A linear relationship between left ventricular mass (LVM) and signal-averag
ed QRS duration (QRSD) in healthy subjects has been previously demonstrated
. To complete this evaluation, the aim of this study was to analyze the rel
ation between LVM and QRSD (assessed, respectively, by echocardiography and
signal-averaged EGG) using several regression models (linear, quadratic, c
ubic, logarithmic, exponential, and double-logarithmic regression models) a
nd varying the choice of the band-pass filtering. Although a linear relatio
n between LVM and QRSD has been confirmed, other regression models have dem
onstrated better fittings to the data: The best was provided by the cubic r
elation model. Moreover, the use of a 40- to 250-Hz band-pass filtering pro
vided the best model fittings, In conclusion, our analysis pointed out the
cubic regression as a specific relationship model between LVM and QRSD in h
ealthy subjects, Further studies could verify the effectiveness of this rel
ation model in subjects with a wider range of LVM.