Optimizing the aldosterone-to-renin ratio cut-off for screening primary aldosteronism based on cardiovascular risk: a collaborative study

ABSTRACT

Objectives

Aldosterone-to-renin ratio (ARR) based screening is the first step in the diagnosis of primary aldosteronism (PA). However, the guideline-recommended ARR cutoff covers a wide range, from the equivalent of 1.3 to 4.9 ng·dl⁻¹/mIU∙l⁻¹. We aimed to optimize the ARR cutoff for PA screening based on the risk of cardiovascular diseases (CVD).

Methods

Longitudinally, we included hypertensive participants from the Framingham Offspring Study (FOS) who attended the sixth examination cycle and followed up until 2014. At baseline (1995–1998), we used circulating concentrations of aldosterone and renin to calculate ARR (unit: ng·dl⁻¹/mIU∙l⁻¹) among 1,433 subjects who were free of CVD. We used spline regression to calculate the ARR threshold based on the incident CVD. We used cross-sectional data from the Chongqing Primary Aldosteronism Study (CONPASS) to explore whether the ARR cutoff selected from FOS is applicable to PA screening.

Results

In FOS, CVD risk increased with an increasing ARR until a peak of ARR 1.0, followed by a plateau in CVD risk (hazard ratio 1.49, 95%CI 1.19–1.86). In CONPASS, when compared to essential hypertension with ARR < 1.0, PA with ARR ≥ 1.0 carried a higher CVD risk (odds ratio 2.24, 95%CI 1.41–3.55), while essential hypertension with ARR ≥ 1.0 had an unchanged CVD risk (1.02, 0.62–1.68). Setting ARR cutoff at 2.4 ~ 4.9, 10% ~30% of PA subjects would be unrecognized although they carried a 2.45 ~ 2.58-fold higher CVD risk than essential hypertension.

Conclusions

The CVD risk-based optimal ARR cutoff is 1.0 ng·dl⁻¹/mIU∙l⁻¹ for PA screening. The current guideline-recommended ARR cutoff may miss patients with PA and high CVD risk.

Clinical Trial Registration

ClinicalTrials.gov (NCT03224312)

KEYWORDS:

• Aldosterone-to-renin ratio
• cardiovascular risk
• cutoff
• primary aldsoteronism
• screening

Introduction

Primary aldosteronism is an adrenal disorder characterized by autonomous aldosterone production and suppressed renin (1). Compared with essential hypertension, patients with primary aldosteronism have an increased risk of cardiovascular events such as stroke, coronary artery disease and chronic heart failure (2). Targeted treatment for primary aldosteronism, either with adrenal surgery or mineralocorticoid receptor antagonists, can reduce the risk of cardiovascular events (3). Hence, early detection and treatment of primary aldosteronism is important for long-term cardiovascular health (4).

Screening with an aldosterone-to-renin ratio (ARR) is the most widely used first-step in the diagnosis of primary aldosteronism (5–7). However, the recommended ARR cutoffs vary considerably between guidelines: the US Endocrine Society clinical practice guideline suggested cutoffs of 2.4, 3.7, or 4.9 ng·dl⁻¹/mIU∙l⁻¹, while the European Society of Hypertension recommended values of 1.3, 2.0, or 2.7 ng·dl⁻¹/mIU∙l⁻¹. A recent meta-analysis reported that it is hard to recommend a single ARR threshold and one of the most important reasons was the heterogenous confirmatory criteria for primary aldosteronism (8) without a gold standard for diagnosis (1,9). There is a need for a more evidence-based ARR cutoff as it can significantly impact clinical practice and patient care.

The major objective for the management of primary aldosteronism is to reduce cardiovascular risk. As an indicator of autonomous aldosterone secretion, ARR was shown to be significantly associated with cardiovascular events in the general population (10) or patients with essential hypertension (11). In the current study, we aimed to explore and validate the optimal ARR cutoff based on the risk of cardiovascular events. We firstly used longitudinal data of Framingham Offspring Study (FOS) to establish the optimal ARR cutoff based on the non-linear relationship between ARR and cardiovascular events among hypertensive patients. In the cross-sectional analysis of the Chongqing Primary Aldosteronism Study (CONPASS), we enrolled hypertensive participants who were not taking interfering medications and received at least one confirmatory test for primary aldosteronism. We further explored whether the optimal ARR cutoff selected from FOS is applicable to the screening of primary aldosteronism in hypertensive patients from CONPASS.

Methods

Study population

Study design of the Framingham Offspring Study has been reported since 1971–1975 (12). In the current analysis, we included participants attending the sixth examination cycle (1995–1998, current baseline) of the Framingham Offspring Study (13–15). Participants were eligible for the current analyses if they had baseline measurements of serum aldosterone and plasma renin concentration, and were followed up on until December 31, 2014, and had access to medical records for baseline and follow-up cardiovascular events. Subjects with a baseline history of cardiovascular diseases were not included in the current analysis. The institutional review board of Boston University Medical Center approved the protocol.

CONPASS is a prospective cohort which was established in 2016 and aimed to explore long-term cardiovascular risk in patients with primary aldosteronism and essential hypertension (ClinicalTrials.gov Identifier: NCT03224312). We used CONPASS database to perform a cross-sectional analysis and validate the ARR cutoff set by the Framingham Offspring Study. In order to assess the long-term cardiovascular outcomes among individuals with primary aldosteronism and essential hypertension, CONPASS enrolled a diverse group of patients, ranging from individuals who had newly diagnosed hypertension in primary care to individuals who had resistant hypertension in the referral center. The ethical committee of Chongqing Medical University approved CONPASS and written informed consent was provided by all study participants.

Physical examination and laboratory assessment

Evaluations of anthropometric data, physical examinations, and other laboratory tests in FOS were previously published (13–17). We described these details in the supplementary materials (18). Circulating renin and aldosterone concentrations were measured in the sixth examination cycle. An immunochemiluminometric assay (Nichols Advantage ® Direct Renin assay) was used to measure plasma renin concentration (PRC) (15), and a radioimmunoassay (Quest Diagnostics) was used to measure serum aldosterone concentration (SAC), without withdrawal of antihypertensive medications (13,14). In CONPASS, we also reported anthropometric information, physical examinations, and laboratory tests in previous publications (4,19–21). We collected blood samples in the morning and used automated chemiluminescence immunoassays (LIAISON; DiaSorin, Italy) to measure plasma aldosterone concentration (PAC) and PRC. We withdrew all antihypertensive medications that affected the renin-angiotensin-aldosterone system and corrected hypokalaemia before the assessment of renin and aldosterone concentrations.

Diagnosis of primary aldosteronism and essential hypertension in CONPASS

For CONPASS, we asked participants to undergo at least one confirmation tests if they had ARR ≥ 1.0 ng·dl⁻¹/mIU∙l⁻¹, or ARR < 1.0 ng·dl⁻¹/mIU∙l⁻¹ but were willing to receive confirmatory tests. Confirmation tests included saline infusion test (positive: PAC-posttest ≥8.0 ng·dl⁻¹), captopril challenge testing (positive: PAC-posttest ≥11.0 ng·dl⁻¹), and fludrocortisone suppression test (positive: PAC-posttest ≥6.0 ng·dl⁻¹) (20,22,23). Subjects were diagnosed as primary aldosteronism if they had one or more positive confirmatory tests. Subjects were diagnosed as essential hypertension if they had a negative confirmatory test, or ARR < 1.0 ng·dl⁻¹/mIU∙l⁻¹ without any feature of primary aldosteronism that recommended by the Endocrine Society guideline (5,6). For the diagnosis of essential hypertension, we further excluded other secondary hypertension, including Cushing’s syndrome, pheochromocytoma and paraganglioma, Graves’s disease and renal artery stenosis (4,19,24).

Assessment of outcomes

For FOS, the primary outcome was the incidence of cardiovascular diseases (CVD) which consisted of coronary heart disease, congestive heart failure, stroke and transient ischemic attack. For CONPASS, we identified CVD based on a diagnosis of coronary heart disease, stroke (including transient ischemic attack), or congestive heart failure. Two senior physicians from the First Affiliated Hospital of Chongqing Medical University independently confirmed the diagnosis of CVD in CONPASS participants. Further information on the outcomes were described in the supplementary data (18).

Statistical analysis

We used the Kolmogorov – Smirnov test to assess normality for continuous variables. Categorical variables were presented as percentage. Normal distributed variables were presented as means and 95% confidence interval (CI) and skewed normal distributed variables were presented as the median (interquartile range). The categorical data was analyzed using the chi-square test.

In FOS, we used restricted cubic spline regression models to fit the relationship between ARR and CVD, with curves delineated based on four equally spaced knots at the 25th, 50th, 75th, and 95th percentiles. We investigated the associations between different ARR cutoff values and the risk of CVD incidence by using Cox proportional hazards regression. In CONPASS, we used multivariable logistic regression models to calculate the odds ratios (ORs) and 95%CI of CVD risk across varied ARR cutoff values. We adjusted for multiple confounders which included age, gender, body mass index, systolic blood pressure, current smoking status, alcohol consumption, total cholesterol, the presence or absence of diabetes, antihypertensive medication use (for FOS). We further conducted a stepwise selection procedure to explore the importance of CVD risk factors in the multivariable model. We used the analysis of receiver operator characteristic curve to calculate the diagnostic parameters, including sensitivity, specificity, positive and negative likelihood ratio of the selected ARR cutoff. Stata version 15, SPSS (version 22.0) and MedCalc (Version 11.4.2.0) were used for all statistical analyses.

Results

Demographic and biochemical characteristics of included participants

The flow chart for the prospective analysis of Framingham Offspring Study (FOS) and the cross-sectional analysis of the Chongqing Primary Aldosteronism Study (CONPASS) were shown in Supplementary Figure S1 (18). There were 1,433 participants with hypertension included in FOS at baseline, and 2,556 hypertensive participants were included in CONPASS. We summarized characteristics of the study participants from FOS (baseline) and CONPASS according to their ARR tertiles in Table 1.

Table 1. Demographic, clinical and biochemical characteristics according to tertiles of ARR among hypertensive participants from FOS and CONPASS.

	ARR Tertiles in FOS			ARR Tertiles in CONPASS
	Tertile 1	Tertile 2	Tertile 3	Tertile 1	Tertile 2	Tertile 3
Male/Female	282/191	260/235	179/286	602/242	401/467	295/549
Median age (yr)	59.0 (53.0,66.0)	62.0 (54.0,68.0)	63.0 (56.0,69.0)	46.0 (35.0,56.0)	51.0 (43.0,60.0)	52.0 (44.0,61.0)
Body-mass index (kg/m2	28.5 (25.7,32.1)	28.2 (25.3,31.9)	28.5 (25.5,31.7)	25.0 (22.6,27.5)	25.3 (23.2,27.7)	24.4 (22.4,26.7)
SBP (mmHg)	134 (122,145)	138 (127,149)	141 (132,154)	150 (140,162)	152 (141,163)	150 (139,162)
DBP (mmHg)	79.0 (71.0,85.0)	80.0 (72.0,86.0)	81.0 (74.0,86.0)	94.5 (85.1,102)	92.5 (85.2,100)	91.0 (82.0,100)
History of Diabetes (%)	76 (15.7%)	65 (13.5%)	67 (14.4%)	94 (11.1%)	102 (11.8%)	111 (13.2%)
Triglyceride (mg/dl)	128 (88.0,175)	126 (91.5,180)	131 (98.0,200)	139.0 (91.2,212.5)	139.0 (98.3,199.4)	110.7 (78.8,173.5)
HDL-c(mg/dl)	46.0 (37.0,57.0)	47.0 (38.0,60.0)	48.0 (40.0,60.0)	47.6 (40.2,56.8)	48.0 (39.8,57.2)	46.8 (37.5,56.1)
LDL-c(mg/dl)	125 (106,147)	129 (110,149)	126 (106,150)	116.4(92.9,139.9)	114.5 (94.7,138.1)	100.9 (80.8,123.4)
FPG (mg/dl)	100 (94.0,113)	100 (93.0,110)	100 (92.0,111)	99.0 (90.0,111.6)	99.0 (90.0,109.8)	95.4 (86.4,104.4)
Aldosterone concentration (ng/dl)^a	8.00 (6.00,11.0)	10.0 (7.00,15.0)	12.0 (9.00,17.0)	8.54 (5.71,13.2)	10.9 (7.59,16.9)	19.9 (12.1,33.6)
Plasma renin concentration (mIU/l)^b	32.0 (18.0,73.0)	12.0 (8.00,17.0)	5.00 (3.00,8.00)	29.9 (18.6,48.7)	9.76 (6.06,15.0)	1.98 (0.94,3.60)
ARR (ng·dl^−1/mIU·l^−1)	0.30 (0.12,0.44)	0.88 (0.71,1.07)	2.25 (1.67,3.50)	0.31 (0.19,0.43)	1.10 (0.78,1.71)	7.88 (4.39,24.1)

Data were expressed as median (interquartile range) and the number (%); Abbreviations: ARR, Aldosterone-renin ratio; FOS, the Framingham Offspring Study; CONPASS: Chongqing Primary Aldosteronism Study; SBP, systolic blood pressure; DBP, diastolic blood pressure; LDL-c, low density lipoprotein cholesterol; HDL-c, high density lipoprotein cholesterol; FPG, fasting plasma glucose.

^aserum aldosterone concentration in FOS patients was determined by radioimmunoassay, plasma aldosterone concentrations in CONPASS patients was determined by automated chemiluminescent immunoassay, ^bplasma renin concentration in FOS patients was determined by the immunochemiluminometric assay, and by automated chemiluminescence immunoassays in CONPASS patients. Aldosterone-renin ratio was calculated as renin divided by aldosterone.

ARR cut-off and CVD risk in FOS

Among the 1,433 hypertensive participants from FOS who were free of CVD at baseline, 398 subjects (27.78%) developed CVD during a median of 18 years of follow-up. The restricted cubic spline regression showed that the risk of CVD increased with increasing ARR, reaching a peak at an ARR of 1.0 (unit: ng·dl⁻¹/mIU∙l⁻¹) and plateaued thereafter (Figure 1a). When compared to persons with ARR < 1.0 (reference), participants whose ARR ≥ 1.0 showed a significantly higher risk of CVD (HR 1.49 [95%CI 1.19–1.86]) (Figure 1b, Table 2).

Figure 1. Relationships between aldosterone-renin-ratio and risk of cardiovascular diseases, among hypertensive participants from FOS. panel a shows the hazard ratios (HR) and 95% CIs of aldosterone-renin-ratio (ARR) and CVD incidence. The HRs and 95% CIs were delineated on the basis of restricted cubic spline regression with four equally spaced knots at 25th, 50th, 75th, and 95th percentiles. Red dot line of vertical axis (HR = 1.0) was used as reference (a). Panel B shows the cumulative probability of CVD according to baseline ARR classification (<1 ng·dl⁻¹/mIU∙l⁻¹ or ≥ 1 ng·dl⁻¹/mIU∙l⁻¹) (b). All these effects were based on multivariate models adjusted for age, sex, body mass index, current smoking, alcohol consumption, systolic blood pressure, with or without diabetes, total cholesterol, and use of antihypertensive drugs (for FOS).

Table 2. Different aldosterone-renin-ratio cutoffs and risk of cardiovascular diseases among hypertensive participants from FOS.

ARR	Incident/Total	HR (95%CI)
Cutoff: 1.0 ng·dl^−1/mIU·l^−1
<1.0	183/774	reference
≥1.0	215/659	1.49(1.19,1.86)^a
Cutoff: 2.4 ng·dl^−1/mIU·l^−1
<2.4	333/1215	reference
≥2.4	65/218	1.13(0.85,1.50)
<1.0	183/774	reference
1.0–2.4	150/441	1.53(1.20,1.95)^a
≥2.4	65/218	1.39(1.02,1.88)^a
Cutoff: 3.7 ng·dl^−1/mIU·l^−1
<3.7	363/1335	reference
≥3.7	35/98	1.40(0.97,2.02)
<1.0	183/774	reference
1.0–3.7	180/561	1.70(1.16,2.51)^a
≥3.7	35/98	1.45(1.15,1.83)^a
Cutoff: 4.9 ng·dl^−1/mIU·l^−1
<4.9	378/1378	reference
≥4.9	20/55	1.43(0.86,2.28)
<1.0	183/774	reference
1.0–4.9	195/604	1.76(1.08,2.86)^a
≥4.9	20/65	1.46(1.16,1.83)^a

All these effects were based on multivariate models adjusted for age, sex, body mass index, current smoking, alcohol consumption, systolic blood pressure, with or without diabetes, total cholesterol, and use of antihypertensive drugs.

Abbreviations: ARR, aldosterone-renin-ratio.

^aP value was less than 0.05.

We further calculated CVD risk when setting ARR cutoff values at 2.4, 3.7 and 4.9 (unit: ng·dl⁻¹/mIU∙l⁻¹), as recommended by the Endocrine Society clinical practice guideline. Compared to persons with ARR < 2.4 (reference), participants with ARR ≥ 2.4 showed a similar CVD risk (HR 1.13 [95%CI 0.85–1.50]). After dividing ARR < 2.4 into ARR < 1.0 (reference) and ARR 1.0–2.4, when compared to the reference, participants whose ARR ranged 1.0–2.4 or ARR ≥ 2.4 had a significantly higher risk of CVD (HR 1.53 [95%CI 1.20–1.95] for ARR ranged 1.0–2.4, 1.39 [1.02–1.88] for ARR ≥ 2.4). Similarly, when compared to persons with ARR < 3.7 (or 4.9), participants with ARR ≥ 3.7 (or 4.9) had unchanged risk of CVD. After dividing ARR < 3.7 (or ARR < 4.9) into ARR < 1.0 (reference) and ARR 1.0–3.7 (or 1.0–4.9), when compared to the reference, we observed a significantly higher risk of CVD among participants whose ARR ranged between 1.0–3.7 or 1.0–4.9, or participants whose ARR ≥ 3.7 or ARR ≥ 4.9 (Table 2).

ARR cut-off and CVD risk in CONPASS

Among the 2,556 hypertensive participants from CONPASS, 158 participants (6.18%) had a history of CVD. As participants in CONPASS had a final diagnosis of essential hypertension or primary aldosteronism based on confirmatory tests, we further explored the relationship between ARR cutoffs and CVD risk according to the final diagnosis. When compared to patients with essential hypertension and ARR < 1.0 (unit: ng·dl⁻¹/mIU∙l⁻¹) (reference), those with essential hypertension and ARR ≥ 1.0 showed unchanged CVD risk (OR 1.02 [95%CI 0.62–1.68]), while those with primary aldosteronism and ARR ≥ 1.0 exhibited a significantly higher risk of CVD (OR 2.24 [95%CI 1.41–3.55]). Among 12 participants with primary aldosteronism and ARR < 1.0, no history of CVD was reported. When compared to patients with essential hypertension and ARR < 2.4 (reference), those with essential hypertension and ARR ≥ 2.4 showed unchanged CVD risk, while those with primary aldosteronism showed a significantly higher risk of CVD whether they had an ARR < 2.4 (OR 2.58 [95%CI 1.02–6.49]) or ARR ≥ 2.4 (OR 2.07 [95%CI 1.35–3.18]). Setting ARR cutoffs at 3.7 or 4.9 resulted in CVD risk profiles similar to those observed at an ARR cutoff of 2.4 (Table 3).

Table 3. Different aldosterone-renin-ratio cutoffs and risk of cardiovascular diseases among participants from CONPASS.

ARR	Final Diagnosis^b	Case/Total	OR (95%CI)
Cutoff: 1.0 ng·dl^−1/mIU·l^−1
<1.0	Essential Hypertension	59/1213	reference
<1.0^a	Primary Aldosteronism	0/12	-
≥1.0	Essential Hypertension	41/729	1.02(0.62,1.68)
≥1.0	Primary Aldosteronism	58/602	2.24(1.41,3.55)^c
Cutoff: 2.4 ng·dl^−1/mIU·l^−1
<2.4	Essential Hypertension	80/1600	Reference
<2.4	Primary Aldosteronism	6/67	2.58(1.02,6.49)^c
≥2.4	Essential Hypertension	20/342	0.90(0.50,1.63)
≥2.4	Primary Aldosteronism	52/547	2.07(1.35,3.18)^c
Cutoff: 3.7 ng·dl^−1/mIU·l^−1
<3.7	Essential Hypertension	86/1732	Reference
<3.7	Primary Aldosteronism	11/127	2.45(1.21,4.96)^c
≥3.7	Essential Hypertension	14/210	1.30(0.66,2.57)
≥3.7	Primary Aldosteronism	47/487	2.21(1.43,3.41)^c
Cutoff: 4.9 ng·dl^−1/mIU·l^−1
<4.9	Essential Hypertension	88/1792	Reference
<4.9	Primary Aldosteronism	16/178	2.48(1.35,4.54)^c
≥4.9	Essential Hypertension	12/150	1.68(0.82,3.43)
≥4.9	Primary Aldosteronism	42/436	2.24(1.43,3.50)^c

All these effects were based on multivariate models adjusted for age, sex, body mass index, current smoking, alcohol consumption, systolic blood pressure, with or without diabetes, and total cholesterol.

Abbreviations: ARR, aldosterone-renin-ratio; primary aldosteronism was diagnosed if one of confirmatory tests was positive.

^aOR (95%CI) cannot be calculated because the case of CVD was 0.

^bIn the current analysis, the reference standard of primary aldosteronism was established based on a clinically comprehensive evaluation: we asked participants who had ARR ≥1.0 ng∙dl⁻¹/mIU∙l⁻¹, or whose ARR <1.0 ng∙dl⁻¹/mIU∙l⁻¹ but were strongly suspected as primary aldosteronism to undergo at least one confirmation tests (saline infusion test, captopril challenge testing, or fludrocortisone suppression test), and subjects were diagnosed as primary aldosteronism if they had a positive confirmatory test. We considered subjects as essential hypertension in the condition of ARR <1.0 ng∙dl⁻¹/mIU∙l⁻¹ or ARR ≥1.0 ng∙dl⁻¹/mIU∙l⁻¹ with a negative confirmatory test.

^cP value was less than 0.05.

Stepwise selection of CVD risk factors in FOS and CONPASS

Supplementary Table S1 showed the selected risk factors with p values from the stepwise multivariate Cox models (for FOS) or Logistic models (for CONPASS). Dependent variable is time to first CVD in FOS and CVD history in CONPASS, respectively. The significance of the risk factor after accounting for all risk factors in models adjusted by age and sex is represented by the P-value. In FOS, in addition to a history of diabetes, being a current smoker and having dyslipidaemia, the parameter of ARR ≥ 1.0 ng·dl⁻¹/mIU∙l⁻¹ at baseline (screening) was the fourth most significant variable in the model (p = .006). In CONPASS, the parameter of ARR ≥ 1.0 ng·dl⁻¹/mIU∙l⁻¹ (in the context of confirmed primary aldosteronism) was third most significant in the model (p = .005), following a history of diabetes and dyslipidaemia (Supplementary Table S1) (18).

Diagnostic accuracy of the optimized ARR cut-off for screening in CONPASS

We evaluated the diagnostic accuracy of different ARR cutoffs for screening primary aldosteronism. Setting the ARR cutoff at 1.0 ng·dl⁻¹/mIU∙l⁻¹, the sensitivity and specificity for screening were 0.98(0.97,0.99) and 0.63(0.60,0.65), respectively (Table 4), suggesting that less than 2% of patients with primary aldosteronism would be missed using this ARR threshold. In addition, there was no CVD history in the 12 participants with primary aldosteronism who had baseline ARR < 1.0 ng·dl⁻¹/mIU∙l⁻¹ (Table 3). Setting ARR cutoffs at 2.4, 3.7 and 4.9 ng·dl⁻¹/mIU∙l⁻¹, the sensitivity for screening was 0.89(0.86,0.91), 0.79(0.76,0.82) and 0.71(0.67,0.74), respectively (Table 4), suggesting that 10%~30% of participants with primary aldosteronism would be missed at the higher ARR cutoffs.

Table 4. Diagnostic accuracy of aldosterone-to-renin ratio for the screening of primary aldosteronism in CONPASS ^a..

ARR Cutoffs (ng·dl^−1/mIU·l^−1)	Sensitivity	Specificity	Positive LR	Negative LR
1.0	0.98(0.97,0.99)	0.63(0.60,0.65)	2.62 (2.50,2.70)	0.03(0.02,0.05)
2.4	0.89(0.86,0.91)	0.82(0.81,0.84)	5.07 (4.90,5.30)	0.13(0.10,0.20)
3.7	0.79(0.76,0.82)	0.89(0.88,0.91)	7.34 (7.00,7.70)	0.24(0.20,0.30)
4.9	0.71(0.67,0.74)	0.92(0.91,0.93)	9.17 (8.70,9.70)	0.32(0.30,0.40)

The ARR cutoff value of 1.0 ng∙dl⁻¹/mIU∙l⁻¹ was selected by the prospective analysis of ARR and CVD in FOS, while ARR cutoffs for 2.4, 3.7 and 4.9 ng∙dl⁻¹/mIU∙l⁻¹ were recommended by the US Endocrine Society clinical practice guideline.

Abbreviations: ARR, aldosterone-to-renin ratio; LR, Likelihood Ratio.

^aIn the current analysis, the reference standard of primary aldosteronism was established based on a clinically comprehensive evaluation: participants were asked to undergo at least one confirmation tests if they had ARR ≥1.0 ng∙dl⁻¹/mIU∙l⁻¹, or ARR <1.0 ng∙dl⁻¹/mIU∙l⁻¹ but were willing to receive confirmatory tests. Confirmation tests included saline infusion test (positive: PAC-posttest ≥ 8.0 ng∙dl⁻¹), captopril challenge testing (positive: PAC-posttest ≥ 11.0 ng∙dl⁻¹), and fludrocortisone suppression test (positive: PAC-posttest ≥ 6.0 ng∙dl⁻¹). Subjects were diagnosed as primary aldosteronism if they had one or more positive confirmatory tests. Subjects were diagnosed as essential hypertension if they had a negative confirmatory test, or ARR <1.0 ng∙dl⁻¹/mIU∙l⁻¹ without any feature of primary aldosteronism that recommended by the Endocrine Society guideline. For the diagnosis of essential hypertension, we further excluded other secondary hypertension, including Cushing’s syndrome, pheochromocytoma and paraganglioma, Graves’s disease and renal artery stenosis.

Discussion

Beyond previous studies which selected ARR cutoffs based on heterogenous confirmatory test thresholds for diagnosing primary aldosteronism, we have now established the ARR cutoff for screening primary aldosteronism based on the risk of CVD. From the longitudinal data of FOS, we illustrated that CVD risk peaked once ARR exceeded 1.0 (unit: ng·dl⁻¹/mIU∙l⁻¹). In the CONPASS cohort, we demonstrated that primary aldosteronism with ARR ≥ 1.0 carried a higher CVD risk than participants with essential hypertension and ARR < 1.0. Using guideline-recommended higher ARR cutoffs (2.4 ~ 4.9) for screening, 10%~30% of people with primary aldosteronism would be missed and continue to carry a higher risk of CVD without appropriate management. Although the lowest cutoff (1.3) recommended by the European Society of Hypertension was not supported by robust evidence, our data suggest that lowering the ARR cutoff to 1.0 ng·dl⁻¹/mIU∙l⁻¹ for screening primary aldosteronism may improve detection and benefit long-term cardiovascular health.

In current guidelines, ARR is recommended as the most reliable screening method for primary aldosteronism (5,6). Although ARR is not perfect, many studies have demonstrated that ARR is superior to any other single parameter such as aldosterone, renin, and/or potassium (5). However, the ARR cutoff for screening primary aldosteronism has been controversial for a long time. In a recent meta-analysis which included 10 observational studies involving 4110 participants, the reported ARR sensitivity ranged from 10% to 100%, and specificity ranged from 70% to 100%. Apart from differences in ethnicity and laboratory assays, the highly heterogenous diagnostic criteria for primary aldosteronism made it difficult to recommend a single ARR threshold(8). It is widely known that there is no gold standard to confirm primary aldosteronism, especially the bilateral subtypes(1,9). As cardiovascular health is a major concern in patients with primary aldosteronism, defining an ARR cutoff based on the relationship between ARR and CVD represents a meaningful approach.

The relationship between ARR and CVD has been investigated in the general population and people with essential hypertension. A prospective study recruited 883 adults from the general population of Ohasama, and after a mean of 11-year follow-up, 45 subjects had incident strokes. Among adults with high sodium intake, the risk of stoke was significantly increased with an increment of baseline ARR after adjusting for potential confounders(10). Another prospective cohort recruited 125 patients with essential hypertension at baseline, and followed them for nearly 20 years. Compared to the group with low-ARR (<50th percentile of ARR), participants with high-ARR (≥50th percentile) showed a significantly higher cardiovascular morbidity(11). Using the longitudinal data from FOS, we demonstrated that the risk of CVD increased with increments in the ARR, reaching a peak and plateau once ARR exceeded 1.0 (unit: ng·dl⁻¹/mIU∙l⁻¹). The increased CVD risk was not different at higher ARR cutoffs, suggesting that cardiovascular risk became stably elevated once ARR ≥ 1.0. In the CONPASS cohort, we further demonstrated that only patients with primary aldosteronism whose ARR ≥ 1.0 exhibited an increased risk of CVD, while patients with essential hypertension and ARR ≥ 1.0 had an unchanged CVD risk. In addition, multivariable stepwise models in FOS and CONPASS indicated that ARR ≥ 1.0 ranked as one of top four cardiovascular risk factors, together with diabetes, dyslipidaemia and smoking.

Previously, a meta-analysis indicated that the aldosterone/direct renin concentration ratio is an effective screening test for primary aldosteronism(25). Using the largest sample size from a single center, our analysis of the CONPASS cohort demonstrated that setting the ARR cutoff at 1.0 ng·dl⁻¹/mIU∙l⁻¹ obtained a sensitivity of 0.98, although the specificity was only 0.63. As a screening test, a high sensitivity and a moderate specificity are acceptable, because even if subjects received a false positive screening result, they would be excluded after confirmatory testing. Based on an ARR cutoff of 2.4 ~ 4.9 ng·dl⁻¹/mIU∙l⁻¹, which is currently recommended by primary aldosteronism Management Guidelines(5,6), 10%~30% of participants with primary aldosteronism in the CONPASS cohort would not be recognized. These people would continue to carry a higher CVD risk than those with essential hypertension.

The main strengths of our analysis consisted of long-term follow-up data, and a sufficient number of people with CVD who underwent standardized evaluation in the FOS cohort. Furthermore, the longitudinal data were supported by data from 2,556 hypertensive participants enrolled in CONPASS. However, one limitation of the data from FOS is that none of the participants were formally tested for primary aldosteronism using dynamic tests. Hence, the association between ARR and CVD is purely based on the ARR, unlike in the CONPASS cohort when the ARR was only associated with CVD in participants with confirmed primary aldosteronism. In addition, FOS participants took a range of medications, including those which interfered with the renin-angiotensin-aldosterone system and could therefore affect the ARR. Although we cannot explore the influence of individual classes of interfering medications on the relationship between ARR and CVD, mainly because of the limited sample size of participants who received antihypertensive monotherapy in FOS, we have tried to adjust the use of antihypertensive drugs as the confounder in FOS. Furthermore, in CONPASS, the ARR threshold derived from FOS was tested on participants who were off all interfering medications and still found to be valid. Future prospective studies with larger sample size, where participants have their ARR measured both on and off interfering medications, may better define the influence of antihypertensive drugs on the relationship between ARR and CVD. Other limitations included the use of cross-sectional baseline data from CONPASS rather than longitudinal data. This is because CONPASS participants who received a diagnosis of primary aldosteronism were treated with surgery or mineralocorticoid receptor antagonists, hence their longitudinal data would not accurately reflect the actual CVD risk associated with primary aldosteronism. Moreover, the FOS and CONPASS cohorts had many differences in their ethnicity, age and laboratories assays for aldosterone and renin. Despite these differences, our analyses consistently showed that cardiovascular risk was elevated once ARR exceeded 1.0 ng·dl⁻¹/mIU∙l⁻¹. Finally, from the perspective of health resource and public health policy, screening for primary aldosteronism with ARR is cost-effective(26–29). However, whether it is feasible or cost-effective lower the ARR cutoff to 1.0 ng·dl⁻¹/mIU∙l⁻¹ requires more evidence. Modified management strategies that bypass confirmatory testing or subtyping should be tested to support the implementation of the lower ARR cutoff(30).

Conclusion

In conclusion, we recommend a ARR cutoff of 1.0 ng·dl⁻¹/mIU∙l⁻¹ for screening primary aldosteronism based on its association with CVD. The current guideline-recommended ARR cutoff values may lead to a high rate of unrecognized primary aldosteronism while affected people continue to carry the burden of a highly modifiable disease. Having a single ARR cutoff based on cardiovascular risk may simplify the diagnostic pathway for primary aldosteronism whilst increasing its diagnosis and alerting patients of their elevated but treatable cardiovascular risk.

Abbreviations

ARR, aldosterone-to-renin ratio; CI, confidence interval; CVD, cardiovascular disease; FOS, Framingham Offspring Study; ORs, odds ratios; PAC, plasma aldosterone concentration; PRC, plasma renin concentration; SAC, serum aldosterone concentration.

Author contributions

Conception and design: J.B.H., Q.F.L., S.M.Y. Analysis and interpretation of the data: J.B.H., R.L.L, X.J.C., Q.L.Z., and W.J.L. Drafting of the article: C.X.H., J.Y., F.F.W. and H.S. Critical revision of the article for important intellectual content: L.Q.M., Y.S., and Y.W. Obtaining of funding: J. H., Q.L., L.Q.M., and S.M.Y. Administrative, technical, or logistic support: Q. C., Z.W., Z. F., and Y. S. Collection and assembly of data: Z.H.W., Q.F.C., Y.S., Y.W. and L.Q.M.

Ethics approval and consent to participate

Framingham Offspring Study (FOS) was approved by the institutional review board of Boston University Medical Center and conducted in accordance with the declaration of Helsinki. All the participants in FOS provided written informed consent. Chongqing Primary Aldosteronism Study (CONPASS) was approved by the ethical committee of Chongqing Medical University (No: 201411) and conducted in accordance with the declaration of Helsinki. All the participants in CONPASS provided written informed consent.

Acknowledgments

We thank the participants and staff of the Framingham Offspring Study. We thank other members of the Chongqing Primary Aldosteronism Study (CONPASS) Group: Mei Mei, MD, PhD; Suxin Luo, MD, PhD; Kangla Liao, MD; Yao Zhang, MD, PhD; Yunfeng He, MD, PhD; Yihong He, MD; Ming Xiao, PhD; and Bin Peng, PhD for suggestions of study design and revision.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Additional information

Funding

This work was supported by National Natural Science Foundation of China, Major Joint Project (U21A20355); National key research & development plan of China, major project of prevention and treatment for common diseases (2022YFC2505300, sub-project: 2022YFC2505301, 2022YFC2505302, 2022YFC2505306); National Natural Science Foundation of China (82270878, 81870567, 81970720, 82000810, 82170825, 82100833); National key research & development plan of China, major project of prevention and treatment for common diseases (2021YFC2501600, sub-project: 2021YFC2501603); Joint Medical Research Project of Chongqing Science and Technology Commission & Chongqing Health and Family Planning Commission (Major Project, 2022ZDXM003); Chongqing Outstanding Youth Funds (cstc2019jcyjjq0006); The China Postdoctoral Science Foundation (2019M663499, 2020T130760, 2021MD703927) ;Program for Youth Innovation in Future Medicine, Chongqing Medical University. The Natural Science Foundation of Chongqing, China (cstc2020jcyj-bshX0081, cstc2021jcyj-bshX0207); Intelligent Medicine Research Project of Chongqing Medical University (NO: ZHYX202113).