Background
In adults, left ventricular (LV) systolic twist is an important factor that determines LV filling, both at rest and during exercise. In children, lower LV twist has been demonstrated at rest, but its adaptation during exercise and its functional consequences on LV filling are unknown.
Methods
Using speckle-tracking echocardiography, LV twist-untwist mechanics were studied in 25 children (aged 10–12 years) and 20 young adults (aged 18–44 years) at rest and during three exercise workloads performed at 20%, 30%, and 40% of their maximal aerobic power.
Results
At rest, LV twist was lower in children, because of a higher temporal dispersion of peak rotation between base and apex. During exercise, the increase of basal rotation was blunted in children compared with adults (−6.7 ± 2.7° vs −9.0 ± 2.0° at 40% of maximal aerobic power, P < .05). Consequently, LV twist increased to a lesser extent (13.0 ± 5.0° vs 15.8 ± 4.5° at 40% of maximal aerobic power, P < .05). The increase in LV untwisting rates during exercise was also lower in children, leading to a lower percentage of untwisting during early diastole (8 ± 8% vs 29 ± 20% at 40% of maximal aerobic power, P < .001). Consequently, during early diastole, the normal timing of diastolic events observed in young adults, with untwist occurring before radial displacement, was blunted in children. Nevertheless, children exhibited normal LV filling due to higher diastolic radial and longitudinal strain rates.
Conclusions
Twist-untwist mechanics may evolve with advancing age. In children, early diastolic LV untwisting appears to be less important than in adults. Their better LV intrinsic myocardial relaxation may ensure adequate LV filling during exercise without dependence on the additional effect of suction resulting from LV energy recoil.
Because of the helical orientation of the myofibers, cardiomyocyte shortening during systole induces not only left ventricular (LV) normal strains but also shear strains. LV twist is one of these shear strains, characterized by simultaneous basal clockwise and apical counterclockwise rotation. LV untwisting occurs very early in diastole (i.e., predominantly during isovolumic relaxation), thus promoting LV suction by increasing the intraventricular pressure gradient from the apex to the base. During exercise, the increase in heart rate dramatically shortens diastole duration, and a recent study demonstrated a major increase in LV untwisting resulting from an increase in both systolic basal and apical rotation. This “systolic-diastolic” coupling via the LV twist-untwist sequence constitutes one key element maintaining LV filling and stroke volume during exercise in healthy individuals.
It was recently proposed that LV twist mechanics were age related. Aging is accompanied by increased LV twist at rest, which limits its ability to increase during exercise. In children, both basal and apical LV rotation and thus LV twist were lower at rest compared with young adults. Moreover, children exhibit a greater diastolic untwisting rate and a greater amount of untwisting during isovolumic relaxation. Thereby, these specific “LV twist mechanics” in children may contribute to different responses to exercise than those observed in adults. Yet data regarding the kinetics of LV twist and untwist mechanics during exercise in children are still lacking.
We used speckle-tracking echocardiography (STE) to assess kinetics of LV myocardial mechanical events in children at rest and during an incremental exercise test. STE constitutes a reliable and suitable method to study ventricular myocardial function at rest and during exercise and is suitable for pediatric investigations. We hypothesized that the time course of LV twisting and untwisting mechanics during exercise will be altered in children because of their lower LV systolic twist. LV untwisting velocity should be altered during exercise because of lower systolic twist. Because normal LV diastolic function has previously been reported during exercise in children compared with adults, we hypothesized that the blunted LV twist-untwist mechanics might be counterbalanced by a better intrinsic myocardial relaxation or ventricular compliance properties.
Methods
Study Population
We evaluated 25 young, prepubertal healthy boys (mean age, 11 ± 1 year) and 20 healthy male adults (mean age, 25 ± 9 years). None reported regular training habits or had any clinical or anamnestic evidence of cardiovascular disease or arterial hypertension. Subjects were excluded if resting echocardiography demonstrated an ejection fraction <50%, significant valvular disease, or abnormal right ventricular function. Boys were examined clinically, and their pubertal status was assessed using Tanner stages. All boys were at Tanner stage 1. This study received approval from the local ethics committee, and written informed consent was obtained from the adults and from all children and their parents. The study conformed to the latest revision of the Declaration of Helsinki.
Experimental Protocol
Body height and mass were assessed. For both children and adults, maximal aerobic power was initially estimated via the Wasserman equation, {body mass × [(50.72 − 0.372 × age)] − 350}/10.3, and corrected for the semisupine position (20% were deduced from calculated values). For children, the estimated maximal aerobic power was checked by an additional incremental exercise protocol (stages of 1 min on a E-Bike ergometer; GE Healthcare, Milwaukee, WI).
For the cardiovascular evaluation, subjects were positioned on the dedicated semisupine cycling ergometer (E-Bike). After a 15-min to 20-min resting period, each subject underwent an incremental exercise test that included three workloads of 6 min at 20% (W1), 30% (W2), and 40% (W3) of maximal aerobic power, followed by incremental workloads of 1 min until exhaustion. During the test, the pedaling rate was kept constant at 70 to 80 rpm for all subjects.
Echocardiographic Data Acquisition
Two-dimensional and Doppler echocardiographic data were recorded at the end of the resting period and during the last 4 min of the W1, W2, and W3 stages, with commercially available systems (Vivid; GE Vingmed Ultrasound AS, Horten, Norway) using a 3.5-MHz sector-scanning electronic transducer (M4S; GE Vingmed Ultrasound AS). All measurements were taken by a single observer who was well experienced in exercise echocardiography. We recorded cine loops in parasternal short-axis (basal and apical levels) and apical four-chamber views during a 2-sec to 3-sec breath-holding period at the end of breathing out. Two-dimensional grayscale harmonic images were obtained at a rate of 65 to 90 frames/sec. Images were acquired in cine loops triggered by the QRS complex and saved for offline analysis, which was performed using dedicated software (EchoPAC version 6.0; GE Vingmed Ultrasound AS). Systemic arterial blood pressure was measured at rest and during each stage of exercise in the left arm using manual sphygmomanometry and auscultation. During the last 30 sec of each workload, we measured aortic blood flow velocity in the ascending aorta with a 2.0-MHz continuous-wave Doppler transducer (Pedof P2D CW; GE Vingmed Ultrasound) placed at the suprasternal notch for stroke volume determination, as previously described.
Data Analysis
M-mode measurements were obtained offline from the parasternal short-axis view. LV inflow E and A waves were recorded using pulsed Doppler in the apical four-chamber view. From electrocardiographic recordings, we measured the time delays from the onset of the QRS complex to the onset of aortic blood flow (aortic valve opening delay), to the end of aortic blood flow (aortic valve closure [AVC] delay), to the onset of early filling blood flow (mitral valve opening delay), and to the peak of early filling blood flow (peak E). Isovolumetric relaxation time (IVRT) was calculated as mitral valve opening minus AVC.
Speckle-tracking analysis of LV strain and twist was conducted as previously described. After manually tracing the endocardial border on the end-systolic frame of the 2D sequence, the software automatically tracked myocardial motion in this region of interest. Whenever the software signaled poor tracking efficiency, the observer readjusted the endocardial trace line and/or the width of the region of interest until a satisfying tracking score was obtained. All measurements were averaged over three to five cardiac cycles. LV longitudinal strain and strain rate (SR) were assessed in an apical four-chamber view. Radial strain, circumferential strain and SR, and LV rotation and rotational rate were assessed from short-axis views at the basal and apical levels. Particular care was taken to ensure that the basal short-axis plane contained the mitral valve and that the apical plane was acquired with the transducer in a caudal position (i.e., below the papillary muscles) to improve LV apical rotation measurement.
Two-dimensional strain and SR data were processed using a specific toolbox developed in our laboratory with Scilab version 4.1 (Scilab; INRIA, Le Chesnay, France). For temporal analysis, this software adjusted all strain and SR variables for intersubject differences in heart rate and transducer frame rate acquisition. The time sequence was normalized by interpolation to the percentage of systolic duration (i.e., AVC represented 100% of systole). After normalization, the software averaged the data from three to five cardiac cycles and performed the detection of peak strain and SR events and their timing (expressed as a percentage of systolic duration). Net LV twist was calculated as the instantaneous difference between LV apical and basal rotation. The following indices of diastolic function were calculated during IVRT: the untwisting angle (°UT IVRT = twist at AVC − twist at the end of IVRT), the percentage of untwisting during IVRT (°UT IVRT /[twist at AVC] × 100), and the mean untwisting rate during IVRT (°UT IVRT /IVRT). To assess the dynamics of global LV twist and its coupling with LV mean radial displacement (reflecting volumetric changes of the left ventricle) throughout the cardiac cycle (systole and diastole), graphical loops were built using time-aligned values of LV twist and radial displacement throughout the cardiac cycle. Mean radial displacement values were averaged from six segments obtained in basal and apical short-axis planes.
Statistical Analysis
Values are expressed as mean ± SD in tables and text and as mean ± SE in figures. Statistical analysis was performed using StatView version 5.0 (SAS Institute Inc., Cary, NC). For comparison of subject characteristics, unpaired t tests were performed. For cardiac variables, one-way analysis of variance (i.e., age group) with repeated measures (i.e., resting or exercise intensities) was performed with post hoc Bonferroni correction as appropriate. Linear regressions were used to determine the relationships between LV twist and LV untwisting rate during exercise. Statistical significance for all analyses was considered at P < .05. Intraobserver reproducibility of speckle-tracking echocardiographic indices has been assessed previously in our laboratory on 12 subjects, and coefficients of variation were <8% for both strain and rotation.
Results
Subjects’ characteristics at rest are presented in Table 1 . Briefly, children showed lower height and mass than adults ( P < .05). Conventional morphologic echocardiographic measurements showed lower LV mass relative to body surface area as well as lower LV relative wall thickness ( P < .05), suggesting more compliant ventricles in children for the latter parameter.
| Variable | Children ( n = 25) | Adults ( n = 20) |
|---|---|---|
| Age (y) | 11 ± 1 | 25 ± 9 ∗ |
| Height (cm) | 144 ± 7 | 177 ± 5 ∗ |
| Weight (kg) | 37 ± 8 | 72 ± 8 ∗ |
| BSA (m 2 ) | 1.22 ± 0.13 | 1.88 ± 0.12 ∗ |
| Resting mean arterial pressure (mm Hg) | 76 ± 5 | 99 ± 7 ∗ |
| LV mass/BSA (g/m 2 ) | 53.8 ± 9.5 | 92.5 ± 14.4 ∗ |
| Relative wall thickness | 0.25 | 0.35 ± 0.05 ∗ |
During exercise, the W1, W2, and W3 stages were performed at similar relative intensities (percentages of maximal aerobic power) in both groups, that is, at 19 ± 1%, 28 ± 1%, and 38 ± 2% of maximal aerobic power in adults and 18 ± 3%, 28 ± 4%, and 37 ± 5% in children ( P > .05). Nevertheless, heart rates were higher in children at each stage: 112 ± 9, 118 ± 9, and 128 ± 10 beats/min in children versus 100 ± 12, 110 ± 12, and 121 ± 12 beats/min in adults ( P < .05). In both groups, normalized stroke volume increased significantly from rest to W2 and remained constant between W2 and W3. At maximal exercise ( Table 2 ), mechanical power and mean arterial pressure were significantly lower in children, whereas maximal relative oxygen uptake was significantly higher ( P < .05). Nevertheless, indexed cardiac output and estimated LV filling pressure (E/e′ ratio) were similar in the two groups ( P > .05).
| Variable | Children ( n = 25) | Adults ( n = 20) |
|---|---|---|
| Maximal aerobic power (W) | 114 ± 26 | 221 ± 33 † |
| Maximal oxygen uptake (mL/min/kg) | 45.7 ± 9.3 | 36.0 ± 5.7 ∗ |
| Indexed maximal cardiac output (L/min/m 2 ) | 11.1 ± 3.0 | 10.9 ± 1.9 |
| Maximal mean arterial pressure (mm Hg) | 92 ± 9 | 128 ± 9 † |
| E/e′ ratio | 9.0 ± 1.7 | 8.8 ± 1.3 |
In both groups, longitudinal strain increased during exercise (from −18.0 ± 3.0% at rest to −20.4 ± 2.6% at W3 in children vs from −17.6 ± 2.1% to −23.2 ± 1.9% in adults, P < .05). However, longitudinal diastolic SR was higher in children at each stage (from 2.05 ± 0.33 sec −1 at rest to 2.46 ± 0.40 sec −1 at W3 in children vs from 1.41 ± 0.34 to 2.22 ± 0.38 sec −1 in adults, P < .05). Similar results were obtained for radial diastolic SR (from −2.48 ± 0.57 sec −1 at rest to −3.49 ± 0.79 sec −1 at W3 in children vs from −1.72 ± 0.33 to −2.79 ± 0.49 sec −1 in adults, P < .05).
LV Systolic Rotation and Twist during Exercise
Kinetics of peak LV basal and apical rotation and twist from rest to W3 are presented in Figure 1 , and values at rest and at W3 are presented in Table 3 . During exercise, apical rotation increased progressively in both groups. However, the increase in basal rotation was higher in adults, and LV twist was therefore lower in children at each workload. Kinetics of time to peak rotation and twist expressed as a percentage of systolic duration are presented in Figure 2 . From rest to maximal exercise, peak basal rotation was delayed in children. Peak apical rotation occurred earlier in children, but only at rest and at W1. Consequently, the delay between time to peak basal and apical rotation was higher in children at rest and at W1.
| Variable | Rest | Exercise | ||
|---|---|---|---|---|
| Children | Adults | Children | Adults | |
| Apical rotation | ||||
| Peak (°) | 5.9 ± 2.7 | 6.1 ± 2.7 | 9.6 ± 3.2 § | 9.6 ± 3.2 § |
| Time to peak (%) | 71.6 ± 18.9 | 91.4 ± 6.9 † | 74.4 ± 21.3 | 72.8 ± 18.4 ‡ |
| Basal rotation | ||||
| Peak (°) | −5.0 ± 1.9 | −5.3 ± 2.4 | −6.7 ± 2.7 § | −9.0 ± 2.0 † , § |
| Time to peak (%) | 120.4 ± 23.2 | 108.0 ± 20.2 ∗ | 117.3 ± 22.7 | 111.2 ± 17.2 ∗ |
| Twist | ||||
| Peak (°) | 8.5 ± 4.9 | 10.4 ± 3.2 † | 13.0 ± 5.0 § | 15.8 ± 4.5 † , § |
| Time to peak (%) | 97.1 ± 6.4 | 92.4 ± 4.5 | 94.8 ± 9.8 | 91.5 ± 6.0 |
∗ P < .05 and † P < 01. versus children.
LV Diastolic Untwist during Exercise
In both groups, significant ( P < .05) and similar correlations ( Figure 3 ) were found between LV systolic twist and untwisting rate ( R 2 = 0.43 in children vs R 2 = 0.50 in adults, P > .05) from rest to W3. Kinetics of LV twist and twisting rate as well as basal and apical rotation at rest and at W3 are presented in Figure 4 . During exercise, the increase in peak LV untwisting rate was lower in children than in adults. Therefore, peak untwisting rate was lower at W3 in children ( Figure 4 ). LV diastolic variables are presented in Table 4 . Time to peak untwisting rate did not significantly change during exercise in children or in adults. There was no significant change in °UT IVRT during exercise, but it was significantly lower in children. Mean untwisting rate during IVRT remained constant in children, whereas it increased during exercise in adults. The percentage of untwisting during IVRT decreased during exercise in both groups, but children showed lower values than adults.
| Variable | Rest | W3 | ||
|---|---|---|---|---|
| Children | Adults | Children | Adults | |
| Diastolic duration (msec) | 522 ± 113 | 608 ± 128 ∗ | 230 ± 30 | 272 ± 31 ∗ , § |
| IVRT (msec) | 41.2 ± 13.7 | 65.6 ± 18.0 | 26.1 ± 10.8 | 35.1 ± 10.2 |
| Systolic time duration (%) | 13.3 ± 4.5 | 20.3 ± 5.6 † | 11.9 ± 5.1 ‡ | 15.7 ± 4.9 † , ‡ |
| Mitral valve opening | ||||
| Systolic time duration (%) | 113.3 ± 4.5 | 120.4 ± 5.6 † | 111.9 ± 5.1 ‡ | 115.7 ± 4.9 † , § |
| Peak E | ||||
| Peak (cm/sec) | 96 ± 12 | 88 ± 19 | 128 ± 17 § | 126 ± 17 § |
| Systolic time duration (%) | 138.4 ± 5.1 | 148.6 ± 9.0 † | 142.8 ± 7.7 | 151.2 ± 10.4 † |
| Untwisting rate | ||||
| Peak (°/sec) | −78.7 ± 33.8 | −88.7 ± 34.3 | −140.5 ± 51.7 § | 182.9 ± 53.5 ∗ , § |
| Time to peak (%) | 113.3 ± 44.7 | 117.4 ± 6.7 | 132.0 ± 38.9 | 121.9 ± 7.2 |
| °UT IVRT (°) | 1.4 ± 1.1 | 4.0 ± 1.5 † | 1.0 ± 1.1 | 3.9 ± 1.9 † |
| Mean untwisting rate during IVRT (°/sec) | 34.3 ± 29.5 | 64.1 ± 28.9 ∗ | 38.3 ± 37.7 | 110.7 ± 47.6 † , § |
| Percentage of untwisting during IVRT | 25.7 ± 28.6 | 45.3 ± 22.5 † | 8.8 ± 9.2 § | 29.8 ± 20.4 † , § |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
