Key result
Machine learning clustering identifies 4 hypertensive phenotypes with distinct cardiovascular risks and LV remodeling responses.
Why the study?
The authors sought to identify hypertensive phenotypes with distinct clinical profiles, comorbidities, remodeling trajectories, and outcomes using unsupervised machine learning clustering of multidimensional data including strain imaging.
Can machine-learning-derived clustering using clinical and echocardiographic strain data identify distinct phenotypes with divergent outcomes and remodeling trajectories in hypertensive patients?
Can machine-learning-derived clustering using clinical and echocardiographic strain data identify distinct phenotypes with divergent outcomes and remodeling trajectories in hypertensive patients?
Machine-learning clustering of clinical and echocardiographic strain data identifies four distinct hypertensive phenotypes with divergent cardiovascular risks and remodeling trajectories, potentially enabling tailored management.
No takes yet. Share an insight, caveat, or question.
ML phenotypes may refine risk stratification in hypertension; hypothesis-generating and requires prospective validation before guiding care.
Aims We applied unsupervised machine learning (ML) clustering to a large cohort of hypertensive patients undergoing echocardiography with strain imaging to identify phenotypes with distinct clinical profiles, comorbidities, remodeling trajectories, and outcomes. Methods and Results We analyzed 1,607 patients from the STRATS-HHD registry who underwent echocardiography at baseline and after 6–18 months of therapy. Twenty clinical, laboratory, and echocardiographic variables—including left atrial and left ventricular strain—underwent principal component analysis and K-means clustering (K=4). Clusters were derived in the SNUBH cohort (n=1,204) and validated in the CAUH cohort (n=403), two institutional subsets of the registry. Remodeling trajectories were assessed using baseline-adjusted models, and associations with outcomes were evaluated using multivariable Cox regression. Four clusters emerged: (1) AF-predominant, with advanced remodeling and the highest event risk; (2) elderly, with metabolic–renal comorbidities but preserved function; (3) middle-aged, with prevalent coronary disease and relatively preserved function; and (4) younger, with severe hypertension, marked strain impairment, and the greatest remodeling regression with therapy. Prognosis varied: cluster 1 had the highest risk of cardiovascular death, heart failure hospitalization (HHF), stroke, and major adverse cardiovascular events (MACE); cluster 2 exhibited increased cardiovascular death and intermediate HHF risk; cluster 3 showed elevated coronary risk; and cluster 4 the most favorable outcomes. Associations between medication and remodeling varied, with renin–angiotensin blockade linked to LV mass regression in cluster 4. Conclusions ML-based clustering incorporating strain identified four distinct HHD phenotypes with divergent remodeling, therapeutic responses, and outcomes. Data-driven phenotyping may improve risk stratification and enable tailored management in hypertension.
Hwang et al. (2026) studied this question. ML clustering of hypertensive patients identified 4 distinct phenotypes with different remodeling, outcomes, and responses; cluster 1 had highest risk and cluster 4 showed greatest LV mass regression .
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: