Background
The aim of this study was to compare speckle-tracking echocardiography–derived left ventricular (LV) systolic mechanics and their relationships with LV diastolic properties in young patients with hypertension and in young competitive athletes in relation to their respective alterations of LV structure.
Methods
Nineteen sedentary controls, 22 top-level rowers, and 18 young newly diagnosed, never-treated patients with hypertension, all male, underwent Doppler echocardiography including pulsed tissue Doppler of the mitral annulus and speckle-tracking echocardiography. Peak longitudinal strain was calculated in apical long-axis, four-chamber, and two-chamber views, and values of the three views were averaged (global longitudinal strain [GLS]). Regional circumferential and radial strain were calculated at the LV basal, middle, and apical levels, and values were averaged (global circumferential strain and global radial strain). LV torsion was determined as the net difference in the mean rotation between the apical and basal levels.
Results
The three groups were comparable for age, whereas body mass index and blood pressure were higher in patients with hypertension, and heart rate was lower in rowers. LV mass index was higher in rowers and in patients with hypertension than in controls, without differences in relative wall thickness, ejection fraction, and midwall shortening. Left atrial volume index was greater in rowers than in controls and patients with hypertension. Annular systolic velocity (s′) ( P < .001) and early diastolic velocity (e′) ( P < .0001) were lower and the E/e′ ratio was higher ( P < .0001) in patients with hypertension. GLS was lower in patients with hypertension (−17.5 ± 2.8%) than in rowers (−22.2 ± 2.7%) and in controls (−21.1 ± 2.0%) ( P < .0001). Global circumferential strain, global radial strain, and torsion were similar among the three groups. In the pooled population, GLS was an independent contributor to E/e′ ratio ( P < .0001) after adjusting for age, heart rate, meridional end-systolic stress, LV mass index and left atrial volume index. By receiver operating characteristic curve analyses, both GLS and E/e′ ratio appeared to be accurate in discriminating patients with hypertension from healthy controls, with the E/e′ ratio being more sensitive (77.8%) and GLS more specific (89.5%).
Conclusions
The hearts of young patients with hypertension are characterized by reduced GLS, whereas global circumferential strain, global radial strain, and torsion are similar to those of athletes’ hearts. The extent of GLS is strongly associated with LV diastolic function, independently of afterload changes and the degree of LV hypertrophy.
Athlete’s heart is a left ventricular (LV) adaption to long-term intensive endurance training characterized by increases in chamber size, wall thickness, and LV mass. The increased afterload associated with arterial hypertension induces apparently similar changes in the left ventricle, and LV hypertrophy is considered the main hallmark of cardiac involvement in the hypertensive population. Differential LV mass increase in athlete’s and hypertensive heart involve the myocardial composition, because it is sustained by an increase of active muscular mass in athletes, whereas myocardial fibrosis and the associated alterations of the extracellular matrix are the major determinants of hypertensive heart, even before the development of clear-cut LV hypertrophy. In general, it is widely accepted that LV diastolic properties are the principal functional aspect by which athlete’s and hypertensive heart can be discriminated at rest: Doppler echocardiography allows the detection of a pattern of delayed LV relaxation in patients with hypertension, while both LV relaxation and filling are supernormal in trained athletes. Pulsed tissue Doppler can be further useful for this distinction: although early diastolic velocity of the mitral annulus (e′) is one of the earliest markers for impaired relaxation, the ratio of transmitral early diastolic velocity to e′ (E/e′) correlates well with LV filling pressures and has been shown to be a prognostic indicator for cardiovascular disease, including arterial hypertension.
LV ejection fraction (EF) at rest cannot distinguish the structural modifications of patients with hypertension and athletes. However, when EF is still normal, structural changes of hypertensive heart (LV concentric geometry) are accompanied by alterations of diastolic properties, a parallel depression of midwall mechanics of the circumferential fibers, and a subclinical reduction of tissue Doppler–derived systolic longitudinal function. Today, both regional and global function may be further analyzed using speckle-tracking echocardiography (STE), a non-Doppler ultrasound technique that allows a semiautomated, quantitative analysis of myocardial deformation in three spatial directions—longitudinal, radial, and circumferential—all contributing to global LV mechanics, and offers an evaluation of the entity and direction of LV rotation. There is an ongoing debate regarding the most sensitive deformation measure to detect early LV impairment in different cardiac pathologies. Accordingly, the present study was designed to compare STE-derived LV systolic mechanics and their relationships with LV diastolic properties in young patients with native hypertension and in top-level, competitive rowers in relation to their respective alterations of LV structure and with reference to a healthy, sedentary reference group.
Methods
Study Population
Exclusion criteria were diabetes mellitus, congestive heart failure, coronary artery disease (history of angina and inducible myocardial ischemia in stress echocardiography or single photon-emission computed tomography), valvular heart disease, atrial fibrillation, the use of any cardioactive medications, inadequate echocardiographic quality (inadequate acquisition of basal and/or middle and/or apical short-axis view for STE and/or LV torsion calculation in two rowers, three controls, and four patients with hypertension). After exclusions, the study population consisted of 19 untrained healthy controls; 22 top-level, competitive rowers; and 18 patients with native arterial hypertension, all male. Healthy controls were recruited from the personnel of our university and their relatives. To be eligible, rowers had to be engaged in a mixed training program (isotonic and isometric) ≥30 hours/week in the past 5 years. All the rowers had been awarded medals at national and/or international competitions. They were examined during an intensive training period, ≥24 hours after their last athletic activity. Patients with hypertension were recruited according to the diagnostic criteria of the current European Society of Hypertension and European Society of Cardiology recommendations. To be eligible, patients had to be newly diagnosed, never treated pharmacologically, and aged ≤ 45 years. All participants gave written informed consent, and the study was approved by the local ethics committee.
Procedures
Standard Doppler echocardiographic, pulsed tissue Doppler, and speckle-tracking echocardiographic examinations were performed using a Vivid 7 ultrasound scanner (GE Healthcare, Milwaukee, WI) using a 2.5-MHz transducer with harmonic capability, according to the standards of our laboratory. At the end of echocardiographic examination, heart rate and blood pressure (BP) (average of three measurements by a cuff sphygmomanometer) were recorded.
Standard Echocardiographic Examination
The quantitative analysis of the left ventricle was performed according to the current recommendations of the American Society of Echocardiography and the European Association of Echocardiography. LV EF and stroke volume were derived from LV end-diastolic and end-systolic volumes obtained as the average of the measurements in apical four-chamber and two-chamber views according to the modified Simpson’s rule. Midwall fractional shortening was also calculated as a further measurement of global systolic function. LV mass was indexed to height 2.7 (LV mass index [LVMi]). Relative diastolic wall thickness was determined as twice the posterior wall thickness divided by LV end-diastolic diameter. Circumferential end-systolic stress and meridional end-systolic stress (ESSm) were computed according to standardized methodology. Left atrial volume (area-length method) was determined as the average of measurements in apical four-chamber and two-chamber views and indexed to body surface area (left atrial volume index [LAVi]). Transmitral pulsed Doppler and tissue Doppler were recorded in the apical four-chamber view. The pulsed tissue Doppler sample volume was placed at the level of both the basal lateral and the basal septal mitral annulus. Annular peak systolic velocity (s′) and annular peak early diastolic relaxation velocity (e′) were measured at both the annular sites and averaged (average s′ and average e′, respectively). The ratio of transmitral peak early velocity (E) to average e′ (E/e′), an index shown to be a reliable estimate of invasively determined LV filling pressure in heart failure and a prognosticator for cardiovascular disease, was also calculated.
STE
STE was performed according to validated methods on three consecutive cardiac cycles of two-dimensional LV images in apical (long-axis and four-chamber and two-chamber) and parasternal short-axis views (at the base, just below the mitral level; at the papillary muscle level; and at the apex, just proximal to the level with LV cavity obliteration at end-systole). Reliable recording of two-dimensional images for speckle-tracking echocardiographic analysis requires a minimum frame rate of 30 Hz, but in the present study, frame rates of 50 to 70 Hz were obtained by acquiring the LV cavity with the narrowest scan and at the lowest possible depth to display on the screen the left ventricle as large as possible. The same field depth was kept for all the chamber views. Images were digitally stored on hard disk. Custom acoustic tracking software allowing semiautomated, two-dimensionally derived strain analysis (EchoPAC Advanced Analysis Technologies; GE Healthcare) was applied to two-dimensional grayscale images by tracking movements of “speckles” in myocardial tissue, frame by frame, throughout the cardiac cycle. The software is interactive, because the endocardial-cavity interface is traced manually while a second epicardial tracing is automatically generated. The software automatically divides each image into six myocardial segments and accepts segments of good tracking quality while rejecting poorly tracked segments, allowing the observer to manually override its decision at the same time using visual assessment. The aortic valve closure time is defined in the apical long-axis view and used subsequently as a reference point in the other views. For the aim of the present study, the peak negative systolic longitudinal strain was assessed from six segments in the posterior and anteroseptal walls (apical long-axis view), posterior septal and lateral walls (four-chamber view), and inferior and anterior walls (two-chamber view). Global longitudinal strain (GLS) was calculated by averaging all values of regional peak longitudinal strain obtained in each apical view before aortic valve closure. Global circumferential strain (GCS) and global radial strain (GRS) were obtained as the average of the regional values measured in the six myocardial segments of basal, middle, and apical parasternal short-axis views. Basal-to-apical torsion was calculated as the net difference in LV rotation angle at the apical (counterclockwise, positive angles) and basal (clockwise, negative angles) short-axis plane occurring at end-systole. In addition, longitudinal strain was calculated by levels (average of six basal, six middle, and six apical LV segments) and also circumferential and radial strain calculated separately at the basal, middle, and apical levels of the left ventricle (average of the six segments at each level).
Statistical Analysis
Statistical analysis was performed using SPSS version 12 (SPSS, Inc, Chicago, IL). Data are presented as mean ± SD. Descriptive statistics were obtained using one-factor analysis of variance (post hoc analysis by Bonferroni’s test) and χ 2 distribution with the computation of exact P values using the Monte Carlo method. Least squares linear regression was used to evaluate univariate and multivariate correlates of a given variable. Multiple linear regression analysis was used to identify the independent contributors of E/e′ ratio in the pooled population. In this model, potential multicollinearity (computation of in-model tolerance) was also taken into account and collinearity considered acceptable for tolerance > 0.70. The null hypothesis was rejected at P < .05.
Results
The characteristics of the study population are listed in Table 1 . Age was not significantly different among the three groups, whereas body mass index and BP values (as measured at the end of the echocardiographic examination) were higher in patients with hypertension, and heart rate was lower in rowers.
| Healthy controls | Rowers | Patients with hypertension | ||
|---|---|---|---|---|
| Variable | ( n = 19) | ( n = 22) | ( n = 18) | P |
| Age (years) | 28.5 ± 6.6 | 27.7 ± 8.4 | 32.7 ± 6.1 | NS |
| BMI (kg/m 2 ) | 24.5 ± 3.0 | 24.6 ± 1.9 | 27.0 ± 2.7 ∗ § | <.005 |
| Systolic BP ∗ (mm Hg) | 124.5 ± 10.9 | 129.9 ± 11.6 | 146.9 ± 10.5 ‡ ‖ | <.0001 |
| Diastolic BP ∗ (mm Hg) | 77.9 ± 8.2 | 73.3 ± 9.4 | 96.9 ± 6.9 ‡ ‖ | <.005 |
| Mean BP ∗ (mm Hg) | 93.4 ± 8.3 | 92.2 ± 9.2 | 113.6 ± 6.8 ‡ ‖ | <.0001 |
| HR (beats/min) | 62.7 ± 11.7 | 55.9 ± 10.2 ¶ | 67.3 ± 7.4 † | <.005 |
‡ P < .0001 for rowers vs patients with hypertension.
‖ P < .0001 for healthy controls vs patients with hypertension.
Doppler echocardiographic analysis is summarized in Table 2 . LVMi was significantly higher in rowers and in patients with hypertension than in controls, without a significant difference of relative wall thickness among the three groups. The prevalence of clear-cut LV hypertrophy (LVMi > 44 g/m 2.7 in women and LVMi > 48 g/m 2.7 in men ) was 20.3% (12 of 59) in the pooled population: 36.4% (8 of 22) in rowers and 22.2% (4 of 18) in patients with hypertension (χ 2 test, P = .53). Only one healthy control and one patient with hypertension had relative wall thicknesses > 0.42, indicative of LV concentric geometry (data not shown). ESSm and circumferential end-systolic stress were higher in patients with hypertension than in the other two groups (both P values < .01). In presence of similar EF and midwall fractional shortening, stroke volume was higher in rowers. Standard Doppler-derived transmitral E/A ratios were significantly lower in patients with hypertension than in rowers, reflecting the higher contribution of A velocity, without a difference in deceleration time and isovolumic relaxation time. Tissue Doppler-derived s′ average ( P < .001) and e′ average ( P < .0001) were lower and the E/e′ ratio was higher ( P < .0001) in patients with hypertension. According to the recommendations for the evaluation of LV diastolic dysfunction by echocardiography, 15 of the 28 patients with hypertension had normal diastolic function, two had a grade I diastolic dysfunction (pattern of delayed relaxation: E/A ratio < 1 and E/e′ ratio < 8), and only one patient had an E/e′ ratio > 13, a recognized cutoff point for increased LV filling pressure (data not shown). LAVi was higher in rowers and patients with hypertension than in controls.
| Variable | Healthy controls | Rowers | Patients with hypertension | P |
|---|---|---|---|---|
| IVSTd (mm) | 8.4 ± 1.2 | 9.9 ± 1.0 ¶ | 9.1 ± 2.0 § | <.01 |
| PWTd (mm) | 7.7 ± 1.1 | 9.0 ± 1.0 ∗∗ | 8.3 ± 1.1 § | <.001 |
| LVEDV (mL) | 133.9 ± 33.0 | 161.3 ± 33.3 ¶ | 131.4 ± 44.0 ∗ | <.02 |
| LVESV (mL) | 49.8 ± 18.0 | 60.3 ± 21.5 | 52.3 ± 33.9 | NS |
| LVMi (g/m 2.7 ) | 30.8 ± 5.6 | 41.1 ± 8.1 # | 37.0 ± 11.2 § | <.0001 |
| RWTd | 0.30 ± 0.06 | 0.32 ± 0.04 | 0.33 ± 0.07 | NS |
| EF (%) | 61.7 ± 6.8 | 66.3 ± 6.9 | 61.6 ± 5.9 | NS |
| Midwall FS (%) | 18.9 ± 2.5 | 18.3 ± 2.0 | 17.6 ± 3.4 | NS |
| ESSm (g/cm 2 ) | 43.8 ± 14.3 | 40.0 ± 9.9 | 50.4 ± 17.1 ∗ § | <.05 |
| ESSc (g/cm 2 ) | 82.4 ± 24.0 | 78.8 ± 16.8 | 96.9 ± 30.9 ∗ § | <.05 |
| SV (mL) | 84.2 ± 19.3 | 101.0 ± 16.1 ¶ | 79.1 ± 20.0 † | <.001 |
| LAVi (mL/m 2 ) | 26.7 ± 7.3 | 36.6 ± 8.7 ∗∗ | 29.4 ± 6.3 ∗ | <.0001 |
| E peak velocity (m/s) | 0.78 ± 0.12 | 0.80 ± 0.13 | 0.78 ± 0.16 | NS |
| A peak velocity (m/s) | 0.52 ± 0.12 | 0.44 ± 0.10 | 0.62 ± 0.13 ‡ § | <.0001 |
| E/A ratio | 1.61 ± 0.48 | 1.92 ± 0.53 # | 1.31 ± 0.40 † | <.001 |
| DT (ms) | 172.8 ± 23.3 | 184.1 ± 33.7 | 185.0 ± 32.8 | NS |
| IVRT (ms) | 80.1 ± 17.2 | 79.4 ± 15.8 | 80.4 ± 17.8 | NS |
| Sa average (cm/s) | 9.3 ± 1.7 | 10.6 ± 2.0 | 8.1 ± 1.7 ‡ | <.001 |
| Ea average (cm/s) | 13.9 ± 2.7 | 15.5 ± 3.0 | 10.8 ± 3.7 ‡ § | <.0001 |
| E/Ea ratio | 5.8 ± 1.1 | 5.2 ± 7.4 | 7.9 ± 3.0 ‡ ‖ | <.0001 |
‡ P < .0001 for rowers vs patients with hypertension.
‖ P < .001 for healthy controls vs patients with hypertension.
Table 3 reports speckle-tracking echocardiographic analysis of GLS, GCS, GRS, and torsion, and Table 4 shows the average values of longitudinal, circumferential, and radial strain for levels (basal, middle, and apical) of the left ventricle. Although GLS was lower in patients with hypertension than in rowers and in controls ( P < .0001), GCS and GRS were similar in the three groups, and despite a trend toward higher values in patients with hypertension, LV torsion also did not differ significantly among the three groups ( Table 3 ). The average values of circumferential and radial strain at the different levels (basal, middle, apical) did not differ among the three groups, while the reduction of longitudinal strain of patients with hypertension in comparison with both normal controls and athletes was greater at the basal and middle levels than at the apical level ( Table 4 ).
| Variable | Healthy controls | Rowers | Patients with hypertension | P |
|---|---|---|---|---|
| GLS (%) | −21.1 ± 2.0 | −22.2 ± 2.7 | −17.5 ± 2.8 ∗ † | <.0001 |
| GLE (%) | −10.0 ± 0.8 | −10.7 ± 1.1 | −7.5 ± 0.8 ∗ † | <.0001 |
| GLA (%) | −4.4 ± 1.1 | −3.6 ± 1.1 | −7.1 ± 1.6 ∗ † | <.0001 |
| GCS (%) | −17.6 ± 2.9 | −17.7 ± 2.5 | −16.4 ± 4.3 | NS |
| GRS (%) | 46.4 ± 15.8 | 47.6 ± 19.1 | 46.0 ± 17.3 | NS |
| Basal rotation (°) | 3.7 ± 0.5 | 2.9 ± 1.5 | 3.7 ± 2.0 | NS |
| Apical rotation (°) | 6.2 ± 1.4 | 6.1 ± 2.3 | 7.5 ± 1.9 | NS |
| Torsion (°) | 9.7 ± 1.8 | 9.2 ± 2.0 | 10.8 ± 3.4 | NS |
∗ P < .0001 for rowers vs patients with hypertension.
† P < .0001 for healthy controls vs patients with hypertension.
| Variable | Healthy controls | Rowers | Patients with hypertension | P |
|---|---|---|---|---|
| Basal longitudinal strain (%) | −20.9 ± 2.3 | −22.0 ± 2.9 | −17.5 ± 2.8 † ‡ | <.0001 |
| Middle longitudinal strain (%) | −20.9 ± 1.9 | −22.1 ± 2.8 | −17.4 ± 2.8 † § | <.0001 |
| Apical longitudinal strain (%) | −21.2 ± 2.1 | −22.4 ± 2.6 | −17.6 ± 3.1 ∗ ‡ | <.001 |
| Basal circumferential strain (%) | −16.7 ± 2.7 | −16.8 ± 2.4 | −13.5 ± 8.6 | NS |
| Middle circumferential strain (%) | −18.2 ± 3.3 | −18.8 ± 2.6 | −17.3 ± 3.2 | NS |
| Apical circumferential strain (%) | −17.8 ± 2.9 | −17.8 ± 2.6 | −16.7 ± 3.3 | NS |
| Basal radial strain (%) | 46.7 ± 13.6 | 48.1 ± 17.0 | 49.6 ± 19.4 | NS |
| Middle radial strain (%) | 44.9 ± 13.9 | 45.8 ± 17.0 | 46.8 ± 19.1 | NS |
| Apical radial strain (%) | 44.0 ± 12.2 | 44.0 ± 16.9 | 46.9 ± 19.0 | NS |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
