Statistical Analysis

Sex-specific clinical and CMR-derived measures were assessed. The distribution of Ecc was individually assessed according to sex, age, treatment group (intensive/conventional), study cohort (primary/secondary), smoking status, alcohol use, obesity status (BMI ≥ or <30 kg/m²), renal dysfunction parameters, hypertension, lipids profile, metabolic syndrome criteria, and presence/type of myocardial scar (Supplementary Data). Spearman correlation was used to evaluate the association between diffuse fibrosis assessed by T1 values and Ecc, adjusting for sex. For myocardial scar, the presence (binary: yes/no) and the type of scar (two patterns: ischemic and nonischemic) were assessed as independent variables. The associations between covariates and circumferential deformation (Ecc as the dependent variable) were assessed using minimally adjusted linear regression models, which included a basic model and minimally adjusted risk factor models. Finally, a multivariate linear regression model that tested the influence of cardiovascular risk factors and myocardial scar on Ecc and LV ejection fraction values was assessed, adjusting for anthropometrics, cardiovascular risk factors, and presence/type of myocardial scar.