Reference Values for Myocardial Two-Dimensional Strain Echocardiography in a Healthy Pediatric and Young Adult Cohort

VENTRICULAR AND ATRIAL MECHANICS

Reference Values for Myocardial Two-Dimensional Strain Echocardiography in a Healthy Pediatric and Young Adult Cohort Karen A. Marcus, MD, Annelies M. C. Mavinkurve-Groothuis, MD, Marlieke Barends, BSc, Arie van Dijk, MD, PhD, Ton Feuth, MSc, Chris de Korte, PhD, and Livia Kapusta, MD PhD, Nijmegen, The Netherlands; Petach Tikvah, Israel

Background: The accurate evaluation of intrinsic myocardial contractility in children with or without congenital heart disease (CHD) has turned out to be a challenge. Two-dimensional strain echocardiographic (2DSTE) imaging or two-dimensional speckle-tracking echocardiographic imaging appears to hold significant promise as a tool to improve the assessment of ventricular myocardial function. The aim of this study was to estimate left ventricular myocardial systolic function using 2DSTE imaging in a large cohort consisting of healthy children and young adults to establish reference strain values. Methods: Transthoracic echocardiograms were acquired in 195 healthy subjects (139 children, 56 young adults) and were retrospectively analyzed. Longitudinal, circumferential, and radial peak systolic strain values were determined by means of speckle tracking. Nonlinear regression analysis was performed to assess the effect of aging on these 2DSTE parameters. Results: There was a strong, statistically significant second-order polynomial relation (P < .001) between global peak systolic strain parameters and age. Global peak systolic strain values were lowest in the youngest and oldest age groups. Conclusion: This is the first report to establish age-dependent reference values per cardiac segment for myocardial strain in all three directions assessed using 2DSTE imaging in a large pediatric and young adult cohort. There is a need to use age-specific reference values for the adequate interpretation of 2DSTE measurements. (J Am Soc Echocardiogr 2011;24:625-36.) Keywords: Child, Healthy, Reference value, Speckle tracking, Two-dimensional strain echocardiography Despite impressive developments in echocardiographic technology during the past 50 years, there are still important challenges that new techniques try to meet. One important yet partially unresolved challenge that remains is the quantification of ventricular myocardial function in children with congenital heart disease (CHD). Traditional methods, such as the one-dimensional M-mode technique (e.g., fractional shortening) and two-dimensional imaging (e.g., ejection fraction), are used to evaluate left ventricular function. However, these methods are not always applicable to pediatric hearts with complex From the Children’s Heart Centre (K.A.M., M.B., L.K.), the Department of Pediatric Hematology and Oncology (A.M.C.M.-G.), the Department of Cardiology (A.D.), the Department of Epidemiology, Biostatistics, and Health Technology Assessment (T.F.), and the Clinical Physics Laboratory (C.K.), Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands; and the Heart Institute, Schneider Children’s Medical Centre of Israel, Petach Tikvah, Israel (L.K.). Reprint requests: Karen A. Marcus, MD, Children’s Heart Centre, Radboud University Nijmegen Medical Centre, M325.04.004 KG, PO Box 9101, 6500 HB Nijmegen, The Netherlands (E-mail: [email protected]). 0894-7317/$36.00 Copyright 2011 by the American Society of Echocardiography. doi:10.1016/j.echo.2011.01.021

congenital malformations. The cardinal reason for this is that they are based on geometric assumptions and overlook the complex characteristics of CHD.1 Moreover, those current indices provide no information about regional alterations in ventricular myocardial contraction.2 These disadvantages complicate the assessment of systolic and diastolic function in children with CHD and underscore the need for better quantitative techniques to assess ventricular function in this subset of patients. New modalities, such as strain imaging, have been developed to address the previously mentioned shortcomings. Myocardial strain is a dimensionless assessment of regional ventricular deformation (i.e., a percentage of deformation), whereas myocardial strain rate (SR), which is a time derivate of strain, indicates the rate of deformation of a defined myocardial segment. Multiple studies have proven the accuracy and reliability of strain (and SR) techniques in the assessment of (regional) ventricular myocardial function.3,4 Moreover, these techniques appear to be able to detect subclinical myocardial dysfunction at an earlier stage compared with conventional imaging modalities.5-7 Several diagnostic imaging modalities to assess myocardial strain and SR have been developed in recent years. In addition to tagged magnetic resonance imaging (MRI) and Doppler tissue imaging (DTI), a third method to determine myocardial strain was introduced some years ago. This relatively novel technique is 625

626 Marcus et al

Journal of the American Society of Echocardiography June 2011

Abbreviations

CHD = Congenital heart disease

CI = Confidence interval DTI = Doppler tissue imaging LVETc = Left ventricular ejection time corrected for heart rate

LVM = Left ventricular mass MRI = Magnetic resonance imaging SR = Strain rate 2DSTE = Two-dimensional strain echocardiographic VCFc = Rate-corrected velocity of circumferential fiber shortening

based on two-dimensional echocardiographic images, hence the name: two-dimensional strain echocardiographic (2DSTE) imaging. It determines myocardial deformation by means of frameby-frame tracking and motion analysis of speckles within B-mode images using optical flow algorithms. Validation studies with tagged MRI and sonomicrometry in the adult population have provided evidence that 2DSTE imaging is a reliable method to determine ventricular myocardial function.3,4 Despite its limitations, such as sensitivity to noise, it has the advantages over DTI-based strain imaging that the strain estimates are not angle dependent and that strain is obtained

in two dimensions.4,8-11 Intraobserver and interobserver reliability scores have been established previously and are high for almost all parameters.12 Because this novel technique could be easily accessible and applicable, and moreover is not based on geometric assumptions, it may serve as an important tool in the future evaluation of regional and global ventricular function in pediatric subjects with or without CHD. The establishment of normal values in a healthy pediatric cohort is a mandatory prerequisite for its use in evaluating (pathologic) changes in ventricular function. Data in this area are scarce and include only peak systolic strain in one or two directions.13,14 For these reasons, we aimed to evaluate left ventricular myocardial strain in all three directions (longitudinal, circumferential, and radial) assessed by means of 2DSTE imaging in a large, healthy pediatric and young adult cohort to establish 2DSTE reference values and to determine the influence of age and growth on these values. MATERIALS AND METHODS Study Population Subjects who were routinely referred for echocardiographic evaluation of an asymptomatic, innocent heart murmur between May 1,

Table 1 Demographic and anthropometric characteristics of study subjects (n = 195) categorized by age group Age group

Variable

1 0y

2 1–4 y

3 5–9 y

4 10–14 y

5 15–19 y

6 20–24 y

7 25–29 y

8 30–40 y

Number Male Age (y) Height (m) Weight (kg) BSA (m2) BMI (kg/m2) HR (beats/min) SBP (mm Hg) DBP (mm Hg)

24 13 (54%) 0.3 6 0.3 0.62 6 0.11 6.3 6 2.6 0.32 6 0.10 15.9 6 2.2 118 6 12 82 6 8 56 6 6

34 19 (56%) 2.9 6 1.0 0.95 6 0.10 14.6 6 3.4 0.62 6 0.10 15.9 6 1.4 101 6 14 98 6 10 62 6 10

36 25 (69%) 7.2 6 1.2 1.26 6 0.09 24.7 6 4.4 0.93 6 0.12 15.4 6 1.3 84 6 13 104 6 8 70 6 8

29 16 (55%) 12.8 6 1.6 1.59 6 0.13 46.3 6 12.0 1.43 6 0.23 17.9 6 2.1 77 6 14 110 6 10 72 6 8

21 9 (43%) 17.0 6 1.3 1.76 6 0.09 66.4 6 12.0 1.81 6 0.20 21.2 6 2.3 65 6 9 116 6 11 75 6 9

25 16 (64%) 21.7 6 1.2 1.78 6 0.10 70.3 6 11.9 1.86 6 0.20 22.1 6 2.0 63 6 11 118 6 12 75 6 8

13 8 (62%) 27.3 6 1.3 1.82 6 0.08 76.5 6 11.9 1.96 6 0.19 23.1 6 2.2 60 6 14 121 6 11 77 6 8

13 6 (46%) 35.6 6 2.6 1.77 6 0.10 77.8 6 11.0 1.95 6 0.18 24.9 6 2.9 65 6 8 123 6 12 78 6 9

Data are expressed as mean 6 SD or as number (percentage). BSA, Body surface area; BMI, body mass index; DBP, diastolic blood pressure; HR, heart rate; SBP, systolic blood pressure.

Figure 1 Short-axis imaging at the level of the papillary muscle with segmentation used for 2DSTE imaging. Ant, Anterior wall; AntSept, anterior septal wall; Inf, inferior wall; Lat, lateral wall; LV, left ventricle; Post, posterior wall; RV, right ventricle; Sep, septal wall.

Figure 2 Radial and circumferential strain and strain rate curves of the septal wall at the level of the papillary muscle. ECG, Electrocardiogram; ES, end-systole; SAX-PM, short-axis view at the level of the papillary muscle.

Marcus et al 627

Journal of the American Society of Echocardiography Volume 24 Number 6

Table 2 Conventional echocardiographic parameters of study subjects (n = 195) categorized by age group Age group

Variable

1 0y

2 1–4 y

3 5–9 y

4 10–14 y

5 15–19 y

6 20–24 y

7 25–29 y

8 30–40 y

LVET (sec) LVETc (sec) VCFc (circ/sec) Tei index FS Biplane EF ESWS (g/cm2) LVM/BSA (g/m2)

0.19 6 0.01 0.28 6 0.02 1.32 6 0.22 0.37 6 0.01 0.37 6 0.05 0.73 6 0.07 30.7 6 11.0 36.2 6 12.1

0.23 6 0.01 0.30 6 0.02 1.28 6 0.20 0.37 6 0.01 0.38 6 0.05 0.71 6 0.06 42.8 6 13.9 48.5 6 11.6

0.26 6 0.01 0.30 6 0.02 1.25 6 0.19 0.38 6 0.01 0.37 6 0.04 0.67 6 0.05 43.6 6 10.7 57.2 6 12.3

0.28 6 0.01 0.32 6 0.03 1.19 6 0.19 0.38 6 0.03 0.37 6 0.04 0.70 6 0.06 45.3 6 11.6 59.9 6 14.2

0.29 6 0.01 0.30 6 0.02 1.31 6 0.20 0.40 6 0.03 0.39 6 0.06 0.67 6 0.07 44.4 6 13.4 71.8 6 16.0

0.29 6 0.01 0.29 6 0.03 1.21 6 0.43 0.39 6 0.03 0.39 6 0.07 0.74 6 0.04 40.0 6 10.4 78.6 6 18.0

0.29 6 0.01 0.29 6 0.03 1.20 6 0.43 0.38 6 0.01 0.37 6 0.05 0.77 6 0.04 45.7 6 16.2 77.3 6 19.1

0.30 6 0.01 0.31 6 0.03 1.20 6 0.41 0.38 6 0.02 0.40 6 0.04 0.75 6 0.05 41.8 6 10.1 75.4 6 17.2

Data are expressed as mean 6 SD. BSA, Body surface area; EF, ejection fraction; ESWS, end-systolic wall stress; FS, fractional shortening; LVET, left ventricular ejection time; LVETc, left ventricular ejection time corrected for heart rate; VCFc, heart-rate corrected velocity of circumferential fiber shortening.

2005, and November 1, 2009, were retrospectively analyzed for their eligibility for inclusion in the study. All consecutive healthy subjects age from birth to 40 years were identified from our echocardiographic database from the outpatient clinic at both the Children’s Heart Centre and the Adult Heart Centre (Nijmegen, The Netherlands). Patients with structural (congenital) heart disease or abnormal cardiac rhythms were excluded. Other exclusion criteria consisted of hypertension and/or chronic or recent acute illness. Demographic characteristics, including age and gender, were collected at the time the echocardiographic studies were performed. Informed consent was obtained from each participant. This study was approved by the local medical ethics committee. Two-Dimensional Strain Echocardiographic Data Acquisition A complete physical examination was performed, including weight, height, and blood pressure measurements (see Table 1). Subsequently, all subjects underwent a detailed transthoracic echocardiographic examination in the left lateral position according to the recommendations of the American Society of Echocardiography.15 Every examination was performed at rest, without using sedation. Images were obtained with a 3.0-MHz (S3) or a 5.0-MHz (S5) phased-array transducer using a commercially available system, the Vivid 7 echocardiographic scanner (GE Vingmed Ultrasound AS, Horten, Norway). The choice of an S3 or S5 transducer depended on the age and posture of the child. Twodimensional multiframe B-mode (grayscale) images were obtained in the apical four-chamber and parasternal midcavity short-axis view (at the level of the papillary muscle) and parasternal basal short-axis view (at the level of the mitral valve). A sector scan angle of 30
to 60
was chosen, and frame rates of 70 to 90 Hz were used, because these rates are considered to be optimal for twodimensional speckle tracking.4,11,16 Data were stored at the same frame rate as the acquisition frame rate. Preferably images from three cardiac cycles triggered by the R wave of the QRS complex were digitally saved in cine loop format. Offline speckle-tracking analysis was performed using software for echocardiographic quantification (EchoPAC version 6.1.0; GE Vingmed Ultrasound AS). The timing of aortic valve closure and mitral valve opening with respect to peak systolic strain were manually obtained, using single gated pulsed-wave Doppler or continuous-wave Doppler images of the left ventricular outflow tract. For these measurements, special care was taken to keep the heart rate in the same range as during the

two-dimensional grayscale imaging used for two-dimensional strain calculations. Endomyocardial borders of the left ventricle were manually traced within the end-systolic frame. The second, epicardial tracing was automatically generated by the computer algorithm and, when necessary, manually adjusted to cover the whole myocardial wall. The tracking algorithm then followed the myocardial speckles during the cardiac cycle. Tracking was accepted only if both visual inspection as well as the EchoPAC software indicated adequate tracking. This means that tracking of any given segment was accepted only when it was indicated with a green box. The software automatically divided the cross-sectional image into six segments, which were named and identified according to the statement of the Cardiac Imaging Committee of the Council on Clinical Cardiology of the American Heart Association.17 The left ventricular segments to be analyzed were the apical, middle, and basal segments of the septal and the lateral wall of the four-chamber view, as well as the anteroseptal, anterior, lateral, posterior, inferior, and septal segments of the basal and midcavity short-axis views. Strain curves of the three consecutive cardiac cycles and values of the manual timing were imported into a custom-made software package for further analysis. The Q-Q interval was estimated from the electrocardiographic signal to obtain cardiac cycle length. It is known that the systolic phase of the cardiac cycle does not change much with small changes in heart rate, in contrast to the diastolic phase.18 Therefore, the diastolic phase of the three cardiac cycles was automatically extended and adjusted by the software package to the longest of the three cardiac cycles. This intervention prevents a shift of the peak systolic strain while averaging the three consecutive cardiac cycles. Cardiac cycles with lengths >10% different from the mean length of the three cardiac cycles were excluded from averaging and thus from further analysis. Myocardial longitudinal, radial, and circumferential strain values were obtained. To determine global strain, the strain values of the six segments were averaged for the four-chamber as well as the short-axis views. Strain values are dimensionless and are expressed as percentages. Negative strain values reflect shortening, while positive strain values reflect lengthening or thickening. All offline measurements with EchoPAC were performed by a single observer (K.A.M.). Interobserver and intraobserver variability was determined by measurement of left ventricular myocardial strain in 40 randomly selected subjects (five in each age group). To assess intraobserver variability, the same observer (K.A.M.) measured the left ventricular segments again with an interval >2 months to avoid recall bias. To assess interobserver variability, strain measurements were performed by

628 Marcus et al

Data are expressed as mean 6 SD percentage. S-L, Longitudinal peak systolic strain. *P < .05 vs with age groups denoted by superscript numerals, determined by one-way analysis of variance with Bonferroni’s correction for multiple comparisons.

20.5 6 1.2 19.1 6 1.1 17.7 6 1.2 18.9 6 1.0*2-7 16.9 20.9 22.4 6 1.8 20.3 6 1.0 19.7 6 1.2 20.6 6 1.2*1,5,8 18.2 23.0 22.7 6 1.8 20.3 6 1.6 19.4 6 1.3 20.9 6 1.3*1,5,8 18.3 23.5 22.5 6 1.7 20.2 6 1.6 19.4 6 1.5 20.7 6 1.3*1,4,5,8 18.1 23.3 19.6 6 1.5 16.2 6 1.1 15.8 6 1.3 18.3 6 1.9*2-7 14.5 22.1

22.6 6 1.5 20.1 6 1.8 19.3 6 1.8 21.0 6 1.3*1,5,8 18.4 23.6

23.9 6 1.7 21.7 6 1.6 20.0 6 1.2 21.8 6 1.3*1,2,8 19.2 24.4

25.0 6 1.4 22.4 6 1.7 20.5 6 1.7 22.5 6 1.3*1-3,6-8 19.9 25.1

19.2 6 1.3 20.3 6 1.4 22.0 6 2.0 19.6 6 1.5 20.3 6 1.8 22.8 6 1.9 20.4 6 1.6 22.5 6 1.4 25.1 6 1.2 20.0 6 1.7 21.5 6 1.6 23.8 6 1.3 20.1 6 1.8 21.1 6 1.4 22.6 6 1.5 19.6 6 1.5 20.9 6 1.3 22.7 6 1.7

6 20–24 y 5 15–19 y 2 1–4 y

3 5–9 y

4 10–14 y

Age group 1 0y

17.0 6 2.6 18.2 6 2.7 20.4 6 1.9

A total of 226 subjects were evaluated for inclusion in the study. Of those subjects, 31 (13.7%) were subsequently excluded in light of incomplete echocardiographic data or suboptimal imaging quality. In total, 195 healthy subjects (139 children and 56 young adults) were enrolled in the study. Subject characteristics and anthropometric parameters are described in Table 1. Conventional echocardiographic parameters of the study subjects are presented in Table 2. All standard echocardiographic findings were within previously described normal values. Myocardial strain parameters are presented in Tables 3 to 5 and Figures 3 to 5. Tracking was feasible in 91% of all segments in the four-chamber view, in 96% of all segments in the short-axis

Variable

RESULTS

Table 3 Longitudinal peak systolic strain of study subjects (n = 195) categorized by age group

Statistical Analysis Patients were divided into eight different age groups (see Table 1) for further analysis. All demographic, conventional echocardiographic, and two-dimensional strain values are expressed as mean 6 SD. The relations between age and global strain parameters were determined using scatterplots, one-way analysis of variance, and second-order polynomial regression analysis. Multiple linear regression analysis was performed to determine the additional effect of anthropometrics (including heart rate, weight, and height) and conventional echocardiographic parameters on the myocardial global strain variables. Intraobserver and interobserver agreement was calculated using the Bland-Altman approach, including the calculation of mean bias (average difference between measurements), the statistical significance of the mean bias on paired t tests (the null hypothesis was zero bias), and the lower and upper limits of agreement (95% limits of agreement of mean bias). In addition, the coefficient of variation was determined (i.e., the standard deviation of the difference of paired samples divided by the average of the paired samples). P values < .05 was considered to indicate significance. Statistical analyses were performed using SPSS for Windows version 16.0 (SPSS, Inc., Chicago, IL).

7 25–29 y

Conventional Echocardiographic Parameters Quantification of cardiac chamber size, ventricular mass, and systolic and diastolic left ventricular function were measured in accordance with the recommendations for chamber quantification of the American Society of Echocardiography.15 Left ventricular systolic function was indicated using fractional shortening, ejection fraction, left ventricular myocardial performance index (Tei index), end-systolic wall stress, and rate-corrected velocity of circumferential fiber shortening (VCFc). Ejection fraction was calculated using the modified Simpson’s rule. The pulsed-wave Doppler–derived myocardial performance index was calculated by adding the isovolumetric contraction time and the isovolumetric relaxation time and dividing the sum by the ejection time.19 Left ventricular end-systolic wall stress was calculated using the modified formula of Rowland and Gutgesell.20 VCFc was calculated using the formula obtained from Colan et al.21 Left ventricular mass (LVM) was calculated using the formula for estimation of LVM according to Devereux and Reichek22 and was subsequently indexed to body surface area.

Septum Basal Mid Apical Lateral wall Apical Mid Basal Global S-L 5th percentile global S-L 95th percentile global S-L

8 30–40 y

a second observer (M.B.), who was blinded to the results of the first observer (K.A.M.). Figures 1 and 2 show an illustration of twodimensional strain imaging.

17.5 6 1.5 18.8 6 1.4 20.7 6 1.4

Journal of the American Society of Echocardiography June 2011

Marcus et al 629

Journal of the American Society of Echocardiography Volume 24 Number 6

Table 4 Radial peak systolic strain of study subjects (n = 195) categorized by age group Age group

Variable

PM Anteroseptal Anterior Lateral Posterior Inferior Septal Global RP 5th percentile global S-RP 95th percentile global S-RP MV Anteroseptal Anterior Lateral Posterior Inferior Septal Global S-RM 5th percentile global S-RM 95th percentile global S-RM

1 0y

2 1–4 y

3 5–9 y

4 10–14 y

43.3 6 10.7 50.0 6 14.8 58.5 6 12.0 63.2 6 11.6 56.6 6 12.6 40.1 6 7.8 52.0 6 9.9*4,5 32.2 71.8

44.9 6 5.8 51.9 6 7.8 58.8 6 8.6 61.1 6 8.9 58.5 6 9.0 45.9 6 6.5 53.5 6 6.7 40.1 66.9

44.5 6 7.2 52.8 6 8.5 59.4 6 7.8 64.8 6 7.1 61.2 6 6.1 46.6 6 5.4 54.9 6 5.5 43.9 65.9

46.5 6 6.1 57.1 6 5.7 62.8 6 5.8 66.3 6 5.5 65.2 6 6.5 50.1 6 5.9 58.0 6 5.4*1 47.2 68.8

39.6 6 4.2 46.7 6 5.6 55.5 6 6.1 60.5 6 6.0 56.3 6 7.0 41.1 6 5.0 49.9 6 4.3*4-6 41.3 58.5

41.6 6 5.1 48.3 6 7.0 54.1 6 8.2 58.5 6 7.7 54.7 6 6.9 42.9 6 5.3 50.0 6 5.7*4-6 38.6 61.4

43.4 6 5.1 49.2 6 5.1 56.6 6 5.2 61.2 6 7.4 57.5 6 5.4 45.8 6 4.9 52.3 6 4.5*1,2 43.3 61.3

46.8 6 5.3 52.7 6 5.8 58.6 6 7.1 62.2 6 6.5 59.0 6 6.4 50.3 6 5.6 54.9 6 5.4 44.1 65.7

5 15–19 y

6 20–24 y

7 25–29 y

8 30–40 y

48.7 6 3.9 55.9 6 4.4 61.2 6 4.6 66.8 6 4.1 64.3 6 4.3 51.7 6 5.3 58.1 6 4.0*1 50.1 66.1

47.7 6 6.9 55.5 6 5.8 62.0 6 5.9 66.4 6 5.5 60.9 6 4.8 51.1 6 5.8 57.3 6 5.0 47.3 67.3

47.9 6 6.5 51.5 6 6.4 57.0 6 6.3 63.1 6 5.8 60.2 6 5.3 48.0 6 5.7 54.6 6 5.3 44.0 65.2

44.3 6 5.7 52.9 6 5.8 56.4 6 5.9 62.2 6 5.6 60.5 6 6.4 49.1 6 4.2 54.2 6 4.7 44.8 63.6

49.9 6 3.4 54.7 6 3.2 57.6 6 4.0 63.0 6 5.8 60.3 6 5.2 51.1 6 3.7 56.1 6 3.8*1,2 48.5 63.7

48.2 6 5.1 52.4 6 5.5 57.3 6 6.9 62.5 6 6.0 59.3 6 6.5 49.6 6 4.8 54.9 6 5.4*1,2 44.1 65.7

46.0 6 5.3 51.8 6 3.8 55.0 6 5.9 59.5 6 4.3 57.8 6 4.8 46.7 6 5.2 52.8 6 4.1 44.6 61.0

43.2 6 5.1 51.4 6 5.2 56.1 6 5.4 60.1 6 4.4 57.2 6 3.8 45.4 6 5.8 52.2 6 4.3 43.6 60.8

Data are expressed as mean 6 SD percentage. MV, Mitral valve; PM, papillary muscle; S-RM, radial peak systolic strain at the level of the mitral valve; S-RP, radial peak systolic strain at the level of the papillary muscle. *P < .05 vs with age groups denoted by superscript numerals, determined by one-way analysis of variance with Bonferroni’s correction for multiple comparisons.

view at the level of the papillary muscle, and in 91% of all segments in the short-axis view at the level of the mitral valve. The most striking observation was a quadratic relationship (P < .001) between global peak systolic parameters and age. This means that global peak systolic strain values were lowest in the youngest and oldest age groups in contrast to teens, who showed the highest global peak systolic values. There were statistically significant differences (P < .05) between most of the age groups for all global peak systolic parameters as assessed by means of one-way analysis of variance with Bonferroni’s correction for multiple comparisons (Tables 3–5). A second-order polynomial regression analysis revealed that age (together with its square) accounted for 37% of variation in global longitudinal peak systolic strain (R2 = 0.368 P < .05). Age and its square accounted for 22% and 29% of the variation in global circumferential peak systolic strain at the levels of the papillary muscle (R2 = 0.222, P < .05) and mitral valve (R2 = 0.293, P < .05), respectively. In the radial direction, age and its square accounted for 12% and 14% of the variation in global radial peak systolic strain at the levels of the papillary muscle (R2 = 0.117, P < .05) and mitral valve (R2 = 0.142, P < .05), respectively. The results for intraobserver and interobserver variability results are shown in Table 6. There were no important differences in variability scores between the various age groups or various segments. Multiple linear regression analyses among the anthropometric and standard echocardiographic parameters were performed to determine the additional effect of these parameters on the myocardial global strain variables (Table 7). Of these parameters, only VCFc and left ventricular ejection time corrected for heart rate (LVETc) appeared to be significantly associated with global strain values after

adjustment for age. VCFc was associated with global circumferential strain (P < .05). A 1-unit (i.e., circ/sec) increase in VCFc corresponded to a 1.5% decrease in global circumferential peak systolic strain at the level of the papillary muscle (95% confidence interval [CI], 0.3%– 2.8%) and a 1.7% decrease in global circumferential peak systolic strain at the level of the mitral valve (95% CI, 0.5%–2.9%). LVETc was associated with both global circumferential and global longitudinal peak systolic strain (P < .05). A 1-msec increase in LVETc corresponded to a 0.015% decrease in global longitudinal peak systolic strain (95% CI, 0.006%–0.023%). As with global peak systolic circumferential strain, a 1-msec increase in LVETc corresponded to a 0.05% decrease in global peak systolic circumferential strain at the level of the papillary muscle (95% CI. 0.034%–0.067%) and a 0.024% decrease at the level of the mitral valve (95% CI, 0.009%–0.039%).No gender-based differences were present. Figures 6A and 6B illustrate peak systolic strain in the longitudinal direction versus some of the key morphometric variables.

DISCUSSION Reference values for 2DSTE measurements in the pediatric and young adult population are scarce. Two previously published reports have focused on this subject. Lorch et al.14 investigated 2DSTE imaging in a large healthy pediatric population. Their study provided reference values for two-dimensional longitudinal strain outcomes, but circumferential and radial strain measurements were not included. Bussadori et al.13 described reference values for circumferential and

630 Marcus et al

Table 5 Circumferential peak systolic strain of study subjects (n = 195) categorized by age group Age group

Variable

2 1–4 y

3 5–9 y

4 10–14 y

5 15–19 y

6 20–24 y

7 25–29 y

8 30–40 y

25.2 6 5.7 18.2 6 1.6 13.8 6 2.3 10.7 6 4.0 17.0 6 6.5 26.5 6 6.0 18.6 6 3.3*2-8 12.0 25.2

26.8 6 3.6 21.3 6 4.1 18.1 6 2.0 16.0 6 1.9 20.4 6 2.0 25.3 6 3.6 21.3 6 2.0*1,3-5 17.3 25.3

26.3 6 2.3 24.4 6 2.2 20.5 6 1.5 19.4 6 1.1 22.7 6 2.0 27.2 6 2.6 23.4 6 1.7*1,2,7,8 20.0 26.8

26.1 6 3.0 24.1 6 2.0 21.0 6 1.7 20.0 6 1.4 22.8 6 1.7 26.9 6 2.8 23.5 6 1.8*1,2,7,8 19.9 27.1

27.5 6 2.8 22.9 6 2.0 21.1 6 2.2 19.2 6 2.5 22.4 6 2.1 28.6 6 2.5 23.6 6 2.0*1,2,7,8 19.6 27.6

25.2 6 2.1 22.5 6 1.7 19.7 6 1.5 18.0 6 1.6 20.9 6 1.9 24.8 6 2.8 21.8 6 1.5*1 18.8 24.8

24.7 6 2.5 20.9 6 2.6 18.6 6 2.1 18.0 6 2.1 19.6 6 1.5 24.6 6 3.2 21.1 6 1.9*1,3-5 17.3 24.9

24.1 6 2.0 20.4 6 2.3 17.8 6 2.0 17.3 6 2.4 20.7 6 2.8 23.5 6 2.7 20.6 6 2.2*1,3-5 16.2 25.0

25.3 6 3.8 21.1 6 2.1 17.4 6 2.9 15.3 6 2.9 20.1 6 2.1 26.4 6 3.4 20.9 6 2.0*1 16.9 24.9

24.3 6 3.3 21.8 6 2.0 19.2 6 1.5 18.2 6 1.2 20.7 6 1.8 24.7 6 2.3 21.5 6 1.7*1,2 18.1 24.9

24.0 6 2.5 22.4 6 1.9 19.3 6 2.4 17.7 6 2.0 21.9 6 2.7 26.0 6 2.5 21.9 6 2.1*1,2 17.7 26.1

24.8 6 2.6 20.8 6 2.0 18.2 6 1.4 18.2 6 2.1 20.3 6 1.8 24.0 6 2.2 21.1 6 1.3*1 18.5 23.7

25.1 6 4.1 20.6 6 1.5 18.8 6 1.2 16.9 6 2.0 20.1 6 1.6 24.3 6 3.4 21.0 6 1.6*1 17.8 24.2

23.2 6 2.3 19.7 6 1.6 18.3 6 1.1 16.2 6 1.2 20.2 6 2.2 23.4 6 3.9 20.2 6 1.4*1 17.4 23.0

25.1 6 3.8 27.1 6 3.9 16.5 6 4.2 20.3 6 4.4 15.4 6 3.6 11.5 6 3.6 11.4 6 4.0 12.4 6 2.7 16.0 6 7.6 17.8 6 3.4 24.5 6 5.2 24.9 6 4.7 17.5 6 2.5*2-8 19.7 6 2.0*1,4,5 12.5 15.7 22.5 23.7

Data are expressed as mean 6 SD percentage. MV, Mitral valve; PM, papillary muscle; S-CM, circumferential peak systolic strain at the level of the mitral valve; S-CP, circumferential peak systolic strain at the level of the papillary muscle. *P < .05 vs with age groups denoted by superscript numerals, determined by one-way analysis of variance with Bonferroni’s correction for multiple comparisons.

Journal of the American Society of Echocardiography June 2011

PM Anteroseptal Anterior Lateral Posterior Inferior Septal Global S-CP 5th percentile global S-CP 95th percentile global S-CP MV Anteroseptal Anterior Lateral Posterior Inferior Septal Global S-CM 5th percentile global S-CM 95th percentile global S-CM

1 0y

Marcus et al 631

Journal of the American Society of Echocardiography Volume 24 Number 6

Figure 3 Global peak systolic longitudinal strain (GlobalSL) (%) versus age among study subjects (scatterplot). Lines indicate regression line (mean) in the middle and 95% individual prediction interval.

Figure 5 Midventricular global peak systolic radial strain at the level of the papillary muscle (GlobalSRP) (%) versus age among study subjects (scatterplot). Lines indicate regression line (mean) in the middle and 95% individual prediction interval.

Table 6 Intraobserver and interobserver variability of global peak systolic strain Variable

Intraobserver Global S-L Global S-CP Global S-CM Global S-RP Global S-RM Interobserver Global S-L Global S-CP Global S-CM Global S-RP Global S-RM

Figure 4 Midventricular global peak systolic circumferential strain at the level of the papillary muscle (GlobalScP) (%) versus age among study subjects (scatterplot). Lines indicate regression line (mean) in the middle and 95% individual prediction interval. longitudinal strain and SR measurements in a small pediatric (n = 15) and adult (n = 30) population. Data on circumferential and radial two-dimensional strain measurements in a large, healthy pediatric and young adult population, to establish their reference values and investigate the influence of age on them, were still lacking. In the present study, we collected 2DSTE data in different left ventricular wall segments and in all three directions, longitudinal, radial, and circumferential, in a large, healthy cohort consisting of children and young adults. Our findings are in disagreement with the previous study performed by Lorch et al. showing a relative invariability of longitudinal peak systolic strain with age. Regression analysis performed on our data showed a strong, statistically significant relationship between age and global peak systolic strain values in all directions. Subjects at both ends of the age spectrum displayed the lowest peak systolic strain values, in contrast to teenagers, whose strain values were signif-

Bias

LLA

ULA

CV

0.03 0.39 0.02 0.32 0.23

2.89 3.51 3.19 .01 9.19

2.95 2.74 3.14 9.36 8.73

7.15 7.22 7.89 8.91 8.52

0.04 0.46 0.09 0.39 0.28

3.35 3.61 4.52 14.78 13.21

3.27 4.54 4.3 13.22 13.78

8.12 9.44 11.04 14.01 12.83

CV, Coefficient of variation; LLA, 95% lower limit of agreement; S-CM, circumferential peak systolic strain at the level of the mitral valve; S-CP, circumferential peak systolic strain at the level of the papillary muscle; S-L, Longitudinal peak systolic strain; S-RM, radial peak systolic strain at the level of the mitral valve; S-RP, radial peak systolic strain at the level of the papillary muscle; ULA, 95% upper limit of agreement.

icantly higher. In several previous studies, the growth and function of the human heart were analyzed by means of echocardiography and ergometry, which gave evidence for an increase in cardiac contractility during puberty.23,24 Increased cardiac contractility is probably one of the most important factors that enables a substantial increase in physical working capacity during puberty.25 The systolic thickness of the ventricular walls increases disproportionally when related to the rise in afterload during pubertal development. With this observation, it seems that cardiac systolic function prepares a reserve, probably for future increase in myocardial energy expenditure due to an increase in functional working capacity. The same relationship between age on one hand and LVM on the other has been reported in healthy, normotensive humans. A study performed by Cain

632 Marcus et al

Journal of the American Society of Echocardiography June 2011

Table 7 Coefficient of determination (R2) Outcome variable Predictor

Global S-L

Global S-CP

Global S-CM

Global S-RP

Global S-RM

Age//age2 Age//age2//LVETc Age//age2//VCFc Age//age2//Tei index Age//age2//FS Age//age2//biplane EF Age//age2//ESWS Age//age2//LVIDs Age//age2//LVIDd Age//age2//LVEDV Age//age2//HR Age//age2//weight Age//age2//height Age//age2//BMI Age//age2//BSA Final model{ Predictors used in final modelk

0.37§ 0.41* 0.37 0.37 0.38 0.37 0.38 0.37 0.37 0.38 0.37 0.38 0.38 0.37 0.37 0.41* Age//age2 (P < .0001)

0.22§ 0.25* 0.24* 0.23 0.22 0.22 0.23 0.22 0.22 0.23 0.22 0.22 0.22 0.22 0.22 0.28‡ Age//age2 (P < .0001)

0.29§ 0.34† 0.31* 0.29 0.29 0.29 0.29 0.29 0.30 0.29 0.30 0.29 0.29 0.29 0.30 0.37‡ Age//age2 (P < .0001)

0.12§ 0.12 0.12 0.12 0.12 0.13 0.12 0.12 0.12 0.12 0.12 0.12 0.12 0.12 0.12 0.12§ Age//age2 (P < .0001)

0.14§ 0.14 0.14 0.14 0.14 0.14 0.14 0.14 0.14 0.15 0.14 0.15 0.14 0.14 0.15 0.14§ Age//age2 (P < .0001)

LVETc (P = .001)

LVETc (P < .001) VCFc (P = .025)

LVETc (P < .001) VCFc (P = .031)

Each R2 value is based on the linear model that uses all subjects with known values for all predictors used in that particular model. BMI, Body mass index; BSA, body surface area; EF, ejection fraction; ESWS, end-systolic wall stress; FS, fractional shortening; HR, heart rate; LVEDV, left ventricular end-diastolic volume; LVETc, left ventricular ejection time corrected for heart rate; LVIDd, left ventricular internal diameter during diastole; LVIDs, left ventricular internal diameter during systole; S-CM, circumferential peak systolic strain at the level of the mitral valve; SCP, circumferential peak systolic strain at the level of the papillary muscle; S-L, Longitudinal peak systolic strain; S-RM, radial peak systolic strain at the level of the mitral valve; S-RP, radial peak systolic strain at the level of the papillary muscle; VCFc, heart-rate corrected velocity of circumferential fiber shortening. *P < .01, †P < .001, and ‡P < .0001 (based on partial F test comparing model with indicated predictors to model using only age and its square as predictors); §P < .0001 (based on global F test for linear model using age and its square as predictors). k P values based on partial F test comparing the final model with the model based on all variables in final model minus the variable of concern. {The final model uses as predictors age and its square and the variables that contribute significantly over age and its square, as shown in the upper part of the table, and remained significant in multivariate analysis.

et al.26 identified age as the major determinant of LVM in healthy subjects. LVM strongly increased when children reached pubertal age and subsequently declined during adult life to a level that is approximately 80% of early adulthood despite increasing blood pressure and exercise activity with increasing age in their study population. The underlying biologic trigger that causes the sudden surge in cardiac contractility and LVM has not been identified. A possible explanation could be a physiologic response as an adaptation process to an increased oxygen demand due to an increasing body mass and functional working tasks. Other factors, such as genetic and hormonal factors, should also be considered. There is evidence that hormones, such as growth hormone and sex hormones, whose concentrations surge during puberty, influence cardiac growth and systolic function.27-35 A remarkable result of our study was the observation of low global peak systolic strain values in infants. Previous reports have not reached consensus on the cardiac contractile state in this age group compared with older children. Some have concluded that newborn infants have a higher basal contractile state that cannot be accounted for completely by a lower afterload.36-38 Pena et al.,39 for instance, found higher peak systolic radial and longitudinal strain assessed by means of DTI in healthy neonates during their first day of life compared with a second measurement at the age of 1 month. Others have reported lower DTI velocities in infants compared with older children, which is in accordance with the results of

invasive animal studies.40 These studies describe a lower myocardial contractile state in the newborn age group.41 Besides lower global systolic strain results, we also found a greater range of variation in the youngest age group in comparison with older subjects. This variation could be explained by a previously reported greater sensitivity of myocardial performance to changes in afterload.38 Indeed, several important hemodynamic changes occur during early neonatal life. Preload increases because of the closure of the ductus arteriosus, with subsequent increase of pulmonary blood flow. A postnatal increment in afterload is related to the removal of the low-resistance placental circulation and increased systemic arterial blood pressure. When taking these alterations in loading conditions into account, it seems likely that they exert at least some influence on myocardial deformation. Therefore, it is important to note that our study population did not include infants aged <1 month. Besides hemodynamic and other biologic factors, it is possible that (part) of the large variations in myocardial strain in neonates reflect a technical shortcoming of 2DSTE analysis. Thin left ventricular walls of infants with small absolute deformation distances could be more susceptible to error. After adjustment for age, of the anthropometric and conventional echocardiographic parameters tested in this study, only VCFc and LVETc appeared to be significantly associated with global strain values. Unlike Lorch et al.,14 we did not find a strong relation between cardiac size (e.g., left ventricular end-diastolic volume) and peak

Journal of the American Society of Echocardiography Volume 24 Number 6

Marcus et al 633

Figure 6 Global peak systolic longitudinal strain (GlobalSL) (%) versus various morphometric variables among study subjects (scatterplots). Scatterplots with locally weighted line smoothers (loess regression lines) to indicate curved linear relationships. BSA, Body surface area (m2); EF_biplane, left ventricular ejection fraction (%); HR, heart rate (beats/min); LVEDV, left ventricular end-diastolic volume (mL); LVESV, left ventricular end-systolic volume (mL); LV_mass, left ventricular mass (g); LV_mass_BSA, left ventricular mass corrected for body surface area (g/m2).

634 Marcus et al

systolic strain parameters. Although our findings did not indicate one specific anthropometric or hemodynamic variable that explains the observed differences in myocardial deformation between the various age groups, it is likely that each individual morphometric variable (and their alterations during growth and development) contributes to the overall changes in peak systolic strain we have observed. Our results indicate a significant gradient of deformation from base to apex for longitudinal strain in the left ventricular septum and lateral free walls. In addition, circumferential and radial strain increased from base to midventricle as well. These findings are in agreement with those of Bussadori et al.13 However, considering that in this latter study, a different two-dimensional strain method was used that analyzed only the left side of the septum for longitudinal strain and only the subendocardial layer for circumferential strain, it is interesting to observe that in our study population, this base-to-apex gradient was significant even using speckle tracking that uses kernels distributed throughout the whole thickness of the myocardial wall and septum. In previous studies, different imaging techniques have given variable results regarding the uniformity of left ventricular strain from base to apex. Tagged MRI and 2DSTE studies reported higher strain in the apex, in contrast to tissue velocity imaging studies, which did not show significant variation from base to apex.42-45 The latter could be a reflection of the implications of the angle of insonation on apical measurements when using DTI. In addition to technical factors contributing to the base-to-apex gradient of deformation, we consider that it is possible that there exists a physiologic substrate in which the base-to-apex gradient is the result of the torsional mechanism of the left ventricular system and the direction of contraction of the descending fibers in the internal loop of helical ventricular myocardial band. Study Limitations This study was limited by its retrospective nature. Our study population did not include infants aged <1 month. Furthermore, the dependence of 2DSTE imaging on frame-by-frame tracking of the myocardial pattern makes it dependent on image factors, including reverberation artifacts and attenuation. Indeed, technical proficiency remains important in image processing. Also, we did not investigate radial and circumferential strain at the apical level of the left ventricle. Comparison with an independent external technique, such as tagged MRI, was not performed in the current study, primarily because the validation of speckletracking software has already been compared with MRI previously.46,47 Overall, the variability of radial peak systolic strain measurement was higher compared with circumferential and longitudinal measurement, which is in accordance with another study.48 However, the reproducibility of peak systolic strain measurements in all three directions of deformation (expressed by intraobserver and interobserver variability scores) was somewhat higher compared with this previous study. In addition, we used custom-made software that is not commercially available. The custom-made software was specially developed to improve the estimation of timing of peak systolic strain measurements while averaging strain curves, as well as to include peak systolic strain measurements that occur (shortly) before aortic valve closure. Although the custom-made software is not commercially available, this method can be implemented in generally available software such as MATLAB (The MathWorks Inc., Natick, MA; see the Appendix for instructions).

CONCLUSIONS With this study, we present age-specific reference values for 2DSTE parameters, which are essential for its use in evaluating (pathologic)

Journal of the American Society of Echocardiography June 2011

changes in ventricular myocardial function. This study shows that age is a major determinant of peak systolic strain values in healthy subjects and therefore emphasizes the need for age-specific reference values for adequate interpretation of 2DSTE measurements.

ACKNOWLEDGMENTS We thank Gert Weijers of the Clinical Physics Laboratory for his technical support in the study. We also express our gratitude to all the sonographers at our center for their invaluable participation.

REFERENCES 1. Pacileo G, Di Salvo G, Limongelli G, Miele T, Calabro R. Echocardiography in congenital heart disease: usefulness, limits and new techniques. J Cardiovasc Med 2007;8:17-22. 2. Marwick T. Techniques for comprehensive two dimensional echocardiographic assessment of left ventricular systolic function. Heart 2003; 89(suppl):iii2-8. 3. Admunsen BH, Helle-Valle T, Edvardsen T, Torp H, Crosby J, Lyseggen E, et al. Noninvasive myocardial strain measurement by speckle tracking echocardiography: validation against sonomicrometry and tagged magnetic resonance imaging. J Am CollCardiol 2006;47:789-93. 4. Korinek J, Wang J, Sengupta PP, Miyazaki C, Kjaergaard J, McMahon E, et al. Two-dimensional strain—a Doppler-independent ultrasound method for quantitation of regional deformation: validation in vitro and in vivo. J Am SocEchocardiogr 2005;18:1247-53. 5. Aurigemma G, Silver K, Priest M, Gaash W. Geometric changes allow normal ejection fraction despite depressed myocardial shortening in hypertensive left ventricular hypertrophy. J Am Coll Cardiol 1995;26: 195-202. 6. El-Menyar A, Galzerano D, Asaad N, Al-Mulla A, Arafa S, Al Suwaidi J. Detection of myocardial dysfunction in the presence of normal ejection fraction. J Cardiovasc Med 2007;8:923-33. 7. Vinereanu D, Ionescu AA, Fraser AG. Assessment of left ventricular long axis contraction can detect early myocardial dysfunction in asymptomatic patients with severe aortic regurgitation. Heart 2001;85:30-6. 8. Artis NJ, Oxborough DL, Williams G, Pepper CB, Tan LB. Two-dimensional strain imaging: a new echocardiographic advance with research and clinical applications. Int J Cardiol 2008;123:240-8. 9. Dandel M, Hetzer R. Echocardiographic strain and strain rate imaging— clinical applications. Int J Cardiol 2009;132:11-24. 10. Pavlopoulos H, Nihoyannopoulos P. Strain and strain rate deformation parameters: from tissue Doppler to 2D speckle tracking. Int J Cardiovasc Imaging 2008;24:479-91. 11. Leitman M, Lysyansky P, Sidenko S, Shir V, Peleg E, Binenbaum M, et al. Two-dimensional strain—a novel software for real-time quantitative echocardiographic assessment of myocardial function. J Am Soc Echocardiogr 2004;17:1021-9. 12. Mavinkurve-Groothuis AMC, Weijers G, Groot-Loonen J, Pourier MS, Feuth T, de Korte CL, et al. Interobserver, intraobserver and intrapatient reliability scores of myocardial strain imaging with two-dimensional echocardiography in patients treated with anthracyclines. Ultrasound Med Biol 2009;35:697-704.
13. Bussadori C, Moreo A, Di Donato M, De Chiara B, Negura D, DallAglio E, et al. A new 2D-based method for myocardial velocity strain and strain rate quantification in a normal adult and paediatric population: assessment of reference values. Cardiovasc Ultrasound 2009;7:8-19. 14. Lorch SM, Ludomirsky A, Singh GK. Maturational and growth-related changes in left ventricular longitudinal strain and strain rate measured by two-dimensional speckle tracking echocardiography in a healthy pediatric population. J Am SocEchocardiogr 2008;21:1207-15. 15. Lang RM, Bierig M, Devereux RB, Flachskampf FA, Foster E, Pellikka PA, et al; Chamber Quantification Writing Group; American Society of

Journal of the American Society of Echocardiography Volume 24 Number 6

Echocardiography’s Guidelines and Standards Committee; European Association of Echocardiography. Recommendations for chamber quantification: a report from the American Society of Echocardiography’s Guidelines and Standards Committee and the Chamber Quantification Writing Group, developed in conjunction with the European Association of Echocardiography, a branch of the European Society of Cardiology. J Am Soc Echocardiogr 2005;18:1440-63. 16. Delgado V, Tops LF, van Bommel RJ, van der Kley F, Marsan NA, Klautz RJ, et al. Strain analysis in patients with severe aortic stenosis and preserved left ventricular ejection fraction undergoing surgical valve replacement. Eur Heart J 2009;30:3037-47. 17. Cerqueira MD, Weissman NJ, Dilsizian V, Jacobs AK, Kaul S, Laskey WK, et al. American Heart Association Writing Group on Myocardial Segmentation and Registration for Cardiac Imaging. Standardized myocardial segmentation and nomenclature for tomographic imaging of the heart. A statement for healthcare professionals from the Cardiac Imaging Committee of the Council on Clinical Cardiology of the American Heart Association. Int J Cardiovasc Imaging 2002;18:539-42. 18. Berne RM, Levy MN. Physiology. 3rd ed. St. Louis, MO: Mosby; 1993. 19. Tei C, Ling LH, Hodge DO, Bailey KR, Oh JK, Rodeheffer RJ, et al. New index of combined systolic and diastolic myocardial performance: a simple and reproducible measure of cardiac function—a study in normals and dilated cardiomyopathy. J Cardiol 1995;26:357-66. 20. Rowland DG, Gutgesell HP. Use of mean arterial pressure for noninvasive determination of left ventricular end-systolic wall stress in infants and children. Am J Cardiol 1994;74:98-9. 21. Colan SD, Borow KM, Neumann A. Left ventricular end-systolic wall stress-velocity of fiber shortening relation: a load-independent index of myocardial contractility. J Am Coll Cardiol 1984;4:715-24. 22. Devereux RB, Reichek N. Echocardiographic determination of left ventricular mass in man. Anatomic validation of the method. Circulation 1977;55:613-8. 23. Burke GL, Arcilla RA, Culpepper WS, Webber LS, Chiang YK, Berenson GS. Blood pressure and echocardiographic measures in children: the Bogalusa Heart Study. Circulation 1987;75:106-14. 24. Milicevic G, Narancic NS, Steiner R, Rudan P. Increase in cardiac contractility during puberty. Coll Antropol 2003;27:335-41. 25. Cumming GR, Everatt D, Hastman L. Bruce treadmill test in children: normal values in a clinic population. Am J Cardiol 1978;41:69-75. 26. Cain PA, Ahl R, Hedstrom E, Ugander M, Allansdotter-Johnsson A, Friberg P, et al. Physiological determinants of the variation in left ventricular mass from early adolescence to late adulthood in healthy subjects. Clin Physiol Funct Imaging 2007;27:225-62. 27. Bisi G, Podio V, Valetto MR, Broglio F, Bertuccio G, Del Rio G, et al. Acute cardiovascular and hormonal effects of GH and hexarelin, a synthetic GHreleasing peptide, in humans. J Endocrinol Invest 1999;22:266-72. 28. Donath MY, Jenni R, Brunner HP, Anrig M, Kohli S, Glatz Y, et al. Cardiovascular and metabolic effects of insulin-like growth factor 1 at rest and during exercise in humans. J Clin Endocrinol Metab 1996;81:4089-94. 29. Cittadini A, Ishiguro Y, Str€ omer H, Spindler M, Moses AC, Clark R, et al. Insulin-like growth factor-1 but not growth hormone augments mammalian myocardial contractility by sensitizing the myofilament to calcium through a wortmannin-sensitive pathway: studies in rat and ferret isolated muscles. Circ Res 1998;83:50-9. 30. Freestone NS, Ribaric S, Mason WT. The effect of insulin-like growth factor1 on adult rat cardiac contractility. Mol Cell Biochem 1996;163–164:223-9. 31. Golden KL, Marsh JD, Jiang Y, Moulden J. Acute actions of testosterone on contractile function of isolated rat ventricular myocytes. Eur J Endocrinol 2005;152:470-83.

Marcus et al 635

32. Golden KL, Marsh JD, Jiang Y, Brown T, Moulden J. Gonadectomy of adult male rats reduces contractility of isolated cardiac myocytes. Am J Physiol Endocrinol Metab 2003;285:E449-53. 33. Zengin K, Tokac M, Duzenli MA, Soylu A, Aygul N, Azdemir K. Influence of menstrual cycle on cardiac performance. Maturitas 2007;58:70-4. 34. Pelzer T, Jazbutyte V, Hu K, Segerer S, Nahrendorf M, Nordbeck P, et al. The estrogen receptor-alpha agonist 16alpha-LE2 inhibits cardiac hypertrophy and improves hemodynamic function in estrogen-deficient spontaneously hypertensive rats. Cardiovasc Res 2005;67:604-12. 35. Gallinelli A, Angioni S, Matteo ML, Montaldo PL, Fenu MA, Volpe A. Variations of cardiac performance and inotropism in healthy postmenopausal women treated with estroprogestin replacement therapy. Menopause 1999;6:49-55. 36. Colan SD, Parness IA, Spevak PJ, Sanders SP. Developmental modulation of myocardial mechanics: age- and growth-related alterations in afterload and contractility. J Am Coll Cardiol 1992;19:619-29. 37. Crepaz R, Pitscheider W, Radetti G, Gentili L. Age-related variation in left ventricular myocardial contractile state expressed by the stress velocity relation. PediatrCardiol 1998;19:463-7. 38. Rowland DG, Gutgesell HP. Noninvasive assessment of myocardial contractility, preload, and afterload in healthy newborn infants. Am J Cardiol 1995;75:818-21. 39. Pena JL, da Silva MG, Faria SC, Salemi VM, Mady C, Baltabaeva A, Sutherland GR. Quantification of regional left and right ventricular deformation indices in healthy neonates by using strain rate and strain rate imaging. J Am Soc Echocardiogr 2009;22:369-75. 40. Eidem BW, McMahon CJ, Cohen RR, Wu J, Finkelshteyn I, Kovalchin JP, et al. Impact of cardiac growth on Doppler tissue imaging velocities: a study in healthy children. J Am Soc Echocardiogr 2004;17:212-21. 41. Nakanishi T, Jarmakani JM. Developmental changes in myocardial mechanical function and subcellular organelles. Am J Physiol 1984;246: H615-25. 42. Marwick TH, Leano RL, Brown J, Sun JP, Hoffmann R, Lysyansky P, et al. Myocardial strain measurement with 2-dimensional speckle-tracking echocardiography: definition of normal range. JACC Cardiovasc Imaging 2009;2:80-4. 43. Bogaert J, Rademarkers FE. Regional nonuniformity of normal adult human left ventricle. Am J Physiol Heart Circ Physiol 2001;280:H610-20. 44. Leitman M, Lysiansky M, Lysyansky P, Friedman Z, Tyomkin V, Fuchs T, et al. Circumferential and longitudinal strain in 3 myocardial layers in normal subjects and in patients with regional left ventricular dysfunction. J Am Soc Echocardiogr 2010;23:64-70. 45. Saito K, Okura H, Watanabe N, Hayashida A, Obase K, Imai K, et al. Comprehensive evaluation of left ventricular strain using speckle tracking echocardiography in normal adults: comparison of three-dimensional and two-dimensional approaches. J Am Soc Echocardiogr 2009;22:1025-30. 46. Becker M, Bilke E, Kuhl H, Katoh M, Kramann R, Franke A, et al. Analysis of myocardial deformation based on pixel tracking in two dimensional echocardiographic images enables quantitative assessment of regional left ventricular function. Heart 2006;92:1102-8. 47. Cho GY, Chan J, Leano R, Strudwick M, Marwick TH. Comparison of two dimensional speckle and tissue velocity based strain and validation with harmonic phase magnetic resonance imaging. Am J Cardiol 2006;97: 1661-6. 48. Koopman LP, Slorach C, Hui W, Manlhiot C, McCrindle BW, Friedberg MK, et al. Comparison between different speckle tracking and color tissue Doppler techniques to measure global and regional myocardial deformation in children. J Am Soc Echocardiogr 2010;23: 919-28.

636 Marcus et al

APPENDIX Three consecutive cardiac cycles were analyzed. Results from all views and segments were separately digitally stored in text files on the local hard drive of the GE workstation (with disabled drift compensation). The data files then were exported from the system and stored on the network for further analysis in a custom-made software package, Cardiac Curve Analysis Tool, using MATLAB version 7.4.0.287 (r2007a).  The step-by-step procedure of the custom-made software was as follows:  Automatic load of the three consecutive cardiac cycles results.  Up-sampling (cubic spline) to 2,000 samples per cardiac cycle (to be able to select precise time stamps for the QQ definition).  Interactive QQ onset defining (on the three consecutive cardiac cycles).  Curve length check (if length difference is >10%, delete the longest cycle).

Journal of the American Society of Echocardiography June 2011

 Deletion of data before and after the selected QQ time stamps.  Padding zeros at the end of the diastolic phase for the shortest cycles (to generate equal data lengths).  Drift compensation (using MATLAB’s detrend [linear] function).  Curve averaging (using MATLAB’s mean function).  Maximum detection.  Estimation of averaged peak values per segment.

The custom-made software used in this study was specially developed to improve the timing of systolic deformation (rate) while averaging multiple cardiac cycles. This method does not affect the (average) maximum systolic strain (rate) measurements compared with commercially available EchoPAC software to determine the average peak systolic strain (rate) values (per cardiac segment).