Association of Left Atrial Myocardial Function With Left Ventricular Diastolic Dysfunction in Subjects With Preserved Systolic Function: A Strain Rate Imaging Study

Abstract

Background:

The purpose of this study was to investigate the relationship between left atrial (LA) myocardial function and left ventricular (LV) diastolic dysfunction in subjects with preserved LV ejection fraction (LVEF).

Methods:

The study included a group of 118 hypertensive patients and normal subjects. LV diastolic dysfunction was classified into 4 groups: none, mild, moderate, and severe. Peak strain rates in systole (S‐Sr), early diastole (E‐Sr), and late diastole (A‐Sr) were obtained from Doppler‐derived strain rate imaging to evaluate LA myocardial deformation.

Results:

No significant difference in LA dimension was observed in subjects with different degrees of LV diastolic dysfunction, although LA myocardial strain rate parameters were all significantly different across the 4 groups (all with P < 0.001). Compared with patients of normal diastolic function, the mild diastolic dysfunction group had significantly lower E‐Sr (0.62 ± 0.18 s⁻¹ vs 1.20 ± 0.38 s⁻¹, P < 0.001) and S‐Sr (0.78 ± 0.16 s⁻¹ vs 0.94 ± 0.22 s⁻¹, P < 0.001) but increased A‐Sr (1.14 ± 0.29 s⁻¹ vs 1.00 ± 0.23 s⁻¹, P = 0.05).

Conclusions:

By using strain rate imaging, significant changes of LA deformation in response to different stages of LV diastolic dysfunction were detected in subjects with preserved LVEF. Quantification of LA myocardial function rather than LA size may have the potential to predict early LV diastolic dysfunction in subjects with preserved LVEF. Copyright © 2010 Wiley Periodicals, Inc.

The authors have no funding, financial relationships, or conflicts of interest to disclose.

Introduction

Diastolic dysfunction plays an important role in the pathophysiology of heart failure, especially in patients with preserved systolic function.1, 2 It has been demonstrated that different degrees of left ventricular (LV) diastolic dysfunction are related to long‐term mortality compared with normal patients.3 Contributing up to 30% of total LV stroke volume in normal individuals, the left atrium (LA) is of critical importance to LV diastolic dysfunction.4 Assessment of global LA function parameters such as LA size and volume has been established as a prognostic marker for LV diastolic dysfunction.5, 6 However, there is a paucity of literature regarding the association of LA myocardial function with LV diastolic dysfunction.

In recent years, tissue Doppler imaging (TDI)‐derived strain rate imaging has been used for noninvasive evaluation of LA myocardial function.7, 8, 9 With strain rate imaging, LA segmental myocardial function can be analyzed quantitatively. The present study was designed to investigate the relationship between LA myocardial function and LV diastolic dysfunction by using strain rate imaging.

Methods

Subjects

A total of 95 hypertensive patients and 29 normal subjects were enrolled in the present study. Subjects were recruited from the general clinic of the First Affiliated Hospital of Guangxi Medical University. All the enrolled patients were diagnosed with essential hypertension and had been treated with antihypertensive drugs regularly for ≥1 year. We excluded patients with secondary hypertension or underlying systemic diseases such as diabetes mellitus. Based on standard echocardiography, subjects showing significant abnormalities such as valvular diseases, hypertrophic cardiomyopathy, LV systolic dysfunction, or pericardial effusion were excluded. The final study population consisted of 91 hypertensive patients and 27 normal subjects. Electrocardiograms were interpreted by the same experienced reader (F. Zhang) masked to other clinical information.

The present study was conducted in accordance with the ethical standards stated in the Declaration of Helsinki with informed consent obtained.

Blood Pressure Measurements

After 5 to 10 minutes of rest in a sitting position, blood pressure was measured with a mercury sphygmomanometer using the appropriate cuff size. Three measurements were taken ≥10 minutes apart and were averaged to determine the systolic and diastolic blood pressure.

Echocardiography

Echocardiographic examination was performed according to the recommendations of the American Society of Echocardiography10 using a Philips iE33 ultrasound system (Philips Healthcare, the Netherlands) with a 2.0‐MHz transducer in the partial left lateral decubitus position on each subject. Three cardiac cycles were stored in cineloop format for offline analysis. Isochronous electrocardiographic recording was obtained during the echocardiographic measurements. Data were analyzed by 2 independent experienced observers blinded to the clinical information.

The left atrial diameter in end‐systole, left ventricular end‐systolic and end‐diastolic diameter, and intraventricular and posterior wall thickness in end‐diastole were obtained. Left ventricular ejection fraction (LVEF) was calculated using Simpson's method. Left ventricular mass was estimated by the Devereux formula.11 Left ventricular mass index was then calculated as left ventricular mass/body surface area. The LA volume was measured by using the biplane area‐length method.12 The indexed volume of LA (LAVI) was then calculated as LA volume/body surface area. Mitral inflow velocities were evaluated with a 1‐ to 2‐mm sample volume placed at the mitral valve tip by pulse‐wave Doppler in the apical 4‐chamber view. Diastolic peak early (E) and peak late (A) transmitral flow velocity, peak E to peak A velocities (E/A), and deceleration time of peak E velocity (EDT) were measured from the average of 3 beats.

Tissue Doppler Imaging and Strain Rate Imaging

After the conventional standard echocardiography examination, the echocardiographic settings were changed to the application of TDI with the patient in the same position. The color TDI was performed with a frame rate >100 frames/second. Mitral annulus velocities were measured by TDI using the pulse‐wave Doppler mode. Systolic (Sm), early diastolic (Em), and late diastolic (Am) annular velocity were obtained in the apical 4‐chamber view with a 2‐ to 5‐mm sample volume placed at the septal corner of the mitral annulus. The E/Em was calculated for assessment of the LV diastolic dysfunction.

Digitally stored loops of TDI were also used for offline analysis by the proper software (Strain Quantification; Philips Healthcare, the Netherlands). The recorded wall was positioned in the center of the sector to minimize artifactual data and realigned so that the direction of motion interrogated was as near as possible to parallel to the direction of the insonating beam. For tissue Doppler atrial longitudinal velocities and strain rates, analysis was performed from the apical 4‐chamber view and 2‐chamber view. The sample was positioned between the middle and the inferior edge of the interarterial septum, of the left atrial lateral wall, of the anterior wall, and of the posterior wall. Analysis was performed on the 4 LA walls by measuring 2 segments of each wall. The peak strain rate was obtained in each segment in systole (S‐Sr), early diastole (E‐Sr), and late diastole (A‐Sr). As no regional differences in strain rate measurements were seen in the LA walls, the average values from 8 segments of the 4 LA walls were used to simplify the analysis. Measurements were excluded if a smooth SR curve could not be obtained or if the angles of interrogation exceeded 30°.

Classification of LV Diastolic Dysfunction

By using conventional echocardiography and TDI, the subjects were divided into the following 4 groups: (1) the no LV diastolic dysfunction group, consisting of 36 subjects who had normal E/A ratios (0.75 < E/A < 2) with EDT that was not reduced (>140 ms) and E/Em ratio that was not increased (<10); (2) the mild LV diastolic dysfunction group, consisting of 20 subjects who had mitral inflow patterns with reduced E/A ratios (≤0.75) and E/Em ratio that was not increased; (3) the moderate LV diastolic dysfunction group, consisting of 58 subjects with normal E/A ratios, increased E/Em ratios (≥10), and EDT that was not reduced; and (4) the severe LV diastolic dysfunction group, consisting of 4 subjects with increased E/A ratios (>2), increased E/Em ratios (≥10), and reduced EDT (<140 ms).13

Statistical Analysis

Statistical analysis was performed using the SPSS 13.0 statistical software package (SPSS Inc., Chicago, IL). A Kruskal‐Wallis test was performed to evaluate the difference between groups of continuous variables, and then a Mann‐Whitney U test was used to calculate difference between groups in case of a significant overall difference. The likelihood ratio χ ² test was used to evaluate the difference between groups of categorical variables. Association between LA myocardial deformation and E/Em ratio was assessed by Pearson correlation coefficient. Interobserver and intraobserver variabilities for strain rate measurement were assessed by the coefficient of variation in 10 subjects (5 patients and 5 normal subjects). Interobserver variability was assessed by 2 independent observers and intraobserver variability by 1 observer twice within 1 week. A P value ≤0.05 was considered statistically significant.

Results

The study included 91 essential hypertensive patients (39 women, 52 men) and 27 normal subjects (9 women, 18 men). Of the 118 subjects, the mean age was 50.7 ± 13.8 years and LVEF was 69.8% ± 5.5%.

Based on the above criteria for classification of different LV filling patterns, the subjects were all divided into 4 groups. Table 1 presents the demographics in the different groups. There was no significant difference between the 4 groups regarding body mass index and the presence of hypertension. Subjects with normal diastolic function were younger (P < 0.001 for all other subgroups), whereas subjects with severe diastolic dysfunction were older (P < 0.001 for all other subgroups). However, there was no significant difference in age between the other 2 diastolic categories. Significant difference of hypertension duration was detected between the 4 groups (Table 1). About 73% of hypertensive patients were on only 1 type of antihypertensive medication (angiotensin‐converting enzyme inhibitor, angiotensin II receptor blocker, calcium channel blocker, and diuretic); the medications in the 4 groups are summarized in Table 1.

Table 1.

Demographics in Different LVDD Groups

Variable	No DD (n = 36)	Mild DD (n = 20)	Moderate DD (n = 58)	Severe DD (n = 4)	P Value Overall
Age (y)	43.3 ± 12.6a	58.6 ± 10.5	51.1 ± 12.0b	73.3 ± 20.6c	<0.001
BMI (kg/m2)	23.8 ± 3.1	24.9 ± 3.5	25.2 ± 3.5	23.8 ± 2.3	0.216
Heart rate (beats/min)	69.7 ± 10.2	75.8 ± 10.5d	70.5 ± 8.9e	60.8 ± 10.8e	0.021
SBP (mm Hg)	129.6 ± 16.4	141.7 ± 19.8	145.7 ± 26.2d	141.3 ± 18.3	0.028
DBP (mm Hg)	85.8 ± 13.6	92.3 ± 16.2	92.3 ± 17.2	71.8 ± 13.4f	0.045
Hypertension	24 (66.7%)	16 (80%)	47 (81%)	4 (100%)	0.192
Hypertension duration	2.6 ± 2.1a	6.7 ± 6.1	4.6 ± 4.3	16.0 ± 4.5a	0.003
Antihypertensive medications
ACEI or ARB	5 (20.8%)	4 (25%)	5 (10.6%)	0 (0%)	0.293
CCB	12 (50%)	3 (18.8%)d	26 (55.3%)	0 (0%)	0.008
Diuretics	3 (12.5%)	2 (12.5%)	5 (10.6%)	1 (25%)	0.896
Others	4 (16.7%)	7 (43.8%)	11 (23.4%)	3 (75%)	0.05

Abbreviations: ACEI, angiotensin‐converting enzyme inhibitor; ARB, angiotensin II receptor blocker; BMI, body mass index; CCB, calcium channel blocker; DD, diastolic dysfunction; DBP, diastolic blood pressure; LVDD, left ventricular diastolic dysfunction; SBP, systolic blood pressure.

Data are expressed as n (%) and mean ± SD.

^a P < 0.05 vs other 3 groups.

^b P < 0.05 vs No DD and Mild DD groups.

^c P < 0.05 vs No DD and Moderate DD groups.

^d P < 0.05 vs No DD group.

^e P < 0.05 vs Mild DD group.

^f P < 0.05 vs Mild DD and Moderate DD groups.

The echocardiographic variables are listed in Table 2. All the subjects included in the present study had normal LVEF, and there was no significant difference in LVEF across the 4 groups. Compared with other groups, the mild group had significantly higher peak mitral A wave and lower peak mitral E wave.

Table 2.

Echocardiographic Variables in the 4 Study Groups

Variable	No DD (n = 36)	Mild DD (n = 20)	Moderate DD (n = 58)	Severe DD (n = 4)	P Value Overall
LVEDd (mm)	46.2 ± 3.9	45.3 ± 3.8	45.5 ± 3.9	50.8 ± 2.1a	0.046
LVESd (mm)	27.2 ± 3.3	27.0 ± 3.2	27.3 ± 3.1	32.8 ± 2.9a	0.044
LVEF (%)	70.5 ± 4.7	68.6 ± 6.5	69.8 ± 5.6	69.8 ± 6.2	0.762
IVSt (mm)	9.6 ± 1.5	11.3 ± 1.6b	10.5 ± 1.5b	13.0 ± 1.2c	<0.001
LVPWt (mm)	9.3 ± 1.1a	11.0 ± 1.4	10.3 ± 1.4	13.3 ± 1.5a	<0.001
LVMI (g/m2)	88.4 ± 16.9	109.6 ± 30.8b	97.2 ± 22.9	155.9 ± 19.3a	0.001
LAd (mm)	34.7 ± 3.8	36.5 ± 3.0	35.3 ± 3.4	36 ± 2.7	0.438
LAVI (mL/m2)	37.6 ± 4.8	38.6 ± 4.1	37.5 ± 4.4	50.5 ± 12.2a	0.095
E (cm/s)	73.8 ± 14.4	61.5 ± 10.5b	83.9 ± 15.7d	98.4 ± 21.1d	<0.001
A (cm/s)	64.7 ± 11.7	93.8 ± 16.3b	77.9 ± 15.7d	47.4 ± 10.3a	<0.001
E/A	1.2 ± 0.4	0.7 ± 0.1b	1.1 ± 0.3e	2.1 ± 0.1a	<0.001
EDT (ms)	185.1 ± 30.7	205.4 ± 26.8b	177.5 ± 18.4e	122.8 ± 16.3a	<0.001
Sm (cm/s)	8.0 ± 1.3	7.7 ± 1.5	7.4 ± 1.1	6.5 ± 1.0b	0.013
Em (cm/s)	9.0 ± 1.9a	6.2 ± 1.3	7.0 ± 1.4	6.2 ± 1.6	<0.001
Am (cm/s)	9.3 ± 1.5	9.6 ± 1.5	9.4 ± 1.2	8.4 ± 0.8	0.269
E/Em	8.3 ± 1.2a	10.2 ± 2.1	12.1 ± 1.8d	16.6 ± 5.4	<0.001

Abbreviations: A, diastolic peak late transmitral flow velocity; Am, late diastolic annular velocity; DD, diastolic dysfunction; E, diastolic peak early transmitral flow velocity; E/A, peak E to peak A velocity; EDT, deceleration time of peak E velocity; Em, early diastolic annular velocity; IVSt, intraventricular wall thickness; LAd, left atrial diameter in end‐systole; LAVI, indexed volume of left atrium; LVEDd, left ventricular end‐diastolic diameter; LVEF, left ventricular ejection fraction; LVESd, left ventricular end‐systolic diameter; LVMI, left ventricular mass index; LVPWt, left ventricular posterior wall thickness; Sm, systolic annular velocity.

Data are expressed as mean ± SD.

^a P < 0.05 vs other 3 groups.

^b P < 0.05 vs No DD group

^c P < 0.05 vs No DD and Moderate DD group.

^d P < 0.05 vs No DD and Mild DD group.

^e P < 0.05 vs Mild DD group.

The LA myocardial parameters are presented in Table 3. Significant differences were observed across the 4 groups. The LA myocardial strain rate parameters were all significantly reduced in patients with severe diastolic dysfunction compared with patients with no diastolic dysfunction (all P < 0.001). Compared with patients of normal diastolic function, the mild diastolic dysfunction group had significantly lower E‐Sr (0.62 ± 0.18 s⁻¹ vs 1.20 ± 0.38 s⁻¹, P < 0.001) and S‐Sr (0.78 ± 0.16 s⁻¹ vs 0.94 ± 0.22 s⁻¹, P < 0.001) but increased A‐Sr (1.14 ± 0.29 s⁻¹ vs 1.00 ± 0.23 s⁻¹, P = 0.05).

Table 3.

Measurements of the LA Myocardial Deformation in the 4 Groups

Variable	No DD (n = 36)	Mild DD (n = 20)	Moderate DD (n = 58)	Severe DD (n = 4)	P Value Overall
S‐Sr (s−1)	0.94 ± 0.22a	0.78 ± 0.16	0.74 ± 0.22	0.59 ± 0.26	<0.001
E‐Sr (s−1)	1.20 ± 0.38a	0.62 ± 0.18	0.92 ± 0.29b	0.76 ± 0.34	<0.001
A‐Sr (s−1)	1.00 ± 0.23	1.14 ± 0.29c	0.94 ± 0.26d	0.62 ± 0.33b	0.005

Abbreviations: A‐Sr, peak strain rate in late diastole; DD, diastolic dysfunction; E‐Sr, peak strain rate in early diastole; LA, left atrial; S‐Sr, peak strain rate in systole.

Data are expressed as mean ± SD.

^a P < 0.05 vs other 3 groups.

^b P < 0.05 vs No DD and Mild DD groups.

^c P = 0.05 vs No DD group.

^d P < 0.05 vs Mild DD group.

The LA myocardial parameters including S‐Sr, E‐Sr, and A‐Sr were significantly correlated with E/Em (r = −0.381, −0.467, and −0.229, respectively, all with P < 0.013). Of these, the correlation between the E‐Sr and E/Em was the strongest (Figure 1).

Figure 1.

Correlation of LA myocardial deformation with the E/Em ratio in all subjects. E‐Sr was significantly correlated with E/Em (Pearson correlation analysis, r = −0.467, P < 0.001). Abbreviations: E, diastolic peak early transmitral flow velocity; Em, early diastolic annular velocity; E‐Sr, peak strain rate in early diastole; LA, left atrial.

Figure 2.

Representative strain rate curves of subjects from no LV diastolic dysfunction group (A) mild LV diastolic dysfunction group, (B) moderate LV diastolic dysfunction group, (C) and (D) severe LV diastolic dysfunction group. Abbreviations: LV, left ventricular.

The feasibility of LA myocardial deformation measurement has been reported in previous studies.7, 9 By using the Philips iE33 ultrasound system, LA myocardial deformation can be measured successfully in about 95% of subjects. Two subjects in the no LV diastolic dysfunction group, 1 subject in the mild LV diastolic dysfunction group, and 3 subjects in the moderate LV diastolic dysfunction group were excluded due to inadequate image quality (if a smooth SR curve could not be obtained or if the angles of interrogation exceed 30°). Coefficients of variance of the interobserver variability for S‐Sr, E‐Sr, and A‐Sr were 7.1%, 8.2%, and 9.8%, respectively.

Discussion

The present study demonstrated that all LA myocardial parameters, including S‐Sr, E‐Sr, and A‐Sr, were significantly correlated with E/Em. No significant difference of LA dimension and LAVI was observed in subjects with different degrees of LV diastolic dysfunction (except for the severe group), whereas LA myocardial strain rate parameters were all significantly different across the 4 groups. Compared with patients of normal diastolic function, the mild diastolic dysfunction group had significantly lower E‐Sr and S‐Sr but increased A‐Sr.

It is known that left atrial size can be estimated by assessment of LA dimension and that E/Em has been shown to be an excellent predictor of LV filling pressure in previous studies using the TDI technique.14, 15 In the present study, significant correlations were observed between LA deformation parameters and E/Em. Our findings imply that LA myocardial deformation may have the potential to predict LV diastolic dysfunction.

Previous studies reported that impaired ventricular relaxation was associated with decreased conduit function.16, 17 By investigating the LA myocardial function in type 2 diabetes using velocity vector imaging, Jarnert et al18 reported a significant reduction of early diastolic SR observed in the mild diastolic dysfunction group. Consistent with the previous reports, the present study shows that the LA E‐Sr decreased significantly in subjects with LV diastolic dysfunction. Of the 4 groups, the E‐Sr was minimized in subjects with mild diastolic dysfunction. However, except for the severe group, we could not detect any difference of left atrial size and LAVI across the other 3 groups. Although the LA size and volume were not changed, our findings indicated that mild LV diastolic dysfunction results in a reduction of LA myocardial deformation, which was reflected by a significant reduction in E‐Sr (by an average reduction of 48.3%). Our results suggest that the decrease of LA myocardial function occurs even prior to the development of overt LA enlargement.

The LA functions as a reservoir, pump, and conduit, which has been demonstrated to play an important role in modulating LV filling.19, 20 According to the Frank‐Starling law, the LA pump function has been shown to increase in the presence of mild LV diastolic dysfunction and then significantly reduce when LV diastolic dysfunction develops to moderate and severe status.17, 19 In this study, A‐Sr was significantly reduced in the moderate and severe dysfunction group, suggesting decrease of LA myocardial pump function in response to moderate and severe LV diastolic dysfunction. In subjects with mild LV diastolic dysfunction, the A‐Sr was found to be increased significantly. Correspondingly, peak A velocity (the global LA pump function) was significantly increased in the mild group in standard echocardiographic examination, indicating an increase of pump function.

According to a study using standard echocardiographic imaging, LA reservoir function has been shown to increase in the presence of mild LV diastolic dysfunction and decrease in subjects of moderate and severe LV diastolic dysfunction. However, by using strain rate imaging, Jarnert et al reported that no difference of LA myocardial reservoir function (reflected by S‐Sr) was detected between normal, mild, and moderate diastolic function subjects.18 In the present study, our results show that LA myocardial reservoir function decreased even in the stage of mild LV diastolic dysfunction, though no difference was detected between different degrees of LV diastolic dysfunction. Thus, further longitudinal studies are warranted to investigate LA reservoir function in patients with early LV diastolic dysfunction.

Study Limitations

The present study population had a relatively small sample size of patients with severe diastolic dysfunction, and only hypertensive patients and normal subjects were included, thus the present results may not be applied to other populations. Subjects with normal diastolic function were significantly younger than those with diastolic dysfunction. A negative correlation between age and all LA myocardial deformation parameters was found in the population (data not shown), which may confound the present findings. But of note, the mild diastolic dysfunction group (subjects were older than the normal diastolic function group) had increased A‐Sr. Thus, it seems that the changes of LA myocardial deformation may not be due solely to the confounding effect of age, but were the response to different degrees of LV diastolic dysfunction.

The TDI‐derived strain rate imaging technique is angle‐dependent, which has been described in previous studies.21, 22 In our study, all measurements were performed with an angle of interrogation of <30° to improve accuracy.

Conclusion

TDI‐derived SR imaging is a useful noninvasive method for quantifying regional LA myocardial function, which is closely associated with different degrees of LV diastolic dysfunction in subjects with preserved LVEF. By using strain rate imaging, significant changes of LA deformation in response to different stages of LV diastolic dysfunction were detected in subjects with preserved LVEF. Quantification of LA myocardial function may have the potential to predict early LV diastolic dysfunction in subjects with preserved LVEF before the development of overt LA enlargement. Future longitudinal studies are warranted to further our understanding of the progressive changes of LA myocardial deformation in response to LV diastolic dysfunction.