Metabolic and inflammatory phenotypic divergence between primary aldosteronism and essential hypertension: a propensity score-matched cohort study

Background/Introduction Patients with primary aldosteronism (PA) face a significantly higher risk of cardiovascular target organ damage compared to those diagnosed with essential hypertension (EH). However, the systemic metabolic and inflammatory characteristics distinguishing PA from EH remain poorly defined. Purpose This study aims to systematically compare population-level differences in emerging proatherosclerotic and inflammatory markers between patients with PA and those with EH within a large, clinically matched cohort. Methods This retrospective cohort study utilises data from a comprehensive clinical database. To reduce baseline confounding, 1:4 propensity score matching (PSM) is performed for PA and EH patients based on age, gender, body mass index (BMI), history of tobacco and alcohol, and blood pressure. Missing laboratory values are addressed using multiple imputation. Multivariable logistic regression, restricted cubic splines (RCS), and standardised Z-scores are utilised to evaluate the independent associations, non-linear dose-response patterns, and population-level phenotypic shifts of non-high-density lipoprotein cholesterol (non-HDL-C), serum uric acid to high-density lipoprotein (SUA/HDL) ratio, C-reactive protein and triglyceride-glucose (CRP-TyG), and CRP to triglyceride (CRP-TG) indices with PA risk. Results The final matched cohort includes 2291 patients (459 PA; 1832 EH). The standardised Z-score dumbbell plot demonstrates that PA patients exhibit a lower burden across most metabolic and inflammatory markers, with heterogeneity observed in certain indices (e.g., atherogenic index of plasma, AIP). In fully adjusted models, elevated non-HDL-C (odds ratio (OR) 0.866; 95% confidence interval (CI) 0.777-0.965), CRP-TyG (OR 0.784; 95% CI 0.661-0.929), CRP-TG (OR 0.855; 95% CI 0.741-0.987), and SUA/HDL ratio (OR 0.772; 95% CI 0.684-0.872) are independently associated with lower odds of PA (all P < 0.05). RCS modelling reveals a linear inverse association for non-HDL-C (Poverall = 0.016), alongside marked non-linear inverse relationships for the remaining three indicators (Poverall < 0.001; Pnon-linear ≤ 0.010) that are particularly pronounced across the low-to-moderate metabolic burden spectrum. Conclusion(s) Compared to baseline-matched patients with EH, those with PA manifest a lower systemic metabolic and inflammatory burden. These findings suggest that the elevated cardiovascular risk in PA may be partially decoupled from traditional lipid abnormalities and low-grade systemic inflammation. Despite the retrospective design, our findings underscore that an apparently unremarkable metabolic profile must not deter clinicians from pursuing PA screening in hypertensive individuals.