Systolic-diastolic functional coupling in healthy children and in those with dilated cardiomyopathy

Systolic and diastolic function affect dilated cardiomyopathy (DCM) outcomes. However, systolic:diastolic (S:D) coupling, as a distinct characteristic, may itself affect function but is poorly characterized. Using tissue Doppler echocardiography, we found that S:D coupling becomes weaker in DCM with left ventricular remodeling and dysfunction and that the S:D coupling ratio may be useful to assess coupling, warranting study in relation to patient outcomes.

Abstract

Systolic and diastolic function affect dilated cardiomyopathy (DCM) outcomes. However, systolic-diastolic coupling, as a distinct characteristic, may itself affect function but is poorly characterized. We hypothesized that echocardiographic left ventricular (LV) longitudinal systolic tissue velocities (S') correlate with diastolic longitudinal velocities (E') and that their relationship is associated with ventricular function and that this relationship is impaired in pediatric DCM. We analyzed data from the Pediatric Heart Network Ventricular Volume Variability study, using linear regression and generalized additive modeling to assess relationships between S' and E' at the lateral and septal mitral annulus. We explored relationships between the systolic:diastolic (S:D) coupling ratio (S':E' relative to age) and ventricular function. Up to 4 echocardiograms from 130 DCM patients (mean age: 9.3 ± 6.1 yr) and 1 echocardiogram from each of 591 healthy controls were analyzed. S' and E' were linearly related in controls (r = 0.64, P < 0.001) and DCM (r = 0.83, P < 0.001). In DCM, the magnitude of association between S' and E' was reduced with progressive ventricular remodeling. The S:D ratio was more strongly associated with LV function in controls vs. DCM. The septal S:D ratio was higher (presumed worse) in DCM vs. controls (0.69 ± 0.13 vs. 0.62 ± 0.12, P = 0.001). A higher septal S:D ratio was associated with worse LV dimensions (parameter estimate: 0.0061, P = 0.004), mass (parameter estimate: 0.0074, P = 0.002), ejection fraction (parameter estimate: −0.0303, P = 0.024), and inflow propagation (parameter estimate: −0.3538, P < .001). S:D coupling becomes weaker in DCM with LV remodeling and dysfunction. The S:D coupling ratio may be useful to assess coupling, warranting study in relation to patient outcomes.

NEW & NOTEWORTHY

Systolic and diastolic function affect dilated cardiomyopathy (DCM) outcomes. However, systolic:diastolic (S:D) coupling, as a distinct characteristic, may itself affect function but is poorly characterized. Using tissue Doppler echocardiography, we found that S:D coupling becomes weaker in DCM with left ventricular remodeling and dysfunction and that the S:D coupling ratio may be useful to assess coupling, warranting study in relation to patient outcomes.

both systolic and diastolic dysfunctions are present in dilated cardiomyopathy (DCM) and both contribute to morbidity and mortality in this condition (28). In clinical practice, systolic and diastolic functions are traditionally considered as two separate and distinct phases. However, the systolic and diastolic behaviors of the left ventricle (LV) are physiologically linked through several mechanisms. Diastolic influence on systole (diastolic → systolic coupling) is the most familiar and is mediated through the Frank-Starling and other mechanisms enhance systolic contraction. Systolic influence on diastole (systolic → diastolic coupling) is mediated through enhancement of early diastolic filling by myocardial elastic recoil in early diastole (24, 25).

Although in normal hearts reciprocal coupling between systole and diastole enhances cardiac performance (37), experimental data have yielded conflicting results on whether the Frank-Starling mechanism is disrupted in DCM (17, 34, 35). The profound impairment of ventricular molecular and myocardial systolic and diastolic properties that is characteristic of DCM may disrupt diastolic → systolic coupling through reduced preload reserve and impair systolic → diastolic coupling secondary to the diminished end-systolic deformation that is associated with reduced fiber shortening. Therefore, assessment of systolic and diastolic function, and of the coupling between them, the subject of this study, is of interest. Defining abnormal linking between systolic and diastolic function as a distinct entity may itself hold important information regarding patient cardiac status and potentially regarding clinical outcomes beyond evaluation of systolic and diastolic function alone. While several papers demonstrate that systolic and diastolic function are linked (24, 25, 37), systolic-diastolic (S:D) coupling as distinct functional entities has not been systematically evaluated in either adult or pediatric DCM, and although theoretically predicted, it remains unknown whether impaired coupling is associated with cardiac dysfunction. Moreover, S:D coupling is not routinely investigated or accounted for clinically and to date there is no accepted method for assessing S:D coupling in clinical practice. Gold standard methods for assessing systolic and diastolic function in this population include measures of contractility such as end-systolic elastance and ventricular filling pressures, respectively. However, these necessitate invasive catheterization and even then end-systolic elastance is not routinely measured. Echocardiography is commonly used to assess cardiac function in DCM and tissue Doppler imaging can provide measures of systolic and diastolic function in the same cardiac sample. Although twist-untwist is a key mechanism linking systolic and diastolic function (7), assessment of twist-untwist is relatively cumbersome by two-dimensional echocardiography. A more simple and practical approach, and presumably more feasible in clinical practice, is to study the longitudinal vector of the overall twist-untwist action of the LV, readily achievable using tissue Doppler echocardiography.

Using tissue Doppler echocardiography to assess systolic and diastolic longitudinal function and their coupling, we previously found that mitral annular systolic longitudinal velocities correlate linearly with diastolic velocities in healthy children and in those with hypertrophic or DCM (22). This relationship varied between the groups, implying that S:D coupling may be differentially affected by the specific disease (22). We also found that the ratio of peak systolic (S') to peak early diastolic (E') longitudinal myocardial velocities (=S'/E'), a putative index of S:D coupling, correlated with mitral inflow deceleration time (22). As there is no currently accepted index of S:D coupling used in clinical routine, a direct ratio between indexes of systolic and diastolic function may be useful. As correlations between systolic and diastolic velocities cannot be used to assess S:D coupling in the individual patient (as correlations cannot be derived from a single measurement in one individual), the S:D coupling ratio z-score (defined as the z-score relative to age of the ratio of peak S' divided by peak E') may be a potential clinically applicable parameter to indicate S:D coupling (22). This ratio is based on the following concept: if S:D coupling is reduced, and the energy built up in systole is not fully transferred into diastole, we would expect less diastolic response for a given systolic velocity and thus a higher S:D coupling ratio. However, our initial single-center study was limited by its relatively small size and its retrospective nature. The aim of the current exploratory study was to utilize a large, multicenter, prospectively collected, relatively homogenous cohort of children with DCM, analyzed in a core echocardiographic laboratory, to: 1) investigate the relation of systolic to diastolic tissue velocities as an indicator of S:D coupling; 2) investigate the relation of S:D coupling to parameters of ventricular size and function in pediatric DCM; and 3) compare these results with normal controls. We hypothesized that LV longitudinal systolic velocities are normally highly correlated with diastolic longitudinal velocities, that their relationship is related to ventricular size and function, and that S:D coupling is impaired in DCM compared with healthy controls.

METHODS

This study was a subanalysis of the Pediatric Heart Network Ventricular Volume Variability (VVV) database, which aimed to determine the interstudy variability of echocardiographically derived LV measurements in pediatric DCM patients and the variance of change in z-score on serial follow-up (5). The z-scores reflect the number of SDs from the parameter mean for a particular population or age group relative to the distribution in a normal population, thereby allowing comparison of values for children over the pediatric age range by adjusting for the effects of age and body size. The z-score for the mean of a normal population distribution is 0, and the normal range is typically defined as −2 to +2 SD. In brief, children or young adults age <22 yr with known or suspected DCM disease duration >2 mo, anticipated longitudinal follow-up to occur at the same institution, and informed consent were prospectively enrolled at eight centers between the years 2005 and 2007 (5). Echocardiographic eligibility criteria included LV dilation [end-diastolic dimension (EDD) >5.5 cm or z-score >2] and LV dysfunction [ejection fraction (EF) <50%, shortening fraction <28%, or z-score for either >−2]. Exclusion criteria included other forms of cardiomyopathy, noncompaction, congenital heart disease, nonsinus rhythm, and hemodynamic instability (5). Repeat echocardiographic assessment was performed on return visits up to 18 mo following enrollment. The mean time between the first and follow-up echocardiogram (performed closest to a 12-mo follow-up target) was 9.1 ± 3.5 (median 9.6, range 2 to 18) mo. In some patients a third, and in two cases a fourth, follow-up echocardiogram was available and included in the analysis.(5)

Normal Data

As the VVV study did not include a normal control cohort, we used z-scores derived from a large database of normal controls collected at a single center (Boston Children's Hospital) as a reference for the S:D coupling ratio and to explore associations between S:D coupling and ventricular function (3). The characteristics of this normal population and z-score methodologies have been previously described (3, 4). In brief, subjects were either normal volunteers or were referred for screening for conditions such as murmur, mitral valve prolapse, or bicuspid aortic valve. These subjects had normal echocardiograms, no history of disease, normal height, weight, and blood pressure for age. In normal controls, the echocardiogram was obtained at a single visit. Pulsed tissue velocities were prospectively acquired at the mitral lateral and septal annuli, as described for the DCM cohort. Peak S' and E' were measured in the same core laboratory that measured echo parameters for the VVV study. For the current analysis, we used existing measurements from the core laboratory database.

Echocardiography

Echocardiographic methodologies have been previously detailed and were performed according to recently published pediatric guidelines (5, 19). The current analysis used LV volumes and mass calculated using the (5/6) × area × length method and dimensions from two-dimensional echocardiography as the (5/6) × area × length method had better reproducibility than other methods in recent pediatric studies (19, 21). LV mass was expressed as a z-score. Wall stress was derived from the Laplace equation, provided by the ultrasound software, using the blood pressure, LV dimension, and wall thickness. Pulsed tissue Doppler (TDI) measurements were acquired from the apical four-chamber view at the mitral lateral and septal annulus and the peak systolic (S'), early diastolic (E'), and late diastolic (A') TDI velocities recorded. When TDI E' and A' waves were fused, the “summation” wave was taken as the peak diastolic wave. Peak velocities were averaged over three beats. For the current study, echocardiographic data from the primary image acquisition obtained at each time point, averaged across three beats using the core laboratory primary reader's interpretation from the VVV database, were used.

Assessing Relationships Between Systolic and Diastolic Velocities

Relationships between systolic and diastolic function were explored by assessing correlations between S' and E' at the lateral and septal mitral annulus.

S:D Coupling Ratio

Expressing the S:D coupling ratio as a z-score allowed for investigation of its association with parameters of ventricular size, remodeling, and function, which are related to age. The S:D coupling ratio was calculated at the lateral mitral annulus and the septal annulus, as the average of three successive beats.

Statistical Analysis

Descriptive statistics include the median with interquartile range for skewed variables, mean ± SD for other continuous variables, and frequency with percentage for categorical variables. Peak systolic and peak early diastolic velocity and S:D coupling index z-score were compared by t-test or Wilcoxon rank sum test as appropriate to assess differences between the two cohorts. The within-subject correlation coefficient (the proportion of variation in systolic velocity explained by diastolic velocity after removing the variation due to measurement on different subjects) was calculated to assess whether the increase in systolic velocity over time is associated with diastolic velocity within the same individual. Between-subject correlation was calculated by using a weighted average of the number of repeated measurements available for each subject (33). Linear regression with a random effect for subject and a compound symmetry covariance structure to account for correlation among repeated measures and a fixed continuous time effect denoting time since the baseline echo was used to model the serial measures of S:D coupling index to identify the impact of LV size and function on the S:D coupling index. Linear regression, piecewise linear regression, and generalized additive modeling (GAM) were used to assess the association between systolic and diastolic velocities. Tests of interaction were used to assess the impact of LV size and function on the strength of association between systolic and diastolic velocities. LV size and function measures (e.g., LV EDD z-score) were categorized using quartiles to report estimated slopes by magnitude of LV size and function. In testing associations, S' and E' were sometimes placed on the “dependent” (y-axis) and sometimes the “independent” variable (x-axis). In reality, systole and diastole are reciprocally linked both being dependent on one another.

Linear regression was also used to identify the impact of LV size and function on the S:D coupling index. As an absolute LV EDD >5.5 cm was an inclusion criterion (and may correspond to a z-score of <2 in some patients), and as regression analyses included data from all visits, with some patients improving LV dimensions over time, data with LV EDD z-score <2 are included. The number and type of significant findings from the two cohorts were compared to assess whether S:D coupling is weaker in DCM subjects relative to normal controls. All analyses were conducted using SAS version 9.2 (Statistical Analysis System, Cary, NC) and R version 2.13.0. The study was approved by the institutional review committees at each institution. Subjects gave informed consent.

RESULTS

The current study included 130 fully eligible patients for the VVV study, 130 of whom had baseline echocardiography, at a mean age of 9.3 ± 6.1 yr (5). Of these, 23 patients had echocardiography only at baseline, 70 had one follow-up echocardiogram at a mean interval of 7.1 ± 3.5 mo, 28 had 2 follow-up echocardiograms at a mean interval of 10.9 ± 3.1 mo, 7 had 3 follow-up echocardiograms at a mean interval of 11.3 ± 2.5 mo, and 2 had 4 follow-up echocardiograms. Patient characteristics are presented in Table 1. The majority of subjects had known or suspected idiopathic or familial DCM and were in New York Heart Association or Ross classification class 1 or 2. TDI velocities and the S:D ratio are summarized in Table 2. At baseline, diastolic summation waves were present in 19 (14.6%) patients.

Table 1.

Dilated cardiomyopathy subject characteristics at baseline

Variable	n
Age at echocardiography, yr	130	9.3 ± 6.1 (median = 9.3)
Height, cm	128.0 ± 36.7
Weight, kg	37.7 ± 27.6
Gender
Male	58	44.6%
Female	72	55.4%
Primary causes of dilated cardiomyopathy
Idiopathic or familial	118	90.7%
Metabolic disorder	2	1.5%
Mitochondrial disorder	2	1.5%
Neuromuscular disease	4	3.1%
Single gene defect	4	3.1%
Congestive heart failure classification
Ross classification (age <5 yr)	45
Class I	35	77.8%
Class II	6	13.3%
Class III	4	8.9%
Class IV	0	0%
New York Heart classification (age ≥5 yr)	85
Class I	50	58.8%
Class II	33	38.8%
Class III	2	2.4%
Class IV	0	0%
Medications at baseline echo
Antithrombotic	36	28.0%
β-blocker	68	52.0%
Angiotensin converting enzyme inhibitor	109	84.0%
Diuretic	68	52.0%
Glycoside	77	59.0%

Table 2.

Systolic and diastolic tissue Doppler velocities and their ratio at the lateral mitral and septal annulus

	Dilated Cardiomyopathy Baseline Echocardiogram (n = 130)			Controls (n = 591)			t-Test	Wilcoxon
	Raw value, cm/s	z-Score (means ± SD)	Median	Raw value, cm/s	z-Score (means ± SD)	Median	P value	P value
Lateral mitral annulus
Diastolic velocity (E')	12.69 ± 4.59	−1.60 ± 1.66	−1.68	17.29 ± 4.18	−0.02 ± 1.06	−0.09	<.0001	<.0001
Systolic velocity (S')	6.44 ± 2.60	−1.49 ± 1.43	−1.75	9.65 ± 2.51	0.03 ± 0.98	−0.04	<.0001	<.0001
Septal annulus
Diastolic velocity (E')	8.48 ± 2.70	−2.19 ± 1.35	−2.01	12.74 ± 2.63	0.01 ± 0.96	−0.04	<.0001	<.0001
Systolic velocity (S')	5.73 ± 1.77	−1.79 ± 2.22	−2.00	7.63 ± 1.27	0.12 ± 1.03	0.05	<.0001	<.0001
S:D ratio index z-score
Left lateral annulus	0.58 ± 0.14	−0.20 ± 1.33	−0.55	0.54 ± 0.19	0.01 ± 1.02	−0.15	0.128	0.001
Septal annulus	0.69 ± 0.13	0.87 ± 1.40	0.85	0.62 ± 0.12	0.00 ± 1.06	−0.13	<.0001	<.0001

Normal controls included 591 subjects (283 females). Their demographics were as follows: age 10.38 ± 5.47 yr (range: 0.01- 20.50); 62 were <2 yr old; weight: 41.10 ± 22.66 kg (range: 1.23-99.70); height: 139.73 ± 33.96 cm (range: 46.00–196.00); and body surface area: 1.23 ± 0.51 cm² (range: 0.11–2.30); all had normal body mass index: 19.04 ± 3.49 (range: 10.80–29.45).

Relationship Between Systolic and Diastolic Velocities in Normal Controls

Diastolic and systolic velocities were positively correlated at the lateral mitral and septal annuli. A one-unit increase in systolic velocity was associated with a 0.96 unit increase in diastolic velocities at the lateral mitral annulus (r = 0.58; P < 0.001) and a 1.32 unit increase (r = 0.64, P < 0.001) in diastolic velocities at the septal annulus. However, at both locations, a nonlinear relationship between these variables was present (GAM P <0.001 for both) with little association at high velocities, i.e., ∼10 cm/s or higher (Fig. 1). Results were unchanged when controls younger than age 2 mo were excluded to have the normal cohort more closely resemble the DCM cohort.

Fig. 1.

Relationship between the tissue Doppler peak early diastolic velocity E' (y-axis) and the peak systolic velocity S' (x-axis) at the lateral mitral (top) and septal (bottom) annulus in normal controls. A 1-unit increase in systolic velocity was associated with a 0.96 unit increase in diastolic velocities at the lateral mitral annulus (r = 0.58; P < 0.001) and a 1.32 unit increase (r = 0.64, P < 0.001) in diastolic velocities at the septal annulus. Nonlinearity P < 0.001. AVV, atrioventricular valve.

Relationship Between Systolic and Diastolic Velocities in DCM Patients

Relationships between systolic and diastolic velocities in the DCM cohort are presented in Fig. 2. Moderate correlations between systolic and diastolic tissue velocities were observed at the lateral mitral annulus (Fig. 2A). This relationship was relatively constant over the duration of follow-up. The relationship between systolic and diastolic velocities was highest at the septal annulus (r = 0.83, 0.76, 0.66, and 0.80 at baseline through third follow-up visit, respectively; P < 0.001 for all; Fig. 2B).

Fig. 2.

Relationship between the tissue Doppler peak early diastolic velocity E' (y-axis) and the peak systolic velocity S' (x-axis) at the lateral mitral (A) and septal (B) annulus in dilated cardiomyopathy (DCM) patients. These are positively correlated at baseline and at follow-up visits 1–3. The correlation coefficients account for the repeated measurements per subject. Data points from different study visits are displayed with different colors.

Within-subject correlation analysis showed that an increase in the lateral mitral annulus peak systolic velocity was associated with an increase in peak early diastolic velocity (r = 0.52, P < 0.001). Between-subject correlation coefficient analysis showed that higher S' values correlated with higher E' velocities (r = 0.71, P < 0.001). Similarly, within-subject Pearson's correlation for the septal annulus was r = 0.62 and between-subjects correlation was r = 0.80 (P < 0.001 for both). Nonlinear modeling did not reveal stronger associations between the two parameters.

Impact of LV Size and Function on Systolic-Diastolic Coupling

To examine whether LV size and function impact the association between longitudinal systolic-diastolic coupling, an interaction term was included in the mixed effects regression model. This term assesses if the association between diastolic and systolic velocities is stronger at higher or lower levels of ventricular function. We found that both EDD and end-systolic dimension (ESD) impacted the relationship between peak early systolic and diastolic velocities at the lateral mitral annulus (interaction P = 0.016 and 0.049, respectively). Patients with larger LV EDD had weaker correlations between systolic and diastolic velocities with the slope decreasing by 0.03 for every one-unit increase in peak early diastolic velocity (interaction P = 0.006; Fig. 3A). The slope estimates of peak S' vs. peak E' varied by size of the LV EDD z-score: a one-unit increase in E' was associated with a 0.37 unit increase in peak systolic velocity when the LV EDD z-score was 1.87 (25th percentile), 0.33 when LV EDD z-score was 3.19 (median), and 0.28 when the LV EDD z-score was 4.96 (75th percentile). Similarly, among patients with an average ESD z-score in the top quartile (>75th percentile), overall S:D coupling was weaker (but still positive) compared with patients with average LV ESD z-score in the lowest quartile (<25th percentile; Fig. 3B). There was no interaction with LV EF (P = 0.77) or with raw LV mass (P = 0.39) or mass-to-volume ratio (P = 0.11).

Fig. 3.

Relationships between the tissue Doppler peak early diastolic velocity E' (y-axis) and the peak systolic velocity S' (x-axis) at the lateral mitral annulus in DCM patients stratified by severity of left ventricular (LV) dilation in quartiles (Q1-4) of severity. Relationships stratified by end-diastolic dimension (EDD) z-score (A) and by end-systolic dimension (ESD) z-score (B). For end-diastolic dimension z-score the 25th; 50th and 75th percentiles are 1.9, 3.3, and 5.4, respectively. For end-systolic dimension z-score the 25th; 50th and 75th percentiles are 4.4, 6.4, and 9.9, respectively. Data points from different study visits are displayed with different colors. The slope of the S:D relationship decreases with increasing LV remodeling.

Effect of Myocardial Shortening and Wall Stress

We identified a significant interaction (P = 0.027) between E' and velocity of fiber shortening z-score at the lateral mitral annulus. The association between systolic and diastolic velocities strengthens as velocity of fiber shortening z-score increases (increase in slope for peak S' vs. E' of 0.011 per one-unit increase fiber shortening velocity z-score).

We also identified a significant interaction (P = 0.027) between E' and end-systolic fiber stress z-score. That is, the prediction of systolic velocity by diastolic velocity gets weaker as end-systolic fiber stress z-score increases (slope for peak S' vs. E' decreases by 0.02 per each one-unit increase in end-systolic fiber stress z-score). However, as the velocity of fiber shortening itself depends on end-systolic wall stress, which in turn is influenced by ESD, we examined the two significant interactions after adjusting for ESD z-score. After adjustment, the interaction between E' and fiber shortening velocity z-score becomes nonsignificant (P = 0.90). Similarly, after adjustment for ESD z-score, the interaction between E' and end-systolic fiber stress z-score also becomes nonsignificant (P = 0.99). Thus the dependence of coupling magnitudes for both velocity of shortening and end-systolic stress is influenced predominantly by underlying LV systolic size.

Systolic-Diastolic Relationships Expressed as a Systolic-Diastolic Coupling Ratio: Normal Controls

In healthy controls the S:D coupling ratio was (mean ± SD) 0.58 ± 0.14 at the mitral lateral annulus (n = 591) and 0.62 ± 0.12 at the septal annulus (n = 590).

Relationships between the S:D coupling ratio and functional parameters in normal controls.

MITRAL LATERAL ANNULUS.

The results from all tested associations are presented in the appendix (see Table A1). Linear associations between the mitral lateral S:D coupling ratio and other functional indexes with P < 0.1 in normal controls are summarized in Table 3. A higher S:D ratio z-score was associated with: older age at echocardiography, male gender, higher systolic blood pressure, larger ventricular dimensions and mass, smaller sphericity index, higher velocity of fiber shortening, and lower peak early diastolic velocity.

Table A1.

Univariate linear regressions of the S:D ratio raw value and z-score on patient and echocardiographic characteristics at the lateral mitral annulus among normal controls

		Outcome = Raw Ratio		Outcome = Ratio z-Score
Covariate	n	Estimate*	P value	Estimate*	P value
Age at echocardiography, yr	591	0.0012	0.232	0.0456	<.001
Gender (male vs. female)	591	0.0292	0.012	0.1784	0.034
SBP-for-age z-score	558	0.0214	<.001	0.1276	0.003
DBP-for-age z-score	557	0.0121	0.035	0.0564	0.183
ESD-for-BSA z-score	531	−0.0130	0.033	−0.0933	0.036
End diastolic volume, ml	575	0.0001	0.243	0.0050	<.001
EDV-for-BSA z-score	574	−0.0185	0.002	−0.0970	0.025
End systolic volume, ml	574	0.0003	0.325	0.0117	<.001
ESV-for-BSA z-score	574	−0.0314	<.001	−0.2409	<.001
Ventricular mass, g	575	0.0003	0.038	0.0063	<.001
Mass-for-BSA z-score	574	−0.0066	0.267	−0.0323	0.456
Mass-to-volume ratio	575	0.1743	<.001	0.9910	0.002
MV-for-age z-score	575	0.0215	<.001	0.1389	<.001
Sphericity index	523	−0.3563	<.001	−3.0216	<.001
Sphericity-for-age z-score	523	−0.0229	<.001	−0.1792	<.001
Shortening fraction, % [2D]	531	0.0019	0.312	−0.0048	0.729
Shortening fraction-for-age z-score	531	−0.0017	0.801	−0.0052	0.915
Velocity of fiber shortening, circ/s	476	0.2335	<.001	1.0542	0.013
VCFc-for-age z-score	476	0.0171	0.005	0.1240	0.005
Ejection fraction, %	575	−0.0458	0.724	−0.1971	0.834
EF-for-age z-score	575	−0.0006	0.915	0.0041	0.924
End-systolic meridional fiber stress, g/cm2	481	−0.0004	0.384	0.0041	0.198
End-systolic circumferential fiber stress, g/cm2	477	−0.0002	0.463	0.0025	0.145
Peak early velocity, m/s	459	−0.2194	<.001	−1.2407	<.001
Peak early velocity-for-age z-score	459	−0.0345	<.001	−0.2487	<.001
Peak atrial velocity, m/s	422	0.1371	0.008	0.5649	0.139
Peak A-wave velocity-for-age z-score	420	0.0076	0.166	0.0397	0.327
Early deceleration time, ms	419	0.0000	0.964	0.0023	0.063
Deceleration time-for-age z-score	419	−0.0049	0.461	−0.0435	0.370
Left ventricular flow propagation velocity, cm/s	224	−0.0001	0.756	0.0001	0.964
LV flow propagation-for-age z-score	224	−0.0031	0.747	−0.0074	0.915
Duration of flow reversal during atrial systole, ms	79	−0.0011	0.087	−0.0031	0.468
Pulmonary vein A-wave duration-for-age z-score	79	−0.0123	0.473	−0.0701	0.546
R-R interval, ms	591	−0.0001	<.001	0.0000	0.839
Heart rate, beats/min	591	0.0014	<.001	0.0009	0.614

EF, ejection fraction; R-R interval, electrocardiogram R-R interval; LV, left ventricular.

^*Slope for continuous predictors and mean group difference for categorical predictors.

Table 3.

Univariate linear regressions of the S:D ratio raw value and z-score on echocardiographic characteristics at the lateral mitral annulus among normal controls

		Outcome = Raw Ratio		Outcome = Ratio z-Score
Covariate	n	Estimate*	P value	Estimate*	P value
Age at echocardiography, yr	591	0.0012	0.232	0.0456	<.001
Gender (male vs. female)	591	0.0292	0.012	0.1784	0.034
SBP-for-age z-score	558	0.0214	<.001	0.1276	0.003
DBP-for-age z-score	557	0.0121	0.035	0.0564	0.183
EDV-for-BSA z-score	574	−0.0185	0.002	−0.0970	0.025
ESV-for-BSA z-score	574	−0.0314	<.001	−0.2409	<.001
Ventricular mass, g	575	0.0003	0.038	0.0063	<.001
Mass-to-volume ratio	575	0.1743	<.001	0.9910	0.002
MV-for-age z-score	575	0.0215	<.001	0.1389	<.001
Sphericity-for-age z-score	523	−0.0229	<.001	−0.1792	<.001
VCFc-for-age z-score	476	0.0171	0.005	0.1240	0.005
Peak early velocity (E)-for-age z-score	459	−0.0345	<.001	−0.2487	<.001
Peak atrial (A) velocity, m/s	422	0.1371	0.008	0.5649	0.139
Heart rate, beats/min	591	0.0014	<.001	0.0009	0.614

n Is the total number of observations, not subjects. S:D, systolic and diastolic ratio; SBP, systolic blood pressure; DBP, diastolic blood pressure; BSA, body surface area; EDV, end-diastolic volume; ESV, end systolic volume; MV, mass-to-volume ratio; VCFc, velocity of fiber shortening.

^*Slope for continuous predictors and mean group difference for categorical predictors.

Septal annulus.

Linear associations between the septal S:D coupling ratio and other functional indices with P < 0.1 in normal controls are summarized in Table 4. The results from all tested associations are presented in the appendix (see Table A2). A higher S:D ratio z-score at the septal annulus in healthy children was associated with: older age at echocardiography, male gender, higher systolic blood pressure, larger ventricular dimensions and mass, smaller sphericity index, and smaller peak early diastolic velocity.

Table 4.

Univariate linear regressions of the S:D ratio raw value and z-score at the septal annulus among normal controls

		Outcome = Raw Ratio		Outcome = Ratio z- Score
Covariate	n	Estimate*	P value	Estimate*	P value
Age at echocardiography, yr	590	−0.0029	<.001	0.0418	<.001
Gender (male vs. female)	590	0.0249	0.010	0.1781	0.042
SBP-for-age z-score	557	0.0196	<.001	0.1339	0.002
DBP-for-age z-score	556	0.0135	0.003	0.0681	0.113
End diastolic volume, ml	574	−0.0003	0.002	0.0046	<.001
EDV -for- BSA z-score	573	−0.0116	0.019	−0.0406	0.367
ESV-for-BSA z-score	573	−0.0150	0.003	−0.1621	<.001
Ventricular mass, g	574	−0.0003	0.014	0.0055	<.001
Mass-to-volume ratio	574	0.1222	0.001	0.5750	0.090
MV-for-age z-score	574	0.0137	0.003	0.0892	0.035
Sphericity-for-age z-score	522	−0.0108	0.063	−0.1177	0.028
Shortening fraction, % [2D]	530	0.0016	0.299	−0.0236	0.093
VCFc-for-age z-score	476	0.0080	0.098	0.0677	0.124
End-systolic meridional fiber stress, g/cm2	480	−0.0009	0.014	0.0051	0.112
End-systolic circumferential fiber stress, g/cm2	476	−0.0004	0.016	0.0029	0.092
Peak early velocity (E) vs. age z-score	458	−0.0224	<.001	−0.2004	<.001
Peak atrial (A) velocity, m/s	422	0.1092	0.006	0.2172	0.572
Early deceleration time vs. age z-score	419	−0.0064	0.215	−0.0840	0.078
Duration of flow reversal during atrial systole, ms	79	−0.0007	0.070	0.0017	0.598
Heart rate, beats/min	590	0.0018	<.001	−0.0001	0.957

n Is the total number of observations, not subjects.

^*Slope for continuous predictors and mean group difference for categorical predictors.

Table A2.

Univariate linear regressions of the S:D ratio raw value and z-score at the septal annulus among normal controls

		Outcome = Raw Ratio		Outcome = Ratio z-Score
Covariate	n	Estimate*	P value	Estimate*	P value
Age at echocardiography, yr	590	−0.0029	<.001	0.0418	<.001
Gender (male vs. female)	590	0.0249	0.010	0.1781	0.042
SBP-for-age z-score	557	0.0196	<.001	0.1339	0.002
DBP-for-age z-score	556	0.0135	0.003	0.0681	0.113
End diastolic volume, ml	574	−0.0003	0.002	0.0046	<.001
EDV-for-BSA z-score	573	−0.0116	0.019	−0.0406	0.367
End systolic volume, ml	573	−0.0007	0.005	0.0113	<.001
ESV-for-BSA z-score	573	−0.0150	0.003	−0.1621	<.001
Ventricular mass, g	574	−0.0003	0.014	0.0055	<.001
Mass-for-BSA z-score	573	−0.0066	0.186	−0.0396	0.379
Mass-to-volume ratio	574	0.1222	0.001	0.5750	0.090
MV-for-age z-score	574	0.0137	0.003	0.0892	0.035
Sphericity index	522	−0.1429	0.094	−2.0370	0.010
Sphericity-for-age z-score	522	−0.0108	0.063	−0.1177	0.028
Shortening fraction, %	530	0.0016	0.299	−0.0236	0.093
Shortening fraction-for-age z-score	530	−0.0083	0.119	−0.0775	0.116
Velocity of fiber shortening, circ/s	476	0.2045	<.001	0.6608	0.119
VCFc-for-age z-score	476	0.0080	0.098	0.0677	0.124
Ejection fraction, %	574	−0.1014	0.346	−0.7649	0.433
EF-for-age z-score	574	−0.0046	0.361	−0.0307	0.497
End-systolic meridional fiber stress, g/cm2	480	−0.0009	0.014	0.0051	0.112
End-systolic circumferential fiber stress, g/cm2	476	−0.0004	0.016	0.0029	0.092
Peak early velocity, m/s	458	−0.1802	<.001	−0.9902	<.001
Peak early velocity-for-age z-score	458	−0.0224	<.001	−0.2004	<.001
Peak atrial velocity, m/s	422	0.1092	0.006	0.2172	0.572
Peak A-wave velocity-for-age z-score	420	0.0030	0.477	−0.0013	0.974
Early deceleration time, ms	419	−0.0003	0.007	0.0010	0.424
Deceleration time-for-age z-score	419	−0.0064	0.215	−0.0840	0.078
Left ventricular flow LV propagation velocity, cm/s	224	−0.0001	0.682	0.0007	0.833
LV flow propagation-for-age z-score	224	−0.0032	0.679	−0.0041	0.954
Duration of flow reversal during atrial systole, ms	79	−0.0007	0.070	0.0017	0.598
Pulmonary vein A-wave duration-for-age z-score	79	−0.0009	0.931	0.0221	0.801
R-R interval, ms	590	−0.0002	<.001	0.0001	0.584
Heart rate, beats/min	590	0.0018	<.001	−0.0001	0.957

^*Slope for continuous predictors and mean group difference for categorical predictors.

Systolic-Diastolic Coupling Ratio in DCM Patients

In comparison to normal controls, at the baseline visit DCM patients had a similar S:D coupling ratio at the lateral mitral annulus (0.58 ± 0.14 vs. 0.54 ± 0.19, P = 0.11) and higher S:D coupling ratio at the septal annulus (0.69 ± 0.13 vs. 0.62 ± 0.12, P = 0.001). At the lateral mitral annulus, the mean S:D coupling ratio z-score of DCM patients (−0.20 ± 1.33) was not different from the normal mean of 0, while at the septal annulus the S:D coupling ratio z-score of DCM was significantly higher than the normal mean (0.87 ± 1.40 vs. 0.00 ± 1.06, P < 0.0001) (Table 2).

Association of the S:D coupling ratio with other functional indexes in DCM.

LATERAL MITRAL ANNULUS.

The results from all tested associations are presented in the appendix (see Table A3). Linear associations between the S:D coupling ratio and other functional indexes in DCM patients at the baseline visit with a P < 0.1 are shown in Table 5. LV EF, age, gender, and blood pressure did not affect the statistical association between S' and E' at the lateral mitral annulus. The S:D coupling ratio at the mitral lateral annulus was positively associated with diastolic blood pressure. LV EDD was positively related to the S:D ratio raw score, while LV mass z-score was positively related to the S:D ratio raw score and z-score. LV flow propagation velocity was negatively associated with S:D ratio raw score and z-score.

Table A3.

Univariate linear regressions of the S:D ratio raw value and z-score at the lateral mitral annulus among DCM patients at baseline

		Outcome = Raw Ratio		Outcome = Ratio z Score
Covariate	n	Estimate	P value	Estimate	P value
Age at echocardiography, yr	103	−0.005	0.133	−0.013	0.581
Gender (male vs. female)	103	0.085	0.022	0.670	0.014
SBP-for-age z-score	102	0.0214	0.162	0.1344	0.216
DBP-for-age z-score	102	0.0410	0.009	0.2704	0.016
End diastolic volume, ml	103	0.0000	0.948	0.0009	0.547
EDV-for-BSA z-score	103	0.0151	0.010	0.1046	0.012
End systolic volume, ml	103	0.0001	0.659	0.0018	0.324
ESV-for-BSA z-score	102	0.0220	0.004	0.1450	0.007
Ventricular mass, g	103	−0.0001	0.716	0.0007	0.734
Mass-for-BSA z-score	103	0.0331	<.001	0.2188	<.001
Mass-to-Volume ratio	103	0.0110	0.933	−0.0033	0.997
MV-for-age z-score	103	0.0014	0.934	−0.0048	0.967
Sphericity index	103	0.1951	0.312	1.1943	0.383
Sphericity-for-age z-score	103	0.0111	0.415	0.0816	0.397
Shortening fraction, %	103	−0.0064	0.027	−0.0426	0.040
Shortening fraction-for-age z-score	103	−0.0248	0.009	−0.1559	0.022
Velocity of fiber shortening, circ/s	102	−0.1248	0.164	−0.7754	0.223
VCFc-for-age z-score	102	−0.0162	0.056	−0.0932	0.120
Ejection fraction, %	103	−0.0038	0.023	−0.0266	0.025
EF-for-age z-score	103	−0.0136	0.071	−0.1027	0.055
End-systolic fiber stress, g/cm2	102	0.0008	0.085	0.0064	0.051
Peak early velocity, m/s	82	−0.0944	0.334	−0.7791	0.269
Peak early velocity-for-age z-score	82	−0.0133	0.449	−0.1281	0.313
Peak atrial velocity, m/s	82	0.2740	0.064	1.5772	0.142
Peak A-wave velocity-for-age z-score	82	0.0249	0.104	0.1608	0.146
Early deceleration time, ms	82	0.0000	0.972	0.0007	0.754
Deceleration time-for-age z-score	82	0.0031	0.772	0.0249	0.746
Left ventricular flow propagation velocity, cm/s	100	−0.0018	0.073	−0.0130	0.068
LV flow propagation-for-age z-score	100	−0.0385	0.098	−0.2897	0.078
Duration of flow reversal during atrial systole, ms	100	−0.0001	0.844	−0.0004	0.846
Pulmonary vein A-wave duration-for-age z-score	100	0.0011	0.883	0.0022	0.967
R-R interval, ms	103	0.0000	0.906	0.0004	0.689
Heart rate, beats/min	103	0.0002	0.885	−0.0043	0.654

DCM, dilated cardiomyopathy.

Table 5.

Univariate linear regressions of the S:D ratio raw value and z-score at the lateral mitral annulus among dilated cardiomyopathy patients at baseline

		Outcome = Raw Ratio		Outcome = Ratio z-Score
Covariate	n	Estimate	P value	Estimate	P value
Gender (male vs. female)	103	0.085	0.022	0.670	0.014
DBP-for-age z-score	102	0.0410	0.009	0.2704	0.016
EDV-for-BSA z-score	103	0.0151	0.010	0.1046	0.012
ESV-for-BSA z-score	102	0.0220	0.004	0.1450	0.007
Mass-for-BSA z-score	103	0.0331	<.001	0.2188	<.001
VCFc-for-age z-score	102	−0.0162	0.056	−0.0932	0.120
Ejection fraction, %	103	−0.0038	0.023	−0.0266	0.025
End-systolic fiber stress, g/cm2	102	0.0008	0.085	0.0064	0.051
Peak atrial velocity, m/s	82	0.2740	0.064	1.5772	0.142
Inflow propagation-for-age z-score	100	−0.0385	0.098	−0.2897	0.078

To further assess these relationships and to incorporate data from serial visits we used a mixed-effects model with piecewise linear terms for the covariate to investigate if the slopes were different before and after the cut-point suggested by the data. Results with P < 0.1 are presented in Table 6. Full results are presented in the appendix. We found that LV mass z-score is significantly associated with S:D ratio z-score when mass z-score is ≥5, but the slope is not different from zero when LV mass z-score is <5. Other variables were not significantly associated with S:D ratio z-score at the mitral lateral annulus.

Table 6.

Mixed model regressions of the S:D ratio z-score at the lateral mitral annulus in dilated cardiomyopathy patients, serial data

		Outcome = Raw Ratio		Outcome = Ratio z-Score
Covariate	n	Estimate*	P value	Estimate*	P value
Gender (male vs. female)	224	0.0512	0.084	0.3794	0.069
DBP-for-age z-score	222	0.0228	0.031	0.1514	0.052
Functional status (Ross/NYHA class I, II vs. III, IV)	224	0.0854	0.041	0.6496	0.039
Mass-for-BSA z-score	223	0.0175	0.008	0.1136	0.020
Inflow propagation-for-age z-score	218	−0.0254	0.048	−0.2017	0.035
Heart rate, beats/min	224	−0.0011	0.229	−0.0119	0.070

n Is the total number of observations, not subjects. NYHA, New York Health Association.

^*Slope for continuous predictors and mean group difference for categorical predictors.

Since scatter plots between the S:D coupling ratio and predictor variables were curvilinear, a quadratic model was also applied to the data. EDD z-score, ESD z- score, ventricular mass z-score, and velocity of fiber shortening z-score showed a significant quadratic relation with the S:D coupling ratio (see Fig. A1). With regard to LV size (EDD z-score and ESD z-score), the model indicated that S:D coupling ratio does not change with LV size until the dimension z-score exceeds ∼5 SD.

Fig. A1.

Systolic:diastolic (S:D) ratio z-score at the lateral mitral annulus vs. parameters of LV size and function. EDDz, end-diastolic dimension z-score; ESDz, end-systolic dimension z-score; VCFCz, velocity of circumferential fiber shortening z-score.

SEPTAL ANNULUS.

Interactions between the S:D coupling ratio and other functional predictors were stronger at the septal than at the lateral mitral annulus. Linear relationships for the baseline visit with P < 0.1 are shown in Table 7 and a mixed-effects model with piecewise linear terms for the covariate incorporating data from all visits are shown in Table 8. The results from all tested associations are presented in the appendix (see Tables A5 and A6). A higher S:D ratio raw score at the septal annulus was associated with larger ventricular dimensions and mass, lower shortening fraction, lower ejection fraction, lower LV flow propagation velocity, shorter duration of flow reversal during atrial systole, and higher heart rate. A piecewise linear model showed that the peak A-wave velocity-for-age z score is significantly associated with septal S:D ratio z-score when peak A-wave velocity z-score is above ≥−0.5 (slope = 0.31, P = 0.018). Likewise, the E/A ratio for age z-score was negatively associated with septal S:D ratio z-score when E/A for age z-score was below 0.5 (slope = −0.55, P = 0.007), but the relationship was not significant when E/A for age z-score is ≥0.5.

Table 7.

Univariate linear regressions of the S:D ratio raw value and z-score at the septal annulus among dilated cardiomyopathy patients at baseline

		Outcome = Raw Ratio		Outcome = Ratio z-Score
Covariate	n	Estimate*	P value	Estimate*	P value
Age at echocardiography, yr	94	0.0015	0.563	0.0631	0.016
SBP-for-age z-score	93	−0.0200	0.078	−0.2352	0.045
End diastolic volume, ml	94	0.0002	0.162	0.0043	0.008
EDV-for-BSA z-score	94	0.0082	0.060	0.0730	0.110
End systolic volume, ml	94	0.0004	0.059	0.0061	0.004
ESV-for-BSA z-score	93	0.0118	0.039	0.0922	0.122
Ventricular mass, g	94	0.0004	0.085	0.0074	0.002
Mass-for-BSA z-score	94	0.0164	0.015	0.1315	0.062
Ejection fraction, %	94	−0.0031	0.017	−0.0303	0.024

^*Slope for continuous predictors and mean group difference for categorical predictors.

Table 8.

The S:D coupling ratio mixed model regressions of the S:D ratio raw value and z-score at the septal annulus in dilated cardiomyopathy patients

		Outcome = Raw Ratio		Outcome = Ratio z-Score
Covariate	n	Estimate*	P value	Estimate*	P value
Age at baseline echocardiogram, yr	212	−0.0002	0.920	0.0437	0.053
DBP-for-age z-score	210	−0.0176	0.032	−0.2013	0.017
EDV-for-BSA z-score	212	0.0096	0.011	0.0893	0.021
ESV-for-BSA z-score	209	0.0133	0.005	0.1181	0.016
Mass-for-BSA z-score	212	0.0169	0.002	0.1489	0.009
VCFc-for-age z –score	210	−0.0092	0.044	−0.0779	0.096
Ejection fraction, %	211	−0.0030	0.002	−0.0319	0.001
Peak atrial (A)-wave velocity-for-age z-score	166	0.0144	0.089	0.1138	0.194
Inflow propagation-for-age z-score	208	−0.0336	<.001	−0.3538	<.001
Duration of pulmonary vein A-wave flow reversal-for-age z-score	208	−0.0093	0.012	−0.0995	0.009
Heart rate, beats/min	211	0.0017	0.019	0.0092	0.204

n Is the total number of observations, not subjects.

^*Slope for continuous predictors and mean group difference for categorical predictors.

Since scatter plots between the S:D coupling ratio and A and E/A ratio were curvilinear, a quadratic model was also applied to the data, with both demonstrating a significant quadratic term.

Comparison of Associations Between S:D Coupling Ratio and Other Parameters of LV Function in DCM vs. Healthy Controls

There were a similar number of associations between the S:D coupling ratio and LV size and function in the DCM patient dataset compared with normal controls for the septal annulus (see Table A7). In both groups, the nonlinear analyses of the component and ratio measures suggest that there is less coupling at the extremes of velocities (lowest values in the DCM patients and highest values in healthy controls).

DISCUSSION

Systolic function contributes to diastolic filling, which in turn influences subsequent systolic contraction through the Frank-Starling effect, even though the coupling between them is not routinely assessed. However, in the presence of myocardial dysfunction, S:D coupling may itself be impaired.

In this exploratory study of a relatively new concept (Fig. 4), we used tissue velocities to investigate the association between systolic and diastolic function in a large group of prospectively collected DCM patients and compared results to a large group of normal controls. The major findings of our study are that:

Fig. 4.

Diagram representing the concepts explored in this study and its main results. On the left a normal LV is depicted with normal systolic and diastolic motion, as assessed by tissue Doppler velocities S' and E', respectively. The solid lines depict the LV in diastole; the dotted lines represent the LV in systole. Based on literature, factors such as the protein titin act as a spring to couple systolic and diastolic function. When the spring's recoil characteristics are intact, depicted as the tightly coiled spring on the left, the ratio between systolic and diastolic velocities (depicted by up and down arrows) is preserved indicating normal systolic-diastolic coupling. In this circumstance, diastolic velocities are relatively close to systolic velocities and consequently the ratio between them (S'/E'; systolic: diastolic coupling index) relatively low. In dilated cardiomyopathy (right), the heart remodels, becomes dysfunctional and with a putative loss of the spring's recoil characteristics (represented by the “stretched” spring on the right). Thus, although S' is lower due to systolic dysfunction, the diastolic coupling is itself impaired resulting in a lower E' for a given S'. This leads to progressive loss of systolic-diastolic coupling and an increasing S'/E' ratio.

1) Systolic and diastolic tissue velocities are positively related.

2) Based on results at the septal annulus, this relationship is impaired in children with DCM compared with that in normal controls.

3) The magnitude of the S:D coupling ratio, a putative index of systolic-diastolic coupling, is related to LV size and global systolic function.

4) Assessment of tissue velocities at the septal annulus appears to be more useful than the lateral annulus when assessing systolic-diastolic coupling.

Relationships Between Systolic and Diastolic Velocities in Normal Children

LV restoring forces, lengthening load, and myocardial relaxation are independent determinants of E' (26). In turn, tension produced in systole, through myocardial contraction and torsion, influences both lengthening load and restoring forces (15).

The relationship between systolic contraction and diastolic filling is through recoil and suction forces (11). The central role of ventricular end-systolic volume in determining coupling is associated with the mechanical action of the heart where release of energy built up in systole creates the early diastolic restoring forces and suction that contribute to rapid early diastolic filling (30). This component of S:D coupling is particularly important in augmenting cardiac output during exercise (11, 24, 31, 32, 36). In disease, the loss of S:D coupling may be particularly detrimental, as the mitral inflow propagation velocity, which reflects in part early diastolic suction, is independently linked to mortality in the same DCM population we studied (23). Likewise, diastole impacts systolic function, primarily through the preload effect of end-diastolic fiber length. Among the factors that determine end-diastolic fiber length, a higher ventricular mass:volume ratio is associated with decreased coupling (higher S:D coupling ratio) (9, 20, 22, 29, 37).

Association Between Systolic and Diastolic Function in Children with DCM

As we previously found in a smaller study (22), systolic and diastolic velocities were related to each other and associated with functional parameters in children with DCM, but coupling was reduced, as reflected by a higher S:D coupling ratio, in DCM patients. As depicted in Fig. 4, lower coupling would lead to a lower E' for a given S' and therefore a higher S:D coupling ratio. The reduction in S:D coupling in DCM patients is likely multifactorial, stemming from both systolic and diastolic dysfunction and from the impaired coupling itself, as suggested by our data. Among the parameters we investigated, LV remodeling is a key candidate to explain decreased coupling, as detailed previously for healthy controls. Our current results indicate that with increasing LV EDD there was a decrease in the strength of S:D coupling. Indeed, studies in patients at the time of cardiopulmonary bypass have shown that improvements in LV function correlate with improvements in LV relaxation and end-systolic volume and are associated with increased early diastolic intraventricular pressure gradients, consistent with improved LV suction and elastic recoil (12). It is also possible that impaired coupling worsens or contributes to LV remodeling and the S:D ratio may in this case be an early marker for LV remodeling in DCM. This requires further investigation. In addition, impaired torsion (6, 27), diastolic dysfunction in various forms (14), and mechanical dyssynchrony (13, 14) may contribute to reduced coupling in DCM patients.

The molecular mediator of S:D coupling is thought to be titin (16). This protein acts as a spring that develops potential energy during systole that is released in early diastole. At the same time, titin protects against excessive sarcomere elongation in late diastole, helping to maintain ventricular dimensions and promote optimal coupling. While our results suggest an important role of LV remodeling in disruption of S:D coupling in DCM patients, the molecular mechanism may be alterations in the titin protein, including shift in isoforms and genetic mutations, among others (2, 18). This requires further study.

Our data suggest that the S:D ratio assessment at the septal annulus may be more useful as an indicator of S:D coupling than the lateral mitral annulus. Likewise, S:D coupling at the septal rather than lateral mitral annulus was more strongly associated with other LV functional parameters. This result was somewhat surprising in that we expected velocities at the lateral mitral annulus to be more strongly associated to LV parameters because the lateral annulus in contrast to the septum is exclusively a LV structure. While the explanation for this result is not obvious, we postulate that septal velocities reflect a combination of right ventricle and LV function and perhaps are more indicative of the sum of RV and LV coupling or perhaps reflect a component of ventricular-ventricular interactions that may be important in DCM (10). Interestingly, lower septal peak systolic TDI velocity z-score and not lateral peak systolic TDI velocity z-score was recently found to be an independent predictor of disease progression in this population (23). Varying activation such as in bundle branch block and other regional wall motion abnormalities may also affect septal vs. lateral velocities.

Clinical Implications

Various echocardiographic parameters have been proposed as associated with outcomes in pediatric DCM. Among these, LV size is important and our results suggest that this may in part be due to loss of S:D coupling (1, 23). Therefore, the S:D coupling ratio warrants further study in relation to clinical outcomes such as brain natriuretic peptide levels and exercise capacity. As diastolic dysfunction is challenging to assess in pediatric DCM (8), this ratio may potentially differentiate between “intrinsic” myocardial diastolic abnormalities and those “secondary” to systolic dysfunction. Treatment of systolic dysfunction would presumably help improve the latter. As this exploratory study did not investigate these concepts, these remain purely conjectures worthy of further study. In this context, it would be interesting to study the relationship in adults with diastolic dysfunction with preserved ejection fraction, an uncommon entity in children (31). Perhaps most importantly, the prognostic value of S:D coupling in clinical practice in the pediatric DCM population requires further study.

Limitations

We did not study the clinical implications of S:D coupling and how its assessment may contribute to functional assessment in general or to prognostication, independent of systolic or diastolic functional assessment alone. Myocardial tissue velocities are segmental in nature and thus velocities sampled at any given point may not reflect the global association between systolic and diastolic function. Radial and rotational motions may also affect systolic-diastolic coupling but were not studied and the S':E' ratio accounts only for longitudinal motion. Likewise, this ratio only assesses early diastole and late diastolic function is not assessed. Tissue velocities are influenced by translational motion that does not reflect S:D coupling. As disease severity increases, tissue Doppler velocities may be harder to measure, with more variability in measurements, which may affect assessment of coupling. Nonetheless, in the VVV study tissue Doppler had acceptable reproducibility in the very same population (5). We did not account for the effect of medications on S:D coupling. In this regard, β-blockers may affect heart rate and thus S:D coupling. Rotational mechanics were not available in the database, and twist-untwist and particularly the rate of twist-untwist may provide a more physiological and/or comprehensive reflection of S:D coupling. Nonetheless, tissue velocities are simpler to obtain and potentially more applicable in routine practice.

Conclusion

In conclusion, systolic and diastolic tissue velocities are positively correlated, reflecting S:D functional coupling. S:D coupling is related to multiple indices of LV size and function but is weaker in DCM as reflected by higher S:D ratios vs. controls, fewer associations between the S:D ratio and parameters of ventricular function and weaker associations between S' and E' with LV remodeling. The S:D coupling ratio, especially when measured at the septal annulus, may be a useful index of coupling performance and warrants further study in relation to outcomes and patient management in children with DCM.

GRANTS

This work was supported by National Heart, Lung, and Blood Institute U01 Grants HL-068269, HL-068270, HL-068279, HL-068281, HL-068285, HL-068292, HL-068290, and HL-068288.

DISCLAIMERS

The contents of this publication are solely the responsibility of the authors and do not necessarily represent the official views of the National Heart, Lung, and Blood Institute.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

M.K.F., R.M., L.M.-R., H.T.H., A.N., K.F., K.M.M., K.A., C.C., L.A.S., and S.D.C. conception and design of research; M.K.F., R.M., M.L., L.M.-R., H.T.H., A.N., K.F., K.M.M., K.A., C.C., L.A.S., and S.D.C. interpreted results of experiments; M.K.F., M.L., L.A.S., and S.D.C. prepared figures; M.K.F. drafted manuscript; M.K.F., R.M., M.L., L.M.-R., H.T.H., A.N., K.F., K.M.M., K.A., C.C., L.A.S., and S.D.C. edited and revised manuscript; M.K.F., R.M., M.L., L.M.-R., H.T.H., A.N., K.F., K.M.M., K.A., C.C., L.A.S., and S.D.C. approved final version of manuscript; R.M., M.L., L.A.S., and S.D.C. analyzed data.

APPENDIX

Figure A1 depicts the systolic:diastolic (S:D) ratio z-score at the lateral mitral annulus vs. parameters of LV size and function. Tables A1, A2, A3, A4, A5, and A6 provide univariate regressions between the S:D ratio z-score and parameters of ventricular function. Table A7 provides the model significance for the S:D ratio at the lateral mitral and septal annulus.

Table A4.

Univariate mixed model regressions of the S:D ratio z-score at the lateral mitral annulus in dilated cardiomyopathy patients

		Outcome = Raw Ratio		Outcome = Ratio z-Score
Covariate	n	Estimate*	P value	Estimate*	P value
Age at baseline echocardiogram, yr	224	−0.0004	0.877	0.0196	0.289
Gender (male-for-female)	224	0.0512	0.084	0.3794	0.069
SBP-for-age z-score	223	0.0031	0.758	0.0162	0.828
DBP-for-age z-score	222	0.0228	0.031	0.1514	0.052
Functional status (Ross/NYHA class I, II-for-III, IV)	224	0.0854	0.041	0.6496	0.039
End diastolic volume, ml	223	0.0000	0.858	0.0012	0.315
EDV-for-BSA z-score	223	0.0068	0.153	0.0437	0.205
End systolic volume, ml	222	0.0001	0.595	0.0018	0.234
ESV-for-BSA z-score	220	0.0095	0.104	0.0579	0.173
Ventricular mass, g	223	0.0001	0.813	0.0019	0.263
Mass-for-BSA z-score	223	0.0175	0.008	0.1136	0.020
Mass-to-volume ratio	223	0.1004	0.248	0.7641	0.235
MV-for-age z-score	223	0.0120	0.271	0.0888	0.272
Cardiac index, l/min/m2 [5/6*area*length]	221	−0.0043	0.707	−0.0690	0.422
Sphericity index	223	−0.0403	0.768	−0.5136	0.606
Sphericity-for-age z-score	223	−0.0029	0.762	−0.0228	0.742
Eccentricity index	223	0.2800	0.666	2.9317	0.535
Shortening fraction, % [2D]	223	−0.0023	0.235	−0.0162	0.250
Shortening fraction-for-age z-score	223	−0.0092	0.152	−0.0584	0.214
Velocity of fiber shortening, circ/s	221	−0.0246	0.677	−0.1434	0.740
VCFc-for-age z-score	221	−0.0042	0.463	−0.0193	0.643
Ejection fraction, %	222	−0.0017	0.147	−0.0124	0.140
EF-for-age z-score	222	−0.0065	0.212	−0.0543	0.153
End-systolic fiber stress, g/cm2	220	0.0002	0.600	0.0017	0.434
Peak early velocity, m/s	173	−0.0696	0.248	−0.6087	0.172
Peak early velocity-for-age z-score	173	−0.0112	0.299	−0.1095	0.171
Peak atrial velocity, m/s	173	0.0629	0.468	0.1257	0.844
Peak A-wave velocity-for-age z-score	173	0.0053	0.555	0.0174	0.791
Early deceleration time, ms	173	0.0000	0.892	0.0001	0.924
Deceleration time-for-age z-score	173	−0.0007	0.911	−0.0064	0.893
Early velocity/atrial velocity (E/A)	173	−0.0195	0.139	−0.1313	0.179
E/A-for-age z-score	173	−0.0115	0.168	−0.0887	0.153
Inflow propagation velocity, cm/s	218	−0.0011	0.045	−0.0087	0.039
Inflow propagation-for-age z-score	218	−0.0254	0.048	−0.2017	0.035
Myocardial performance index	212	0.0667	0.289	0.5080	0.277
Duration of flow reversal during atrial systole, ms	218	0.0000	0.959	−0.0001	0.924
Pulmonary vein A-wave duration-for-age z-score	218	0.0007	0.876	0.0003	0.992
R-R interval, ms	224	0.0001	0.336	0.0010	0.132
Heart rate, beats/min	224	−0.0011	0.229	−0.0119	0.070

^*Slope for continuous predictors and mean group difference for categorical predictors.

Table A5.

Univariate linear regressions of the S:D ratio raw value and z-score at the septal annulus among DCM patients at baseline

		Outcome = Raw Ratio		Outcome = Ratio z-Score
Covariate	n	Estimate*	P value	Estimate*	P value
Age at echocardiography, yr	94	0.0015	0.563	0.0631	0.016
Gender (male vs. female)		0.0186	0.504	0.1684	0.562
SBP-for-age z-score	93	−0.0200	0.078	−0.2352	0.045
DBP-for-age z-score	93	−0.0062	0.604	−0.1044	0.399
End diastolic volume, ml	94	0.0002	0.162	0.0043	0.008
EDV-for-BSA z-score	94	0.0082	0.060	0.0730	0.110
End systolic volume, ml	94	0.0004	0.059	0.0061	0.004
ESV-for-BSA z-score	93	0.0118	0.039	0.0922	0.122
Ventricular mass, g	94	0.0004	0.085	0.0074	0.002
Mass-for-BSA z-score	94	0.0164	0.015	0.1315	0.062
Mass-to-Volume ratio	94	−0.0030	0.976	0.0168	0.987
MV-for-age z-score	94	0.0002	0.991	−0.0013	0.992
Sphericity index	94	0.1501	0.288	1.0254	0.487
Sphericity-for-age z-score	94	0.0135	0.176	0.1346	0.196
Shortening fraction, %	94	−0.0030	0.192	−0.0238	0.315
Shortening fraction-for-age z-score	94	−0.0096	0.196	−0.0568	0.465
Velocity of fiber shortening, circ/s	93	−0.1001	0.145	−0.8537	0.234
VCFc-for-age z-score	93	−0.0092	0.159	−0.0546	0.421
Ejection fraction, %	94	−0.0031	0.017	−0.0303	0.024
EF-for-age z-score	94	−0.0146	0.012	−0.1613	0.008
End-systolic fiber stress, g/cm2	93	0.0001	0.773	0.0020	0.577
Peak early velocity, m/s	78	0.0823	0.281	0.7803	0.326
Peak early velocity-for-age z-score	78	0.0150	0.278	0.1158	0.421
Peak atrial velocity, m/s	78	0.1894	0.116	1.1247	0.372
Peak A-wave velocity-for-age z-score	78	0.0198	0.106	0.1572	0.218
Early deceleration time, ms	78	−0.0001	0.588	−0.0004	0.880
Deceleration time-for-age z-score	78	−0.0054	0.523	−0.0576	0.509
Inflow propagation velocity, cm/s	92	−0.0010	0.169	−0.0105	0.171
Inflow propagation-for-age z-score	92	−0.0243	0.156	−0.2718	0.124
Duration of flow reversal during atrial systole, ms	92	−0.0001	0.674	−0.0009	0.684
Pulmonary vein A-wave duration-for-age z-score	92	−0.0020	0.716	−0.0318	0.574
R-R interval, ms	94	0.0000	0.822	0.0008	0.483
Heart rate, beats/min	94	0.0002	0.871	−0.0094	0.374

^*Slope for continuous predictors and mean group difference for categorical predictors.

Table A6.

Univariate mixed model regressions of the S:D ratio raw value and z-score at the septal annulus in dilated cardiomyopathy patients

		Outcome = Raw Ratio		Outcome = Ratio z-Score
Covariate	n	Estimate*	P value	Estimate*	P value
Age at baseline echocardiogram, yr	212	−0.0002	0.920	0.0437	0.053
Gender (male-for-female)	212	0.0269	0.285	0.2679	0.296
SBP-for-age z-score	211	−0.0060	0.461	−0.0891	0.289
DBP-for-age z-score	210	−0.0176	0.032	−0.2013	0.017
Functional status (Ross/NYHA class I, II-for-III, IV)	212	−0.0503	0.128	−0.5033	0.139
End diastolic volume, ml	212	0.0002	0.111	0.0042	0.003
EDV-for-BSA z-score	212	0.0096	0.011	0.0893	0.021
End systolic volume, ml	211	0.0004	0.022	0.0061	<.001
ESV-for-BSA z-score	209	0.0133	0.005	0.1181	0.016
Ventricular mass, g	212	0.0003	0.112	0.0065	0.002
Mass-for-BSA z-score	212	0.0169	0.002	0.1489	0.009
Mass-to-volume ratio	212	0.0300	0.684	0.3466	0.647
MV-for-age z-score	212	0.0031	0.737	0.0315	0.743
Cardiac index, l/min/m2	210	0.0001	0.989	−0.0663	0.480
Sphericity index	212	0.1444	0.210	1.0753	0.362
Sphericity-for-age z-score	212	0.0104	0.192	0.1059	0.195
Eccentricity index	212	−0.5826	0.284	−4.1792	0.453
Shortening fraction, %	212	−0.0031	0.047	−0.0320	0.048
Shortening fraction-for-age z-score	212	−0.0109	0.040	−0.0982	0.071
Velocity of fiber shortening, circ/s	210	−0.0918	0.050	−0.9139	0.058
VCFc-for-age z-score	210	−0.0092	0.044	−0.0779	0.096
Ejection fraction, %	211	−0.0030	0.002	−0.0319	0.001
EF-for-age z-score	211	−0.0131	0.003	−0.1513	<.001
End-systolic fiber stress, g/cm2	209	0.0001	0.720	0.0016	0.504
Peak early velocity, m/s	166	−0.0040	0.944	−0.1806	0.762
Peak early velocity-for-age z-score	166	−0.0003	0.975	−0.0472	0.660
Peak atrial velocity, m/s	166	0.1505	0.080	0.9013	0.309
Peak A-wave velocity-for-age z-score	166	0.0144	0.089	0.1138	0.194
Early deceleration time, ms	166	−0.0003	0.142	−0.0021	0.253
Deceleration time-for-age z-score	166	−0.0093	0.126	−0.1011	0.107
Early velocity/atrial velocity	166	−0.0158	0.329	−0.0978	0.558
E/A-for-age z-score	166	−0.0093	0.364	−0.0918	0.386
Left ventricular flow propagation velocity, cm/s	208	−0.0015	<.001	−0.0154	<.001
LV flow propagation-for-age z-score	208	−0.0336	<.001	−0.3538	<.001
Myocardial performance index	202	−0.0475	0.311	−0.5026	0.303
Duration of flow reversal during atrial systole, ms	208	−0.0004	0.010	−0.0038	0.012
Pulmonary vein A-wave duration-for-age z-score	208	−0.0093	0.012	−0.0995	0.009
R-R interval, ms	211	−0.0001	0.043	−0.0008	0.262
Heart rate, beats/min	211	0.0017	0.019	0.0092	0.204

^*Slope for continuous predictors and mean group difference for categorical predictors.

Table A7.

Model significance for the S:D ratio z-score in dilated cardiomyopathy patients and normal controls at the lateral mitral and septal annulus

	Lateral Mitral Annulus		Septal Annulus
Covariate	DCM	Normal	DCM	Normal
Age at baseline echocardiogram, yr		+		+
Gender (male vs. female)		+		+
SBP-for-age z-score		+		+
DBP-for-age z-score	+		–
End-diastolic short axis dimension, cm		+	+	+
EDD-for-for-BSA z-score		–
End-systolic short axis dimension, cm		+	+	+
ESD-for–for-BSA z-score		–	+
End diastolic volume, ml		+	+	+
EDV for BSA\-z-score		–	+
End systolic volume, ml		+	+	+
ESV for BSA z-score		–	+	–
Ventricular mass, g		+	+	+
Mass-for-BSA z-score	+		+
Mass-to-volume ratio		+
MV-for-age z-score		+		+
Sphericity index		–		–
Sphericity-age-z		–		–
Shortening fraction, %			–
SF-for-age z-score
Velocity of fiber shortening, circ/s		+
VCFc-for-age z-score		+
Ejection fraction, %			–
EF-for-age z-score			–
Peak mitral early velocity (E), m/s		–		–
Peak mitral E velocity z-score		–		–
Peak atrial velocity
Peak A velocity z-score
E-wave deceleration time, ms
E-wave deceleration time z-score
Inflow propagation velocity, cm/s	–
Inflow propagation z-score	–
Duration of flow reversal during atrial systole, ms			–
Pulmonary vein A-wave duration z-score			–
R-R interval, ms		–
Heart rate, beats/min		+

+ or–Indicates P < 0.05.