Assessment of Left Atrial Function in Essential Hypertension Patients With Normal Left Ventricle Function by Volume-derived Values and Two-dimensional Strain

Background: To investigate left atrial(LA) function in essential hypertension(EHT) patients by using volume-derived values, two-dimensional strain. Methods: 51 normal subjects and 95 EHT patients(without LVH: 50, and with LVH: 45) were enrolled for this research. LA Volume-derived index was measured in apical 4-, 2-chamber views by Simpson′s method. LA strain(S-reservoir, S-conduit, S-booster pump) and strain rate (SR-reservoir, SR-conduit, SR-booster pump), representing the reservoir, conduit and booster pump functions, respectively, were measured by two-dimensional speckle tracking echocardiography(STE). Results: Volume-derived values(contain Total LAEF, passive LAEF and active LAEF) and strain-derived values(S-reservoir, S-conduit, Sr-reservoir, Sr-conduit and Sr-booster pump) in EHT patients were signicantly lower than normal subjects. Correlation test showed LA stiffness had a strong correlation with LA conduit function in EHT patients. The AUC values were higher for detection LA conduit function than LA reservoir and booster pump function in EHT patients. LA expansion index and LA stiffness also have the higher AUC with higher sensitivity and specicity values for detection the LA dysfunctions in EHT patients. Conclusions: In this study, we concluded that LA functions were damaged in EHT patients, which could be detect by LA volume-derived values and two-dimensional strain sensitively and reproductively. LA stiffness maybe a predictor for LV remodelling in EHT patients. test was used to detect the normality of all values. Differences between the EHT patients and normal subjects were compared with an independent Student′s t-test for the data distribution was normal. For variables with a non-normal distribution, the nonparametric Mann-Whitney test was used. Differences among the EHT patients without LVH, with LVH and normal subjects about LA volume-derived values, two-dimensional strain and strain rate were compared with one-way analysis of variance (ANOVA). Comparisons of two samples were using the Student-Newman-Keuls (SNK) test. Correlations between variables were tested using Pearson or Spearman correlation tests as appropriate. The values for LA volume-derived values, LA strain, strain rate and LA stiffness of EHT patients were determined from receiver operating characteristic (ROC) curve analysis. Yoden′s index was used to determine the cut-off point with the best composite of specicity and sensitivity. Data were presented as the mean ± standard deviation (SD). Difference was considered statistically signicant in all tests when the P-value was <0.05.


Introduction
Essential hypertension (EHT) is a common risk factor for cardio-cerebrovascular disease, and its prevalence is increasing year by year. However, the pathophysiology of EHT is not completely found [1,2].
Left atrial (LA) function plays an important role in cardiac cycle: as a reservoir during left ventricle (LV) systole, a conduit from pulmonary veins to LV during early diastole and a booster pump in late diastole [3,4]. Earlier assessment of LA function in EHT patients is particularly crucial for its prognosis. An increasing LA size in EHT patients is a common nding in the clinical practice. LA function always evaluated by measuring LA volumes, pulmonary vein ow, and the strain of LA walls by tissue Doppler imaging (TDI) [2]. However, these approaches are limited by a number of shortcomings. Volume-derived values are limited by the irregular LA geometric [5,6], pulmonary vein ow is limited by the image quality, and TDI is limited for its angle-dependence [7].
In this research, we mainly study three functions of LA without and with left ventricle hypertrophy (LVH) in EHT patients with normal left ventricle ejection fraction (LVEF) using 2D strain and strain rate, and then compared these values with LA volume-derived values and LA stiffness. Furthermore, provide a simple and accurate technique to detect LA dysfunction in EHT patients without or with LV remodel in preserved LVEF.

Subjects And Methods
Study sample 55 normal subjects and 100 EHT patients were enrolled consecutively for the research. The inclusion criteria for EHT patients were: Meet the World Health Organization and International Society of Hypertension criteria [1]. LVEF > 50%, A LVEF > 50% was just considered that the EHT patients had a normal LV systolic function. All patients had no history of coronary heart disease, rheumatic heart disease, hyperlipidemia and diabetes mellitus. According to LV mass index, men with a LV mass index >115 g/m 2 and women with a LV mass index >95 g/m 2 were considered to have LV hypertrophy (LVH, 48 subjects). Patients without LVH were contained 52 subjects. EHT patients were then divided into two group: EHT patients without LVH and with LVH.
The normal subjects had no evidence of hypertension and any other cardiovascular diseases. All of the physical examination, electrocardiogram, and echocardiography were showed normal.

Two-dimensional Doppler echocardiography
All EHT patients and normal subjects underwent conventional 2D echocardiography (Vivid E9, GE Healthcare, Horten, Norway). The echocardiography was performed when the patient was in hospital. LA diameter (LAd), interventricular septal thickness at the end-diastolic period (IVSd) and LV posterior wall thickness at the end-diastolic period (LVPWd) were measured in the parasternal long-axis view of the LV using M-mode.
The peak early and late diastolic mitral valves velocities (E and A, respectively) were measured by pulsedwave Doppler, and E/A was calculated. The peak early (e′) and late (a′) diastolic mitral annular velocities were obtained by averaging the values at the septum and lateral walls using pulsed wave TDI, and then E/e′ was calculated.
ECG leads were connected to all EHT patients and normal subjects. The standard high frame rate (>36 /s) of the apical 3-, 4-and 2-chamber views (all contained LA) of three consecutive cycles were stored for offline analysis.

Data analysis for LA function
The apical 4-, 3-and 2-chamber views were analysed using the EchoPAC software (2D-Strain, EchoPAC PC version 203, GE Healthcare, Horten, Norway). Used the buttons LAX, A4C, and A2C to sketch the endomyocardial of LA, respectively. The software will automatically create a region of interest (ROI) that matched LA walls. Once the ROI was approved, the software divided LA walls into six segments. However, in the apical 3-chamber view, we included only the posterior wall of LA because the opposing wall included the ascending aorta [11]. LA strain (S-reservoir, S-conduit and S-booster pump) and strain rate (Sr-reservoir, Sr-conduit and Sr-booster pump) in the apical 4-, 3-and 2-chamber views were measured ( Figure 1).
The ratio of E/e′ to LA peak strain (S-reservoir) was used to estimate LA stiffness [14].

Reproducibility and repeatability
Intraobserver and interobserver variability for LA reservoir, conduit and booster pump strain and strain rate were determined by repeating measurements in random selected 30 patients among all enrolled patients. For the second intraobserver measurements, the observer was "blinded" to results of the initial measurements.

Statistical analysis
All data analyses were performed using SPSS 21.0 software (SPSS, Chicago, IL, USA). Shapiro-Wilk′s test or Kolmogorov-Smirnov′s test was used to detect the normality of all values. Differences between the EHT patients and normal subjects were compared with an independent Student′s t-test for the data distribution was normal. For variables with a non-normal distribution, the nonparametric Mann-Whitney test was used. Differences among the EHT patients without LVH, with LVH and normal subjects about LA volumederived values, two-dimensional strain and strain rate were compared with one-way analysis of variance (ANOVA). Comparisons of two samples were using the Student-Newman-Keuls (SNK) test. Correlations between variables were tested using Pearson or Spearman correlation tests as appropriate. The values for LA volume-derived values, LA strain, strain rate and LA stiffness of EHT patients were determined from receiver operating characteristic (ROC) curve analysis. Yoden′s index was used to determine the cutoff point with the best composite of speci city and sensitivity. Data were presented as the mean ± standard deviation (SD). Difference was considered statistically signi cant in all tests when the P-value was <0.05.

Results
155 patients satis ed the baseline inclusion criteria. 9 patients were excluded from strain and LA volume analysis because of inadequate image quality (n = 5), tachycardia (n = 2) and irregular heartbeat (n = 2). A total of 146 patients (mean age, 51.55 ± 15.19 years, 82 men) were therefore evaluated in the study and were initially divided into two groups, normal controls (n = 51, mean age, 49.45 ± 17.56 years, 27 men) and EHT patients (n = 95, mean age, 52.68 ± 13.71 years, 55 men). EHT patients were then further subdivided into two groups, EHT without LVH (n = 50, mean age, 51.04 ± 14.50 years, 30 men) and EHT with LVH (n = 45, mean age, 54.51 ± 12.69 years, 25 men) ( Table 1).  Basic information in EHT patients and normal subjects (Table 1 and Table 2) There were signi cant differences between EHT patients and normal subjects in BSA, SBP, SDP, LAD, IVSd, LVPWd, LVM, LVMI, A, E/A, e′, E/e′ and S/D (p < 0.05). However, there were no signi cant differences in LVEDV, LVESV, LVEF, E, a′, S, D and Ar (p > 0.05).  LA volume characteristics in EHT patients and normal subjects ( Table 3, Fig. 2) The values of LAVmax, LAVpre, LAVmin, total LASV and active LASV in EHT patients were signi cantly larger than normal subjects (p < 0.05). However, the values of total LAEF, passive LAEF, active LAEF and LA expansion index were signi cantly lower than normal subjects (p < 0.05). The values of passive LASV were larger than normal subjects, however, there were no signi cant difference between the two groups (p > 0.05). AVmax: maximum LA volume, LAVpre: precontraction LA volume, LAVmin: minimum LA volume, Total LASV: total LA stroke volume, Active LASV: active LA stroke volume, Passive LASV: passive LA stroke volume, Total LAEF: total LA ejection fraction, Active LAEF: active LA ejection fraction, Passive LAEF: passive LA ejection fraction, S-reservoir: LA strain corresponding to reservoir function, Sconduit: LA strain corresponding to conduit function, S-booster pump: LA strain corresponding to booster pump function, SR-reservoir: LA strain rate corresponding to reservoir function, SR-conduit: LA strain rate corresponding to conduit function, SR-booster pump: LA strain rate corresponding to booster pump function. *Signi cantly different (p < 0.05) when EHT patients compared with the normal controls.
Subgroup analysis among EHT patients without LVH, with LVH and normal subjects. The values of LAVmax, LAVpre, LAVmin, total LASV and active LASV in EHT patients with LVH were larger than EHT patients without LVH, and larger than normal subjects. However, the values of total LAEF, passive LAEF, active LAEF and LA expansion index in EHT patients with LVH were lower than EHT patients without LVH, and lower than normal subjects. All above mentioned values had signi cant differences among normal controls, EHT patients without LVH and with LVH (p < 0.05).
LA strain, strain rate and LA stiffness in EHT patients and normal subjects ( Table 3, Fig. 3) The absolute values of S-reservoir, S-conduit, Sr-reservoir, Sr-conduit and Sr-booster pump in EHT patients were signi cantly lower than normal subjects (p < 0.05).
The values of LA stiffness in EHT patients were signi cantly larger than normal subjects (p < 0.05).
Subgroup analysis among EHT patients without LVH, with LVH and normal subjects. The absolute values of S-reservoir, S-conduit, S-booster pump, Sr-reservoir, Sr-conduit and Sr-booster pump in EHT patients with LVH were lower than EHT patients without LVH, and lower than normal subjects, however, the values of LA stiffness in EHT patients with LVH were larger than EHT patients without LVH, and larger than normal subjects. All above mentioned values exclude S-booster pump had signi cant differences among normal controls, EHT patients without LVH and with LVH (p < 0.05).
S-reservoir and Sr-reservoir versus 2D echocardiographic parameters in EHT patients (Table 4) S-reservoir in EHT patients without LVH was negatively correlated with LAVmax index.
Sr-reservoir in EHT patients with LVH was negatively correlated with LAVmax.  Fig. 4) S-conduit in EHT patients without LVH was positively correlated with E, e′, and negatively correlated with LA stiffness.
Sr-conduit in EHT patients without LVH was positively correlated with S/D, LA stiffness, and negatively correlated with E, e′.
S-conduit in EHT patients with LVH was positively correlated with E, e′, Passive LAEF, LA Expansion Index, D, and negatively correlated with S/D, LA stiffness.
Sr-conduit in EHT patients with LVH was positively correlated with LA stiffness, and negatively correlated with e′, E/e′, passive LAEF, LA Expansion Index, D. S-booster pump in EHT patients without LVH was positively correlated with A, and negatively correlated with LA stiffness.
Sr-booster pump in EHT patients without LVH was negatively correlated with A and active LAEF.
S-booster pump in EHT patients with LVH was positively correlated with a′.
Sr-booster pump in EHT patients with LVH was positively correlated with Active LASV, LA stiffness, and negatively correlated with a′.  Table 7, Fig. 6.
ROC curve analysis was used to discriminate whether LA volume-derived values, LA strain and strain rate, LA stiffness were able to predict LA dysfunction. Intra-and Interobserver Variability was presented in Table 8.
30 patients were randomly selected and remeasured by two observers (experienced cardiac sonographer and cardiologist) blinded to patient clinical data and each other's results. Intra-observer variability was performed by the experienced cardiac sonographer on off-line data at different points in time.
Interobserver variability was performed by the cardiologist repeating measurements from the same images. Intra-and interobserver variabilities were calculated by intraclass correlation coe cient (ICC). All LA reservoir, conduit and booster pump strain and strain rate parameters exhibited excellent intra-and interobserver correlation, with ICC values > 0.92.  [18] had demonstrated that in EHT patients, an elevation in systolic blood pressure (SBP) leads to increased LA stiffness during ventricular systole and LV diastolic stiffness, in association with continued and further advanced LV diastolic dysfunction. Liu Y, et al [19] had evaluated the LA physic function of hypertensive patients with or without coexisting diabetes by using STE, and found abnormal LA reservoir and conduit functions. They concluded that STE-derived strain and strain rate imaging were sensitive methods for evaluating LA function. Xu TY, et al [20]had investigated LA function in relation to hypertension using STE in patients with preserved LVEF, and found that hypertension was associated with impaired LA function, as assessed by STE, even before LA enlargement or after LV remodel. Jarasunas J, et al [21] had found that patients with paroxysmal AF and primary arterial hypertension have decreased reservoir, conduit and pump LA functions even in the absence of echocardiographic signs of LV diastolic dysfunction.
Our results were accordance with previous studies. Volume-derived values in EHT patients revealed impaired LA functions compared with normal subjects. Strain and strain rate showed LA reservoir, conduit and booster pump function were damaged, and LA reservoir and conduit function were more serious in EHT patients with LVH than without LVH. LA stiffness in EHT patients with LVH was signi cantly larger than without LVH and normal subjects. EHT not only leads to LV myocardial hypertrophy and myocardial brosis, but also leads to LA myocardial brosis. In normal aging process, there was a decrease in conduit volumes together with an increase in active atrial emptying. These changes were probably because of a compensatory mechanism to overcome the normal age-related decrease in LV relaxation [8]. The increased LA stiffness could also rise LV end diastolic pressure and lling pressure, at last, the LA conduit function was impaired more serious. In EHT patients, the decreased reservoir, passive and booster pump function maybe a composite measure of LV diastolic function.
Subgroup analysis also showed that LA volumed-derived values and LA strain and strain rate for detecting LA dysfunction were complement to each other.
Correlation test showed that LA stiffness had a strong correlation with LA conduit function both in EHT patients without LVH or LVH. ROC analysis showed the AUC values were higher for detection LA conduit function than LA reservoir and booster pump function in EHT patients. LA expansion index and LA stiffness also have the higher AUC values for detection LA dysfunctions in EHT patients. The results demonstrated that LA stiffness can assess the LA conduit dysfunction in EHT patients, also LA stiffness can re ect LV remodelling with preserved LVEF in EHT patients without LVH or LVH.

Conclusions
In this study, increased LA stiffness, combined LA volume-derived values, strain and strain rate may conclude that LA reservoir, conduit and booster pump function were impaired in EHT patients without LVH and with LVH. 2D strain is a sensitive and reproductive tool for detecting the impairment of LA dysfunction. LA stiffness maybe a predictor for LV remodelling in EHT patients without LVH and with LVH.

Declarations Funding Sources
This study was funded by Changzhou Science and Technique Program (Grant No.: CJ20190098).

Availability of data and materials
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Ethics approval and consent to participate
This study was approved by the ethics committee of Changzhou No.2 People′s Hospital and the First A liated Hospital of Soochow University.

Consent for publication
All participants provided their written, informed consent. Figure 1 LA functions in apical 4-chamber views in normal subjects and EHT patients (without LVH and with LVH) by using EchoPAC. S-reservoir, S-conduit and S-booster pump, as well as SR-reservoir, SR-conduit and SRbooster pump, corresponded to the LA reservoir, conduit and booster pump functions, respectively. The strain and strain rate index decreased in EHT patients.   Correlation tests showed S-conduit in EHT patients without LVH (A) and with LVH (B) were negatively correlated with LA stiffness. Correlation tests showed Sr-conduit in EHT patients without LVH (C) and with LVH (D) were positively correlated with LA stiffness.

Figure 5
A: Correlation test showed S-booster pump in EHT patients without LVH was negatively correlated with LA stiffness, B: Correlation test showed Sr-booster pump in EHT patients with LVH was positively correlated with LA stiffness. Figure 6 ROC analysis showed the AUC for detection LA reservoir, conduit, booster pump functions (A, B, C) and LA dysfunction (D).