Torsional correlates for End Systolic Volume Index in adult healthy subjects.

3-dimensional (3D) speckle tracking-based motion-detecting echocardiography (STE) was used in order to evaluate left ventriculum (LV) rotation, twist and torsion in 16 segments (base, medium, apex and global) in 32 healthy subjects. All parameters were indexed by body surface area. Then, gender-related differences were removed by gender-centering the entire dataset. Linear models were run to predict ESVI by using single torsional parameters and by evaluating the degree of their collinearities. Given the strong collinearities on predictors, we performed a principal component regression and we found an axis that predicts ESVI better than the single most correlated original variables. We statistically assessed that this axis is a significant best predictor of ESVI than original torsional variables yet containing strict relationships with them. The physiological implications of our results are discussed in the context of heart contractility theory as well as of the actual difficulties to measure electro-mechanical determinants of healthy subjects in vivo.