Abstract
Previous cross-sectional studies have demonstrated that fat-free mass (FFM) is an important determinant of left ventricular mass (LVM) in athletes. However, cross-sectional investigations have not the ability to detect the dynamic adaptation occurring with training. We hypothesized that LVM adapts concurrently with the increase of FFM induced by exercise conditioning. We sought to study the relationship between the variations of LVM and of FFM occurring in top-level soccer players during the season.
Twenty-three male top-level athletes were recruited. LVM was assessed by echocardiography and FFM by dual-energy X-ray absorptiometry. Serial measurements were performed pre-season, after 1 month, at mid- and end-season, and after 2 months of detraining.
LVM significantly increased at mid-season versus pre-season values, reaching the highest value at the end of the season (p < 0.05). While body weight did not vary during the study period, FFM significantly increased (p < 0.05 for mid-/end-season vs. pre-season data). After the detraining, no significant differences were observed between pre-season and detraining echocardiographic data. The only independent predictors of LVM were left ventricular stroke volume and FFM (R = 0.36, p = 0.005; R = 0.35, p = 0.005, respectively). When ΔLVM index was set as dependent variable, the only independent predictor was ΔFFM (R = 0.87, p = 0.002).
Changes in LVM occur in close association with changes in FFM, suggesting that the left ventricle adapts concurrently with the increase of the metabolically active tissue induced by training, i.e. the FFM. Therefore, the dynamic changes in FFM and LVM may reflect a physiological adaptation induced by intensive training.
Introduction
Participation in an intensive training programme is associated with central and peripheral cardiovascular adaptations, and haemodynamic overload on the left ventricle (LV) is the primary stimulus for cardiac morphological adaptation to training.1 Over the past decades, several echocardiographic investigations have been performed in athletes and there is a general consensus that left ventricular mass (LVM), assessed by echocardiography, is increased in the athletic population.2–5 Most of the echocardiographic studies evaluating the athlete’s heart have a cross-sectional design and compare LV parameters of competitive athletes with data obtained in age- and sex-matched sedentary controls. However, the left ventricle is able to adapt rapidly in response to the commencement of vigorous training6,7 and significant training-specific changes can be longitudinally observed in cardiac structure and function in competitive athletes or in healthy subjects after training.8,9 Cross-sectional investigations have not the ability to detect the dynamic nature of cardiac adaptation occurring within the training season and are limited by the assumption made that differences between athletes and sedentary subjects are exclusively related to training status rather than to variations of non-cardiac variables, such as body size. Conversely, the variations of body size and, particularly, of body composition should be considered to clarify the effects of the training status on cardiac remodelling. Indeed, in-season variations of body composition and particularly of fat-free mass (FFM) have previously been reported in athletes,10 and a cross-sectional study has demonstrated that FFM is an important determinant of LVM in athletes.11 Although ideal, FFM is rarely used in research because accurate measurements are not widely available. However, dual-energy X-ray absorptiometry (DXA) is able to overcome the main limitations of the traditional techniques and is considered a valuable tool for the measurement of compositional changes in longitudinal studies.12–14
To date, no longitudinal data are available regarding the impact of an intensive training programme on the changes in both LVM and FFM in elite athletes during the regular season. The aims of this prospective, longitudinal study were to detect the adaptations of the left ventricle in a selected cohort of athletes within the training season and to determine whether changes in LVM are closely related to variations in FFM in professional soccer players.
Methods
Study population and training protocol
Male elite soccer players (Siena Football Club, Italian Premiere League) were recruited for the purpose of this study. Measurements were performed at the beginning of the study, after 1 month, after 5 months, and after 10 months of training, corresponding to the beginning of the training programme, the end of the pre-season conditioning, the middle of the season, and the end of the agonistic season, respectively. In addition, a final measurement was obtained after a 2-month detraining period, performed as per protocol, in the professional soccer players who did not move up to other clubs.
Athletes were engaged in an intensive and closely supervised training programme. During the pre-season conditioning athletes trained for at least 20 h/week. Training sessions mainly consisted of high volume/low intensity running and sprinting conditioning. Athletes also performed 2–3 resistance training sessions per week at high–moderate workload under the supervision of a dedicated coach. At the beginning of the agonistic season, i.e. after 1 month of preseason conditioning, athletes trained for at least 12 h/week and played one/two matches/week during the regular season. They were submitted to training sessions at workloads ranging from 70 to 95% of maximal heart rate, as indicated by individual heart rate monitoring applied during the sessions. Training sessions mainly consisted of technical-tactical drills, low volume/ high intensity running, and sprinting conditioning. Athletes also performed 1–2 resistance training sessions/week at moderate workload under the supervision of a dedicated coach. During the last 4 months of the season, running, sprinting conditioning, and resistance training sessions were gradually reduced and in the final part of the season the training programme consisted only of technical and tactical drills. Because goalkeepers were engaged in a different athletic programme, they were excluded from the present study. All athletes were evaluated at the same stage of their training programme and at the same time of the day, before the training session, at least 48 h after the last strenuous training session.
All participants underwent complete physical examination, ECG, standard echocardiography, and treadmill ECG test with no evidence of pathological findings. Body mass index (BMI) was calculated as weight in kilograms divided by height in metres squared (kg/m2). Body surface area (BSA) was calculated using the Dubois and Dubois formula.15 Athletes were asymptomatic and showed a negative family history of cardiac disease or sudden cardiac death. None of the athletes had cardiovascular structural or functional abnormalities, hypertension, or type I diabetes mellitus. Participants were excluded from the study if they withdrew from training programme for more than 15 days for musculoskeletal injuries.
Based on these criteria, 26 professional soccer players were enrolled in the study. Three were subsequently excluded because they had withdrawn from the training programme >15 days for musculoskeletal injuries. A final population of 23 athletes completed the training programme and was analysed during the study period. The detraining evaluation was available in 10 athletes. After the rationale and the study protocol were explained, the participants gave written informed consent. The investigational protocol was approved by the local ethical committee and was conformed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments.
Determination of body composition
Whole-body composition was measured by DXA (Lunar Prodigy, GE Medical System, Madison, WI, USA), a method using an X-ray tube with a filter to generate low-energy and high-energy photons. Even if the issue of safety has to be always taken into account when using technologies with radiation exposure particularly in young subjects, effective radiation doses for whole-body examinations using GE Lunar Prodigy DXA scanner have been recently estimated in a paediatric population to be lower than 1 µSV and particularly 0.25, 0.22, 0.19, and 0.15 µSV for the neonate, 1 -, 5 -, and 10-year-old child, respectively, with a negligible estimated lifetime cancer risks.16 All athletes were evaluated in the morning before the training, the same day of echocardiographic measurements. Scan time for the total body measurement was approximately 7 minutes. Extended analysis was performed to estimate total mass, FFM, and percentage of fat mass. DXA determination was performed the same day of the echocardiographic acquisition, at the beginning of the training programme, after 1, 5, and 10 months, and after the detraining.
Echocardiographic analysis
Echocardiographic examination was performed by one cardiologist using a high-quality echocardiograph (Vivid 7, GE Medical Systems, Horten, Norway), equipped with an M4S 1.5–4.0 MHz transducer. A one-lead ECG was continuously displayed during the exam. Subjects were studied in the steep left-lateral decubitus position. For all measurements, three beats were stored and analysed off-line (EchoPac, GE, USA). Off-line data analysis was performed by one experienced reader, blinded to the study time point. All echocardiographic data were analysed at the end of the data collection.
Linear measurements of interventricular septum, posterior wall thickness, and LV internal diameters were obtained from the parasternal long-axis acoustic window and measured as recommended.17 LVM was calculated from the ASE recommended formula:18 LVM = 0.8(1.04[(LVIDd + PWTd + SWTd)3–(LVIDd)3]) + 0.6 g, whereas LVIDd is LV internal dimension at end-diastole, PWTd is posterior wall thickness at end-diastole, and SWTd is septal wall thickness at end-diastole. LVM was indexed to BSA, obtaining LVM index (LVMi). Relative wall thickness was calculated by the formula: 2PWT/LVIDd, as recommended.17 LV volume measurements were calculated from the apical views using the biplane method of disk (modified Simpson’s rule), and ejection fraction was calculated as follows: ejection fraction=(EDV – ESV)/EDV, where EDV is end-diastolic volume and ESV is end-systolic volume.17
Statistical analysis
Gaussian distribution of all continuous variables was estimated using the Shapiro–Wilk test and data are presented as mean ± SD or median and interquartile range for continuous variables, as appropriate. Comparison of values obtained during the study period at each time point was performed with Friedman’s test and Dunn’s post-estimation procedure or with repeated-measures ANOVA with Bonferroni post-hoc correction for multiple comparisons, according to data distribution. A p-value < 0.05 was considered significant. Correlation analysis was performed using the Spearman and Pearson methods, according to data distribution. Variables were selected, according to their potential impact on LV structure and function.2,3,7,19 A stepwise linear regression analysis was also used to determine the independent predictors of LVM changes. Reproducibility of LVM measurements was determined in all subjects. Intra-observer concordance of echocardiographic measurements was assessed using paired Student’s t-test and intraclass correlation.19,20 Statistics were performed using Statistical Package for the Social Sciences v. 14.0 for Windows (SPSS, Chicago, IL, USA).
Results
The characteristics of the study population are summarized in Table 1. The mean age of the study population was 26.6 ± 4.1 years. During the study period BSA did not significantly vary. As expected, resting heart rate showed a significant variation between 1-month and pre-season values (p < 0.05); a further decrease was observed at mid-season (p < 0.05); while at the end of the season resting heart rate significantly increased in comparison with mid-season values (p < 0.05), being in an intermediate range between pre-season and 1-month measurements.
Characteristics of professional soccer players observed during the study period (n = 23)
| Variable | Pre season | 1 month | Mid season | End season |
|---|---|---|---|---|
| Height (cm) | 180.8 ± 6.4 | – | – | – |
| Weight (kg) | 79.1 ± 7.3 | 78.2 ± 7.5 | 78.4 ± 8.1 | 78.3 ± 7.0 |
| BSA (m2) | 1.99 ± 0.12 | 1.98 ± 0.12 | 1.98 ± 0.13 | 1.98 ± 0.13 |
| Resting HR (bpm) | 62.1 ± 8.1 | 57.9 ± 8.6* | 53.3 ± 7.4*† | 59.8 ± 8.5*‡ |
| Variable | Pre season | 1 month | Mid season | End season |
|---|---|---|---|---|
| Height (cm) | 180.8 ± 6.4 | – | – | – |
| Weight (kg) | 79.1 ± 7.3 | 78.2 ± 7.5 | 78.4 ± 8.1 | 78.3 ± 7.0 |
| BSA (m2) | 1.99 ± 0.12 | 1.98 ± 0.12 | 1.98 ± 0.13 | 1.98 ± 0.13 |
| Resting HR (bpm) | 62.1 ± 8.1 | 57.9 ± 8.6* | 53.3 ± 7.4*† | 59.8 ± 8.5*‡ |
Values are mean ± SD
p < 0.05 vs. baseline measurements; †p < 0.05 vs. 1-month measurements; ‡p < 0.05 vs. mid-season measurements
BSA, body surface area; HR, heart rate.
Characteristics of professional soccer players observed during the study period (n = 23)
| Variable | Pre season | 1 month | Mid season | End season |
|---|---|---|---|---|
| Height (cm) | 180.8 ± 6.4 | – | – | – |
| Weight (kg) | 79.1 ± 7.3 | 78.2 ± 7.5 | 78.4 ± 8.1 | 78.3 ± 7.0 |
| BSA (m2) | 1.99 ± 0.12 | 1.98 ± 0.12 | 1.98 ± 0.13 | 1.98 ± 0.13 |
| Resting HR (bpm) | 62.1 ± 8.1 | 57.9 ± 8.6* | 53.3 ± 7.4*† | 59.8 ± 8.5*‡ |
| Variable | Pre season | 1 month | Mid season | End season |
|---|---|---|---|---|
| Height (cm) | 180.8 ± 6.4 | – | – | – |
| Weight (kg) | 79.1 ± 7.3 | 78.2 ± 7.5 | 78.4 ± 8.1 | 78.3 ± 7.0 |
| BSA (m2) | 1.99 ± 0.12 | 1.98 ± 0.12 | 1.98 ± 0.13 | 1.98 ± 0.13 |
| Resting HR (bpm) | 62.1 ± 8.1 | 57.9 ± 8.6* | 53.3 ± 7.4*† | 59.8 ± 8.5*‡ |
Values are mean ± SD
p < 0.05 vs. baseline measurements; †p < 0.05 vs. 1-month measurements; ‡p < 0.05 vs. mid-season measurements
BSA, body surface area; HR, heart rate.
Changes in LVM and in FFM observed during the season
During the intensive training programme, significant variations of muscular mass and of LVM occurred. LVM and FFM determinations obtained during the entire season are listed in Table 2. At mid season, FFM significantly increased in comparison with pre-season values (p < 0.05), reaching the highest value of the entire season. At the end of the season, FFM values were relatively stable, maintaining the statistical significance in the comparison with pre-season. Conversely, a significant decrease of the percentage of body fat was observed after intensive training, with the lowest value observed at mid-season. LVM showed a progressive increase during the period of observation, with a Δ = 18 g between pre-season and end-season values, when LVM reached the highest observed value (p < 0.05).
Variations of body composition and of left ventricular mass observed in professional soccer players during the regular season (n = 23)
| Variable | Pre season | 1 month | Mid season | End season |
|---|---|---|---|---|
| FFM (kg) | 64.3 ± 6.0 | 65.5 ± 6.4 | 66.4 ± 6.4* | 66.3 ± 5.0* |
| Body fat (%) | 14.6 ± 4.0 | 11.9 ± 2.9* | 11.0 ± 2.6* | 11.0 ± 3.0* |
| LVM (g) | 195.0 ± 25.8 | 201.8 ± 32.1 | 212.5 ± 32.6*† | 213.5 ± 22.9*† |
| LVMi (g/m2) | 98.3 ± 13.6 | 101.7 ± 15.3 | 107.2 ± 13.5*† | 106.2 ± 13.8* |
| LVM/height (g/m) | 108.2 ± 14.7 | 111.8 ± 17.9 | 117.8 ± 16.4*† | 119.0 ± 13.7*† |
| LVM/height2.7 (g/m2.7) | 39.8 ± 6.6 | 41.1 ± 7.5 | 43.4 ± 7.5*† | 44.2 ± 6.8*† |
| LVM/FFM (g/kg) | 3.1 ± 0.4 | 3.1 ± 0.5 | 3.2 ± 0.4* | 3.2 ± 0.4* |
| Variable | Pre season | 1 month | Mid season | End season |
|---|---|---|---|---|
| FFM (kg) | 64.3 ± 6.0 | 65.5 ± 6.4 | 66.4 ± 6.4* | 66.3 ± 5.0* |
| Body fat (%) | 14.6 ± 4.0 | 11.9 ± 2.9* | 11.0 ± 2.6* | 11.0 ± 3.0* |
| LVM (g) | 195.0 ± 25.8 | 201.8 ± 32.1 | 212.5 ± 32.6*† | 213.5 ± 22.9*† |
| LVMi (g/m2) | 98.3 ± 13.6 | 101.7 ± 15.3 | 107.2 ± 13.5*† | 106.2 ± 13.8* |
| LVM/height (g/m) | 108.2 ± 14.7 | 111.8 ± 17.9 | 117.8 ± 16.4*† | 119.0 ± 13.7*† |
| LVM/height2.7 (g/m2.7) | 39.8 ± 6.6 | 41.1 ± 7.5 | 43.4 ± 7.5*† | 44.2 ± 6.8*† |
| LVM/FFM (g/kg) | 3.1 ± 0.4 | 3.1 ± 0.5 | 3.2 ± 0.4* | 3.2 ± 0.4* |
Values are mean ± SD
p < 0.05 vs. baseline measurements; †p < 0.05 vs. 1-month measurements
FFM, fat-free mass; LVM, left ventricular mass; LVMi, left ventricular mass index.
Variations of body composition and of left ventricular mass observed in professional soccer players during the regular season (n = 23)
| Variable | Pre season | 1 month | Mid season | End season |
|---|---|---|---|---|
| FFM (kg) | 64.3 ± 6.0 | 65.5 ± 6.4 | 66.4 ± 6.4* | 66.3 ± 5.0* |
| Body fat (%) | 14.6 ± 4.0 | 11.9 ± 2.9* | 11.0 ± 2.6* | 11.0 ± 3.0* |
| LVM (g) | 195.0 ± 25.8 | 201.8 ± 32.1 | 212.5 ± 32.6*† | 213.5 ± 22.9*† |
| LVMi (g/m2) | 98.3 ± 13.6 | 101.7 ± 15.3 | 107.2 ± 13.5*† | 106.2 ± 13.8* |
| LVM/height (g/m) | 108.2 ± 14.7 | 111.8 ± 17.9 | 117.8 ± 16.4*† | 119.0 ± 13.7*† |
| LVM/height2.7 (g/m2.7) | 39.8 ± 6.6 | 41.1 ± 7.5 | 43.4 ± 7.5*† | 44.2 ± 6.8*† |
| LVM/FFM (g/kg) | 3.1 ± 0.4 | 3.1 ± 0.5 | 3.2 ± 0.4* | 3.2 ± 0.4* |
| Variable | Pre season | 1 month | Mid season | End season |
|---|---|---|---|---|
| FFM (kg) | 64.3 ± 6.0 | 65.5 ± 6.4 | 66.4 ± 6.4* | 66.3 ± 5.0* |
| Body fat (%) | 14.6 ± 4.0 | 11.9 ± 2.9* | 11.0 ± 2.6* | 11.0 ± 3.0* |
| LVM (g) | 195.0 ± 25.8 | 201.8 ± 32.1 | 212.5 ± 32.6*† | 213.5 ± 22.9*† |
| LVMi (g/m2) | 98.3 ± 13.6 | 101.7 ± 15.3 | 107.2 ± 13.5*† | 106.2 ± 13.8* |
| LVM/height (g/m) | 108.2 ± 14.7 | 111.8 ± 17.9 | 117.8 ± 16.4*† | 119.0 ± 13.7*† |
| LVM/height2.7 (g/m2.7) | 39.8 ± 6.6 | 41.1 ± 7.5 | 43.4 ± 7.5*† | 44.2 ± 6.8*† |
| LVM/FFM (g/kg) | 3.1 ± 0.4 | 3.1 ± 0.5 | 3.2 ± 0.4* | 3.2 ± 0.4* |
Values are mean ± SD
p < 0.05 vs. baseline measurements; †p < 0.05 vs. 1-month measurements
FFM, fat-free mass; LVM, left ventricular mass; LVMi, left ventricular mass index.
A significant impact on left ventricular remodelling was demonstrated during the training period. The changes of LV morphology are listed in Table 3. While LVIDd did not significantly vary during the study period, elite soccer players experienced a significant increase in LVEDV, which was increased after 1 month and at mid-season in comparison with pre-season values (p < 0.05). Conversely, at the end of the season, LVEDV significantly decreased in comparison with mid-season values (p < 0.05). Measurements of interventricular septum and of posterior wall thickness did not significantly vary during the study period and similar results were observed also for relative wall thickness. While LV ejection fraction was relatively unchanged during the study period, LV stroke volume showed a significant increase at 1-month and at mid-season timepoint in comparison with pre-season values (p < 0.05), while a significant reduction was observed at the end of the season as compared with mid-season values (p < 0.05).
Variations of left ventricular echocardiographic parameters observed during the regular season in professional soccer players (n = 23)
| Variable | Pre season | 1 month | Mid season | End season |
|---|---|---|---|---|
| LVIDd (cm) | 5.2 ± 0.3 | 5.2 ± 0.3 | 5.3 ± 0.3 | 5.3 ± 0.3 |
| LVIDs (cm) | 3.3 ± 0.3 | 3.4 ± 0.3 | 3.5 ± 0.2* | 3.6 ± 0.3* |
| IVSTd (cm) | 1.0 ± 0.1 | 1.0 ± 0.1 | 1.1 ± 0.1 | 1.1 ± 0.1 |
| PWTd (cm) | 1.0 ± 0.1 | 1.0 ± 0.1 | 1.1 ± 0.1 | 1.0 ± 0.1 |
| RWT | 0.39 ± 0.05 | 0.39 ± 0.04 | 0.40 ± 0.04 | 0.39 ± 0.03 |
| LVEDV (ml) | 136.3 ± 20.1 | 150.9 ± 20.3* | 152.9 ± 22.8* | 137.5 ± 14.4‡ |
| EF (%) | 64.4 ± 3.8 | 61.9 ± 3.3 | 63.9 ± 3.6 | 64.2 ± 3.7 |
| SV (ml) | 85.5 ± 13.8 | 93.2 ± 12.4* | 98.4 ± 15.4* | 88.4 ± 10.5‡ |
| Variable | Pre season | 1 month | Mid season | End season |
|---|---|---|---|---|
| LVIDd (cm) | 5.2 ± 0.3 | 5.2 ± 0.3 | 5.3 ± 0.3 | 5.3 ± 0.3 |
| LVIDs (cm) | 3.3 ± 0.3 | 3.4 ± 0.3 | 3.5 ± 0.2* | 3.6 ± 0.3* |
| IVSTd (cm) | 1.0 ± 0.1 | 1.0 ± 0.1 | 1.1 ± 0.1 | 1.1 ± 0.1 |
| PWTd (cm) | 1.0 ± 0.1 | 1.0 ± 0.1 | 1.1 ± 0.1 | 1.0 ± 0.1 |
| RWT | 0.39 ± 0.05 | 0.39 ± 0.04 | 0.40 ± 0.04 | 0.39 ± 0.03 |
| LVEDV (ml) | 136.3 ± 20.1 | 150.9 ± 20.3* | 152.9 ± 22.8* | 137.5 ± 14.4‡ |
| EF (%) | 64.4 ± 3.8 | 61.9 ± 3.3 | 63.9 ± 3.6 | 64.2 ± 3.7 |
| SV (ml) | 85.5 ± 13.8 | 93.2 ± 12.4* | 98.4 ± 15.4* | 88.4 ± 10.5‡ |
Values are mean ± SD
p < 0.05 vs. baseline measurements; ‡p < 0.05 vs. mid-season measurements
EF, ejection fraction; IVSTd, interventricular septal thickness at end-diastole; LVEDV, left ventricular end-diastolic volume; LVIDd, left ventricular internal diameter measured at end-diastole; LVIDs, left ventricular internal diameter measured at end-systole; PWTd, posterior wall thickness at end-diastole; RWT, relative wall thickness; SV, stroke volume.
Variations of left ventricular echocardiographic parameters observed during the regular season in professional soccer players (n = 23)
| Variable | Pre season | 1 month | Mid season | End season |
|---|---|---|---|---|
| LVIDd (cm) | 5.2 ± 0.3 | 5.2 ± 0.3 | 5.3 ± 0.3 | 5.3 ± 0.3 |
| LVIDs (cm) | 3.3 ± 0.3 | 3.4 ± 0.3 | 3.5 ± 0.2* | 3.6 ± 0.3* |
| IVSTd (cm) | 1.0 ± 0.1 | 1.0 ± 0.1 | 1.1 ± 0.1 | 1.1 ± 0.1 |
| PWTd (cm) | 1.0 ± 0.1 | 1.0 ± 0.1 | 1.1 ± 0.1 | 1.0 ± 0.1 |
| RWT | 0.39 ± 0.05 | 0.39 ± 0.04 | 0.40 ± 0.04 | 0.39 ± 0.03 |
| LVEDV (ml) | 136.3 ± 20.1 | 150.9 ± 20.3* | 152.9 ± 22.8* | 137.5 ± 14.4‡ |
| EF (%) | 64.4 ± 3.8 | 61.9 ± 3.3 | 63.9 ± 3.6 | 64.2 ± 3.7 |
| SV (ml) | 85.5 ± 13.8 | 93.2 ± 12.4* | 98.4 ± 15.4* | 88.4 ± 10.5‡ |
| Variable | Pre season | 1 month | Mid season | End season |
|---|---|---|---|---|
| LVIDd (cm) | 5.2 ± 0.3 | 5.2 ± 0.3 | 5.3 ± 0.3 | 5.3 ± 0.3 |
| LVIDs (cm) | 3.3 ± 0.3 | 3.4 ± 0.3 | 3.5 ± 0.2* | 3.6 ± 0.3* |
| IVSTd (cm) | 1.0 ± 0.1 | 1.0 ± 0.1 | 1.1 ± 0.1 | 1.1 ± 0.1 |
| PWTd (cm) | 1.0 ± 0.1 | 1.0 ± 0.1 | 1.1 ± 0.1 | 1.0 ± 0.1 |
| RWT | 0.39 ± 0.05 | 0.39 ± 0.04 | 0.40 ± 0.04 | 0.39 ± 0.03 |
| LVEDV (ml) | 136.3 ± 20.1 | 150.9 ± 20.3* | 152.9 ± 22.8* | 137.5 ± 14.4‡ |
| EF (%) | 64.4 ± 3.8 | 61.9 ± 3.3 | 63.9 ± 3.6 | 64.2 ± 3.7 |
| SV (ml) | 85.5 ± 13.8 | 93.2 ± 12.4* | 98.4 ± 15.4* | 88.4 ± 10.5‡ |
Values are mean ± SD
p < 0.05 vs. baseline measurements; ‡p < 0.05 vs. mid-season measurements
EF, ejection fraction; IVSTd, interventricular septal thickness at end-diastole; LVEDV, left ventricular end-diastolic volume; LVIDd, left ventricular internal diameter measured at end-diastole; LVIDs, left ventricular internal diameter measured at end-systole; PWTd, posterior wall thickness at end-diastole; RWT, relative wall thickness; SV, stroke volume.
We sought to further characterize the LVM increase observed during the season in professional soccer players. Correlations between LVM and bodyweight, Δbodyweight, resting HR, Δresting HR, FFM, ΔFFM, LVEDV, ΔLVEDV, LV stroke volume, and Δstroke were performed. Correlations between ΔLVMi and ΔFFM, ΔLVEDV index, ΔLV stroke volume index, and Δbodyweight were also explored. There was a significant positive correlation between LVM and LV stroke volume (R = 0.36, overall p = 0.005) and between LVM and FFM (R = 0.35, overall p = 0.005). Furthermore, a positive correlation was found between ΔLVMi and ΔFFM (R = 0.87, p = 0.002; Figure 1).
Correlation between Δ left ventricular mass index and Δ fat-free mass (end season – pre season) obtained by echocardiography and by dual-energy X-ray absorptiometry in professional soccer players during the regular season.
FFM, fat-free mass; LVMi, left ventricular mass index.
To determine the independent predictors of changes in LVMi, a stepwise linear regression analysis was performed. Input variables included ΔFFM, Δbodyweight, and Δstroke volume. ΔFFM was identified as the only independent predictor of ΔLVMi (β = 0.70, p < 0.05), accounting for the 49% of variability explained by the model.
Detraining
After 2 months of detraining a reverse remodelling of the left ventricle was observed. After detraining LVMi significantly decreased from end-season value of 108.5 ± 16.2 to 92.3 ± 12.5 g/m2 (p < 0.05) with no significant differences with pre-season values (p = 0.76). FFM showed a similar trend and after detraining significantly decreased from end-season value of 67.8 ± 3.6 to 63.3 ± 4.4 kg (p < 0.05) with no significant differences with pre-season values (p = 0.06). Furthermore, no significant differences were observed between pre-season and post-detraining LVEDV (139.1 ± 21.9 vs. 135.4 ± 20.9 ml, p = 0.44), LV stroke volume (87.6 ± 13.3 vs. 87.0 ± 12.6 ml, p = 0.82), and resting HR (59.9 ± 9.3 vs. 60.0 ± 10.1 bpm, p = 0.97).
Intra-observer variability
LVM, as measured by the same observer, was 195.2 ± 18.5 g (range 177–213 g) at first assessment and 193.3 ± 18.7 g (range 175–211 g) at second evaluation; mean difference was 6.8 ± 14 g (range 7–20 g) (p = 0.20).
Discussion
Changes in LVM and in body composition during the season
Longitudinal studies have demonstrated that significant changes in cardiac structure and function occur in trained subjects after a structured programme of exercise training.8,9 Previous cross-sectional studies have reported a close association of LV dimensions to body composition, in particular to FFM, hypothesizing that differences in LVM between athletes and controls depend on variations in FFM.11,21 However, it was unclear from such designs whether the relative increase in LVM is due to physiological training adaptations per se or is simply related to the increased FFM observed in elite athletes.
This prospective, longitudinal study demonstrates that LVM and FFM significantly vary during the season in professional soccer players, as a result of the effects of training. Furthermore, LVM and FFM show a similar trend during the observational period and ΔLVMi demonstrates a strong relationship with ΔFFM. We demonstrated that the left ventricle adapts rapidly in response to the increase of the metabolically active tissue induced by training, suggesting that the increase of LVM observed in elite athletes should be considered a physiological adaptation to the stimulus induced by regular intensive physical exercise. Considering the relationship between FFM and LVM, FFM should be taken into account when indexing LVM. FFM represents the fat-free body compartment with a dominating metabolic demand22 while BSA is affected by fat mass and it may be an inappropriate indexing variable, especially in populations where changes in body composition occur.11 This is the first prospective study with a longitudinal design to contemporary perform DXA and echocardiography to detect LV changes occurring in professional athletes during the regular season.
Batterham and colleagues23 investigated the longitudinal changes in LVM and FFM in army recruits before and after a 10-week military basic training. The authors demonstrated that LV hypertrophy and lean body mass changes do not occur at the same magnitude in response to chronic exercise, reporting that hypertrophy occur to a greater extent in the LV than in total lean body mass.23 However, the study population was different from the elite athletes studied in the present investigation, and the study design by Batterham et al. included only two time points: the baseline and the 10-week measurement. In our study, the use of more time points allowed the investigation of the longitudinal variations of LVM and FFM with greater detail regarding the association between LVM and FFM changes, demonstrating that, as a result of training, there was a gradual and significant increase in LVM and FFM during the season and that other LV parameters, such as LV EDV and LV stroke volume, are sensible to the exercise conditioning and quickly vary during the different periods of the training programme.
Pressler and colleagues21 have recently confirmed that adaptation of LV dimensions is closely related to body composition in athletes, analysed by skinfold thickness with the inherent limitations of this technique. The authors reported that the increase of LVM in elite athletes performing high dynamic exercise demonstrates adaptation beyond the sole influence of body composition, with cardiac dimensions remaining significant independent of the anthropometric scaling variable.21 Even if different athletic populations have been investigated and different techniques have been used in our study compared to the study by Pressler et al. to determine FFM, we demonstrated that other haemodynamic parameters, such as LV stroke volume, show a significant association with LVM, suggesting that variations of LVM occurs as part of a global adaptation of the heart to physical exercise, encompassing not only a modification of body composition but also variations of haemodynamic loads.
The present findings, confirming the variability of morphological adaptations during the training season, highlight the importance to characterize the time point of the training programme during which the athlete is being evaluated and indicate a possible source of error in cross-sectional studies that do not assess athletes at the peak of their training programme and competitive season.
Detraining
A period of detraining results in a reduction of exercise-induced increase in LV stroke volume in conjunction with a regression of LVM and of LV hypertrophy.24–26 These rapid changes are considered a benign adaptation of the athlete’s heart and have been demonstrated after long-term deconditioning in the majority of athletes with cavity enlargement.19,25,26 In the present study, we confirmed the dynamic nature of LV adaptations and the relationship with training and detraining, demonstrating a normalization of LV echocardiographic parameters after 2 months of detraining, performed as per programme. Therefore, our findings support the hypothesis that exercise training-induced adaptive changes in LVM occur rapidly, mimicking the pattern of chronic volume overload, and that the increase of LVM is reversible after cessation of training in professional soccer players.
Limitations
First, we have selected a cohort of subjects homogeneous for age, sex, training regimen, and loading conditions. As this selection represents a key strength of the study, we cannot extrapolate our findings to the whole population of elite athletes composed of subjects engaged in different disciplines with peculiar cardiovascular demands. Although we have studied a small number of athletes, the addition of subjects from other teams may well have presented greater threats to validity. Pooling data from additional teams could result in uncontrollable differences across subjects in factors such as athletic skill, coaching practices, or competitive setting.
The population of the present study was composed only by male athletes: in consideration of a different metabolic demand and hormonal response to training, we cannot exclude that female athletes could experience dissimilar variations of LVM and FFM in response to exercise conditioning in comparison with male subjects. This merits further investigations. A correlation between maximal oxygen uptake and LVM has been previously demonstrated in a cross-sectional study.27 However, in this study we have not determined oxygen uptake. The confirmation of this correlation needs to be investigated in further longitudinal studies. Finally, in the present study a control group has not been included. However, this would be more a theoretical than an actual methodological limitation within the framework of our investigation. Indeed, in this study we chose a longitudinal design with subjects acting as their own controls.
Conclusion
This study demonstrates that changes in LVM occur in close association with changes in FFM in top-level athletes, suggesting that the athlete’s heart adapts concurrently with the increase of the metabolically active tissue induced by training. Therefore, the increase of LVM observed in elite athletes could be considered a physiological adaptation to the stimulus induced by regular intensive physical exercise.
Further research is required to investigate whether this adaptation of the athlete’s heart could be confirmed in different athletic populations engaged in different disciplines and in female athletes.
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Conflict of interest
The authors declare that there is no conflict of interest.
Acknowledgements
The authors wish to thank the Chairman of Siena Football Club, all members of medical and coaching staff, athletes, and managers for their support in the study.
Comments