This study aimed to evaluate left ventricular structure, function and mechanics, in highly-trained, pre-adolescent soccer players compared with age- and sex-matched controls.
The study design was a prospective, cross-sectional comparison of left ventricular structure, function and mechanics.
Twenty-two male soccer players from two professional youth soccer academies (age: 12.0 ± 0.3 years) and 22 recreationally active controls (age: 11.7 ± 0.3 years) were recruited. Two-dimensional conventional and speckle tracking echocardiography were used to quantify left ventricular structure, function and peak/temporal values for left ventricular strain and twist, respectively.
End-diastolic volume index was larger in soccer players (51 ± 8 mm/(m2)1.5 vs. 45 ± 6 mm/(m2)1.5; p = 0.007) and concentricity was lower in soccer players (4.3 ± 0.7 g/(mL)0.667 vs. 4.9 ± 1.0 g/(mL)0.667; p = 0.017), without differences in mean wall thickness between groups (6.0 ± 0.4 mm vs. 6.1 ± 0.5 mm; p = 0.754). Peak circumferential strain at the base (–22.2% ± 2.5% vs. –20.5% ± 2.5%; p = 0.029) and papillary muscle levels (–20.1% ± 1.5% vs. –18.3% ± 2.5%; p = 0.007) were greater in soccer players. Peak left ventricular twist was larger in soccer players (16.92° ± 7.55° vs. 12.34° ± 4.99°; p = 0.035) and longitudinal early diastolic strain rate was greater in soccer players (2.22 ± 0.40 s–1 vs. 2.02 ± 0.46 s–1; p = 0.025).
Highly-trained soccer players demonstrated augmented cardiac mechanics with greater circumferential strains, twist and faster diastolic lengthening in the absence of differences in wall thickness between soccer players and controls.
Introduction
The match-play and training characteristics of soccer1,2 present an ecologically valid model to study the effects of systematic high intensity intermittent exercise training on cardiac structure and function in young athletes.
A recent meta-analysis has reported larger left ventricular (LV) diameter and wall thickness yet similar conventionally derived indices of systolic and diastolic cardiac function in adolescent athletes compared with non-athletes.3 Additionally, cardiac enlargement increased with chronological age, suggesting a potential role of hormones in pubertal adaptation. Importantly, the influence of exercise training on cardiac structure and function in pre-adolescent athletes may be ascertained without the confounding factors of growth and maturation. However, the effects of soccer training in these younger athletes are less clear, with some studies reporting similar absolute and scaled wall thicknesses between pre-adolescent athletes and controls.4,–6
Our group recently documented LV structure and function in highly-trained pre-adolescent athletes, with a particular focus on conventional indices of LV function.7 Speckle tracking echocardiography (STE) comprehensively assesses LV mechanics by quantifying deformation in the longitudinal, circumferential and radial planes, as well as rotation and twist.8,9 Studies using STE to compare young athletes with age-matched controls have been conflicting,10,–12 likely due to variations in maturity status, the sex of the athletes and disparate sports being studied. Some, however, have reported lower longitudinal strain12 and augmented twist.13 Thus, there is a paucity of data detailing LV mechanics in pre-adolescent soccer players which warrants further investigation. Assessment of myocardial strains using STE will further our understanding on the coupling between LV structure and function in this population. Additionally, STE can facilitate temporal mechanical data that will extend our understanding of strains and rotations throughout the cardiac cycle.
Taken together, with the increased professionalisation and subsequent increased training loads in elite youth soccer, at very early ages, there is a need to further interrogate global and regional markers of LV structure, function and mechanics. It was hypothesised that (1) LV structure would not differ between pre-adolescent soccer players and controls; (2) LV longitudinal strain would be lower in soccer players compared with controls, while (3) LV twist mechanics would be greater in soccer players than in controls.
Methods
Participants
Twenty-two highly trained male youth soccer players (age: 12.0 ± 0.3 years) and 22 recreationally active males (controls, mean age: 11.7 ± 0.3 years) were recruited to the study. Two of the soccer players were British African-Caribbean with the remainder Caucasian. Similarly, one of the controls was of British African-Caribbean origin and the remainder were Caucasian.
The soccer players’ training profiles were as follows: 4.5 ± 1.5 years of training, 11 ± 1 months per year training, 4 ± 1 training sessions per week and 9.4 ± 2.4 h per week of training. This volume of exercise training had been consistent for the entirety of their active training years. Soccer players played one competitive match per week and had been engaged in competitive soccer matches for 4 ± 2 years. The soccer players were recruited from two Category one English Premier league youth soccer academies. For one club, 14 boys from the U12 squad and their parents were approached, of which three were not enrolled because of either personal circumstances (n = 2) or a football related injury (n = 1). At the second club, researchers provided information to 15 U12 players and their parents, of which two were recovering from injury, one was released from the club after signing up from the study, and one signed up and simply did not attend the testing. Accordingly, all recruitment was consecutive and included 11 participants from both clubs, with a total of 22 soccer players. Control participants took part in compulsory physical education of 2 h per week (the same as soccer players), were all recreationally active and without engagement of systematic training. The controls self-reported 1.53 ± 1.77 h per week of physical activity.
Written informed parental and participant consent was obtained prior to participation. All procedures performed in the study were in accordance with the Declaration of Helsinki and the study was reviewed and approved by Staffordshire University Ethics Committee.
Protocol/measurements
Participants were asked to refrain from physical activity for 12 h prior to the visit. Physical activity and training questionnaires14 were completed prior to the testing. Following this, stature and body mass were measured. Maturity status was quantified using maturity offset.7,15 Resting arterial blood pressure was recorded in the left arm by an automated blood pressure cuff (Boso, Medicus, Jungingen, Germany) and heart rate was assessed by a 12-lead electrocardiogram (ECG) (CardioExpress SL6, Spacelabs Healthcare, Washington, USA). No abnormalities were detected from the ECG recordings in the participants included within the final analysis. Resting echocardiographic measurements were taken in the left lateral decubitas position. Body surface area (BSA) was calculated by the Mosteller formula.16
Two-dimensional echocardiography
2D echocardiographic procedures were performed by two sonographers (DO (soccer players) and DO + RL (controls)) using a commercially available ultrasound system (VividQ Ultrasound System, GE Ltd, Horton, Norway) and images were analysed offline (EchoPac version 6.0, GE Ltd, Horton, Norway). Conventional measurements of resting LV dimensions and volumes (LV end-diastolic dimension (LVEDd), LV end-systolic dimension, LV end-diastolic volume (LVEDV), LV end-systolic volume) and the subsequent calculations of LV mass and relative wall thickness were made in accordance with American Society of Echocardiography guidelines17 and have been reported previously.7 Linear LV dimensions were scaled to BSA0.5, LV mass to height2.7 and volumes to BSA1.5.18 Concentricity was calculated as LV mass divided by allometrically scaled LVEDV (LVEDV0.667).19 Sphericity index was calculated as LV length divided by LVEDd.20
Peak mitral inflow velocities and pulsed wave tissue Doppler imaging (TDI) were assessed as previously reported.7 E/E′ was calculated as an estimate of LV filling pressure21 from the average of septal and lateral E′.
Stroke volume (SV) and ejection fraction were calculated using Simpson’s biplane method with cardiac output (Q) determined by multiplying SV by the ECG determined heart rate. Both Q and SV were adjusted for BSA (Qindex and SVindex).
LV mechanics were determined from 2D images with frame rates maintained as high as possible within the range of 40 to 90 frames/s. The cardiac cycle with the most defined endocardial border was used for analysis. Adjustments in frequency and gain were used to optimise endocardial delineation, with a single focal zone placed mid LV cavity to reduce the impact of beam divergence. Aortic valve closure was identified from the pulsed wave Doppler of LV outflow tract flow and used to signify end systole. Offline analysis using dedicated speckle tracking software (Echopac V6.0, GE Healthcare, Horton, Norway) provided assessment of LV strains, strain rate, rotations and net twisting.
Longitudinal strain was determined using the apical four-chamber view. Global values were calculated as an average of six myocardial segments from the basal, mid and apical septum and lateral walls. Circumferential and radial strains were determined from the segmental average at the basal and mid-papillary levels. Using the mitral valve and apical levels, LV rotations were determined and twist was calculated as the difference between clockwise basal and counter-clockwise apical rotations during systole. In all instances, peak values and temporal analyses were obtained by importing stored traces into a spreadsheet (Microsoft Corporation, Washington, USA) with a cubic spline add-in (SRS1 software, Boston, USA). Data were normalised to 5% increments during systole and diastole. Good reliability of LV mechanics data has also been established by this research team.9
Statistical analysis
Normality of data was assessed using Shapiro–Wilk. For normally distributed data, a Student’s independent t-test was used to compare LV structure, function and mechanics in soccer players and controls. For non-normally distributed data, a Mann–Whitney U test was employed. The same procedures were used for temporal analysis at each 5% increment independently. A sample size of 22 soccer players provided a (1 -β) of 80% at an alpha level of 0.05. Statistical significance was granted at p < 0.05. Statistical analyses were performed using jamovi (version 0.9).22
Results
Missing data
Data were not obtained in controls (n = 1) due to poor image quality in indices derived from Simpson’s Biplane. Tissue Doppler and longitudinal strain indices were absent for controls (n = 1). Due to >2 segments excluded due to poor tracking in the apical plane, LV twist mechanics were not available for soccer players (n = 1) and controls (n = 4).
Physical characteristics
Soccer players were chronologically slightly older (p < 0.05) than controls (12.0 ± 0.3 and 11.7 ± 0.3 years), but maturity offset (–2.1 ± 0.6 and –2.1 ± 0.6 years) and age at peak height velocity (14.0 ± 0.5 and 13.9 ± 0.6 years) were similar between soccer players and controls (both p > 0.05). Similarly, stature (1.51 ± 0.06 and 1.49 ± 0.07 m) and body mass (40.2 ± 5.8 and 44.0 ± 11.7 kg) were similar between soccer players and controls, respectively (both p > 0.05). There were no inter-group differences for systolic (soccer players: 100 ± 8 mmHg; controls: 105 ± 13 mmHg) and diastolic blood pressure (soccer players: 61 ± 9 mmHg; controls: 61 ± 10 mmHg).
Conventional LV structure and function
Absolute and scaled SV, as well as scaled LVEDV, were greater in soccer players than controls (all p < 0.05). Absolute LV diameters, mean wall thickness and mass were not different between soccer players and controls (all p > 0.05), whereas concentricity was lower and sphericity index greater in soccer players than in controls, respectively (all p < 0.05, Table 1).
Left ventricular structure and volumes in soccer players and controls.
| SP | CON | p-value | |
|---|---|---|---|
| HR (beats/min) | 65 ± 8 | 74 ± 10 | 0.005 |
| LVEDd (mm) | 42 ± 4 | 44 ± 4 | 0.142 |
| LVEDd index (mm/(m2)0.5) | 37 ± 3 | 38 ± 3 | 0.343 |
| LVESd (mm) | 28 ± 3 | 29 ± 3 | 0.403 |
| LVESd index (mm/(m2)0.5) | 25 ± 3 | 25 ± 2 | 0.725 |
| MWT (mm) | 6.0 ± 0.4 | 6.1 ± 0.5 | 0.754 |
| MWT (mm/(m2)0.5) | 5.3 ± 0.4 | 5.3 ± 0.5 | 0.769 |
| RWT | 0.29 ± 0.04 | 0.28 ± 0.04 | 0.387 |
| LVM (g) | 75 ± 14 | 82 ± 18 | 0.204 |
| LVM index (g/m2) | 25 ± 5 | 28 ± 7 | 0.051 |
| LV length (mm) | 76 ± 6 | 74 ± 5 | 0.316 |
| LV length index (mm/(m2)0.5) | 67 ± 5 | 64 ± 5 | 0.140 |
| LVEDV (mL) | 75 ± 10 | 69 ± 15 | 0.106 |
| LVEDV index (mL/(m2)1.5) | 51 ± 8 | 45 ± 6 | 0.007 |
| LVESV (mL) | 26 ± 4 | 26 ± 8 | 0.696 |
| LVESV index (mL/(m2)1.5) | 18 ± 2 | 17 ± 3 | 0.625 |
| Concentricity (g/mL)0.667) | 4.3 ± 0.7 | 4.9 ± 1.0 | 0.017 |
| Sphericity index | 1.8 ± 0.2 | 1.7 ± 0.1 | 0.034 |
| SV (mL) | 49 ± 8 | 43 ± 10 | 0.031 |
| SV index (mL/(m2)1.0) | 38 ± 6 | 32 ± 5 | 0.002 |
| (L/min) | 3.19 ± 0.63 | 3.12 ± 0.67 | 0.741 |
| Cardiac index (L/min per (m2)1.0) | 2.48 ± 0.46 | 2.36 ± 0.42 | 0.391 |
| SP | CON | p-value | |
|---|---|---|---|
| HR (beats/min) | 65 ± 8 | 74 ± 10 | 0.005 |
| LVEDd (mm) | 42 ± 4 | 44 ± 4 | 0.142 |
| LVEDd index (mm/(m2)0.5) | 37 ± 3 | 38 ± 3 | 0.343 |
| LVESd (mm) | 28 ± 3 | 29 ± 3 | 0.403 |
| LVESd index (mm/(m2)0.5) | 25 ± 3 | 25 ± 2 | 0.725 |
| MWT (mm) | 6.0 ± 0.4 | 6.1 ± 0.5 | 0.754 |
| MWT (mm/(m2)0.5) | 5.3 ± 0.4 | 5.3 ± 0.5 | 0.769 |
| RWT | 0.29 ± 0.04 | 0.28 ± 0.04 | 0.387 |
| LVM (g) | 75 ± 14 | 82 ± 18 | 0.204 |
| LVM index (g/m2) | 25 ± 5 | 28 ± 7 | 0.051 |
| LV length (mm) | 76 ± 6 | 74 ± 5 | 0.316 |
| LV length index (mm/(m2)0.5) | 67 ± 5 | 64 ± 5 | 0.140 |
| LVEDV (mL) | 75 ± 10 | 69 ± 15 | 0.106 |
| LVEDV index (mL/(m2)1.5) | 51 ± 8 | 45 ± 6 | 0.007 |
| LVESV (mL) | 26 ± 4 | 26 ± 8 | 0.696 |
| LVESV index (mL/(m2)1.5) | 18 ± 2 | 17 ± 3 | 0.625 |
| Concentricity (g/mL)0.667) | 4.3 ± 0.7 | 4.9 ± 1.0 | 0.017 |
| Sphericity index | 1.8 ± 0.2 | 1.7 ± 0.1 | 0.034 |
| SV (mL) | 49 ± 8 | 43 ± 10 | 0.031 |
| SV index (mL/(m2)1.0) | 38 ± 6 | 32 ± 5 | 0.002 |
| (L/min) | 3.19 ± 0.63 | 3.12 ± 0.67 | 0.741 |
| Cardiac index (L/min per (m2)1.0) | 2.48 ± 0.46 | 2.36 ± 0.42 | 0.391 |
Data are mean ± standard deviation.
SP: soccer players; CON: controls; HR: heart rate; LVEDd: left ventricular end-diastolic diameter; LVESd: left ventricular end-systolic diameter; MWT: mean wall thickness; RWT: relative wall thickness; LVM: left ventricular mass; LV: left ventricular; LVEDV: left ventricular end-diastolic volume; LVESVL: left ventricular end-systolic volume; SV: stroke volume; : cardiac output.
Left ventricular structure and volumes in soccer players and controls.
| SP | CON | p-value | |
|---|---|---|---|
| HR (beats/min) | 65 ± 8 | 74 ± 10 | 0.005 |
| LVEDd (mm) | 42 ± 4 | 44 ± 4 | 0.142 |
| LVEDd index (mm/(m2)0.5) | 37 ± 3 | 38 ± 3 | 0.343 |
| LVESd (mm) | 28 ± 3 | 29 ± 3 | 0.403 |
| LVESd index (mm/(m2)0.5) | 25 ± 3 | 25 ± 2 | 0.725 |
| MWT (mm) | 6.0 ± 0.4 | 6.1 ± 0.5 | 0.754 |
| MWT (mm/(m2)0.5) | 5.3 ± 0.4 | 5.3 ± 0.5 | 0.769 |
| RWT | 0.29 ± 0.04 | 0.28 ± 0.04 | 0.387 |
| LVM (g) | 75 ± 14 | 82 ± 18 | 0.204 |
| LVM index (g/m2) | 25 ± 5 | 28 ± 7 | 0.051 |
| LV length (mm) | 76 ± 6 | 74 ± 5 | 0.316 |
| LV length index (mm/(m2)0.5) | 67 ± 5 | 64 ± 5 | 0.140 |
| LVEDV (mL) | 75 ± 10 | 69 ± 15 | 0.106 |
| LVEDV index (mL/(m2)1.5) | 51 ± 8 | 45 ± 6 | 0.007 |
| LVESV (mL) | 26 ± 4 | 26 ± 8 | 0.696 |
| LVESV index (mL/(m2)1.5) | 18 ± 2 | 17 ± 3 | 0.625 |
| Concentricity (g/mL)0.667) | 4.3 ± 0.7 | 4.9 ± 1.0 | 0.017 |
| Sphericity index | 1.8 ± 0.2 | 1.7 ± 0.1 | 0.034 |
| SV (mL) | 49 ± 8 | 43 ± 10 | 0.031 |
| SV index (mL/(m2)1.0) | 38 ± 6 | 32 ± 5 | 0.002 |
| (L/min) | 3.19 ± 0.63 | 3.12 ± 0.67 | 0.741 |
| Cardiac index (L/min per (m2)1.0) | 2.48 ± 0.46 | 2.36 ± 0.42 | 0.391 |
| SP | CON | p-value | |
|---|---|---|---|
| HR (beats/min) | 65 ± 8 | 74 ± 10 | 0.005 |
| LVEDd (mm) | 42 ± 4 | 44 ± 4 | 0.142 |
| LVEDd index (mm/(m2)0.5) | 37 ± 3 | 38 ± 3 | 0.343 |
| LVESd (mm) | 28 ± 3 | 29 ± 3 | 0.403 |
| LVESd index (mm/(m2)0.5) | 25 ± 3 | 25 ± 2 | 0.725 |
| MWT (mm) | 6.0 ± 0.4 | 6.1 ± 0.5 | 0.754 |
| MWT (mm/(m2)0.5) | 5.3 ± 0.4 | 5.3 ± 0.5 | 0.769 |
| RWT | 0.29 ± 0.04 | 0.28 ± 0.04 | 0.387 |
| LVM (g) | 75 ± 14 | 82 ± 18 | 0.204 |
| LVM index (g/m2) | 25 ± 5 | 28 ± 7 | 0.051 |
| LV length (mm) | 76 ± 6 | 74 ± 5 | 0.316 |
| LV length index (mm/(m2)0.5) | 67 ± 5 | 64 ± 5 | 0.140 |
| LVEDV (mL) | 75 ± 10 | 69 ± 15 | 0.106 |
| LVEDV index (mL/(m2)1.5) | 51 ± 8 | 45 ± 6 | 0.007 |
| LVESV (mL) | 26 ± 4 | 26 ± 8 | 0.696 |
| LVESV index (mL/(m2)1.5) | 18 ± 2 | 17 ± 3 | 0.625 |
| Concentricity (g/mL)0.667) | 4.3 ± 0.7 | 4.9 ± 1.0 | 0.017 |
| Sphericity index | 1.8 ± 0.2 | 1.7 ± 0.1 | 0.034 |
| SV (mL) | 49 ± 8 | 43 ± 10 | 0.031 |
| SV index (mL/(m2)1.0) | 38 ± 6 | 32 ± 5 | 0.002 |
| (L/min) | 3.19 ± 0.63 | 3.12 ± 0.67 | 0.741 |
| Cardiac index (L/min per (m2)1.0) | 2.48 ± 0.46 | 2.36 ± 0.42 | 0.391 |
Data are mean ± standard deviation.
SP: soccer players; CON: controls; HR: heart rate; LVEDd: left ventricular end-diastolic diameter; LVESd: left ventricular end-systolic diameter; MWT: mean wall thickness; RWT: relative wall thickness; LVM: left ventricular mass; LV: left ventricular; LVEDV: left ventricular end-diastolic volume; LVESVL: left ventricular end-systolic volume; SV: stroke volume; : cardiac output.
Conventionally derived systolic and diastolic function in soccer players and controls are presented in Supplementary Material Table A.1 online. LV ejection fraction was greater in soccer players than in controls, whereas absolute and scaled lateral S′ were lower in soccer players (all p < 0.05). E wave deceleration time was longer in soccer players than in controls (p < 0.05). No other functional differences were observed between groups.
LV mechanics
Peak longitudinal strain was not different between groups (p > 0.05). Circumferential strain at the basal and papillary muscle levels were greater in soccer players than in controls (both p < 0.05). Also, peak circumferential and longitudinal diastolic strain rate were greater in soccer players than in controls (both p < 0.05). Apical rotation was higher in soccer players than in controls (p < 0.05), without differences at the basal level (p > 0.05) and, thus, peak LV twist was greater in soccer players (p < 0.05). No further differences in peak LV mechanics were observed between groups (all p > 0.05, Table 2).
Peak left ventricular mechanics in soccer players and controls.
| SP | CON | p-value | |
|---|---|---|---|
| Longitudinal | |||
| Strain (%) | –20.3 ± 1.6 | –19.6 ± 2.5 | 0.283 |
| SRS (s–1) | –1.07 ± 0.11 | –1.11 ± 0.18 | 0.427 |
| SRE (s–1) | 2.22 ± 0.40 | 2.02 ± 0.46 | 0.025 |
| SRA (s–1) | 0.57 ± 0.10 | 0.66 ± 0.15 | 0.105 |
| Mitral valve (base) | |||
| Circumferential strain (%) | –22.2 ± 2.5 | –20.5 ± 2.5 | 0.029 |
| Circumferential SRS (s–1) | –1.29 ± 0.15 | –1.29 ± 0.17 | 0.925 |
| Circumferential SRE (s–1) | 2.16 ± 0.37 | 2.08 ± 0.34 | 0.457 |
| Circumferential SRA (s–1) | 0.36 ± 0.13 | 0.38 ± 0.09 | 0.587 |
| Radial strain (%) | 38.3 ± 13.7 | 33.7 ± 15.4 | 0.304 |
| Radial SRS (s–1) | 2.12 ± 0.50 | 2.22 ± 1.01 | 0.675 |
| Radial SRE (s–1) | –2.54 ± 0.81 | –2.46 ± 0.81 | 0.954 |
| Radial SRA (s–1) | –0.74 ± 0.36 | –1.08 ± 0.67 | 0.065 |
| Papillary muscle (mid-ventricular) | |||
| Circumferential strain (%) | –20.1 ± 1.5 | –18.3 ± 2.5 | 0.007 |
| Circumferential SRS (s–1) | –1.17 ± 0.17 | –1.19 ± 0.18 | 0.732 |
| Circumferential SRE (s–1) | 1.89 ± 0.33 | 1.57 ± 0.39 | 0.001 |
| Circumferential SRA (s–1) | 0.37 ± 0.14 | 0.42 ± 0.12 | 0.215 |
| Radial strain (%) | 61.6 ± 15.8 | 62.2 ± 20.9 | 0.918 |
| Radial SRS (s–1) | 2.38 ± 0.89 | 2.47 ± 1.28 | 0.944 |
| Radial SRE (s–1) | –3.64 ± 1.54 | –3.35 ± 1.47 | 0.569 |
| Radial SRA (s–1) | –1.08 ± 0.54 | –1.60 ± 1.54 | 0.463 |
| Twist mechanics | |||
| Apical rotation (°) | 11.95 ± 5.31 | 7.58 ± 3.55 | 0.005 |
| Basal rotation (°) | –5.69 ± 3.14 | –5.64 ± 2.65 | 0.959 |
| Twist (°) | 16.92 ± 7.55 | 12.34 ± 4.99 | 0.035 |
| SP | CON | p-value | |
|---|---|---|---|
| Longitudinal | |||
| Strain (%) | –20.3 ± 1.6 | –19.6 ± 2.5 | 0.283 |
| SRS (s–1) | –1.07 ± 0.11 | –1.11 ± 0.18 | 0.427 |
| SRE (s–1) | 2.22 ± 0.40 | 2.02 ± 0.46 | 0.025 |
| SRA (s–1) | 0.57 ± 0.10 | 0.66 ± 0.15 | 0.105 |
| Mitral valve (base) | |||
| Circumferential strain (%) | –22.2 ± 2.5 | –20.5 ± 2.5 | 0.029 |
| Circumferential SRS (s–1) | –1.29 ± 0.15 | –1.29 ± 0.17 | 0.925 |
| Circumferential SRE (s–1) | 2.16 ± 0.37 | 2.08 ± 0.34 | 0.457 |
| Circumferential SRA (s–1) | 0.36 ± 0.13 | 0.38 ± 0.09 | 0.587 |
| Radial strain (%) | 38.3 ± 13.7 | 33.7 ± 15.4 | 0.304 |
| Radial SRS (s–1) | 2.12 ± 0.50 | 2.22 ± 1.01 | 0.675 |
| Radial SRE (s–1) | –2.54 ± 0.81 | –2.46 ± 0.81 | 0.954 |
| Radial SRA (s–1) | –0.74 ± 0.36 | –1.08 ± 0.67 | 0.065 |
| Papillary muscle (mid-ventricular) | |||
| Circumferential strain (%) | –20.1 ± 1.5 | –18.3 ± 2.5 | 0.007 |
| Circumferential SRS (s–1) | –1.17 ± 0.17 | –1.19 ± 0.18 | 0.732 |
| Circumferential SRE (s–1) | 1.89 ± 0.33 | 1.57 ± 0.39 | 0.001 |
| Circumferential SRA (s–1) | 0.37 ± 0.14 | 0.42 ± 0.12 | 0.215 |
| Radial strain (%) | 61.6 ± 15.8 | 62.2 ± 20.9 | 0.918 |
| Radial SRS (s–1) | 2.38 ± 0.89 | 2.47 ± 1.28 | 0.944 |
| Radial SRE (s–1) | –3.64 ± 1.54 | –3.35 ± 1.47 | 0.569 |
| Radial SRA (s–1) | –1.08 ± 0.54 | –1.60 ± 1.54 | 0.463 |
| Twist mechanics | |||
| Apical rotation (°) | 11.95 ± 5.31 | 7.58 ± 3.55 | 0.005 |
| Basal rotation (°) | –5.69 ± 3.14 | –5.64 ± 2.65 | 0.959 |
| Twist (°) | 16.92 ± 7.55 | 12.34 ± 4.99 | 0.035 |
Data are presented as means ± standard deviation.
SP: soccer players; CON: controls; SRS: peak systolic strain rate; SRE: peak early diastolic strain rate; SRA: peak late diastolic strain rate.
Peak left ventricular mechanics in soccer players and controls.
| SP | CON | p-value | |
|---|---|---|---|
| Longitudinal | |||
| Strain (%) | –20.3 ± 1.6 | –19.6 ± 2.5 | 0.283 |
| SRS (s–1) | –1.07 ± 0.11 | –1.11 ± 0.18 | 0.427 |
| SRE (s–1) | 2.22 ± 0.40 | 2.02 ± 0.46 | 0.025 |
| SRA (s–1) | 0.57 ± 0.10 | 0.66 ± 0.15 | 0.105 |
| Mitral valve (base) | |||
| Circumferential strain (%) | –22.2 ± 2.5 | –20.5 ± 2.5 | 0.029 |
| Circumferential SRS (s–1) | –1.29 ± 0.15 | –1.29 ± 0.17 | 0.925 |
| Circumferential SRE (s–1) | 2.16 ± 0.37 | 2.08 ± 0.34 | 0.457 |
| Circumferential SRA (s–1) | 0.36 ± 0.13 | 0.38 ± 0.09 | 0.587 |
| Radial strain (%) | 38.3 ± 13.7 | 33.7 ± 15.4 | 0.304 |
| Radial SRS (s–1) | 2.12 ± 0.50 | 2.22 ± 1.01 | 0.675 |
| Radial SRE (s–1) | –2.54 ± 0.81 | –2.46 ± 0.81 | 0.954 |
| Radial SRA (s–1) | –0.74 ± 0.36 | –1.08 ± 0.67 | 0.065 |
| Papillary muscle (mid-ventricular) | |||
| Circumferential strain (%) | –20.1 ± 1.5 | –18.3 ± 2.5 | 0.007 |
| Circumferential SRS (s–1) | –1.17 ± 0.17 | –1.19 ± 0.18 | 0.732 |
| Circumferential SRE (s–1) | 1.89 ± 0.33 | 1.57 ± 0.39 | 0.001 |
| Circumferential SRA (s–1) | 0.37 ± 0.14 | 0.42 ± 0.12 | 0.215 |
| Radial strain (%) | 61.6 ± 15.8 | 62.2 ± 20.9 | 0.918 |
| Radial SRS (s–1) | 2.38 ± 0.89 | 2.47 ± 1.28 | 0.944 |
| Radial SRE (s–1) | –3.64 ± 1.54 | –3.35 ± 1.47 | 0.569 |
| Radial SRA (s–1) | –1.08 ± 0.54 | –1.60 ± 1.54 | 0.463 |
| Twist mechanics | |||
| Apical rotation (°) | 11.95 ± 5.31 | 7.58 ± 3.55 | 0.005 |
| Basal rotation (°) | –5.69 ± 3.14 | –5.64 ± 2.65 | 0.959 |
| Twist (°) | 16.92 ± 7.55 | 12.34 ± 4.99 | 0.035 |
| SP | CON | p-value | |
|---|---|---|---|
| Longitudinal | |||
| Strain (%) | –20.3 ± 1.6 | –19.6 ± 2.5 | 0.283 |
| SRS (s–1) | –1.07 ± 0.11 | –1.11 ± 0.18 | 0.427 |
| SRE (s–1) | 2.22 ± 0.40 | 2.02 ± 0.46 | 0.025 |
| SRA (s–1) | 0.57 ± 0.10 | 0.66 ± 0.15 | 0.105 |
| Mitral valve (base) | |||
| Circumferential strain (%) | –22.2 ± 2.5 | –20.5 ± 2.5 | 0.029 |
| Circumferential SRS (s–1) | –1.29 ± 0.15 | –1.29 ± 0.17 | 0.925 |
| Circumferential SRE (s–1) | 2.16 ± 0.37 | 2.08 ± 0.34 | 0.457 |
| Circumferential SRA (s–1) | 0.36 ± 0.13 | 0.38 ± 0.09 | 0.587 |
| Radial strain (%) | 38.3 ± 13.7 | 33.7 ± 15.4 | 0.304 |
| Radial SRS (s–1) | 2.12 ± 0.50 | 2.22 ± 1.01 | 0.675 |
| Radial SRE (s–1) | –2.54 ± 0.81 | –2.46 ± 0.81 | 0.954 |
| Radial SRA (s–1) | –0.74 ± 0.36 | –1.08 ± 0.67 | 0.065 |
| Papillary muscle (mid-ventricular) | |||
| Circumferential strain (%) | –20.1 ± 1.5 | –18.3 ± 2.5 | 0.007 |
| Circumferential SRS (s–1) | –1.17 ± 0.17 | –1.19 ± 0.18 | 0.732 |
| Circumferential SRE (s–1) | 1.89 ± 0.33 | 1.57 ± 0.39 | 0.001 |
| Circumferential SRA (s–1) | 0.37 ± 0.14 | 0.42 ± 0.12 | 0.215 |
| Radial strain (%) | 61.6 ± 15.8 | 62.2 ± 20.9 | 0.918 |
| Radial SRS (s–1) | 2.38 ± 0.89 | 2.47 ± 1.28 | 0.944 |
| Radial SRE (s–1) | –3.64 ± 1.54 | –3.35 ± 1.47 | 0.569 |
| Radial SRA (s–1) | –1.08 ± 0.54 | –1.60 ± 1.54 | 0.463 |
| Twist mechanics | |||
| Apical rotation (°) | 11.95 ± 5.31 | 7.58 ± 3.55 | 0.005 |
| Basal rotation (°) | –5.69 ± 3.14 | –5.64 ± 2.65 | 0.959 |
| Twist (°) | 16.92 ± 7.55 | 12.34 ± 4.99 | 0.035 |
Data are presented as means ± standard deviation.
SP: soccer players; CON: controls; SRS: peak systolic strain rate; SRE: peak early diastolic strain rate; SRA: peak late diastolic strain rate.
Temporal analyses of left ventricle longitudinal strain (a) and strain rate (b) during the cardiac cycle in soccer players (SP) and controls (CON). Data are means ± standard deviation. Shaded areas indicate statistical significance at p < 0.05.
AVC: aortic valve closure.
Discussion
The major findings from the study were that soccer players had augmented peak LV mechanics (circumferential strain, apical rotation and twist) and faster early diastolic lengthening. Additionally, soccer players had lower concentricity than controls, without concomitant differences in wall thickness or longitudinal strain between soccer players and controls.
Lower concentricity was evident in the presence of larger scaled LVEDV, indicative of heightened preload, in soccer players compared with controls. It is possible that this is the consequence of plasma volume expansion common in well trained adolescents.23 The similar LV structure between groups contradicts previous echocardiographic work that has reported greater LV wall thickness or chamber diameter in pre-adolescent soccer players.12,24 These contrasting findings may be a product of differences in maturation status of the studied populations. Indeed, a recent meta-analysis found pre-adolescent athletes (<14 years) presented cardiac enlargement to a lesser extent than those within the pubertal growth stage.3 Further, Nottin et al.4 found similar wall thicknesses in pre-pubertal cyclists and sedentary children. Taken together, these data support the contention that sufficient maturity is necessary to promote exercise induced increases in LV structure.
Similar longitudinal strain between soccer players and controls disagrees with a similar study in young footballers,12 yet supports the majority of existing literature in young athletes.10,11 Although documentation of longitudinal strain in paediatric athletes is currently in its infancy, these observations corroborate with the adult athlete’s heart that longitudinal strain remains largely unaltered in chronically trained athletes.25 Reduced longitudinal strain has been suggested as an uncommon feature of the adult athlete’s heart26 and this may hold true also for paediatric athletes given the accumulation of evidence to suggest unaltered longitudinal strain in healthy athletic children.
Circumferential strain was greater in soccer players at both the base and papillary muscle levels and highlights a more notable difference in systolic functioning in soccer players compared with the untrained state. Although the greater circumferential strain in soccer players contrasts recent work,11 the increase observed here is likely indicative of greater overall systolic function in soccer players. Yet, the purpose of these adaptations and responsible mechanistic underpinning requires clarification.
LV twist was higher in soccer players mediated through greater apical rotation, without changes in basal rotation, with temporal analysis indicating higher apical rotation and twist through the majority of systole. To our knowledge, this is the first echocardiographic documentation of LV twist mechanics in pre-adolescent athletes, yet concurs with a recent magnetic resonance imaging study.13 These data begin to define the twisting profiles of chronically trained pre-adolescents, which appear to contrast their elite level adult counterparts of reduced net twist compared with untrained controls.25,27 A phasic response in LV twist has been reported in adults with twist increased initially before returning to baseline following structural (true eccentric) remodelling during a more extended period of exercise training.28 Indeed, larger wall thickness appears to drive the reduction in apical rotation and net twist in adults.27 The absence of differences in mean wall thickness between soccer players and controls could explain why twist was higher in soccer players, owing to heightened scaled LVEDV, considering the preload dependence of LV apical rotation and twist.29 Thus, the adaptations in chronically trained pre-adolescents may be more reflective of the acute phase of exercise training in adults.28
Alternatively, in adults LV twist is influenced by muscle fibre orientation and both apical rotation and twist are independently associated, in a parabolic manner, with sphericity index.20 In this study, sphericity index was slightly higher in soccer players and may therefore be placed higher on the ascending arm of the parabolic curve. Irrespective of the mechanistic underpinning, taking these observations together we propose that the heart of paediatric athletes presents a useful model to study the influence of exercise training on LV twist mechanics prior to structural remodelling.
Greater circumferential strains and LV twist mechanics could explain the greater ejection fraction in pre-adolescent soccer players. Subsequently, the likely combination of a larger LVEDV and increased ejection fraction led to a greater SV. The higher ejection fraction observed is not in agreement with other similar work, although the dataset presented within this recent meta-analysis was highly heterogeneous.3 The reasons for discrepancies between studies is unclear. However, ejection fraction is considered a surrogate marker of LV pump function,30 and others have reported that pre-adolescent athletes have similar ejection fraction with comparable or lower LV strain mechanics than controls,10,–12 whereas in this study soccer players had a greater ejection fraction which was accompanied with augmented circumferential strain and twist, and thus the altered mechanics may explain the greater ejection fraction. Together, the combination of LV mechanics and ejection fraction in this study suggests augmented systolic function, yet the functional capacity and potential reserve during exercise warrants further investigation.
Mitral inflow and early diastolic tissue velocities were similar between soccer players and controls, whereas assessment of temporal strain rate indicated that soccer players had faster diastolic lengthening during the early phases of diastole. Strain rate was lower in soccer players at 35-40% diastole during the cardiac cycle, signifying superior lengthening velocities leading to a longer period of diastasis. These data are supported by the greater peak longitudinal and circumferential (papillary muscle level) early diastolic strain rate in soccer players than in controls. Accordingly, it is plausible that novel assessment using STE may facilitate the detection of subtle differences that are not apparent using TDI, with the former being advantageous in being relatively angle independent and less affected by tethering from adjacent segments.8 Collectively, these data suggest improved relaxation may contribute to the enhanced preload/LVEDV and the larger ensuing SV. The functional importance of these observations is yet to be fully realised; however, it is possible that this becomes pertinent to support LV filling especially when diastole shortens during exercise.
Considering that youth soccer players perform at high intensities,1,11 the differentiation between pathological and physiological adaptation is of paramount importance and is of current interest.31 The present study highlights that elite-level soccer training may be a strong enough stimulus to induce LV mechanical adaptations, even in the absence of morphological difference. We report normal values of LV mechanics in asymptomatic pre-adolescents, therefore, the clinical inference is challenging and may warrant further investigation in those with and without known cardiovascular diseases. Additionally, given the small sample sizes in this study, the clinical implications of this work require further study in larger groups, in association with exercise training status. Future longitudinal data are needed to track these players throughout the volatile growth periods and, as a result, these insights may be helpful to diagnose early stages of cardiomyopathies, such as in HCMP phenotype negative but genotype positive individuals, for example.
Limitations
We acknowledge the limitations of cross-sectional studies including, for example, the ability to clearly attribute the cardiac functional changes to soccer training, or the genetic predisposition for pre-adolescents to perform soccer at a high level. Plasma volume was not assessed in this study due to ethical restrictions for blood sampling in the studied population. 2D echocardiography is inherently limited by out-of-plane motion such that apical and basal imaging planes may not be the same through the entire cardiac cycle.8 Limitations in current echocardiographic techniques may explain why LV twist was not obtained in five participants (n = 1 soccer player and n = 4 controls), since optimal acquisition and speckle-tracking are required at both the base and the apex. Accordingly, this resulted in slightly unequal samples sizes between soccer players and controls. While this presents challenges for clinical utility, these parameters are increasingly acknowledged for their potential use in characterising the athlete’s heart, although work is still needed for normative values in LV twist and circumferential strain.32 In this study, however, we applied stringent criteria to both image acquisition and analysis to facilitate confidence in obtaining physiologically meaningful data. Finally, if age-associated increases in LV twist during childhood are related to maturational adaptive modulation,33,34 it would be unlikely that the small differences in chronological age between soccer players and controls would impact results, since biological age (maturity offset) and LV length were similar (i.e. twisting occurs along the same LV length).
Conclusions
Augmented resting LV mechanics (twist, apical rotation and circumferential strain) were observed in highly trained soccer players, including increased circumferential strains, apical rotation and LV twist, with supportive temporal analysis demonstrating faster early diastolic lengthening than untrained, matched controls. These data highlight superior function derived by STE LV mechanics in the absence of LV wall thickness changes yet with lower concentricity.
Supplemental Material
CPR890177 Supplemental Material1 - Supplemental material for Superior cardiac mechanics without structural adaptations in pre-adolescent soccer players
Supplemental material, CPR890177 Supplemental Material1 for Superior cardiac mechanics without structural adaptations in pre-adolescent soccer players by Alexander Beaumont, David Oxborough, Keith George, Thomas W Rowland, Nicholas Sculthorpe, Rachel Lord and Viswanath B Unnithan in European Journal of Preventive Cardiology
Supplemental Material
CPR890177 Supplemental Material2 - Supplemental material for Superior cardiac mechanics without structural adaptations in pre-adolescent soccer players
Supplemental material, CPR890177 Supplemental Material2 for Superior cardiac mechanics without structural adaptations in pre-adolescent soccer players by Alexander Beaumont, David Oxborough, Keith George, Thomas W Rowland, Nicholas Sculthorpe, Rachel Lord and Viswanath B Unnithan in European Journal of Preventive Cardiology
Author contribution
VBU conceived and designed the research study. VBU, TWR, RL and DO conducted the experiments and collected the data. DO and AB analysed data. AB, DO, NS and VBU interpreted the data. AB prepared figures and wrote the manuscript. DO, KG, TWR, NS, RL and VBU edited and revised the manuscript. All authors read and approved the final version of the manuscript.
Acknowledgements
The authors would like to thank Mr Marc Campbell (Wolverhampton Wanders FC) and Dr Russ Wrigley (Blackburn Rovers FC) in the organisation of the soccer players data collection sessions; Ms Katie Davis and Ms Leanne Brittle for their help with the organisation of the testing session of controls (Staffordshire University Academy).
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Comments