Yuqing Liu1†
Zhiheng Zhang2†
Haifeng Zhou3
Yutong Yan1Mei Zhou1Cong Wang1Maoting Gao1
Jun Tao4Meiling Bao5
Tao Yang1*‡Min Sun1*‡
Yuhong Yang1*‡- 1Department of Endocrinology, The First Affiliated Hospital with Nanjing Medical University, Nanjing, China
- 2Division of Hepatobiliary and Transplantation Surgery, Department of General Surgery, Nanjing Drum Tower Hospital, Affiliated Hospital of Medical School, Nanjing University, Nanjing, China
- 3Department of Interventional Radiology, The First Affiliated Hospital with Nanjing Medical University, Nanjing, China
- 4Department of Urology, The First Affiliated Hospital with Nanjing Medical University, Nanjing, China
- 5Department of Pathology, The First Affiliated Hospital with Nanjing Medical University, Nanjing, China
Background: Renin-independent aldosterone secretion contributes to aldosteronism and heightened cardiovascular risk, but renin-independent aldosteronism is highly heterogenous. A refined classification may assist in identifying individuals with distinct cardiovascular risk profiles and guide individualized treatment strategies.
Methods: Unsupervised hierarchical clustering was performed using 12 clinical parameters from patients with renin-independent aldosteronism in our registry cohort (n=404). The cluster centroids derived from the discovery cohort were fixed and applied to the Framingham Heart Study Third Generation cohort (n=417) for subject classification. The identified clusters were evaluated for their association with cardiovascular outcomes, assessed by echocardiographic parameters, serum biomarkers and cardiovascular event rates.
Results: Three replicable clusters of patients with renin-independent aldosteronism were identified. Patients in cluster 2 showed the most severe metabolic abnormalities with the highest lipid and glucose levels, while patients in cluster 3 displayed the highest aldosterone levels. Both clusters 2 and 3 showed elevated baseline blood pressure and left ventricular remodeling compared with cluster 1. Cluster 2 exhibited the highest risk of cardiovascular disease, chronic heart failure and atrial fibrillation, followed by cluster 3, which showed a higher incidence of cardiovascular disease compared with cluster 1.
Conclusions: We identified 3 subgroups with differing degrees of target organ damage and cardiovascular risk. Our findings establish metabolic dysfunction, rather than aldosterone excess, as a potential dominant cardiovascular risk driver in RIA patients, defining a new risk paradigm. Patients with renin-independent aldosteronism with metabolic dysfunction or high aldosterone levels may benefit from mineralocorticoid receptor antagonists with different priorities for metabolic and cardiovascular protection. This new refined classification may help tailor optimal treatment strategies for patients with heterogenous renin-independent aldosteronism.
Introduction
Aldosterone, the predominant mineralocorticoid, is primarily regulated by renin-angiotensin II system and plays a crucial role in maintaining water and electrolyte balance by activating mineralocorticoid receptors (MRs) (1). Excessive aldosterone production that occurs independently of renin and angiotensin II, known as renin-independent aldosteronism (RIA) (2–7), leads to inappropriate activation of MRs, resulting in sodium and fluid retention, and adverse cardiorenal remodeling (8). RIA represents a pathophysiological continuum characterized by pronounced phenotypic heterogeneity. This seamless spectrum ranges from subtle autonomous aldosterone hypersecretion in normotension individuals, to mild hypertension, and ultimately to its most severe and clinically overt form with resistant hypertesion, where the majority of patients reach the diagnostic criteria for primary aldosteronism (9). Formal diagnosis of primary aldosteronism requires standardized confirmatory testing, including the commonly used saline suppression test and captopril challenge test, followed by adrenal venous sampling for subtype classification (10). In contrast, the broader RIA population encompasses all individuals with biochemical evidence of autonomous aldosterone secretion, commonly defined in previous studies (4, 5, 11) as plasma renin activity ≤ 1 µg/L/h and plasma aldosterone concentration ≥ 100 pg/mL.
Accumulating evidence have shown that compared with individuals with renin-dependent aldosteronism, those with RIA face a heightened risk of cardiovascular disease (4, 12, 13), and a higher prevalence of chronic kidney disease (5) and metabolic dysfunction-associated steatotic liver disease (6). Large-scale population studies have examined the relationship between aldosterone, renin and target organ damage, demonstrating that low renin concentrations or an elevated aldosterone-to-renin ratio are associated with incident hypertension (11, 13–16), adverse cardiac remodeling (13, 17) and increased arterial atherosclerosis (12, 13). These findings highlight the potential need for early intervention with MR antagonists in individuals with RIA to mitigate the long-term consequences of excessive aldosterone activity.
Despite these observations, the role of MR antagonists as the optimal first-line therapy for all individuals with RIA remains uncertain. A meta-analysis has demonstrated the superiority of MR antagonists over angiotensin-converting enzyme inhibitors or angiotensin receptor blockers in lowering blood pressure among hypertensive patients with low renin, who often exhibit autonomous aldosterone secretion (9, 18). Other studies have showed that diuretics and epithelial sodium channel inhibitors exhibit greater blood-pressure-lowering efficacy compared with beta-blockers and angiotensin receptor blockers (19, 20). Beyond blood pressure control, MRs activation in extra-renal tissues including the vascular and adipose tissue contributes to vascular and metabolic dysfunction, highlighting the need for therapies that address these effects (21). While angiotensin-converting enzyme inhibitors or angiotensin receptor blockers have demonstrated metabolic and cardiovascular benefits (22–26), steroidal MR antagonists have shown neutral or adverse effects on glucose and lipid metabolism (27–29). In contrast, the non-steroidal MR antagonist finerenone has been shown to reduce the hazard of diabetes in patients with heart failure (30).
The pronounced phenotypic heterogeneity of RIA poses substantial challenges for developing individualized treatment strategies. As demonstrated in other heterogeneous diseases such as type 2 diabetes and metabolic dysfunction-associated steatotic liver disease, data-driven clustering analyses can classify these patients into clinically meaningful subgroups based on multidimensional clinical features, enabling more precise risk assessment and therapeutic decision-making. In RIA, however, such approaches remain limited. Prior studies have largely relied on single biochemical thresholds (e.g., aldosterone or renin levels) or binary definitions of primary aldosteronism to differentiate cardiovascular risk, approaches that fail to capture the continuous spectrum of aldosterone excess (3, 11, 17). Moreover, no systematic effort has integrated clinical, biochemical, and metabolic variables to define distinct subgroups within the broader RIA population. Given the success of clustering-based characterizations in other heterogeneous diseases, applying an unsupervised, data-driven approach to RIA could provide a refined classification framework, identify individuals at highest cardiovascular risk, and better inform targeted use of MR antagonists. To the best of our knowledge, no prior study has applied clustering approaches to delineate clinical subphenotypes in RIA.
To fill this gap, we conducted an unsupervised, data-driven clustering analysis using 12 commonly measured variables in patients with RIA. We aimed to establish a novel RIA classification and to assess its association with specific clinical characteristics, target organ damage and long-term cardiovascular outcomes, thereby enabling the potential for early warning and precise interventions for RIA in the future.
Materials and methods
Study populations
We used data from our registry cohort and Framingham Heart Study Third Generation (Gen III) cohort. The study flow chart of participant selection was shown in Figure 1.
Figure 1. Derivation of study cohort. The study comprised 2 cohorts: the discovery cohort, from the First Affiliated Hospital with Nanjing Medical University, and the validation cohort, derived from the Framingham Heart Study Third Generation cohort. In both cohorts, renin-independent aldosteronism was defined as plasma renin activity ≤ 1 µg/L/h and plasma aldosterone concentration ≥ 100 pg/mL. CVD, cardiovascular disease; n, number.
In our registry cohort, all patients were screened for endocrine hypertension at the First Affiliated Hospital With Nanjing Medical University, Nanjing, China (31). We included patients who met all of the following criteria: plasma renin activity ≤ 1 µg/L/h, plasma aldosterone concentrations ≥ 100 pg/mL, and available medical records including parameters used for clustering, echocardiographic data and serum cardiac parameters at study entry.
The Framingham Heart Study Gen III cohort is a longitudinal cardiovascular epidemiological study with a previously published study design (32–34). It began in 2002, and was funded by the National Institutes of Health and administered by Boston University. We included patients who attended the Gen III cohort examination 1 (2002-2005) and were followed up until December 31, 2019, and met all of the following criteria: no history of cardiovascular disease, plasma renin activity ≤ 1 µg/L/h, plasma aldosterone concentrations ≥ 100 pg/mL, and available medical records including parameters used for clustering at study entry.
Physical examination and laboratory measurement
In our registry cohort, all antihypertensive medications known to interfere with renin-angiotensin-aldosterone system activity (including diuretics, beta-blockers, angiotensin converting enzyme inhibitors, and angiotensin-1 receptor blockers) were withdrawn, or replaced with calcium channel blockers or alpha-adrenergic blockers for 2–4 weeks before blood samples collection. Hypokalemia was corrected to minimize its impact on aldosterone and renin measurements. Blood samples were drawn following an overnight fasting period (≥ 8 hours) with patients in the upright position. Routine biochemical measurements were conducted using enzymatic methods on a chemistry analyzer (Au5800, Olympus Medical Engineering Company). Plasma aldosterone concentrations were measured by a chemiluminescent assay (Mindray), with intra- and inter-batch coefficients of variation ranging from 3.26% to 5.69% and 5.61% to 7.30%, respectively. Plasma renin activity was quantified via radioimmunoassay (Beijing North Institute of Biotechnology), with intra-batch coefficients of variation < 10% and inter-batch coefficients of variation < 15%. Obesity was defined according to the Chinese guidelines for management of obesity as a body mass index (BMI) ≥ 28 kg/m2.
In Gen III cohort, physical examination procedures and laboratory measurements have been described previously (35, 36). Venipuncture was performed on study participants in a supine position after an overnight fast using standard techniques. We reviewed medical history and collected anthropometric data including BMI, systolic and diastolic blood pressure. Plasma aldosterone concentrations were measured using an enzyme-linked immunosorbent assay (ALPCO Diagnostics), and plasma renin activity was measured with a radioimmunoassay (DiaSorin). In both cohorts, the estimated glomerular filtration rate was calculated using Chronic Kidney Disease Epidemiology Collaboration (37). Obesity was defined as a BMI ≥ 30 kg/m2 according to the European Society of Endocrinology Clinical Practice Guideline (38).
Definitions of RIA and primary aldosteronism subtyping
In our registry and Gen III cohorts, RIA was defined in accordance with previous studies (4, 5, 11) as plasma renin activity ≤ 1 µg/L/h and plasma aldosterone concentration ≥ 100 pg/mL. Within the discovery cohort, patients who met the RIA criteria underwent further evaluation for primary aldosteronism. The diagnosis of primary aldosteronism required a positive screening test (aldosterone-to-renin ratio ≥ 20 pg/dL per µg/L/h) together with at least one positive confirmatory test (saline infusion test or captopril challenge test) (10). Patients with confirmed primary aldosteronism were further subtyped as follows: unilateral primary aldosteronism was diagnosed when patients met any of the following criteria: (1) a lateralization ratio ≥ 2.0 following successful catheterization (selectivity index ≥ 2.0) during adrenal venous sampling; (2) complete biochemical success at ≥ 6 months after adrenalectomy; or (3) decision to bypass adrenal venous sampling based on strong clinical and radiological evidence; bilateral primary aldosteronism was diagnosed when patients met either of the following criteria: (1) a lateralization ratio < 2.0 following successful adrenal venous sampling; or (2) a Japan Primary Aldosteronism Study score ≥ 8 (39), predicting a low likelihood of unilateral disease.
Assessment of outcomes
In our registry cohort, we assessed cardiovascular health across RIA clusters using echocardiographic parameters, serum cardiac markers and arterial vascular parameters. The existence of cardiovascular disease was confirmed after rigorous review of hospitalization records, outpatient clinic visits, and diagnostic test results. It was confirmed when meeting the following conditions: (1) coronary heart disease: documented myocardial infarction (confirmed by characteristic cardiac enzyme elevation and electrocardiographic changes) or coronary revascularization procedures (percutaneous coronary intervention or coronary artery bypass grafting); (2) heart failure: clinical diagnosis of congestive heart failure requiring hospitalization or consistent outpatient management, supported by echocardiographic evidence of systolic or diastolic dysfunction; (3) stroke: acute cerebrovascular event confirmed by neuroimaging (computed tomography or magnetic resonance imaging), including both ischemic and hemorrhagic strokes; (4) transient ischemic attack: documented neurological deficits resolving within 24 hours with corresponding neuroimaging findings.
In the Gen III cohort, cardiovascular disease event was defined by the occurrence of a composite endpoint according to validated Framingham Heart Study criteria (34), comprising coronary heart disease, congestive heart failure, cerebrovascular events (including ischemic stroke, hemorrhagic stroke, and transient ischemic attack) and peripheral artery disease (intermittent claudication) (40, 41). Outcomes were ascertained through medical histories, physical examinations, hospitalization records, and communication with participants’ physicians (42). Specific component definitions were as follows: (1) coronary heart disease: coronary death, myocardial infarction, or hospitalized coronary insufficiency (42); (2) stroke: events confirmed by at least two neurologists via medical record review, including definite cerebrovascular accident, atherothrombotic brain infarction, cerebral embolism, intracerebral hemorrhage, or subarachnoid hemorrhage (43); (3) congestive heart failure events were adjudicated by an independent endpoint committee according to the Framingham criteria (44), requiring either: a. ≥ 2 major clinical criteria (e.g., paroxysmal nocturnal dyspnea, jugular venous distension), or b. 1 major criterion + ≥ 2 minor criteria (e.g., ankle edema, dyspnea on exertion); (4) atrial fibrillation was diagnosed if at least 2 cardiologists verified the rhythm abnormality on an electrocardiogram, including Holter electrocardiograms, telemetry or other monitoring data (45, 46).
In both cohorts, left ventricular hypertrophy was defined by a left ventricular mass index ≥115 g/m2 in men and ≥95 g/m2 in women, following the recommendation of the American Society of Echocardiography (47).
Follow-up duration and completeness
Our analysis included Gen III cohort participants from the baseline examination (2002-2005) who had complete follow-up through the study end date. The Framingham Heart Study is characterized by its excellent participant retention, historically achieving a 99% retention rate (40). Participants in this study were followed for a median of 5543 days [4899–6033 days] until December 31, 2019.
Echocardiographic examination
In our registry cohort, all participants underwent 2-dimensional transthoracic echocardiography performed by experienced sonographers using a Phillips EPIQ CVx ultrasound system. Left ventricular mass and left ventricular mass index were calculated using the Devereux and Reichek formula (48). In the Gen III cohort, standardized 2-dimensional echocardiography with color flow and tissue Doppler imaging was performed by experienced sonographers using a Hewlett-Packard Sonos 5500 Ultrasound machine (Phillips Healthcare, Andover, MA) (49). Experienced readers evaluated the digitized echocardiographic images (DICOM format) utilizing Digiview System Software (version 3.7.9.3, Digisonic Inc., Houston, TX, USA).
Cluster analysis and validation
Our registry cohort (hereafter named as discovery cohort) served as the discovery cohort for establishing unsupervised clustering, and the Gen III cohort (hereafter named as validation cohort) was used as the validation cohort to assess the reproducibility of these clusters and their association with cardiovascular risk. Missing data in the discovery cohort were imputed using multiple imputation by the R package “mice” before clustering analysis, while the validation cohort had complete data and did not require imputation. In the discovery cohort, the proportion of missing data for the 12 clustering variables was generally low, with fasting insulin having the highest proportion at 2.48% (10 missing values), followed by cortisol at 0.50% (2 missing values), while the remaining 10 variables had complete data (0% missing) (Supplementary Figure S1). The overall missing rate across all clustering variables in the discovery cohort was 0.25%. We employed the following multiple imputation parameters: 5 imputed datasets (m=5), maximum of 10 iterations, predictive mean matching method, and random seed 12345 to ensure reproducibility. The selection of the 12 clustering parameters was based on their established pathophysiological relevance to RIA. Age at diagnosis, sex, and BMI were included as core demographic and adiposity modifiers. Systolic and diastolic blood pressure were incorporated as principal hemodynamic consequences of aldosterone excess. Plasma renin activity and plasma aldosterone concentrations represented the defining biochemical axis of RIA. Metabolic markers (fasting blood glucose, triglycerides, total cholesterol and fasting insulin) were selected due to the established comorbidity between metabolic dysfunction and aldosterone excess. Cortisol was included to account for its co-secretion and influence on associated comorbidities. Overall, correlations between the selected parameters were weak, as visualized in the correlation matrix (Supplementary Figure S2). The strongest correlation identified was a moderate positive correlation between systolic blood pressure and diastolic blood pressure (r = 0.56). Hierarchical clustering using Ward’s minimum variance method was applied to the discovery cohort based on the distance matrix of 12 clinical parameters. The optimal number of clusters was determined using the cubic clustering criterion, yielding a 3-cluster solution. For external validation, representative subjects were selected from each cluster in the discovery cohort based on proximity to the cluster centroid in the multidimensional feature space. Each subject in the validation cohort was then assigned to the cluster of its nearest representative subject from the discovery cohort, using Euclidean distance, without performing de novo clustering in the validation cohort.
Statistical analysis
Data were analyzed with the use of SPSS version 27.0, GraphPad Prism version 10.1 and R software version 4.0.3. Continuous variables were expressed as mean (standard deviation) for normally distributed data, median (interquartile range) for non-normally distributed data, while categorical variables were presented as numbers and proportions. Normality was assessed using the Kolmogorov-Smirnov test. Differences in continuous variables were evaluated using ANOVA followed by Bonferroni tests or Kruskal-Wallis tests followed by pairwise comparisons. Chi-square and Fisher’s exact tests were used to compare categorical data. Distinct regression models were employed for each cohort based on their study designs to assess the cluster-cardiovascular outcome association. In the discovery cohort (cross-sectional analysis), multivariable logistic regression was used to assess the odds of prevalent cardiovascular disease across clusters (reference: cluster 1), adjusted for smoking, alcohol use, statin use, and diabetes treatment. In the validation cohort (longitudinal analysis), Cox proportional hazards models were used to estimate risks of incident cardiovascular events (reference: cluster 1), with adjustment for the same covariates. Kaplan-Meier curves and log-rank tests were used for visualizing and comparing cumulative event rates. Results are reported as hazard ratios with 95% confidence intervals. A P value < 0.05 was considered as statistically significant.
Results
Clinical characteristics of study populations
A total of 404 patients and 417 patients with RIA, derived from the discovery and validation cohorts, respectively, were included in Figure 1. Patient characteristics were shown in Supplementary Table S1. RIA patients in the validation cohort were younger, and had higher BMI, fasting blood glucose and total cholesterol, and lower systolic blood pressure, diastolic blood pressure, plasma aldosterone concentrations and aldosterone-to-renin ratio, and proportion of diabetes compared with those in the discovery cohort (all, P < 0.05) (Supplementary Table S1).
Clinical characteristics of distinct clusters
A dendrogram showing the patient clustering process is presented in Supplementary Figure S3. We identified 3 major clusters in the discovery cohort data. To visualize these clusters and explore their distinctions, we generated 2D representations using t-distributed stochastic neighbor embedding (t-SNE) (Figure 2). The proportions of the 3 clusters identified in the discovery and validation cohorts are shown in Figure 2 and Supplementary Figure S4B. The distinct variables of 3 clusters are depicted in Figure 2 and 2D. The baseline clinical characteristics of patients stratified by these clusters are shown in Figure 3 and Supplementary Table S2. Patients in cluster 1 (n=262, 64.9%) were the oldest (P<0.001), and displayed the lowest level of systolic (P<0.001) and diastolic blood pressure (P<0.001), and the least severe metabolic dysfunction compared with the other 2 clusters. Cluster 2 (n=33, 8.2%) comprised patients with the most pronounced metabolic abnormalities, characterized by the highest BMI (P<0.001), triglycerides (P = 0.001) and fasting blood glucose (P<0.001), and the highest proportion of obesity (P<0.001) and diabetes (P = 0.002), leading to the greatest use of statins (P = 0.001) and antidiabetic therapy (P = 0.002). Patients in cluster 3 (n=109, 27.0%) showed the highest plasma aldosterone concentrations (P = 0.009) and aldosterone-to-renin ratio (P = 0.042). The proportion of confirmed primary aldosteronism (P = 0.047) was highest in cluster 3, whose patients also showed a high prevalence of unilateral primary aldosteronism (87% of 79, P = 0.173) (Supplementary Figures S5, S6).
Figure 2. Visualization of clustering results and patient distribution in the discovery cohort. (A) t-Distributed Stochastic Neighbor Embedding (t-SNE) plot visualizing high-dimensional data. (B) Patient distribution across clusters. (C) Standardized values (z-scores) of selected variables by cluster. All continuous variables were scaled (mean = 0, SD = 1). (D) The ribbons connect an individual cluster to a variable if the group mean is greater or less than the overall mean for the entire cohort. BMI, body mass index; BP, blood pressure; FBG, fasting blood glucose; FINS, fasting insulin; PAC, plasma aldosterone concentration; PRA, plasma renin activity; RIA, renin-independent aldosteronism; TC, total cholesterol; TG, triglycerides.
Figure 3. Clinical characteristics of patients with RIA stratified by different clusters in the discovery cohort. The distribution of (A) age, (B) BMI, (C) systolic blood pressure, (D) diastolic blood pressure, (E) plasma aldosterone concentration, (F) plasma renin activity, (G) cortisol, (H) total cholesterol, (I) triglycerides, (J) fasting blood glucose, and (K) fasting insulin in different clusters in the discovery cohort. BMI, body mass index; BP, blood pressure; FBG, fasting blood glucose; FINS, fasting insulin; h, hour; PAC, plasma aldosterone concentration; PRA, plasma renin activity; RIA, renin-independent aldosteronism; TC, total cholesterol; TG, triglycerides. *P < 0.05; ***P < 0.001; ****P < 0.0001.
Cardiovascular disease incidences and cardiac parameters of distinct clusters
Cluster 1 was associated with the least severe abnormalities of cardiac structure, while patients in cluster 2 and cluster 3 displayed a comparably elevated echocardiographic parameters indicative of increased left ventricular size and dysfunction, including left ventricular mass (P = 0.001), left ventricular end-systolic (P = 0.032) and end-diastolic diameter (P = 0.012), and interventricular septal thickness (P = 0.017). Patients in cluster 2 exhibited the highest levels of arterial stiffness markers, including lipoprotein-associated phospholipase 2 (P = 0.034) and mean brachial-ankle pulse wave velocity (P = 0.027). The lowest level of estimated glomerular filtration rate was identified in the cluster 1 (P<0.001) (Table 1).
Table 1. Parameters related to target organ damage in different clusters of patients with RIA in the discovery cohort.
As shown in Figure 4, cardiovascular disease incidence was highest in cluster 2 (n=8, 24.2%), followed by cluster 3 (n=13, 11.9%), and lowest in cluster 1 (n=22, 8.4%; P = 0.025). We used multinomial logistic regression to compare the odds of cardiovascular disease incidence across the three clusters. Model 1 was sequentially adjusted in two stages by including different sets of covariates. Results from both stages of adjustment showed that the odds of cardiovascular disease incidence were significantly higher in cluster 2 (odds ratio = 2.770, 95% CI: 1.000-7.135; P = 0.002) and cluster 3 (odds ratio = 2.278, 95% CI: 1.017-5.043; P = 0.042) compared to cluster 1 (reference group) (Figure 5, Supplementary Table S4).
Figure 4. Cardiovascular disease incidence across clusters in RIA individuals. (A) Discovery cohort. (B) Validation cohort. CVD, cardiovascular disease. *P < 0.05; **P < 0.01.
Figure 5. The associations of different clusters and cardiovascular disease incidence in the discovery cohort. Model 1 served as the baseline with no covariate adjustments. Model 2 included adjustments from Model 1, plus factors such as smoking status and alcohol use. Model 3 built upon Model 2 by additionally adjusting for statin use and diabetes treatment. In all models, cluster 1 was used as the reference group.
External validation of clinical characteristics and cardiac outcomes of distinct clusters
Data from validation cohort identified 3 clusters with clinical characteristics similar to those in discovery cohort: cluster 1 (n=162, 38.8%), cluster 2 (n=45, 10.8%) and cluster 3 (n=210, 50.4%). The t-SNE analysis exhibited and verified the polarized distributions of patient characteristics across the clusters (Supplementary Figure S4A). The distribution of aberrant clustering features across different clusters is shown in Supplementary Figures S4C, S4D. Patients in cluster 1 were the youngest (P<0.001), had the highest proportion of female (P<0.001), the lowest systolic (P<0.001) and diastolic blood pressure (P<0.001), and exhibited the least severe metabolic dysfunction relative to the other 2 clusters. Patients in cluster 2 showed the most pronounced metabolic abnormalities, with the highest levels of BMI (P<0.001), total cholesterol (P<0.001), triglycerides (P<0.001), and fasting blood glucose (P<0.001), as well as the highest prevalence of obesity (P<0.001), dyslipidemia (P<0.001), diabetes (P<0.001) and the greatest use of antidiabetic therapy (P<0.001). Patients in cluster 3 exhibited the highest plasma aldosterone concentrations (P = 0.020) and the greatest use of statins (P<0.001) (Supplementary Figure S, Supplementary Table S3).
Cluster 2 exhibited the highest cardiovascular disease incidence (n=7, 15.6%) in Figure 4, followed by cluster 3 (n=13, 6.2%), with cluster 1 showing the lowest rate (n=3, 1.9%; P<0.01). Next, we analyzed baseline cardiovascular parameters across the defined RIA clusters in the validation cohort. These findings were consistent with those from the discovery cohort. Cluster 1 exhibited the least severe cardiac structural abnormalities at baseline, assessed by echocardiographic measures, compared with clusters 2 and 3. Patients in cluster 2 demonstrated the most pronounced left ventricular cardiac remodeling (Table 2).
Table 2. Echocardiographic parameters in different clusters of patients with RIA in the validation cohort.
During a median follow-up of 5543 [4899-6033] days, 23 cardiovascular disease events, 13 coronary heart disease events, and 10 atrial fibrillation events occurred in the validation cohort. The Kaplan-Meier curves revealed that the cumulative incidence of adverse cardiovascular events in cluster 2 were higher than those in cluster 1 and cluster 3 (cardiovascular disease, log rank: P<0.001; coronary heart disease, log rank: P = 0.025; atrial fibrillation, log rank: P = 0.016) (Figure 6). To further examine the relationship between cluster classification and cardiovascular outcomes, Cox regression models were employed using patients from cluster 1 as the reference. Multivariate Cox proportional hazards models adjusted for smoking status, alcohol use, statin use and diabetes treatment demonstrated that patients in clusters 2 had a significantly higher risk of developing cardiovascular disease (hazard ratio = 11.425, 95% CI: 2.218-58.844; P = 0.004), chronic heart failure (hazard ratio = 7.632, 95% CI: 1.563-42.673; P = 0.043) and atrial fibrillation (hazard ratio = 12.879, 95% CI: 1.325-125.158; P = 0.028) compared with the other 2 clusters (Table 3). The risk of cardiovascular disease (hazard ratio = 5.028, 95% CI: 1.124-22.494, P = 0.035) was also higher in cluster 3 than in cluster 1 (Table 3).
Figure 6. Cumulative incidences of cardiovascular events across 3 clusters in the validation cohort. Kaplan-Meier curves were plotted for the cumulative incidence of cardiovascular events among 3 clusters. The log-rank test was used to assess differences among clusters. (A) Cardiovascular disease. (B) Coronary heart diseases. (C) Atrial fibrillation.
Table 3. Cox proportional hazard ratios for cardiovascular events according to clusters in the validation cohort.
Discussion
Using cross-sectional data from our registry cohort and longitudinal data from Gen III cohort, we identified a new clustering of individuals with RIA, which is associated with varying degree of target organ damage and different risks of adverse cardiovascular events. This classification provides valuable insights for guiding therapeutic decisions. We incorporated variables reflective of key clinical features of RIA that are monitored in patients, ensuring that this clustering approach can be applied to clinical practices.
The consistency of clustering results between patients suspected of endocrine hypertension in our registry cohort and the general population enrolled in Gen III study suggests that these clusters are stable and potentially mechanistically distinct, rather than merely reflecting different stages of RIA. Physiologically, renin-independent aldosterone secretion is a response to extracellular hyperkalemia (9), whereas pathological renin-independent secretion is driven by autonomous aldosterone secretion (9) that is typically associated with adrenal abnormalities (50, 51) and aldosterone-driver mutations (52). The phenotypic profile of cluster 3, characterized by the highest aldosterone levels, is consistent with a pathophysiology that may involve adrenal abnormalities and aldosterone-driver mutations (52). We also identified a subgroup of patients with RIA (cluster 2) who exhibit most severe metabolic syndrome but milder aldosterone secretion compared with cluster 3. Although the mechanisms underlying this pattern remain uncertain, metabolic factors could play a role in modulating aldosterone secretion in these individuals. Previous studies have suggested that leptin may stimulate aldosterone production (53), and that lipid abundance can affect steroidogenesis in aldosterone-producing cells by contributing to phospholipid formation (54), altering oxidative stress (31, 55), and acting as lipid second messengers (56) that regulate steroidogenic enzyme activity and transcription. These observations raise the possibility that the dysregulated lipid metabolism observed in cluster 2 may contribute to renin-independent aldosterone secretion. This interpretation is compatible with previous studies showing a higher prevalence of obesity in patients with bilateral adrenal hyperplasia or nonclassical adrenal histopathology, which are subtypes associated with a milder phenotype of primary aldosteronism, compared with those with aldosterone-producing adenomas or classical histopathology in which aldosterone-driver mutations account for approximately 95% (31, 57, 58). The characteristics of cluster 1 shown in both cohorts is consistent with previous reports suggesting a possible association of elevated aldosterone-to-renin ratio with older age, female and chronic kidney disease (59, 60).
Previous population-based studies have reported an increased risk of cardiovascular disease (4, 12, 13), and a higher prevalence of chronic kidney disease (5) and metabolic dysfunction-associated steatotic liver disease (6) in patients with RIA compared with those with renin-dependent aldosteronism. The association between aldosterone and adverse cardiac remodeling or long-term cardiovascular risk has been demonstrated in both normotensive and hypertensive persons across different age groups (12, 13, 17, 61). In line with these findings, our study revealed cardiac structural alteration and an elevated cardiovascular risk in cluster 3 which exhibited higher aldosterone concentrations compared with cluster 1. It has been suggested that the heightened cardiovascular risk in RIA or with increased aldosterone secretion is likely due to enhanced MR activity, as supported by previous studies showing a positive correlation between aldosterone concentrations and MR activation (4, 11). These findings support the use of MR antagonists in patients with RIA to mitigate cardiovascular risk.
However, hypertension practice guidelines at present do not provide recommendations on the treatment of patients with RIA who do not meet the diagnostic criteria for primary aldosteronism (62). Our clustering results suggest that the subgroup of RIA characterized by a high aldosterone concentration (cluster 3) may benefit most from treatment with MR antagonists and potentially aldosterone synthase inhibitors which showed to lower blood pressure by reducing aldosterone production in clinical trials (63). For the subgroup of RIA who exhibit severe metabolic abnormalities and the highest cardiovascular risk (cluster 2), treatment should prioritize both metabolic and cardiovascular protection. While angiotensin-converting enzyme inhibitors and angiotensin receptor blockers have demonstrated benefits in improving insulin sensitivity (26), preserve pancreatic beta-cell function (22, 23) and mitigate metabolic complications (24, 25), their antihypertensive effects might be limited in RIA (18). Among MR antagonists, the steroidal agents spironolactone and eplerenone have metabolic limitations. Spironolactone has been associated with an increase in Hemoglobin A1c, possibly due to its non-selective action which raises cortisol level and subsequently activates glucocorticoid receptors (27). Eplerenone has not demonstrated significant benefits for glucose and lipid metabolism (28, 29). In contrast, finerenone, a non-steroidal MR antagonist with broader distribution in extra-renal tissues (64), shows more potent anti-inflammatory and antifibrotic effects than steroidal MR antagonists (65, 66). Notably, the cardiovascular benefits of finerenone are more pronounced in patients with a higher BMI (67), and it has also been shown to reduce diabetes incidence in patients with heart failure (30). Investigations should be warranted to explore if finerenone, which is currently used in patients with chronic kidney disease and type 2 diabetes (68), can improve blood pressure (69) and cardiovascular health (70) in RIA, especially in cluster 2, where metabolic dysfunction is most severe.
Strength and limitations
The key strengths of the current study included the relatively large sample size of patients with RIA, which is comparable to that reported in previous large-scale population studies, the detailed evaluation of target organ damage in our registry cohort and the confirmation of the clustering results in 2 independent populations. Although patients in the discovery and validation cohorts exhibited significantly different clinical characteristics, the overall consistency of the clustering results indicates the reliability of our findings and supports the concept that RIA may comprise mechanistically distinct subgroups warranting differentiated therapeutic strategies. Nonetheless, several limitations should be acknowledged. First, the presence of confounders such as interfering medications, patient’s posture at blood sampling and electrolyte imbalance in the Gen III study, combined with the known intraindividual variability of aldosterone and renin levels (71), may have introduced misclassification bias due to the reliance on a single measurement. Second, our clustering results cannot be confidently interpreted as representing different etiologies of RIA, as we did not collect data on the cortisol levels after 1 mg dexamethasone test, which is the more accurate indicator of autonomous cortisol production than baseline cortisol levels and may have influenced the clustering results. Third, the blood pressure measurements used in both cohorts were obtained from clinic visits, which may not provide the same level of precision as ambulatory blood pressure monitoring. Fourth, the limited number of outcome events in the Gen III cohort, together with the smaller sample size of the high-metabolic-risk cluster (cluster 2; n=45), contributed to wide confidence intervals in the Cox regression analysis and raised concerns about potential overfitting. Although our power calculation indicated sufficient power (e.g., 80% power to detect a hazard ratio ≥3.5 for cluster 2 versus cluster 1 in the Cox regression analysis), the precision of the risk estimates for this specific subgroup remains limited. Nevertheless, the reproducible identification of cluster 2 in both cohorts, its consistently severe metabolic phenotype, and its numerically higher cardiovascular event rates support its potential clinical relevance, warranting further validation in larger prospective studies. Fifth, the exclusion of overt primary aldosteronism diagnosis, unilateral secretion status, and hypertension duration as clinical variables for stratifying target organ damage may limit precision. These factors were omitted to ensure replication, as they were unavailable in the validation cohort. Nevertheless, the distribution of primary aldosteronism appeared consistent with the clinical profiles of each cluster, and future studies are warranted to determine whether incorporating these variables could further enhance stratification accuracy. Finally, it is important to acknowledge that although cardiovascular outcomes were assessed longitudinally in the validation cohort, the clustering analysis itself was derived exclusively from cross-sectional baseline data. RIA is a dynamic condition, and patient phenotypes may evolve over time in response to disease progression, therapeutic interventions, or lifestyle modifications. Accordingly, we cannot exclude the possibility that the identified clusters represent either stable pathophysiological traits or transient clinical states. Longitudinal studies with repeated measurements of the clustering variables will be essential to determine the temporal stability of these clusters and to elucidate how potential cluster transitions relate to changes in cardiovascular risk and treatment response.
Conclusions and perspectives
In conclusion, our data suggested that integrating clinical variables from patients of RIA can effectively stratify subgroups with varying degree of target organ damage and long-term cardiovascular risks. Our findings indicate that metabolic dysfunction, rather than aldosterone excess, could be an important contributor to cardiovascular risk in RIA patients, though causal inference awaits confirmation in future prospective research. Given the reported association of renin-independent aldosterone secretion and cardiovascular complications (12, 13, 17, 61), clinical trials are warranted to evaluate the impact of early intervention with MR antagonists on long-term cardiovascular health outcome in patients with RIA. Our clustering highlights the need for additional focus in these clinical trials on patients with RIA with severe metabolic dysfunction to guide the optimal medication.
Data availability statement
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s.
Ethics statement
The studies involving humans were approved by the First Affiliated Hospital With Nanjing Medical University, Nanjing, China (approval number 2018-SR-375). The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.
Author contributions
YL: Data curation, Formal analysis, Methodology, Software, Validation, Visualization, Writing – original draft, Writing – review & editing. ZZ: Investigation, Methodology, Software, Writing – review & editing. HZ: Resources, Software, Writing – review & editing. YTY: Data curation, Methodology, Writing – review & editing. MZ: Data curation, Methodology, Writing – review & editing. CW: Data curation, Methodology, Writing – review & editing. MG: Data curation, Software, Writing – review & editing. JT: Project administration, Resources, Writing – review & editing. MB: Resources, Supervision, Writing – review & editing. TY: Investigation, Project administration, Resources, Supervision, Writing – review & editing. MS: Funding acquisition, Investigation, Project administration, Resources, Supervision, Writing – review & editing. YHY: Conceptualization, Funding acquisition, Investigation, Project administration, Resources, Supervision, Writing – original draft, Writing – review & editing.
Funding
The authors declared that financial support was received for this work and/or its publication. This work was supported by the National Natural Science Foundation of China (grant number 82300887); the Natural Science Foundation of Jiangsu Province (grant number BK20220717); the China Postdoctoral Science Foundation (grant number 2022TQ0132); the Key project of Health Department of Jiangsu Province (grant number K2023046).
Acknowledgments
We thank the participants and staffs of the Framingham Heart Study Gen III Cohort.
Conflict of interest
The authors declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fendo.2025.1686480/full#supplementary-material
References
1. Funder JW. Aldosterone and mineralocorticoid receptors-physiology and pathophysiology. Int J Mol Sci. (2017) 18:1032. doi: 10.3390/ijms18051032
2. Ooi A, Khan H, Akram M, Fuller PJ, Milat F, Yang J, et al. Changes in parathyroid hormone across the spectrum of renin-independent aldosteronism. J Clin Endocrinol Metab. (2025) 110:e3694–700. doi: 10.1210/clinem/dgaf151
3. Baudrand R, Guarda FJ, Fardella C, Hundemer G, Brown J, Williams G, et al. Continuum of renin-independent aldosteronism in normotension. Hypertension. (2017) 69:950–6. doi: 10.1161/HYPERTENSIONAHA.116.08952
4. Hu J, Shen H, Huo P, Yang J, Fuller PJ, Wang K, et al. Heightened cardiovascular risk in hypertension associated with renin-independent aldosteronism versus renin-dependent aldosteronism: a collaborative study. J Am Heart Assoc. (2021) 10:e023082. doi: 10.1161/JAHA.121.023082
5. Hu J, Chen X, Luo Y, Yang J, Zeng Q, Luo W, et al. Renin-independent aldosteronism and chronic kidney disease in diabetes: Observational and Mendelian randomization analyses. Metabolism. (2023) 145:155593. doi: 10.1016/j.metabol.2023.155593
6. Zeng Q, Luo X, Chen X, Luo W, Li R, Yang S, et al. Renin-independent aldosteronism and metabolic dysfunction-associated steatotic liver disease and cirrhosis: A genetic association study. Clin Nutr. (2025) 44:193–200. doi: 10.1016/j.clnu.2024.12.011
7. Young WF and Hogan MJ. Renin-independent hypermineralocorticoidism. Trends Endocrinol Metab. (1994) 5:97–106. doi: 10.1016/1043-2760(94)90090-6
8. Buonafine M, Bonnard B, and Jaisser F. Mineralocorticoid receptor and cardiovascular disease. Am J Hypertens. (2018) 31:1165–74. doi: 10.1093/ajh/hpy120
9. Vaidya A, Mulatero P, Baudrand R, and Adler GK. The expanding spectrum of primary aldosteronism: implications for diagnosis, pathogenesis, and treatment. Endocr Rev. (2018) 39:1057–88. doi: 10.1210/er.2018-00139
10. Adler GK, Stowasser M, Correa RR, Khan N, Kline G, McGowan MJ, et al. Primary aldosteronism: an endocrine society clinical practice guideline. J Clin Endocrinol Metab. (2025) 110:2453–95. doi: 10.1210/clinem/dgaf284
11. Brown JM, Robinson-Cohen C, Luque-Fernandez MA, Allison MA, Baudrand R, Ix JH, et al. The spectrum of subclinical primary aldosteronism and incident hypertension: a cohort study. Ann Internal Med. (2017) 167:630–41. doi: 10.7326/M17-0882
12. Inoue K, Goldwater D, Allison M, Seeman T, Kestenbaum BR, and Watson KE. Serum aldosterone concentration, blood pressure, and coronary artery calcium: the multi-ethnic study of atherosclerosis. Hypertension. (2020) 76:113–20. doi: 10.1161/HYPERTENSIONAHA.120.15006
13. Hundemer GL, Agharazii M, Madore F, Vaidya A, Brown JM, Leung AA, et al. Subclinical primary aldosteronism and cardiovascular health: a population-based cohort study. Circulation. (2024) 149:124–34. doi: 10.1161/CIRCULATIONAHA.123.066389
14. Joseph JJ, Pohlman NK, Zhao S, Kline D, Brock G, Echouffo-Tcheugui JB, et al. Association of serum aldosterone and plasma renin activity with ambulatory blood pressure in African Americans: The Jackson Heart Study. Circulation. (2021) 143:2355–66. doi: 10.1161/CIRCULATIONAHA.120.050896
15. Yang J, May Gwini S, Beilin LJ, Schlaich M, Stowasser M, Young MJ, et al. Relationship between the aldosterone-to-renin ratio and blood pressure in young adults: a longitudinal study. Hypertension. (2021) 78:387–96. doi: 10.1161/HYPERTENSIONAHA.121.17336
16. Vasan RS, Evans JC, Larson MG, Wilson PWF, Meigs JB, Rifai N, et al. Serum aldosterone and the incidence of hypertension in nonhypertensive persons. New Engl J Med. (2004) 351:33–41. doi: 10.1056/NEJMoa033263
17. Brown JM, Wijkman MO, Claggett BL, Shah AM, Ballantyne CM, Coresh J, et al. Cardiac structure and function across the spectrum of aldosteronism: the atherosclerosis risk in communities study. Hypertension. (2022) 79:1984–93. doi: 10.1161/HYPERTENSIONAHA.122.19134
18. Shah SS, Zhang J, Gwini SM, Young MJ, Fuller PJ, and Yang J. Efficacy and safety of mineralocorticoid receptor antagonists for the treatment of low-renin hypertension: a systematic review and meta-analysis. J Hum Hypertens. (2024) 38:383–92. doi: 10.1038/s41371-023-00891-1
19. Akintunde A, Nondi J, Gogo K, Jones ESW, Rayner BL, Hackam DG, et al. Physiological phenotyping for personalized therapy of uncontrolled hypertension in Africa. Am J Hypertens. (2017) 30:923–30. doi: 10.1093/ajh/hpx066
20. Turner ST, Schwartz GL, Chapman AB, Beitelshees AL, Gums JG, Cooper-DeHoff RM, et al. Plasma renin activity predicts blood pressure responses to beta-blocker and thiazide diuretic as monotherapy and add-on therapy for hypertension. Am J Hypertens. (2010) 23:1014–22. doi: 10.1038/ajh.2010.98
21. Feraco A, Marzolla V, Scuteri A, Armani A, and Caprio M. Mineralocorticoid receptors in metabolic syndrome: from physiology to disease. Trends Endocrinol Metab: TEM. (2020) 31:205–17. doi: 10.1016/j.tem.2019.11.006
22. Carlsson PO, Berne C, and Jansson L. Angiotensin II and the endocrine pancreas: effects on islet blood flow and insulin secretion in rats. Diabetologia. (1998) 41:127–33. doi: 10.1007/s001250050880
23. Schupp M, Janke J, Clasen R, Unger T, and Kintscher U. Angiotensin type 1 receptor blockers induce peroxisome proliferator-activated receptor-gamma activity. Circulation. (2004) 109:2054–7. doi: 10.1161/01.CIR.0000127955.36250.65
24. Abuissa H, Jones PG, Marso SP, and O’Keefe JH Jr. Angiotensin-converting enzyme inhibitors or angiotensin receptor blockers for prevention of type 2 diabetes: a meta-analysis of randomized clinical trials. J Am Coll Cardiol. (2005) 46:821–6. doi: 10.1016/j.jacc.2005.05.051
25. Pepine CJ, Handberg EM, Cooper-DeHoff RM, Marks RG, Kowey P, Messerli FH, et al. A calcium antagonist vs a non-calcium antagonist hypertension treatment strategy for patients with coronary artery disease. The International Verapamil-Trandolapril Study (INVEST): a randomized controlled trial. JAMA. (2003) 290:2805–16. doi: 10.1001/jama.290.21.2805
26. Tomiyama H, Kushiro T, Abeta H, Ishii T, Takahashi A, Furukawa L, et al. Kinins contribute to the improvement of insulin sensitivity during treatment with angiotensin converting enzyme inhibitor. Hypertension. (1994) 23:450–5. doi: 10.1161/01.hyp.23.4.450
27. Yamaji M, Tsutamoto T, Kawahara C, Nishiyama K, Yamamoto T, Fujii M, et al. Effect of eplerenone versus spironolactone on cortisol and hemoglobin A1(c) levels in patients with chronic heart failure. Am Heart J. (2010) 160:915–21. doi: 10.1016/j.ahj.2010.04.024
28. Johansen ML, Schou M, Rossignol P, Holm MR, Rasmussen J, Brandt N, et al. Effect of the mineralocorticoid receptor antagonist eplerenone on liver fat and metabolism in patients with type 2 diabetes: a randomized, double-blind, placebo-controlled trial (MIRAD trial). Diabetes Obes Metab. (2019) 21:2305–14. doi: 10.1111/dom.13809
29. Preiss D, van Veldhuisen DJ, Sattar N, Krum H, Swedberg K, Shi H, et al. Eplerenone and new-onset diabetes in patients with mild heart failure: results from the Eplerenone in Mild Patients Hospitalization and Survival Study in Heart Failure (EMPHASIS-HF). Eur J Heart Fail. (2012) 14:909–15. doi: 10.1093/eurjhf/hfs067
30. Butt JH, Jhund PS, Henderson AD, Claggett BL, Desai AS, Viswanathan P, et al. Finerenone and new-onset diabetes in heart failure: a prespecified analysis of the FINEARTS-HF trial. Lancet Diabetes Endocrinol. (2025) 13:107–18. doi: 10.1016/S2213-8587(24)00309-7
31. Yang Y, Liu Y, Williams TA, Gao M, Yan Y, Bao M, et al. Metabolic phenotypes and fatty acid profiles associated with histopathology of primary aldosteronism. Hypertens Res. (2025) 48:1363–78. doi: 10.1038/s41440-025-02143-w
32. Andersson C, Johnson AD, Benjamin EJ, Levy D, and Vasan RS. 70-year legacy of the framingham heart study. Nat Rev Cardiol. (2019) 16:687–98. doi: 10.1038/s41569-019-0202-5
33. Mahmood SS, Levy D, Vasan RS, and Wang TJ. The Framingham Heart Study and the epidemiology of cardiovascular disease: a historical perspective. Lancet. (2014) 383:999–1008. doi: 10.1016/S0140-6736(13)61752-3
34. Parikh NI, Hwang S-J, Larson MG, Cupples LA, Fox CS, Manders ES, et al. Parental occurrence of premature cardiovascular disease predicts increased coronary artery and abdominal aortic calcification in the Framingham Offspring and Third Generation cohorts. Circulation. (2007) 116:1473–81. doi: 10.1161/CIRCULATIONAHA.107.705202
35. Hruby A, Dennis C, and Jacques PF. Dairy intake in 2 American adult cohorts associates with novel and known targeted and nontargeted circulating metabolites. J Nutr. (2020) 150:1272–83. doi: 10.1093/jn/nxaa021
36. Cunha J, Chan MV, Nkambule BB, Thibord F, Lachapelle A, Pashek RE, et al. Trends among platelet function, arterial calcium, and vascular function measures. Platelets. (2023) 34:2238835. doi: 10.1080/09537104.2023.2238835
37. Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF 3rd, Feldman HI, et al. A new equation to estimate glomerular filtration rate. Ann Internal Med. (2009) 150:604–12. doi: 10.7326/0003-4819-150-9-200905050-00006
38. Pasquali R, Casanueva F, Haluzik M, van Hulsteijn L, Ledoux S, Monteiro MP, et al. European society of endocrinology clinical practice guideline: Endocrine work-up in obesity. Eur J Endocrinol. (2020) 182:G1–G32. doi: 10.1530/EJE-19-0893
39. Kobayashi H, Abe M, Soma M, Takeda Y, Kurihara I, Itoh H, et al. Development and validation of subtype prediction scores for the workup of primary aldosteronism. J Hypertens. (2018) 36:2269–76. doi: 10.1097/HJH.0000000000001855
40. Tsao CW and Vasan RS. Cohort Profile: The Framingham Heart Study (FHS): overview of milestones in cardiovascular epidemiology. Int J Epidemiol. (2015) 44:1800–13. doi: 10.1093/ije/dyv337
41. Shah RV, Steffen LM, Nayor M, Reis JP, Jacobs DR, Allen NB, et al. Dietary metabolic signatures and cardiometabolic risk. Eur Heart J. (2023) 44:557–69. doi: 10.1093/eurheartj/ehac446
42. D’Agostino RB Sr., Vasan RS, Pencina MJ, Wolf PA, Cobain M, Massaro JM, et al. General cardiovascular risk profile for use in primary care: the Frami ngham Heart Study. Circulation. (2008) 117:743–53. doi: 10.1161/CIRCULATIONAHA.107.699579
43. Mick E, Shah R, Tanriverdi K, Murthy V, Gerstein M, Rozowsky J, et al. Stroke and circulating extracellular RNAs. Stroke. (2017) 48:828–34. doi: 10.1161/STROKEAHA.116.015140
44. McKee PA, Castelli WP, McNamara PM, and Kannel WB. The natural history of congestive heart failure: the Framingham study. New Engl J Med. (1971) 285:1441–6. doi: 10.1056/NEJM197112232852601
45. Schnabel RB, Yin X, Gona P, Larson MG, Beiser AS, McManus DD, et al. 50 year trends in atrial fibrillation prevalence, incidence, risk factors, and mortality in the Framingham Heart Study: a cohort study. Lancet. (2015) 386:154–62. doi: 10.1016/S0140-6736(14)61774-8
46. Segan L, Canovas R, Nanayakkara S, Chieng D, Prabhu S, Voskoboinik A, et al. New-onset atrial fibrillation prediction: the HARMS2-AF risk score. Eur Heart J. (2023) 44:3443–52. doi: 10.1093/eurheartj/ehad375
47. Lang RM, Badano LP, Mor-Avi V, Afilalo J, Armstrong A, Ernande L, et al. Recommendations for cardiac chamber quantification by echocardiography in adults: an update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging. J Am Soc Echocardiogr. (2015) 28:1–39.e14. doi: 10.1016/j.echo.2014.10.003
48. Devereux RB and Reichek N. Echocardiographic determination of left ventricular mass in man. Anat Validation Method Circ. (1977) 55:613–8. doi: 10.1161/01.cir.55.4.613
49. Castro-Diehl C, Song RJ, Sawyer DB, Wollert KC, Mitchell GF, Cheng S, et al. Circulating growth factors and cardiac remodeling in the community: The Framingham Heart Study. Int J Cardiol. (2021) 329:217–24. doi: 10.1016/j.ijcard.2020.12.088
50. Nanba K, Vaidya A, Williams GH, Zheng I, Else T, and Rainey WE. Age-related autonomous aldosteronism. Circulation. (2017) 136:347–55. doi: 10.1161/CIRCULATIONAHA.117.028201
51. Williams TA, Gomez-Sanchez CE, Rainey WE, Giordano TJ, Lam AK, Marker A, et al. International histopathology consensus for unilateral primary aldosteronism. J Clin Endocrinol Metab. (2021) 106:42–54. doi: 10.1210/clinem/dgaa484
52. Williams TA and Reincke M. Pathophysiology and histopathology of primary aldosteronism. Trends Endocrinol Metab: TEM. (2022) 33:36–49. doi: 10.1016/j.tem.2021.10.002
53. Huby A-C, Antonova G, Groenendyk J, Gomez-Sanchez CE, Bollag WB, Filosa JA, et al. Adipocyte-derived hormone leptin is a direct regulator of aldosterone secretion, which promotes endothelial dysfunction and cardiac fibrosis. Circulation. (2015) 132:2134–45. doi: 10.1161/CIRCULATIONAHA.115.018226
54. Holthuis JCM and Menon AK. Lipid landscapes and pipelines in membrane homeostasis. Nature. (2014) 510:48–57. doi: 10.1038/nature13474
55. Pang Y, Gong S, Tetti M, Sun Z, Mir-Bashiri S, Bidlingmaier M, et al. EGR1 regulates oxidative stress and aldosterone production in adrenal cells and aldosterone-producing adenomas. Redox Biol. (2025) 80:103498. doi: 10.1016/j.redox.2025.103498
56. Spaulding SC and Bollag WB. The role of lipid second messengers in aldosterone synthesis and secretion. J Lipid Res. (2022) 63:100191. doi: 10.1016/j.jlr.2022.100191
57. Spyroglou A, Handgriff L, Müller L, Schwarzlmüller P, Parasiliti-Caprino M, Fuss CT, et al. The metabolic phenotype of patients with primary aldosteronism: impact of subtype and sex - a multicenter-study of 3566 Caucasian and Asian subjects. Eur J Endocrinol. (2022) 187:361–72. doi: 10.1530/EJE-22-0040
58. Ohno Y, Sone M, Inagaki N, Yamasaki T, Ogawa O, Takeda Y, et al. Obesity as a key factor underlying idiopathic hyperaldosteronism. J Clin Endocrinol Metab. (2018) 103:4456–64. doi: 10.1210/jc.2018-00866
59. Rifkin DE, Khaki AR, Jenny NS, McClelland RL, Budoff M, Watson K, et al. Association of renin and aldosterone with ethnicity and blood pressure: the Multi-Ethnic Study of Atherosclerosis. Am J Hypertens. (2014) 27:801–10. doi: 10.1093/ajh/hpt276
60. Alnazer RM, Veldhuizen GP, de Leeuw PW, and Kroon AA. The effect of age, sex and BMI on the aldosterone-to-renin ratio in essential hypertensive individuals. J Hypertens. (2023) 41:618–23. doi: 10.1097/HJH.0000000000003377
61. Ananda RA, Gwini S, Beilin LJ, Schlaich MP, Stowasser M, Young MJ, et al. Relationship between renin, aldosterone, aldosterone-to-renin ratio and arterial stiffness and left ventricular mass index in young adults. Circulation. (2024) 150:2019–30. doi: 10.1161/CIRCULATIONAHA.124.070039
62. Unger T, Borghi C, Charchar F, Khan NA, Poulter NR, Prabhakaran D, et al. 2020 International Society of Hypertension global hypertension practice guidelines. Hypertension. (2020) 75:1334–57. doi: 10.1161/HYPERTENSIONAHA.120.15026
63. Marzano L, Merlo M, Martinelli N, Pizzolo F, and Friso S. Efficacy and safety of aldosterone synthase inhibitors for hypertension: a meta-analysis of randomized controlled trials and systematic review. Hypertension. (2025) 82:e47–56. doi: 10.1161/HYPERTENSIONAHA.124.23962
64. Agarwal R, Kolkhof P, Bakris G, Bauersachs J, Haller H, Wada T, et al. Steroidal and non-steroidal mineralocorticoid receptor antagonists in cardiorenal medicine. Eur Heart J. (2021) 42:152–61. doi: 10.1093/eurheartj/ehaa736
65. Grune J, Beyhoff N, Smeir E, Chudek R, Blumrich A, Ban Z, et al. Selective mineralocorticoid receptor cofactor modulation as molecular basis for finerenone’s antifibrotic activity. Hypertension. (2018) 71:599–608. doi: 10.1161/HYPERTENSIONAHA.117.10360
66. Kolkhof P, Jaisser F, Kim S-Y, Filippatos G, Nowack C, and Pitt B. Steroidal and novel non-steroidal mineralocorticoid receptor antagonists in heart failure and cardiorenal diseases: comparison at bench and bedside. Handb Exp Pharmacol. (2017) 243:271–305. doi: 10.1007/164_2016_76
67. Butt JH, Henderson AD, Jhund PS, Claggett BL, Desai AS, Lay-Flurrie J, et al. Finerenone, obesity, and heart failure with mildly reduced/preserved ejection fraction: prespecified analysis of FINEARTS-HF. J Am Coll Cardiol. (2025) 85:140–55. doi: 10.1016/j.jacc.2024.10.111
68. Bakris GL, Agarwal R, Anker SD, Pitt B, Ruilope LM, Rossing P, et al. Effect of finerenone on chronic kidney disease outcomes in type 2 diabetes. N Engl J Med. (2020) 383:2219–29. doi: 10.1056/NEJMoa2025845
69. Hu J, Zhou Q, Sun Y, Feng Z, Yang J, He W, et al. Efficacy and safety of finerenone in patients with primary aldosteronism: a pilot randomized controlled trial. Circulation. (2025) 151:196–8. doi: 10.1161/CIRCULATIONAHA.124.071452
70. Ostrominski JW, Claggett BL, Miao ZM, Filippatos G, Desai AS, Jhund PS, et al. Efficacy and safety of finerenone in type 2 diabetes: a pooled analysis of trials of heart failure and chronic kidney disease. Diabetes Care. (2025) 48:745–55. doi: 10.2337/dc24-1873
Keywords: aldosterone, cardiovascular disease, classification, risk stratification, renin-independent aldosteronism
Citation: Liu Y, Zhang Z, Zhou H, Yan Y, Zhou M, Wang C, Gao M, Tao J, Bao M, Yang T, Sun M and Yang Y (2025) Data-driven cluster analysis and external validation identify phenotypic subgroups in renin-independent aldosteronism with differential cardiovascular risk and therapeutic implications. Front. Endocrinol. 16:1686480. doi: 10.3389/fendo.2025.1686480
Received: 15 August 2025; Accepted: 30 November 2025; Revised: 25 November 2025;
Published: 15 December 2025.
Edited by:
Nina Japundzic-Zigon, University of Belgrade, SerbiaReviewed by:
Wenbo Yang, People’s Hospital of Xinjiang Uygur Autonomous Region, ChinaPan Zhou, Xinjiang Hypertension Institute, China
Copyright © 2025 Liu, Zhang, Zhou, Yan, Zhou, Wang, Gao, Tao, Bao, Yang, Sun and Yang. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Yuhong Yang, eXVob25nLnlhbmdAbmptdS5lZHUuY24=; Min Sun, ZHJzdW5tQG5qbXUuZWR1LmNu; Tao Yang, eWFuZ3RAbmptdS5lZHUuY24=
†These authors have contributed equally to this work and shared first authorship
‡These authors have contributed equally to this work and shared last authorship