Influence of heart rate on left ventricular isovolumic relaxation time: a Doppler study in healthy newborns

Abstract

Left ventricular (LV) isovolumic relaxation time (LVIVRT) obtained by Doppler spectral analysis has emerged as a useful tool to assess diastolic function. 1,2 Correction for heart rate (HR) is an important advance for clinical application of LVIVRT, initially reported in healthy adult subjects by simultaneous phonocardiogram and echocardiogram. 3 A recent investigation concerning the duration of Doppler-derived LVIVRT in infancy and adolescence has been published, and correction for HR by dividing the absolute values by the square root of cardiac cycle length has been presented. 4 The aim of our study was to use such technique in healthy newborns to evaluate the influence of resting HR on LVIVRT and to find a regression equation for the prediction of normal values. We examined 22 healthy full-term neonates (Table). Ductal patency was excluded in all babies by color flow mapping and pulsed Doppler analysis of flow in the main pulmonary artery. Image-directed pulsed and color Doppler echocardiography was performed with the infants in a resting state by means of an ultrasound scanner (Aspen, Acuson, Mountain View, Calif) using a 6-MHz transducer. All examinations included an electrocardiographic trace and were recorded at a screen sweep speed of 75 mm/s on to videotape for later analysis. At each study, from apical window a 4-chamber view was obtained and sample volume (1.5-mm width) was placed in the area of anterior mitral valve leaflet to capture mitral and transaortic flow profiles simultaneously. At least 30 cardiac cycles were recorded for each participant adjusting wall filter settings to the minimum appropriate. Parents received an explanation of the study and gave informed consent in each case. The Doppler traces were read from the videotape with a sample rate of 4 milliseconds. Details regarding the measurement of the LVIVRT have been published. 5 The reliability of the measurement depends on a clear definition of both the closure click of aortic valve and the beginning of LV diastolic filling. The mean of 6 consecutive beats was taken as the representative measurement. HR was calculated from the ratio 60/average cardiac cycle length (RR). The best fitting of LVIVRT versus HR was derived by least squares method. All the participants in this study showed a sharp spike of aortic valve closure and a well-demarcated initial deflection of the early rapid filling (E) wave. In 19 participants the HR was stable, but there were marked differences among babies because of individual variations. In 3 babies, in whom marked changes in HR were observed, the changes of LVIVRT followed the general trend (Figure). It is notable that LVIVRT increases with increasing HR (R 2 0.94, P .01), which is in accordance with previous data in healthy adult man. 3 Repetition of LVIVRT measurement by the same observer (A. D.) in 15 randomly selected babies more than 4 weeks later gave an excellent intraobserver reproducibility (R 0.95). Previous studies of diastolic function in newborns did not specifically focus on the finding of Dopplerderived LVIVRT related to HR. Our results indicate that LVIVRT is HR-dependent and tends to lengthen as HR increases in healthy neonates following a curvilinear relationship. In applying LVIVRT clinically, deviations from the normal regression can be expressed by