].

We did not systematically exclude obstructive sleep apnea (OSA) as a contributory cause of less nocturnal dipping. However, we have previously reported that the prevalence of OSA is extremely high (85%) in patients with resistant hypertension [51], and the large majority of the current cohort would be expected to have OSA.

We and other laboratories have reported a prevalence of PA confirmed by salt loading of approximately 20% in patients with resistant hypertension [1-4]. Based on that percentage, 50 of the 59 patients designated as H-Aldo would have been expected to have classical PA, if confirmation testing had been performed. To avoid risk of severe blood pressure elevations, confirmatory testing with dietary salt loading was not performed as part of this protocol. However, the nine patients who might have been excluded from having classical PA undoubtedly still had some degree of aldosterone excess as indicated by their high urinary excretion of aldosterone and suppressed renin activity.

The results of the present study support the role of aldosterone excess in contributing to resistance to antihypertensive treatment, particularly in elderly patients in whom the effects of aldosterone excess on BP levels were most pronounced. The results further indicate that high aldosterone levels impart increased cardiovascular risk not reflected in office BP measurements. Clinical studies designed to analyze the relationship between aldosterone excess and cardiovascular outcomes should include ABPM to identify aldosterone effects independent of office BP measurements.

 Journal of Hypertension:Volume 25(10)October 2007p 2131-2137