Demographic (gender and age), anthropometric (weight and height) and clinical variables (concomitant risk factors and duration and control of HTN) were recorded, as was current medication.

HTN severity was assessed in accordance with European Society of Hypertension and European Society of Cardiology guidelines11,12 at the first follow-up consultation and on assessment for enrollment in the study. At the first follow-up assessment, mean blood pressure (BP) values on 24-hour ABPM were recorded (n=109), or when these data were unavailable, the highest BP values measured in the first two consecutive consultations (n=110); BP values at the time of assessment for enrollment in the study were also recorded. Brachial BP was measured by sphygmomanometer after a 5-min rest period in a seated position and the mean of two consecutive readings was included in the analysis.

The echocardiographic exams were performed on a Vivid 7 scanner (M4S 1.75–3.5 MHz probe, GE® Medical Systems), and included conventional study, tissue Doppler and analysis of myocardial strain by speckle tracking.

In all exams, cine loops of three cardiac cycles were acquired in two-dimensional mode in left parasternal short-axis (mitral valve, papillary muscles and apex) and apical 4-, 3- and 2-chamber views at a frame rate of >60/s, and stored for subsequent off-line analysis of myocardial strain. The conventional study was performed in accordance with the recommendations of the American Society of Echocardiography and European Association of Echocardiography,13,14 and measurements were indexed to body surface area when appropriate. The mean of three measurements was used in the analysis. Ejection fraction and LV end-diastolic and end-systolic volumes were calculated using Simpson's biplane method.

Assessment of LV geometry was based on LV mass (calculated as 0.8×[LV internal diameter at end-diastole+posterior wall thickness at end-diastole+septal wall thickness at end-diastole]3+0.6 g, indexed to body surface area and gender) and relative wall thickness (calculated as 2×posterior wall thickness at end-diastole/LV internal diameter at end-diastole), and classified as one of four patterns – normal, concentric remodeling, concentric hypertrophy or eccentric hypertrophy – in accordance with the guidelines.13

Diastolic function was characterized by pulsed Doppler study of transmitral flow (assessing peak early filling [E-wave] and late diastolic filling [A-wave] velocities and deceleration time of early filling velocity), pulsed tissue Doppler measurement of early diastolic velocity (E’) at the septal and lateral mitral annulus (determining mean E’ [septal+lateral/2] and mean E/E’ ratio), and assessment of left atrial (LA) end-systolic volume.14–16

The four categories of diastolic function considered were: (1) normal diastolic function, defined as septal E’ ≥8, lateral E’ ≥10 and LA volume <34 l/m2; (2) grade I diastolic dysfunction, defined as septal E’ <8, lateral E’ <10, LA volume ≥34 l/m2, E/A <0.8, deceleration time >200 ms, and mean E/E’ ≤8; (3) grade II diastolic dysfunction, defined as septal E’ <8, lateral E’ <10, LA volume ≥34 l/m2, E/A 0.8–1.5, deceleration time 160–200 ms and mean E/E’ 9–12; and (4) grade III diastolic dysfunction, defined as septal E’ <8, lateral E’ <10, LA volume ≥34 l/m2, E/A ≥2, deceleration time <160 ms and mean E/E’ ≥13.14

Myocardial strain parameters were assessed by blinded off-line speckle tracking analysis on a GE® EchoPAC workstation. A semi-automatic myocardial tracking system was used, with manual demarcation of the endocardial border in end-systole and manual adjustment of the region of interest. Only myocardial segments considered to be of adequate quality by both the automatic system and the operator were included in the analysis.

Global longitudinal myocardial strain was assessed based on analysis of apical views. Systolic function was assessed by measurement of global longitudinal peak systolic strain (GLS [%]) and global longitudinal peak systolic strain rate (GLSR-s [s−1]). Diastolic function was assessed by measurement of global longitudinal peak early diastolic strain rate (GLSR-e [s−1]) and global longitudinal late diastolic strain rate (GLSR-a [s−1]).

Circumferential myocardial strain was assessed based on analysis of short-axis views. Global systolic function was assessed by measurement of global circumferential peak systolic strain (GCS [%]) and global circumferential peak systolic strain rate (GCSR-s [s−1]). Diastolic function was assessed by measurement of global circumferential peak early diastolic strain rate (GCSR-e [s−1]) and global circumferential peak late diastolic strain rate (GCSR-a [s−1]).

A descriptive statistical analysis was performed, and results presented in tables of absolute and relative frequencies. For quantitative variables measures of central location (mean and median) and of dispersion (standard deviation and interquartile range) were calculated. The Student's t test was used to compare the means of continuous variables with a normal distribution, and the Mann-Whitney nonparametric test was used to compare the medians of those with non-normal distribution. The chi-square test was used to compare percentages of qualitative variables. Odds ratios (OR) were calculated with a 95% confidence interval (CI).

In analysis of global myocardial strain parameters, cut-offs for normality were established based on the limit values of longitudinal and circumferential systolic myocardial strain parameters in the group of normotensives. The cut-off for LV systolic dysfunction was the upper limit (mean + 2 standard deviations) for longitudinal and circumferential strain parameters: GLS (%) and GLSR-s (s−1), and GCS (%) and GCSR-s (s−1), respectively; the closer negative values are to the baseline (zero), the greater the LV dysfunction. The degree of concordance was assessed using the kappa coefficient.

SPSS version 19 was used for the statistical analysis, using a level of significance of 5%.

The study included 229 hypertensive individuals (mean age 62±12 years, 53.7% male) with preserved ejection fraction, 68.1% of whom had been diagnosed with HTN more than five years previously. None were in New York Heart Association functional class ≥II. The demographic and clinical characteristics of the population are shown in Table 1.

To determine the limits of normality for global myocardial strain parameters, 20 healthy normotensive individuals were studied, with no significant differences in gender (55% male) or age distribution (mean age 59±7 years) from the hypertensive study population.

In the hypertensive group, 45.8% presented normal LV geometry, 9.7% concentric remodeling, 22.0% concentric hypertrophy and 22.5% eccentric hypertrophy; they also had greater LV mass (107±29 g/m2 vs. 74±13 g/m2, p<0.001) and relative wall thickness (0.40±0.07 vs. 0.31±0.04, p<0.001) than the normotensive group.

Alterations in diastolic function were detected in 52.4% (n=120), with grade I dysfunction in 30.1% (n=69) and grade II in 22.3% (n=51). None presented grade III dysfunction.

The quality of the cine loops acquired in two-dimensional mode was sufficient in all 229 hypertensive subjects to assess longitudinal myocardial strain and in 103 to assess circumferential strain.

The global longitudinal and circumferential myocardial strain parameters that reflect LV systolic function (GLS, GLSR-s, GCS and GCSR-s) showed no significant differences between the two groups. However, global diastolic strain parameters, both longitudinal (GLSR-e and GLSR-a) and circumferential (GCSR-e and GCSR-a), were significantly altered in the hypertensive group (Table 2).

The distribution of global systolic myocardial strain parameters as a function of LV geometry and diastolic function and LA pressure is shown in Table 3.

Mean GLS and GLSR-s in hypertensive individuals were −19.5±2.9% and −1.02±0.18 s−1, respectively. GLS varied significantly according to the presence of altered LV geometry, particularly reduced longitudinal systolic strain in those with concentric LV hypertrophy (p=0.019) (Figure 1). Furthermore, longitudinal systolic strain parameters were significantly reduced in individuals with increased LA pressure, but showed no correlation with severity of diastolic dysfunction.

Figure 1.

Frequency of reduced global longitudinal peak systolic strain according to left ventricular geometry (normal, concentric remodeling, concentric hypertrophy or eccentric hypertrophy) in normotensives and hypertensives. LVH: left ventricular hypertrophy.

(0.13MB).

Mean GCS and GCSR-s in hypertensive individuals were −19.5±3.9% and −1.26±0.3 s−1, respectively, and differed significantly according to the presence of diastolic dysfunction, although paradoxically they were more negative in those with more severe dysfunction (Figure 2).

Figure 2.

Frequency of reduced global longitudinal peak systolic strain according to diastolic function (normal, grade I or II dysfunction) in normotensives and hypertensives.

(0.11MB).

Twist was assessed in 95 hypertensives, with a mean value of 14±6°; it did not vary according to the degree of HTN, changes in LV geometry or diastolic dysfunction. There was however a significant correlation with GCS (Pearson's r −0.37, p<0.001; Spearman's rho −0.40, p<0.0019) and with GCSR-s (Pearson's r −0.28, p=0.003; Spearman's rho −0.30, p=0.001). In addition, twist showed a weak positive correlation with LV ejection fraction (Pearson's r 0.20, p=0.021; Spearman's rho 0.19, p=0.028) (Figure 3).

Figure 3.

A: correlation between twist and ejection fraction; B: correlation between twist and global circumferential peak systolic strain.

(0.2MB).

Myocardial strain parameters showed a normal distribution in the normotensive group and thus limits of normality could be established (−16.50% and −0.82 s−1 for GLS and GLSR-s, respectively). Applying these cut-offs in the hypertensive group revealed impaired longitudinal myocardial strain in 15.3% (n=35).

The limits of normality for GCS and GCSR-s were −12.17% and −0.81 s−1, respectively; applying these cut-offs in the hypertensive group revealed impaired circumferential myocardial strain in 4% (n=9).

The degree of concordance for detection of subclinical LV systolic dysfunction as assessed by the various parameters of myocardial strain was weak to moderate (Table 4).

Given that analysis of circumferential strain was only possible in a small group of individuals, and that the degree of concordance between these parameters for assessing subclinical systolic dysfunction was lower, myocardial strain was assessed based on longitudinal parameters only. Subclinical global systolic dysfunction was thus considered present when GLS was below the cut-off of −16.50% or GLSR-s was below −0.82 s−1.

Hypertensives with subclinical LV dysfunction presented more severe HTN, most of them with uncontrolled BP and LV structural changes; the likelihood of dysfunction was three times greater in those with concentric LV hypertrophy (OR 2.97, 95% CI 1.22–7.22, p=0.016) (Figure 4). Those with subclinical LV dysfunction also presented more severe diastolic dysfunction and increased filling pressures (Table 5).

Figure 4.

Distribution of subclinical ventricular dysfunction according to type of ventricular geometry. LVH: left ventricular hypertrophy.

(0.1MB).

In the hypertensives with subclinical LV dysfunction, ORs were calculated according to the severity and control of HTN. Individuals with grade I (OR 5.22, 95% CI 1.16–23.51, p=0.031) or grade II HTN (OR 9.20, 95% CI 1.84–46.06, p=0.007) on assessment for enrollment in the study, as well as those with uncontrolled BP (OR 3.28, 95% CI 1.56–6.23, p<0.001), showed a greater likelihood of developing subclinical LV dysfunction (Table 5).

One limitation of the study is the fact that measurements were not directly validated by cardiac magnetic resonance imaging (MRI). However, Amundsen et al.28 demonstrated a good correlation between speckle tracking echocardiography and MRI.

A second limitation is that the cut-offs used were established on the basis of a small sample of normotensives, but this method was chosen in order to match the characteristics of the hypertensive group in terms of age, gender and ethnicity.

The fact that HTN had been medicated and controlled for a long period means that the study population may not be truly representative of the whole spectrum of hypertensive heart disease. However, the aim of the study was to assess precisely this subgroup, since detection of subclinical LV dysfunction is particularly important in such patients in order to enable earlier intervention and possibly improve their prognosis.

The authors declare that no experiments were performed on humans or animals for this study.

The authors declare that they have followed the protocols of their work center on the publication of patient data and that all the patients included in the study received sufficient information and gave their written informed consent to participate in the study.

The authors have obtained the written informed consent of the patients or subjects mentioned in the article. The corresponding author is in possession of this document.