Background: Cardiovascular prognosis related to type 2 diabetes may not be adequately captured by information on comorbid conditions such as obesity and hypertension. To inform the cardiovascular prognosis among diabetic individuals, we conducted phenotyping using a clustering approach based on clinical data, echocardiographic indices and biomarkers.
Methods: We performed a cluster analysis on clinical, biochemical and echocardiographic variables from 529 Blacks with diabetes in the Jackson Heart Study. An association between identified clusters and major adverse cardiovascular events (MACE- composite of coronary heart disease, stroke, heart failure and atrial fibrillation) was assessed using Cox proportional hazards modeling.
Results: Cluster analysis separated individuals with diabetes (68% women, mean age 60 ± 10 years) into three distinct clusters (Clusters 1,2 &3 - with Cluster 3 being a hypertrophic cluster characterized by highest LV mass, levels of brain natriuretic peptide [BNP] and high-sensitivity cardiac troponin-I [hs-cTnI]). After a median 12.1 years, there were 141 cardiovascular events. Compared to Cluster1, Clusters 3 had an increased risk of cardiovascular disease (hazard ratio [HR] 1.60; 95% confidence interval [CI] 1.08, 2.37), while Cluster 2 had a similar risk of outcome (HR 1.11; 95% CI 0.73, 168).
Conclusions: Among Blacks with diabetes, cluster analysis identified three distinct echocardiographic and biomarkers phenotypes, with cluster 3 (high LV mass, high cardiac biomarkers) associated with worse outcomes, thus highlighting the prognostic value of subclinical myocardial dysfunction.
Keywords: Biomarkers; Cluster analysis; Comorbidities; Diabetes; Echocardiography; Outcomes.
© 2022. The Author(s).