Article Text


Original article
Echocardiographic findings in 2261 peri-pubertal athletes with or without inverted T waves at electrocardiogram
  1. Leonardo Calò1,
  2. Fabio Sperandii1,2,
  3. Annamaria Martino1,
  4. Emanuele Guerra1,2,
  5. Elena Cavarretta3,4,
  6. Federico Quaranta2†,
  7. Ermenegildo de Ruvo1,
  8. Luigi Sciarra1,
  9. Attilio Parisi2,
  10. Antonia Nigro3,
  11. Antonio Spataro5,
  12. Fabio Pigozzi2
  1. 1Division of Cardiology, Policlinico Casilino, ASL Rome B, Rome, Italy
  2. 2Department of Health Sciences, University of Rome “Foro Italico”, Rome, Italy
  3. 3The FMSI Sport Medicine Institute, Villa Stuart Sport Clinic—FIFA Centre of Excellence, Rome, Italy
  4. 4Department of Medical-Surgical Sciences and Biotechnologies, University of Rome “La Sapienza”, Rome, Italy
  5. 5Institute of Sports Medicine and Science (CONI), Rome, Italy
  1. Correspondence to Professor Leonardo Calò, Division of Cardiology, Policlinico Casilino, ASL Rome B, Via Casilina 1049, Rome 00169, Italy; leonardo.calo{at}


Objective T wave inversion (TWI) has been associated with cardiomyopathies. The hypothesis of this study was that TWI has relevant clinical significance in peri-pubertal athletes.

Methods Consecutive male soccer players, aged 8–18 years, undergoing preparticipation screening between January 2008 and March 2009 were enrolled. Medical and family histories were collected; physical examinations, 12-lead ECGs and transthoracic echocardiogram (TTE) were performed. TWI was categorised by ECG lead (anterior (V1–V3), extended anterior (V1–V4), inferior (DII–aVF) and infero-lateral (DII–aVF/V4–V6/DI-aVL)) and by age.

Results Overall, 2261 (mean age 12.4 years, 100% Caucasian) athletes were enrolled. TWI in ≥2 consecutive ECG leads was found in 136 athletes (6.0%), mostly in anterior leads (126/136, 92.6%). TWI in anterior leads was associated with TTE abnormalities in 6/126 (4.8%) athletes. TWI in extended anterior (2/136, 1.5%) and inferior (3/136, 2.2%) leads was never associated with abnormal TTE. TWI in infero-lateral leads (5/136, 3.7%) was associated with significant TTE abnormalities (3/5, 60.0%), including one hypertrophic cardiomyopathy (HCM) and two LV hypertrophies. Athletes with normal T waves had TTE abnormalities in 4.4% of cases, including one HCM with deep Q waves in infero-lateral leads.

Conclusions In this broad population of peri-pubertal male athletes, TWI in anterior leads was associated with mild cardiac disease in 4.8% of cases, while TWI in infero-lateral leads revealed HCM and LV hypertrophy in 60% of cases. ECG identified all cases of HCM.

Statistics from


The prevalence and clinical significance of T wave inversion (TWI) at ECG in young athletes has been a subject of investigation.1–5 Previous studies have shown that TWI in leads V1–V2 are relatively rare among athletes, with a prevalence ranging from 2.5%, to 4.7%,4 and 0.8% when they extend to V3.5 TWI in infero-lateral leads has been described in 1.5%–1.8% of postpubertal Caucasian athletes,5 ,6 and inverted T waves confined to lateral leads were observed in 0.1%–0.3% of young athletes.4 ,5 Although independent from training-related physiological cardiac remodelling,7 ,8 in the absence of echocardiographic abnormalities and cardiac symptoms, TWI in anterior leads have traditionally been considered benign in children.9 Conversely, their persistence after puberty4 ,10 ,11 or their presence in inferior and lateral leads3 has been associated with cardiomyopathy.

The hypothesis of this study was that TWI has relevant clinical significance in peri-pubertal athletes. Therefore, a systematic application of transthoracic echocardiogram (TTE) during preparticipation screening was adopted in a selected population of peri-pubertal male athletes competing at the regional level, with or without TWI at ECG.


Study population

Consecutive male soccer players undergoing preparticipation screening in the FMSI Sport Medicine Institute—Villa Stuart Sport Clinic in Rome, Italy, were enrolled. Eligible athletes had to be never screened for cardiac disease before. Prior to enrolment, each subject (or the parents of minors) gave written consent to participate in the study. The study was approved by the local institutional review board.

Study procedures

Each participant had a standardised personal and family medical history review, physical examination and resting 12-lead ECG and underwent TTE, performed by an expert cardiologist and sport medicine practitioner.

Medical history and physical examination

Personal medical history was considered positive in case of exertional chest pain or discomfort, syncope or near-syncope during or after exercise, palpitations, and in the presence of shortness of breath or fatigue out of proportion.4 ,12 ,13 Family medical history was considered positive when close relatives had experienced a premature (<40 years) heart attack or sudden death, or had cardiomyopathies, Marfan syndrome, channelopathies, severe arrhythmias, or other disabling cardiovascular diseases.4 ,12 ,13 Positive physical findings included musculoskeletal and/or ocular features suggestive of Marfan syndrome, diminished and/or delayed femoral artery pulses, mid- or end-systolic clicks, a second heart sound that was single or widely split and fixed with respiration, marked heart murmurs (any diastolic or systolic grade >2/6), irregular heart rhythm, and brachial blood pressure >140/90 mm Hg.4 ,12 ,13 Finally, body surface area (BSA) and body mass index (BMI) were calculated.


ECGs were performed according to Italian guidelines.8 Recordings were performed at rest using standard 12-lead placement (Mortara, Milwaukee, USA). ECGs were recorded at a paper speed of 25 mm/s and at a standard gain of 1 mV/cm. Heart rate and QRS axis were manually calculated. T wave voltages, ST segments, QRS duration, PR interval and QT interval were measured in each lead with callipers. Two independent sport medicine physicians and a cardiologist retrospectively examined and interpreted ECG tracings according to European Society of Cardiology (ESC) recommendations.10 Discrepancies were resolved by consensus. Each physician was blinded to history and physical examination data. TWI was diagnosed in the presence of a negative T wave ≥1 mm in ≥2 contiguous leads4 and was localised as follows: anterior leads (V1–V3), extended anterior leads (V1–V4), inferior leads (DII–aVF) and infero-lateral leads (DII–aVF/V4–V6/I–aVL). TWI in anterior leads was diagnosed in the absence of complete right bundle branch block.4

ECG diagnosis of LV hypertrophy (LVH) is challenging in young athletes since the commonly used QRS voltage criteria, applying to adults,14 lack specificity.15 Therefore, for the electrocardiographic assessment of LVH, both the Sokolow–Lyon criteria16 and the Romhilt criteria17 were considered. The QT interval was corrected for heart rate using the Bazett formula.18 ECG abnormalities were divided into common/training-related (ie, possibly related to physiological athlete's heart remodelling) and uncommon/training-unrelated, in line with ESC recommendations.7 Early repolarisation was defined as J point elevation manifested either as QRS slurring or notching, ST segment elevation for more than 0.1 mV in at least two contiguous leads.19

Trans-thoracic echocardiography

Bi-dimensional TTEs were performed in left lateral decubitus by five different experienced cardiologists, using an Acuson ultrasound device (Siemens Healthcare, Erlangen, Germany). Images were acquired from parasternal, apical, subcostal and suprasternal windows and were digitally recorded according to the American Society of Echocardiography guidelines.20 M-mode, bi-dimensional echocardiographic data, as well as pulsed, continuous and colour Doppler flow mapping were recorded. An expert cardiologist reviewed each examination retrospectively, independently and blinded to history; physical examination and ECG findings of the athletes. LV internal diameters and wall thickness measurements were made in M-mode or, alternatively, by using the leading edge convention.20 Hypertrophic cardiomyopathy (HCM) was defined as a hypertrophied (wall thickness >12 mm), non-dilated (end-diastolic diameter <45 mm) LV and one of: impaired diastolic function, enlarged left atrial diameter, systolic anterior motion of the anterior mitral valve leaflet and associated LV outflow tract gradient, asymmetrical pattern of hypertrophy or family history of HCM.5 ,21 Valvular heart diseases were evaluated according to European Association of Echocardiography recommendations.22 ,23 Quantification of the RV size was obtained by measuring the mid-cavity and basal trasversal and longitudinal RV diameter in the apical 4-chamber view.20 RV outflow tract was measured from the parasternal short axis.20 The RV fractional area change and the displacement of the tricuspid annulus toward the apex in systole were also calculated.20 Arrhythmogenic RV cardiomyopathy (ARVC) was diagnosed in accordance to current guidelines.11

Statistical analysis

Continuous variables are reported as the mean±SD and data ranges; categorical variables are reported as number and percentage per category. Continuous variables were analysed with Student's t test and categoric variables with χ2 test, when appropriate. Statistical significance was considered for p<0.05. Computations were performed with SPSS V.19 (IBM, Armonk, New York, USA) and STATA V.11 software (StataCorp LP, Texas, USA).


Study population and clinical findings

Between January 2008 and March 2009, 2261 consecutive male Caucasian athletes were enrolled. Mean age was 12.4±2.6 years. The demographic and clinical characteristics of the population, divided in two subgroups, depending on the presence of TWI, are summarised in table 1.

Table 1

Demographic and clinical characteristics of athletes with or without TWI

No athlete had known cardiac disease and 26 athletes (1.1%) were found to have abnormal personal medical history. In all, 42 athletes (1.9%) reported positive family medical history, mostly for ischaemic heart disease, but also for dilated cardiomyopathy (one athlete), inter-ventricular septal defect (one case) and aortic disease (one case). No athletes reported a family history of sudden cardiac death. A total of 18 athletes (0.8%) had significant abnormal findings on physical examination (table 1). In general, athletes with TWI were younger, with smaller BSA, BMI and less total training hours. No significant differences were found in the family history or physical examination.

Electrocardiographic findings

Electrocardiographic findings are presented in table 2, in which ECG abnormalities are categorised by their likely origin: common/training-related or uncommon/training-unrelated, and the study population is divided in two subgroups, depending on the presence of TWI.

Table 2

Electrocardiographic findings in athletes with or without TWI

The two subgroups were homogeneous for the electrocardiographic parameters, except for heart rate, which was slightly higher in athletes with TWI, and QRS duration, which conversely, was lower (table 2).

TWI in at least two consecutive leads was found in 136 athletes (6.0%), distributed in anterior, extended anterior, inferior and infero-lateral leads in 126/136 (92.6%), 2/136 (1.5%), 3/136 (2.2%) and 5/136 (3.7%) athletes, respectively (table 3).

Table 3

Athletes with negative T waves by age group and ECG lead localisation

TWI was more common among individuals aged 8–10 (75/599, 12.5%) and 11–13 years (45/857, 5.2%) than in those aged 14–16 (14/689, 2%) or 16–18 years (2/116, 1.8%) (table 3). A significant association (p<0.001) was found between age groups and the distribution of T waves in ECG leads. Other training-unrelated ECG abnormalities are listed in table 2.

Echocardiographic findings

Echocardiographic findings in athletes with or without TWI are presented in table 4. Distribution of various parameters is depicted in figures 1 and 2.

Table 4

Echocardiographic findings in athletes with or without TWI

Figure 1

Age, heart rate and LV mass distribution in athletes with or without T wave inversion (TWI). Age and HR distribution are depicted at the top, on the left and on the right, respectively; LV mass and LV mass indexed for height distribution are depicted down, on the left and on the right, respectively. Whiskers plot highlights differences between groups.

Figure 2

End-diastolic and end-systolic diameter, inter-ventricular septum and posterior wall thickness distribution in athletes with or without T wave inversion (TWI). End-diastolic and end-systolic diameters are depicted at the top, on the left and on the right, respectively; inter-ventricular septum and posterior wall thickness distribution are depicted down, on the left and on the right, respectively. Whiskers plot highlights differences between groups.

Athletes with TWI had small ventricular dimensions, walls thickness and atrial diameter, according to their younger age (table 4). No differences were found in TTE measurements indexed by BSA and/or height (table 4). Abnormal TTE was observed in 9/136 (6.6%) athletes with TWI and in 93/ 2125 (4.5%) athletes without TWI. The distribution of TTE findings in athletes, as well as their relationship with TWI localisation in ECG leads, is depicted in figure 3. There were no cases of ARVC. Only 6 out 126 (4.8%) athletes with TWI in the anterior leads had minor TTE abnormalities, including patent foramen ovale (three cases: one age 8, two age 10), mitral valve prolapse (two cases: one age 10, one age 12) and bicuspid aortic valve (one age 12). No athletes with TWI in the extended anterior or inferior leads had any echocardiographic findings. Among the five subjects with infero-lateral TWI, 3 (60.0%) had significant TTE abnormalities, including one HCM (a 13-year-old) and two LVH (one age 16, one age 17); the remaining two had normal TTEs.

Figure 3

Relationship between ECG and transthoracic echocardiogram (TTE) findings in athletes with or without T wave inversion (TWI) at ECG. Arrows indicate subgroups. PFO, MVP and BAV indicate patent foramen ovale, mitral valve prolapse and bicuspid aortic valve, respectively.

The vast majority of athletes without TWI (2032/2125, 95.6%) had normal TTE findings. However, cardiac abnormalities, in some cases serious, were found in 93 children (4.4%), including: HCM (one), mild LVH (six), mild LV dilation (LVD) (six), inter-atrial septal defect (13), bicuspid aortic valve (16), patent foramen ovale (24), inter-atrial septal aneurysm (14), mitral valve prolapse (eight), patent ductus arteriosus (two), and pulmonary artery dilation or subvalvular pulmonic membrane (two and one, respectively). In the athlete with HCM and normal T waves, ECG revealed deep Q waves in infero-lateral leads, suggestive of cardiomyopathy.

Sensitivity, specificity, positive and negative predictive values of TWI for preparticipation screening

The sensitivity and specificity of TWI in identifying cardiac TTE abnormalities mandating sport restriction (ESC guidelines for preparticipation screening12) were 0.7% (95% CI 0.004% to 0.0012%) and 100% (95% CI 0.997% to 1.000%), respectively. The positive and negative predictive values were 50% (95% CI 0.479% to 0.521%) and 94.0% (95% CI 0.929% to 0.949%), respectively. The sensitivity, specificity and predictive values of TWI according to their localisation on the surface ECG, in detecting TTE abnormalities needing clinical follow-up and in identifying cardiac electro-structural abnormalities mandating sport restriction12 or needing clinical follow-up are presented in the online supplementary appendix (table S1A, B respectively).


In this study, we performed comprehensive preparticipation screening including medical and family history, physical examination, ECG and TTE in 2261 young male soccer players competing at the regional level.

When compared with athletes with normal T waves, those with TWI had lower total training times, with consequently higher heart rate values, and were younger, with smaller BSA and BMI. The two groups were homogeneous for family history, physical examination, symptoms, electrocardiographic parameters and indexed echocardiographic measurements. The prevalence of TWI was 6% in our population and decreased with age, being exceptional in athletes ≥14 years. When present, TWI usually appeared in the V1–V3 leads, where it was rarely associated with cardiac abnormalities. TWI in infero-lateral leads was extremely infrequent (five cases, 0.2% of the study population) and was associated with HCM or LVH in a high percentage of cases (3/5, 60.0%). Echocardiographic abnormalities were present in 4.4% (93/2125) of athletes with normal ventricular repolarisation, including one HCM and a number of other conditions potentially related to chronic cardiac disease requiring serial follow-up. The sensitivity of TWI in identifying cardiac structural abnormalities mandating sport restriction was 0.7%, while the specificity was 100%, with positive and negative predictive values of 50% and 94.0%, respectively.

When TWI is present in competitive athletes, ruling out cardiomyopathies is mandatory. The prevalence and distribution of TWI are influenced by age and ethnicity. In fact, the prevalence of TWI ranges from 0.43% in young, Caucasian athletes24 to 39% in adolescent African/Afro-Caribbean ones.25

TWI is rare in postpubertal athletes, in whom it may be associated with cardiomyopathy. In a cohort of 1005 athletes aged 24±6 years (75% men),2 only 27 athletes had TWI. In a heterogeneous population of 1220 athletes aged 22.6±6 years, TWI in anterior and lateral leads was absent in Caucasian subjects, in whom no cardiomyopathies were observed.26 In another study27 involving 510 athletes aged 19±0.3 years, only three subjects with TWI were identified; interestingly, one was diagnosed with HCM.

The prevalence of TWI is less infrequent among postpubertal athletes, where it may be associated with cardiomyopathies. Marek et al reported a 0.43% prevalence of T wave abnormalities in a population of 32 561 high school students (51% boys).24 In their cohort of 32 652 amateur athletes (median age 17 years), Pelliccia et al28 reported 2.3% of TWI in precordial leads. Sharma et al29 observed TWI in 4% of 1000 junior elite athletes aged 15.7±1.4 years. Meanwhile, Di Paolo et al25 compared a different population of 154 adolescent African soccer players aged 15.9±0.7 years with 62 Caucasian players, revealing that TWI was largely present in adolescent Africans and never associated with cardiac disease. Noteworthy, Papadakis et al5 observed a 4% prevalence of TWI among 1710 (age 16±1.7 years) well trained, postpubertal athletes of various ethnicities.5 TWI beyond right precordial leads was observed in only 0.1% of athletes and was never associated with cardiomyopathies.5 TWI in the inferior or lateral leads was rare, and associated with LVH, mitral valve prolapse or atrial septal defects.5

While the prevalence of TWI in postpubertal athletes has been intensively investigated, few studies have explored the prevalence and significance of TWI among younger athletes. In particular, Migliore et al4 observed negative T wave prevalence of 4.7%, 0.9% and 0.1% in the right precordial, inferior, and lateral leads, respectively, in a population of athletes aged 13.9±2.2 years. A 2.5% prevalence of cardiomyopathies (three ARVC and one HCM) was present among athletes with TWI. However, the prevalence of cardiac abnormalities among individuals without TWI was not assessed.

The aim of our paper was not to identify athletes at risk of sudden cardiac death through integrated ECG criteria, but to explore the clinical significance of TWI itself. Therefore, our study, like very few others, investigated repolarisation abnormalities in a selected population of pubertal male athletes by systematically performing TTE in all subjects, regardless of TWI status.

The pubertal phase is challenging because of a potential overlapping between the ‘juvenile pattern’ and repolarisation abnormalities reflecting structural cardiac diseases. In line with Migliore et al,4 athletes with TWI in our study population were younger and less trained. However, they had electro and echocardiographic findings (indexed by BSA) similar to athletes with normal T waves. The higher heart rate and narrower QRS can be explained by the lower training time and younger age. In our population, TWI had a poor ability to identify athletes needing sport restriction, which varied significantly with their distribution across ECG leads (table 4, panel A). Indeed, TWI in anterior leads was not associated with cardiomyopathies, including ARVC, tending to confirm that the benign juvenile pattern extends up to the age of 14 years. However, this might be explained also by a low prevalence of ARVC in the Lazio region where the study was conducted or even by underdiagnosis of ARVC due to an incomplete phenotype in this young population.

Consistent with the findings of Papadakis et al5 and Wilson et al,26 we observed that the prevalence of TWI in infero-lateral leads was very low and was associated with structural cardiac abnormalities, including HCM or LVH in three of five cases; all three were ≥13 years old.


The selected subset of athletes enrolled limited the influence of sex, ethnicity and age on TWI prevalence and distribution, thus differentiating our study from the published ones.4 ,5 However, any conclusion can be drawn about the clinical significance of TWI in athletes of different ages, races and sport disciplines and female gender. The low prevalence of cardiomyopathies observed may be related to the young age of athletes, at which the diseases may not have manifested. A further limitation is represented by the cross-sectional study design without a clinical, electrocardiographic and echocardiographic follow-up. However, a follow-up study of this population will be the object of a future publication.


In our population, TWI in anterior leads was associated with patent foramen ovale, mitral valve prolapse and bicuspid aortic valve in 4.8% of cases. Negative T waves in infero-lateral leads revealed HCM and LVH in 60% of cases. ECG was sufficient to exclude the two cases of HCM. Implementation of echocardiography allowed us to discover cardiac abnormalities, in some cases of a serious nature, requiring serial follow-up because of possible cardiac deterioration over time in athletes with and without TWI.

Key messages

  • What is already known on this subject?

  • T wave inversion (TWI) at ECG is relatively rare among athletes and its clinical significance is not well known.

What might this study add?

  • This study reported a poor association of cardiac structural abnormalities (assessed through systematic transthoracic echocardiography) with TWI in anterior and inferior leads at ECG in a broad population of peri-pubertal Caucasian male soccer players. This study also observed that TWI in infero-lateral leads was associated with transthoracic echocardiogram (TTE) abnormalities in 60% of cases, including one case of hypertrophic cardiomyopathy and two cases of LV hypertrophy. TTE abnormalities, including one case of hypertrophic cardiomyopathy, have been described in 4.4% of athletes with completely normal T waves.

How might this impact on clinical practice?

  • In peri-pubertal Caucasian male soccer players, TWI in anterior leads is benign, while negative T waves in infero-lateral leads raise suspicion of structural cardiac abnormalities, particularly in older boys.


We would like to thank Dr Luigi de Matteis for his assistance during the composition of the paper.


View Abstract
  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Files in this Data Supplement:


  • Contributors LC conceived and designed the paper. FS, AM, EG and EC performed the analysis and interpretation of data and participated in the preparation of the manuscript. QF, EdR, LS, AP, AN, AS and FP performed the interpretation of the data and participated in the preparation of the manuscript. All the authors have read and approved the paper.

  • Competing interests None.

  • Ethics approval This study has been approved by the local institutional review board.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement The authors declare that the data of the research article are original.

Request permissions

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

Linked Articles