highly significantp < 0.001
All intermediate timepoints (3, 6, and 12 months) also showed highly significant improvements compared to baseline (all p < 0.001) with stable distributions compared to 1 month (all p > 0.05) (Fig. 2 ).
All intermediate timepoints (3, 6, and 12 months) also showed highly significant improvements compared to baseline (all p < 0.001) with stable distributions compared to 1 month (all p > 0.05) (Fig. 2 ).
BMI and diabetes mellitus, while approaching significance in univariate analysis ( p = 0.09 and p = 0.08), were excluded from final models due to loss of significance during stepwise selection (final p-values 0.24 and 0.18).
Similarly, SGLT-2 inhibitor use showed a protective trend (HR 0.68) but did not achieve significance ( p = 0.31), likely due to low utilization (12.1%).
With this small number, our ability to reliably identify independent predictors is limited, and variables appearing non-significant (such as LVEF, coronary artery disease) may actually be important predictors but failed to achieve significance due to insufficient power.