Article Text

Download PDFPDF
Hypertension and renovascular disease
Strain rate evaluation of phasic atrial function in hypertension
  1. S Eshoo1,
  2. A C Boyd1,
  3. D L Ross1,
  4. T H Marwick2,
  5. L Thomas3
  1. 1
    University of Sydney, Westmead Hospital, Sydney, Australia
  2. 2
    University of Queensland, Princess Alexandra Hospital, Brisbane, Australia
  3. 3
    Liverpool Hospital, Liverpool, NSW, Australia
  1. Dr S Eshoo, Department of Cardiology, Westmead Hospital, Darcy Road, Westmead, 2145 NSW, Australia; suzannee{at}


Background: Strain (SI) and strain rate (SR) measure regional myocardial deformation and may be a new technique to assess phasic atrial function.

Objective: To examine the feasibility of using SI and SR to evaluate phasic atrial function in patients with mild hypertension (HT).

Patients and methods: The study group comprised 54 patients with mild essential HT (29 women) and 80 age-matched normal controls (47 women). Standard two-dimensional and Doppler echocardiography was performed as well as Doppler tissue imaging. The following left atrial (LA) volumes were measured: (a) maximal LA volume or Volmax; (b) minimal LA volume or Volmin; (c) just before the “p” wave on ECG (Volp). Phasic LA volumes were also calculated. Systolic (S-Sr), early diastolic (E-Sr), late diastolic (A-Sr) strain rate and SI were measured.

Results: Despite no differences in indexed maximal LA volume with only mild increases in left ventricular mass in the HT cohort compared with normal subjects (mean (SD) 86 (18) g/m2 vs 67 (14) g/m2; p = 0.001), E-Sr was significantly lower in the HT cohort. There was a corresponding reduction in indexed conduit volume in the HT cohort compared with normal subjects (10.5 (7.5) ml/m2 vs 13.8 (6.1) ml/m2; p = 0.006). Global E-Sr showed modest negative correlations with LA Volmax and LA ejection fraction. No significant difference was present in S-Sr, A-Sr or global atrial strain between the normal and HT cohorts.

Conclusion: Mild HT results in a reduction in LA conduit volume, although maximal LA volume is unchanged. This is reflected by a reduction in E-Sr with preserved S-Sr and A-Sr.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Left atrial (LA) enlargement is commonly observed by echocardiography in patients with moderate and severe hypertension (HT).1 HT is the most prevalent and modifiable risk factor for the development of atrial fibrillation (AF).2 Hypertensive subjects have about one and a half times the risk of developing AF compared with a normotensive population.3 Thus it is likely that functional changes accompany the observed structural changes of the left atrium in patients with HT.

The development of new echocardiographic techniques, such as Doppler tissue imaging (DTI) has enhanced the ability to assess regional myocardial function non-invasively.4 Strain (SI) and strain rate (SR), derived from DTI, are used to examine myocardial deformation and rate of deformation, respectively. Its initial use has been for quantifying regional myocardial deformation in the ventricle5 6 and more recently in the atrium.7 8 Strain rate imaging has high spatial and temporal resolution and is largely independent of the tethering effect and rotational motion of the myocardium.5 7

SI and SR have been used to evaluate atrial function in normal subjects9 and in conditions likely to cause atrial dysfunction.8 The objective of this study was to examine the feasibility of using SI and SR to evaluate phasic atrial function in patients with mild HT. Thus systolic SR (S-Sr) would serve as a surrogate measure of reservoir function, early diastolic SR (E-Sr) as a measure of conduit function and late diastolic SR (A-Sr) as a measure of active contractile function. We sought to determine whether SR and SI could detect functional changes associated with the changes in phasic LA volumes.


Study approval was obtained from the human research ethics committee at Westmead Hospital, Sydney, Australia. All patients provided informed written consent. The study group comprised 54 patients with mild essential HT (29 women) and 80 age-matched normal controls (47 women). The group with HT was recruited based on a detailed history of HT with all recruited subjects having documented blood pressure (BP) readings of >140/90 mm Hg, but <160/100 mm Hg since diagnosis of HT and during antihypertensive therapy. These details were corroborated from patient hospital records and from records from their treating doctor. Hypertension was idiopathic or “primary” in all patients. None of the patients included had secondary causes of hypertension. Hypertensive patients were recruited from cardiology and renal departments and from the community. Seated BP was measured (two recordings made 10 min after arrival) at the time of the echocardiographic examination. The group with HT was receiving an average (SD) of 1.7 (1.1) antihypertensive drugs; table 1 summarises the medications. Patients were excluded if they had significant valvular disease (more than mild valvular regurgitation and any degree of aortic or mitral stenosis), a history of atrial or ventricular arrhythmias, a left ventricular (LV) ejection fraction of <50%, previous cardiac surgery or implanted devices. All patients were in sinus rhythm with no past history of arrhythmias.

Table 1 Normal and mild hypertension cohorts

The normal population consisted of healthy volunteers recruited from hospital staff and the community. There was no history of HT or documented BP readings >140/90 mm Hg, ischaemic heart disease, significant valvular disease, peripheral vascular disease, cerebrovascular disease or diabetes. Any volunteer who was hypertensive at the time of the echocardiographic examination or who was found to have cardiac pathology on echocardiogram was also excluded from analysis. None of the normal population were receiving antihypertensive or other cardiac drugs. BP for the normal population was recorded twice in the seated position 10 min after arrival.

Standard transthoracic echocardiogram

Doppler M-mode and two-dimensional echocardiography were performed according to established clinical laboratory practice using two commercially available instruments (system 1: General Electric/Vingmed Vivid 7 (Horten, Norway); system 2: General Electric/Vivid 7 (Horten, Norway)) using harmonic 3.5 MHz variable frequency phased-array transducers. Offline measurements were performed on an Echopac workstation (version BT06). The LA linear dimension was estimated by M-mode measurement in the parasternal long-axis view.10 LV wall thickness was measured by M mode using American Society of Echocardiography (ASE) criteria. LV end-diastolic and end-systolic volumes and ejection fraction were determined from the apical four- and two-chamber views using the biplane method of discs. The area–length method was used to determine LV mass.11 All values were indexed to body surface area.

LA volumes and phasic volumes

The following LA volumes were measured: (a) maximal LA volume or Volmax, in ventricular systole just before mitral valve opening; (b) minimal LA volume or Volmin, after mitral valve closure and (c) Volp, just before the “p” wave on ECG. All volumes were calculated from the apical four- and two-chamber zoomed views using the Simpson biplane method of discs.12 LA stroke volume (LASV) was estimated as the difference between Volmax and Volmin. The LA ejection fraction (LAEF) was calculated as (LASV/Volmax)×100.12

The following LA phasic volume parameters were derived13 14: LA passive emptying volume  =  Volmax − Volp; LA passive emptying fraction  =  LA passive emptying volume/Volmax; LA active emptying volume  =  Volp − Volmin; LA active emptying fraction  =  LA active emptying volume/Volp; LA conduit volume  =  LV stroke volume − (Volmax − Volmin).

Traditional measures of LA function

Mitral inflow velocity was obtained by pulsed-wave Doppler examination at a sweep speed of 100 mm/s from the apical four-chamber view by placing the sample volume at the tips of the mitral leaflets. Peak A-wave velocity and the velocity time integral of the A wave was measured as an average of three beats. The atrial fraction was estimated as A-wave velocity time integral divided by the total velocity time integral of mitral inflow.15

LV diastolic parameters

LV diastolic function was determined using standard echocardiographic parameters including peak E velocity, peak A velocity, E/A ratio, deceleration time and isovolumic relaxation time. The early diastolic E′ velocity and late diastolic A′ velocity were measured by DTI. These measurements were acquired by placing the sample volume at the septal annulus, recording at a sweep speed of 100 mm/s and measuring an average from three beats. Care was taken to place the sample volume parallel to the direction of motion. The Nyquist limit was set at 16 cm/s and gains were adjusted to minimise noise.

Strain/strain rate imaging in the left atrium

Colour DTI images were obtained in the apical four- and two-chamber views in all subjects with appropriate changes in sector width to maintain frame rates of >110 frames/s. Data of three cardiac cycles triggered to the QRS complex were saved on a magneto-optical disc (MO-4.8 GB; Imation, Japan) and were analysed offline using commercial software (GE Echopac 6.2). SR was measured from the mid to superior16 17 segments at the septal and lateral walls of the LA in the apical four-chamber view and the inferior and anterior walls from the apical two-chamber view (fig 1A). The sample volume (10×2 mm) was tracked frame by frame to maintain its position within the LA walls. The peak SR was measured in each segment in systole (S-Sr), early diastole (E-Sr) and late diastole (A-Sr) applying Gaussian 60 smoothing. Global S-Sr, E-Sr and A-Sr were also calculated by averaging SR at the septum, lateral, inferior and anterior walls. In addition to velocity, the timing of atrial activity was also measured as the duration from aortic valve closure to the peak A-Sr as a marker of intra-atrial dyssynchrony. Intra-atrial delay was defined by the standard deviation of the timing of atrial activity among the septal, lateral, inferior and anterior walls.18 If the angle of interrogation exceeded 30°, the patient was excluded from the final analysis.

Figure 1

(A) Apical four- and two-chamber views with position of sample volume placement for evaluation of strain and strain rate from septal, lateral, anterior and inferior LA segments. (B) Atrial strain rate and strain curves with sample volume placed at the LA septal segment. Red markers on ECG trace are placed at the end of the T wave indicating diastolic gated strain profile. A-Sr, peak atrial strain rate; AVC, aortic valve closure; E-Sr, early diastolic strain rate; LA, left atrium; MVC, mitral valve closure; SRI, strain rate imaging.

The myocardial strain profiles were calculated by integrating the SR profiles over time and compensating for drift over the cardiac cycle (fig 1B). As active atrial contraction occurs in diastole, the SI curves were gated in diastole by moving the gating marker to the end of the T wave on the ECG. The peak SI was measured in the mid to superior septal, lateral, inferior and anterior segments and the global peak SI was also calculated as an average of the four segments.


Ten subjects from each group were randomly selected for interobserver and intraobserver variability analysis. Peak S-Sr, E-Sr, A-Sr and SI were remeasured by the same observer and by a second independent observer from the stored digital data. LA images were acquired an hour later to assess interstudy reproducibility.

Statistical analysis

All values are expressed as mean (SD). Differences in continuous variables between groups were assessed by an unpaired Student t test. Categorical variables were analysed by the χ2 test, and Fisher exact test, when appropriate. The correlation between two variables was assessed by the Spearman rank correlation coefficient. Repeated measures analysis between and within groups was used to assess the effect of segment position. Data for S-Sr, E-Sr, A-Sr and SI were log transformed to approximate normality before analysis. Bland Altman analysis19 was performed to analyse intra- and interobserver variability. Data were analysed using SPSS (version 15, Chicago, Illinois, USA). Data was considered significant if p⩽0.05.


Clinical and echocardiographic characteristics

Table 1 summarises the demographic and echocardiographic variables of the normal and HT cohort. Although maximum LA volume was larger in the HT group, once corrected for body surface area, this failed to reach significance (p = 0.13). Transmitral peak A velocity (mean (SD)) was significantly higher in the HT cohort than in normal subjects (0.74 (0.2) vs 0.67 (0.17), p = 0.02). Other traditional parameters of atrial function including the A wave VTI, atrial fraction and LAEF were not significantly different in the two groups. Phasic LA volumes in the HT and normal cohorts demonstrated a significant reduction only in conduit volume in the HT group (normal subjects vs patients with HT 13.8 (6.1) vs 10.5 (7.5) ml/m2, p = 0.006) (fig 2 and appendix table A1).

Figure 2

(A and B). Bar graphs demonstrating left atrium phasic volumes, passive and active emptying fractions in hypertension (HT) versus normal groups. *p<0.05 between normal subjects and mild HT.

SR and strain in normal subjects and hypertension

Table 2 summarises the global SR and strain results in the normal and HT cohort. Using repeated measures analysis of variance, there was no effect of position or group for S-Sr (p = 0.25, p = 0.73, respectively), A-Sr (p = 0.17, p = 0.35, respectively) or SI (p = 0.44, p = 0.39, respectively). There was no effect of position with E-Sr (p = 0.33), however the E-Sr in the HT group was significantly lower than the normal cohort by an average of 14.3% across the four positions (95% CI 2 to 25.1%, p = 0.02). As there was no effect of position within the normal and HT cohort, an average of the four segments was calculated to give a global S-Sr, E-Sr, A-Sr and SI profile (fig 3 and appendix fig A1).

Figure 3

Comparison of global S-Sr, E-Sr, A-Sr and SI among normal subjects and subjects with hypertension. A-Sr, peak atrial strain rate; E-Sr, early diastolic strain rate; SI, strain; S-Sr, peak systolic strain rate.

Table 2 Global strain rate and strain in the atrial segments in normal and HT cohort

Additionally, no significant difference was present in intra-atrial delay between the two groups (normal subjects vs patients with HT 34.7 (25.5) ms vs 32.4 (19.5) ms, p = 0.57).

Effect of increased LA size and increased LV mass on E-Sr and A-Sr

Using a cut-off point of 30 ml/m2 to define an enlarged LA, the HT cohort was divided into normal and enlarged LA size. No difference was found in regional S-Sr, E-Sr, A-Sr or SI in all four segments on subgroup analysis. When the HT cohort was divided based on the presence of increased LV mass using the ASE cut-off values of >89 g/m2 in women and >103 g/m2 in men,20 there was no difference in S-Sr, E- Sr, A-Sr or SI in the four segments between groups (data not shown).

Correlation of SR and SI with traditional measures of atrial function

Correlations were performed with global S-Sr, E-Sr, A-Sr, SI and the traditional measures of atrial function. Table 3 summarises the results. Global S-Sr showed a modest negative correlation with LA Volmax. Global E-Sr showed modest negative correlations with LA Volmax, peak A velocity and atrial fraction. E-Sr also correlated with passive emptying fraction (Spearman’s rs = 0.31, p = 0.001) and inversely with indexed active emptying volume (rs = −0.30, p = 0.001).

Table 3 Correlations of S-Sr, E-Sr, A-Sr and strain in normal subjects and the cohort with hypertension

Global A-Sr showed modest correlation with the traditional parameters of LA function including peak A velocity and atrial fraction. Global A-Sr also correlated with septal A′ velocity obtained from DTI (rs = 0.29, p = 0.002). Global strain did not correlate with the traditional parameters of atrial function.

Correlation of SR and SI with clinical and echocardiographic variables in HT

Global E-Sr correlated inversely with age (rs = −0.52, p = 0.001), duration of hypertension (rs = −0.46, p = 0.004), systolic BP (rs = −0.20, p = 0.02) and LV mass (rs = −0.24, p = 0.017). No significant correlation was observed between A-Sr and age, body mass index or LA Volmax. No correlation was observed in S-Sr, A-Sr or SI with systolic or diastolic BP or mean arterial pressure. No association was observed between S-Sr, A-Sr or SI and LV mass (table 3).

Figure 4 shows the Bland-Altman19 analysis for the mean difference and confidence intervals for intraobserver variability. SI had significantly more variability in measurements within and between observers (data not shown) than SR measurements. The interstudy reproducibility (mean (SD)) in this cohort for maximum LA volume was 1.9 (4.2) ml, with a correlation coefficient rs = 0.95. For minimal LA volume, the mean difference was −1.1 (1.8) ml, with rs = 0.98. For pre-P volume, the mean difference was 0.87 (2.2) ml, with rs = 0.97.

Figure 4

Bland-Altman for intraobserver variability and confidence intervals for strain rate and strain (Straight lines represent upper and lower confidence intervals). A-Sr, peak atrial strain rate; E-Sr, early diastolic strain rate; SI, strain; S-Sr, peak systolic strain rate.


We have shown in this pilot study, that strain and strain rate are useful16 21 for assessing phasic LA function in mild HT. Unlike previous studies, which evaluated atrial strain using systolic gating, we have refined the evaluation of LA contractile function using diastolic strain gating.

In the HT cohort studied, no difference was noted in indexed LA Volmax compared with normal subjects. Despite no difference in LA Volmax, the LA conduit volume was reduced in the HT cohort. Correspondingly, the E-Sr, which is a surrogate marker of conduit function, was reduced in the cohort with mild HT. No regional differences were noted in the LA walls, consistent with HT resulting in global changes. No difference was found in S-Sr, a marker of reservoir function or A-Sr, a marker of atrial contractile function.

Phasic atrial changes in mild HT: reservoir function

Several measurements during ventricular systole, as shown in volumetric examination, have been considered as indicators of LA reservoir function.22 In this instance, S-Sr was used as a surrogate marker of LA reservoir function—that is, the passive stretching of the LA wall during LV systole.23 In our cohort, no significant difference was found in LA Volmax, passive emptying volume or fraction between the mild HT and normal cohorts. This is reflected by similar S-Sr values in both the normal and HT cohorts. In contrast, Kokubu et al found that the S-Sr was reduced in a group with mild HT.17 However, their cohort had an increased LV mass (128.5 (33) g/m2) compared with our cohort with a mean (SD) indexed LV mass of 86 (18) g/m2 and the duration of HT in the population studied was also unclear.17

Conduit function

LA conduit function can be assessed by global E-Sr.16 This is an important determinant of LV filling and represents the volume of blood that passes through the LA that cannot be attributed to reservoir or booster pump functions. LA conduit function decreases with advancing age as a result of decreasing LV compliance.24 We have demonstrated an inverse correlation with ageing on LA conduit function in both the hypertensive and normal cohorts. Additionally, the HT cohort had a lower global and regional E-Sr in the four atrial segments than the age-matched normal subjects as reflected by a decrease in LA conduit volume. This has been previously shown in patients with severe HT with LA enlargement,14 but is the first demonstration that even with mild HT in the absence of overt LA enlargement, changes in LA conduit volume are present.

Contractile function

The peak A wave in late diastole has been used as a measure of LA contractile function; however, peak A velocity reflects blood flow due to the atrioventricular pressure gradient between the left ventricle and the LA, rather than intrinsic LA function.25 LA automatic border detection software allows assessment of LA booster function non-invasively.26 LA contractile function as assessed by the traditional parameters of LA function has been shown to be increased in patients with moderate/severe HT and compensates for reduction in LV filling consequent to the effects of HT on diastolic function.27 28 DTI has been used for evaluating LA booster function.29 A-Sr can be used as a marker of atrial contractile function and has recently been used to evaluate longitudinal atrial contractility in a variety of clinical conditions including AF, myocardial infarction and HT.8 16 17

The cohort with mild HT in this study had a normal LA size with an active emptying volume and fraction comparable to that of normal subjects. Thus it is likely that the extent of diastolic impairment consequent to mild HT was insufficient to result in a compensatory increase in atrial booster pump function as demonstrated by similar A-Sr values in both groups. There was a significant correlation between A-Sr and A wave velocity, A wave velocity time integral and atrial fraction, confirming that A-Sr can be used as an index of LA contractile function in HT.

LV diastolic dysfunction in mild HT

There was no difference in indexed maximal LA volumes in the HT cohort. Although LV mass in the HT cohort was higher than in normal subjects, LV mass was within the normal range. Despite this, the transmitral peak A was significantly increased in the HT group compared with normal subjects, with no difference in peak E velocity. On the contrary, ventricular septal E′ was decreased with a corresponding increase in the E/E′ ratio in the HT group compared with healthy normal volunteers. Thus mild HT results in some degree of diastolic dysfunction even before the development of overt LV hypertrophy and LA enlargement.

LA changes: the continuum

Despite the lack of change in traditional markers of atrial function (atrial fraction, LAEF), the E-Sr was reduced in the HT cohort. Diastolic dysfunction consequent to mild HT results in decreased atrial conduit volume manifested by a decrease in E-Sr. With increasing LV hypertrophy and consequent LA enlargement, reservoir function decreases, resulting in reduced S-Sr and E-Sr.16 17 Progressive LA enlargement probably cause a reduction in atrial contractile function (A-Sr) and may then serve as a substrate for the development of AF.8 Thus while treatment strategies in AF are directed primarily towards rhythm control, the effective management of associated HT is additionally important. More detailed study in hypertensive subjects with AF may elucidate the role of atrial phasic function as a guide to anticoagulation therapy or as a determinant of AF recurrence.


The reference standards used for comparison were traditional echocardiographic parameters. The use of invasive LA pressure/volume recordings or CT/MRI to evaluate LA volumes was not considered feasible. As with all Doppler-derived techniques, angle dependency is an important consideration for SR and SI evaluation. All care was taken to ensure all measurements were performed with an angle of interrogation of <30°.

Two-dimensional speckle tracking is a newer technology for evaluating SI and SR, which has recently been used in enlarged atria as it does not rely on angle dependency.30 However, the LA is a difficult chamber for the application of this technique, owing to its thin walls. Moreover, the dropout at the interatrial septum, the LA appendage and origin of the pulmonary veins make the use of this technique technically difficult, particularly in the setting of normal-sized atria as in our patient cohort.


SR is a measure of intrinsic atrial function and can be used to evaluate phasic atrial function. We have shown that mild HT results in reduced E-Sr with no change in S-Sr and A-Sr. No regional differences in SI or SR were seen in the LA walls. This is consistent with HT resulting in global changes to the atria. Thus, the E-Sr may be an early marker of atrial dysfunction before overt LA enlargement or altered contractile function in patients with mild HT.


Table A1 Left atrial phasic volumes in normal subjects and the cohort with hypertension
Figure A1

Comparison of S-Sr, E-Sr, A-Sr and SI among normal subjects and hypertension subjects in the four atrial positions. A-Sr, peak atrial strain rate; E-Sr, early diastolic strain rate; SI, strain; S-Sr, peak systolic strain rate.



  • Funding: SE is a postgraduate medical scholar supported by a University of Sydney postgraduate award.

  • Competing interests: None.

  • Ethics approval: Ethics committee approval from Westmead Hospital Human Research Ethics Committee.