Prevalence and relation to training of electrocardiographic findings in Caucasian children and adolescents 8- to 18-year-old practicing competitive sport

Abstract

Aims

Training frequently induces electrocardiographic (ECG) changes that mimic heart diseases, requiring specific criteria for interpretation. Paediatric athletes represent a unique population as training-induced changes and those due to sexual maturation interact, and specific criteria may be needed. We aimed to assess the prevalence and its relation to training of ECG abnormalities in young athletes aged 8–18 years.

Methods and results

We included 2458 young apparently healthy Caucasian athletes undergoing pre-participation screening. Electrocardiographic abnormalities were classified according to the 2017 International criteria (IC) by adding fragmented QRS and low QRS voltages in limb leads. A subgroup analysis was conducted to test the differences according to a 12-year-old age cut-off, the threshold for IC application. Common (>5%) findings included only mild sinus bradycardia (55–60 b.p.m.), early repolarization, incomplete right bundle branch block, voltage criteria for either left or right ventricular hypertrophy and T-wave inversion in V1–V3 before the age of 12. A heart rate of 50–55 b.p.m., first-degree atrioventricular block, right-axis deviation, and T-wave inversion in V1–V3 in athletes 12- to 16-year-old were uncommon (1–5%) and their prevalence was modulated by age, gender, and/or training status. All other ECG findings were rare (<1%). Ten (0.4%) athletes received an at-risk cardiovascular disease diagnosis including 6 with rare and 1 with uncommon ECG abnormalities: 3 of 7 were classified as normal by IC.

Conclusion

Paediatric athletes exhibit less prominent ECG changes than their adult counterparts, requiring specific criteria. The significance of certain ECG patterns should be evaluated considering age, sex, and training level.

Graphical Abstract

Graphical Abstract.

Introduction

Physical activity induces electrical and structural cardiac remodelling, which can lead to electrocardiographic (ECG) changes. These alterations are common in athletes and are associated with a low probability of underlying disease; however, they resemble patterns that may indicate a pathological condition in sedentary individuals.¹ A typical example is the Sokolow-Lyon voltage criterion for left ventricular hypertrophy (LVH).² Within the context of pre-participation screening (PPS), systematic second-level investigations for such common abnormalities would result in an unacceptable increase in costs with marginal diagnostic benefits. For this reason, specific criteria for interpreting athletes’ ECGs have been developed,³ with the latest version being the 2017 International Criteria (IC).⁴ These criteria classify ECG abnormalities into three categories based on prevalence and likelihood of disease: normal, borderline, and abnormal.

Children and adolescents engaged in competitive sports represent a unique clinical challenge for physicians involved in PPS.⁵ During this stage, pubertal development occurs together with transitioning from a paediatric to an adult ECG pattern. Furthermore, due to incomplete physical development, the nature of the sports practised (predominantly team sports rather than endurance activities), and typically lower training intensity compared to adults, the use of ECG interpretation criteria developed for adult athletes may be inadequate and lead to the classification of potentially pathological findings as normal.^6–9 While the 2017 IC apply to athletes aged 12 years and older, they were mainly developed based on data from older athletic populations. Additionally, considering that many children begin competitive sports before the age of 12, the availability of criteria applicable to younger athletes would be highly beneficial.

The aim of this study is to investigate the prevalence of ECG abnormalities and their correlation with variables such as age, sex, training intensity, and type of sport in a large cohort of apparently healthy children and adolescents engaged in competitive sports who underwent PPS. Furthermore, we performed a subgroup analysis based on age, using 12 years (the threshold at which the 2017 IC start to apply) as the age cut-off.

Methods

Study design and enrolment criteria

This observational, retrospective study included a consecutive series of young competitive Caucasian athletes aged 8–18 years old who underwent PPS between September 2022 and July 2023 at three sports medicine centres of the Italian National Health System: Padova, Noale-Venice, and Cagliari. Moreover, in order to expand the number of highly trained (≥8 h/week) young athletes, we also included a series of 225 athletes referred for PPS to the sports medicine centre of Semmelweis University, Budapest (Figure 1). To be enrolled, athletes needed to train for at least 4 h per week and take part in competitions organized by official national sports federations. Those with previous known cardiovascular diseases were excluded. Because ethnicities are known to influence the ECG findings in athletes, we included only White/Caucasian individuals, as athletes with other ethnic backgrounds represented a small percentage of those who underwent PPS during the study period in our sports medicine centres, thus preventing proper subgroup analysis. The sports practiced were classified into four categories according to the European Society of Cardiology recommendations: skill, power, endurance, and mixed.¹⁶ However, we excluded athletes practicing skill sports because of the limited effect of such disciplines on the cardiovascular system.

Figure 1.

CONSORT-style diagram illustrating the structure of the cohort investigations. The diagram shows the recruitment and clinical work-up of 2458 athletes. Cardiac diagnosis was established in 10 athletes (0.4%). CMR, cardiac magnetic resonance; CT, computed tomography; ECG, electrocardiogram; ES, electrophysiological study; ET, exercise testing; P/E, physical examination.

The local Ethical Committee approved the study: because of the observational and retrospective nature of the study, consent from participants was waived. The data supporting this study are available from the corresponding author upon reasonable request.

Pre-participation screening protocol

In Italy, PPS is carried out by a physician with a postgraduate degree in Sports Medicine, and physical exercise is mandatory for all non-professional competitive athletes. The age at which the PPS starts depends on the sports discipline and is set by each sports federation based on when official competitions begin. The screening protocol is established by Italian law and follows the National guidelines.^10,11 It includes family and personal history, physical examination with blood pressure measurement, spirometry, urine dipstick, visual acuity test, resting 12-lead ECG, and stress testing, particularly aimed at evaluating ventricular arrhythmia inducibility. Further examinations may be required in case of abnormal findings at first-line screening based on ECG findings and other tests.

In Hungary, every competitive athlete must undergo PPS carried out by a physician with specific training in Sports Medicine. It is based on the European Society of Cardiology recommendation,¹² and it includes personal and family history, physical examination with blood pressure measurement, urine test, resting 12-lead ECG, and echocardiography. The athletes with abnormal findings at first evaluation underwent further examinations. In the case of CVD diagnosis, the athlete’s management followed the European guidelines.¹³

For both screening protocols, because of the absence of specific criteria for the interpretation of athletes’ ECG in children under 12 years old, the decision on further investigations in this age group was based on the physician’s individual judgment.

Twelve-lead resting electrocardiogram analysis

ECGs were acquired in the supine position with the limb leads placed at the wrists and ankles at a standard speed of 25 mm/s and a gain of 1 mV/cm. Filters were set at 0.05 and 150 Hz. Electrocardiographic recordings were acquired using Mortara X-Scribe 5 in Cagliari, Philips PageWriter TC70 in Padua and Noale, and GE CardioSoft V6.73 in Budapest. No automated software was used for ECG interpretation, which was performed manually. They were scanned at high definition and QRS amplitudes were measured with digital callipers. Two cardiologists interpreted ECG according to the 2017 International Recommendations for ECG interpretation in athletes⁴ and discrepancies were solved by the senior author. For sinus bradycardia (SB) and premature ventricular beats (PVBs), we also evaluated different cut-offs (<55 b.p.m. and ≥1, respectively) compared to those proposed by the International Recommendations (<60 b.p.m. and ≥2, respectively).

Moreover, two additional parameters were evaluated. The first was the presence of fragmented QRS in at least two contiguous leads, excluding aVR and excluding the rSr’ pattern in V1–V2.¹⁴ The second parameter was low QRS voltage in limb leads, defined as a QRS complex amplitude of <0.5 mV from peak to nadir in each limb lead.¹⁵ The ECG pattern of T-wave inversion (TWI) in lead V1–V4 seen exclusively in Black athletes was not considered, as only White/Caucasian athletes were included in the study.

Statistical analysis

Normal distribution of all continuous variables was examined using the Shapiro–Wilk test and Q–Q plot. Continuous variables are expressed as median and interquartile range (25th–75th percentiles). Categorical variables are expressed as n (%). The study sample was divided into two age subgroups, using 12 years old (the age at which the 2017 International Recommendations for ECG interpretation in athletes⁴ start to apply) as a cut-off. Differences between the two groups were evaluated with the χ² test or the Fisher’s exact test, as appropriate, for categorical variables and with the Student’s t-test or Mann–Whitney U test, as appropriate for continuous variables.

The prevalence of ECG findings in the study sample was classified into three groups: common (>5%), uncommon (1–5%), and rare (<1%). Univariate binary regression models were applied to explore the associations between ECG findings with a prevalence ≥1% and age, sex, type of sport (power/mixed vs. endurance sports), and mean hours of training per week. The results from this model are presented as odds ratio (OR) with accompanying 95% confidence intervals. Variables that showed a significant association in univariate analysis were entered into a multivariable regression model. Finally, a univariate binary regression analysis was performed to assess the association between ECG findings with a prevalence ≥1% and mean hours of training per week in the two pre-specified age subgroups. Interaction testing was performed to understand if the magnitude of the training effect differs significantly between the two age subgroups. A significance level of 0.05 from a two-tailed test was considered statistically significant throughout the study. Data were analysed with SPSS V.29 (IBM).

Results

Demographic and baseline characteristics

The study sample included 2458 athletes aged 8–18, median age 13, 51% males (Table 1). The majority (69%) practiced mixed sports disciplines, 25% endurance sports disciplines, and 6% power sports disciplines. The median number of weekly training hours was 5, with 38% of individuals exercising more than 6 h per week.

Table 1.

Characteristics of the study population in the entire study sample and in the two pre-specified age groups

	Total population (n = 2458)	<12 years old (N = 1384, 56%)	≥12 years old (N = 1074, 44%)	P-value
Age (years)	13 (11–15)	11.0 (10.0–12.0)	15.0 (14.0–17.0)	<0.001
Males	1252 (51%)	653 (47%)	598 (56%)	<0.001
Height (cm)	158 (148–169)	150 (141–157)	170 (163–177)	<0.001
Weight (kg)	50 (39–61)	41.0 (33.5–49.5)	61.0 (54.0–70.0)	<0.001
BSA (Mosteller, m2)	1.48 (1.27–1.66)	1.32 (1.17–1.48)	1.67 (1.55–1.82)	<0.001
BMI (kg/m2)	19.8 (17.4–21.9)	18.2 (16.4–20.6)	21.1 (19.5–23.0)	<0.001
Positive family historya	453 (18%)	249 (18%)	204 (19%)	0.22
Training time (h/week)	5.0 (4.0–6.0)	4.5 (4.0–6.0)	6.5 (4.50–8.00)	<0.001
Training time ≥6 h week	946 (38%)	272 (20%)	674 (63%)	<0.001
Years of competitive sport	2 (1–5)	1 (1–3)	4 (2–6)	<0.001
Endurance sport	623 (25%)	383 (28%)	240 (22%)	<0.001

^aFor premature (<40 years old in men and <50 years old in women) sudden death, inherited cardiomyopathies, or coronary artery disease.

Prevalence of electrocardiographic findings in the overall sample and in the two age subgroups

Electrocardiographic findings classified according to the 2017 IC in the entire study sample and in the two pre-specified age groups are shown in Table 2.

Table 2.

Electrocardiographic findings classified according to the 2017 International Criteria for electrocardiographic interpretation in athletes in the entire study sample and in the two pre-specified age groups

	Overall (n = 2458)	<12 years old (N = 1384, 56%)	≥ 12 years old (N = 1074, 44%)	P
HR (b.p.m.)	75 (67–84)	78.0 (69–85)	73 (65–82)	<0.001
P-wave duration (ms)	90 (80–100)	90 (80–100)	100.0 (90–110)	<0.001
PR interval duration (ms)	134 (120–155)	130 (120–140)	140 (120–160)	<0.001
QRS interval duration (ms)	90 (84–100)	90.0 (80–98)	96.0 (88–100)	<0.001
QT interval (ms)	360 (348–380)	360 (340–374)	372 (354–390)	<0.001
QTc interval, Bazett	406 (389–423)	405 (390–421)	407 (388–424)	0.91
QRS axis (°)	75 (68–83)	75 (60–80)	78 (70–85)	<0.001
International criteria: normal
LVH voltage criteria	333 (13.6%)	206 (14.9%)	127 (11.9%)	0.02
RVH voltage criteria	245 (10.0%)	157 (11.4%)	88 (8.2%)	0.01
Incomplete RBBB	578 (23.5%)	231 (16.7%)	347 (32.4%)	<0.001
Early repolarization	422 (17.2%)	153 (11.1%)	268 (25.0%)	<0.001
TWI in V1–V3 only ≤16 years old	132 (5.4%)	97 (7%)	35 (3.2%)	<0.001
Sinus bradycardia (<60 b.p.m.)a	578 (23.5%)	231 (16.7%)	347 (32.4%)	<0.001
Sinus bradycardia (<55 b.p.m.)	93 (3.8%)	38 (1.5%)	55 (2.2%)	0.002
Sinus bradycardia (50–55 b.p.m.)	75 (3%)	35 (2.5%)	40 (4%)	0.112
Ectopic atrial rhythm	37 (1.5%)	23 (1.7%)	14 (1.3%)	0.47
Junctional rhythm	3 (0.1%)	1 (0.1%)	2 (0.2%)	0.42
1st degree AV block	25 (1.0%)	4 (0.2%)	21 (2.0%)	<0.001
2nd degree Mobitz I AV block	0 (0%)	0 (0.0%)	0 (0.0%)	–
International criteria: borderline
Left-axis deviation	8 (0.3%)	1 (0.1%)	7 (0.7%)	0.01
Left atrial enlargement	7 (0.3%)	3 (0.2%)	4 (0.4%)	0.47
Right-axis deviation	34 (1.4%)	8 (0.6%)	26 (2.4%)	<0.001
Right atrial enlargement	22 (0.9%)	18 (1.3%)	4 (0.4%)	0.02
Complete RBBB	6 (0.2%)	2 (0.1%)	4 (0.4%)	0.33
International criteria: abnormal
Other TWI	20 (0.8%)	3 (0.2%)	17 (1.6%)	<0.001
ST-segment depression	0	0	0	–
Pathological Q waves	0	0	0	–
Complete LBBB	0	0	0	–
QRS duration ≥ 140 ms (no LBBB/RBBB)	1 (0.04%)	1 (0.07%)	0 (0.0%)	1.0
Epsilon wave	0	0	0	–
Ventricular pre-excitation	1 (0.04%)	0	1 (0.09%)	1.0
Prolonged QT interval	1 (0.04%)	1 (0.07%)	0	1.0
Brugada type 1 pattern	0	0	0	–
Profound sinus bradycardia (<30 b.p.m.)	0	0	0	–
PR duration ≥400 ms	0	0	0	–
2nd degree Mobitz II AV block	0	0	0	–
3rd degree AV block	0	0	0	–
Premature ventricular beats ≥2	3 (0.1%)	3 (0.2%)	0	0.262
Premature ventricular beats ≥1a	9 (0.4)	5 (0.6)	4 (0.4)	0.96
Atrial tachyarrhythmias	0	0	0	–
Ventricular tachyarrhythmias	0	0	0	–
Low QRS voltagesb	9 (0.4)	3 (0.2)	6 (0.5)	0.41
Fragmented QRSb	11 (0.5)	4 (0.3)	7 (0.6)	0.30

^aModified cut-off compared to the International Criteria.

^bNot included in the International Criteria.

AV, atrioventricular; b.p.m., beats per minute; LBBB, left bundle branch block; LVH, left ventricular hypertrophy; RBBB, right bundle branch block; RVH, right ventricular hypertrophy; TWI, T-wave inversion.

In the overall population, common (>5%) findings in both the overall study sample and in the two age subgroups were voltage criteria for LVH and right ventricular hypertrophy (RVH), incomplete right bundle branch block (iRBBB), early repolarization (ER), and mild SB (heart rate 55–60 b.p.m.). T-wave inversion in leads V1–V3 was a common pattern in the overall study sample, but correlation analysis between age and the prevalence of TWI in V1–V3 only (Figure 2) showed that the pattern became uncommon (1–5%) between the age of 12 and 17 and rare (<1%) after the age of 17.

Figure 2.

Histograms showing the prevalence of negative T-wave in V1–V3 depending on age.

Sinus bradycardia <55 b.p.m., right-axis deviation (RAD), first-degree atrioventricular block (AVB), and ectopic atrial rhythm (EAR) were uncommon (1–5%) while all the other ECG findings were rare (<1%). Distribution analysis showed that, in both age groups, a heart rate <50 b.p.m. and a PR interval ≥200 ms were rare (<1%) (Figure 3). On the contrary, there was no correlation between age and P-wave duration, QRS duration and axis, and QTc duration (see Supplementary material online, Figure S1).

Figure 3.

Heart rate (HR) and PR interval duration: distribution and age-specific trends. Left panels: Histograms showing the frequency distribution of HR (top) and PR interval (bottom) in the two age groups aged 8–11 years and adolescents aged 12–18 years. Right panels: Boxplots illustrating age-specific trends for HR (top) and PR interval duration (bottom) from 8 to 18 years. Dashed lines indicate proposed reference limits.

The prevalence of each ECG abnormality in the two age subgroups according to sex is shown in the Supplementary material online, Tables. The analysis showed that LVH, RVH, iRBBB, ER, and mild sinus bradycardia were more common in males than in females in both age groups.

Regression analysis

Univariate and multivariable regression analysis for the association between electrocardiographic findings with a frequency ≥1% and age, sex, training, and endurance sport practice is shown in Table 3.

Table 3.

Univariate and multivariable regression analysis for the association between electrocardiographic findings with a frequency ≥1% and age, sex, training, and endurance sport practice

	Univariate analysis			Multivariable analysis
	OR	95% CI	P-value	OR	95% CI	P-value
LVH
Age	0.99	0.94–1.03	0.52
Sex	3.434	2.64–4.45	<0.001	3.47	2.63–4.47	<0.001
Training	0.98	0.94–1.02	0.29
Endurance	0.692	0.52–0.92	0.01	0.89	0.66–1.21	0.48
RVH
Age	0.97	0.92–1.01	0.16
Sex	5.53	3.91–7.81	<0.001	6.32	4.3–9.3	<0.001
Training	0.89	0.84–0.94	<0.001	0.87	0.82–0.93	<0.001
Endurance	0.77	0.56–1.07	0.12
iRBBB
Age	1.15	1.11–1.19	<0.001	1.11	1.07–1.16	<0.001
Sex	1.24	1.03–1.50	0.02	1.17	0.95–1.44	0.13
Training	1.05	1.02–1.09	<0.001	1.01	0.98–1.05	0.35
Endurance	1.02	0.82–1.26	0.86
ER
Age	1.23	1.18–1.28	<0.001	1.08	1.03–1.14	0.001
Sex	2.419	1.93–3.02	<0.001	2.10	1.63–2.71	<0.001
Training	1.17	1.13–1.21	<0.001	1.14	1.10–1.18	0.03
Endurance	0.705	0.55–0.91	0.008	0.78	0.57–1.05	0.11
TWI V1–V3
Age	0.76	0.71–0.82	<0.001	0.76	0.76–0.70	<0.001
Sex	0.90	0.63–1.28	0.55
Training	0.99	0.93–1.05	0.72
Endurance	1.70	1.18–2.46	0.005	1.22	0.83–1.80	0.31
HR <60
Age	1.22	1.15–1.29	<0.001	1.16	1.08–1.24	<0.001
Sex	3.07	2.21–4.28	<0.001	2.83	1.95–4.10	<0.001
Training	1.11	1.07–1.16	<0.001	1.06	1.01–1.11	0.007
Endurance	0.54	0.37–0.80	0.002	0.77	0.50–1.19	0.26
HR < 55
Age	1.22	1.13–1.32	<0.001	1.17	1.07–1.30	0.001
Sex	2.71	1.70–4.31	<0.001	2.38	1.43–3.95	<0.001
Training	1.12	1.06–1.18	<0.001	1.08	1.02–1.14	0.01
Endurance	0.61	0.32–1.05	0.08
Ectopic AR
Age	0.93	0.83–1.06	0.28
Sex	1.42	0.73–2.75	0.30
Training	0.86	0.73–1.01	0.07
Endurance	1.52	0.77–3.0	0.23
1st AVB
Age	1.48	1.22–1.79	<0.001	1.46	1.16–1.82	<0.001
Sex	1.80	0.71–4.52	0.21
Training	1.13	1.02–1.24	0.017	1.05	0.92–1.18	0.44
Endurance	0.49	0.14–1.67	0.25
RAD
Age	1.32	1.15–1.50	<0.001	1.31	1.05–1.64	0.015
Sex	4.40	1.81–10.7	0.001	3.1	1.01–9.32	0.049
Training	1.17	1.08–1.25	<0.001	3.1	1.11–1.26	0.006
Endurance	0.22	0.05–0.93	0.040	0.39	0.08–1.76	0.22

1st AVB, first-degree atrioventricular block; Ectopic AR, ectopic atrial rhythm; ER, early repolarization; HR, heart rate; iRBBB, incomplete right bundle branch block; LVH, voltage criteria for left ventricular hypertrophy; RAD, right-axis deviation; RVH, voltage criteria for right ventricular hypertrophy; TWI V1–V3, T-wave inversion in leads V1–V3 in children ≤16 years old.

In the multivariable regression analysis, age was independently associated with iRBBB, ER, TWI in V1–V3, SB, first-degree AVB, and RAD. LVH, RVH, ER, and SB were independently more prevalent in males. Voltage criteria for RVH, ER, SB, and RAD showed an independent association with mean hours of training per week. No ECG finding demonstrated an independent association with endurance sport practice.

Figure 4 shows the age-specific prevalence of electrocardiographic findings in the paediatric population. For the variables showing a significant association between age and prevalence, 12 years appears to be a reasonable, pragmatic cut-off to differentiate groups with high vs. low prevalence.

Figure 4.

Age-specific prevalence of electrocardiographic findings in the paediatric population. The graph displays the percentage of individuals with electrocardiographic findings with a frequency ≥1% across ages 8–18 years. LVH, left ventricular hypertrophy; RVH, right ventricular hypertrophy; iRRR, incomplete right bundle branch block; ER, early repolarization; SB, sinus bradycardia with subgroups based on heart rate (SB <60, SB 50–55, SB <55); EAR, ectopic atrial rhythm; IAVB, first-degree atrioventricular block. For variables showing an age-dependent prevalence based on the regression analysis (Table 3), dashed red vertical lines indicate the transition between individuals under and over 12 years of age.

Role of training as a modulating factor depending on age

Interaction testing showed that the only ECG parameter that was significantly dependent on training in adolescents ≥12 years old but not in children <12 years old was SB (Table 4).

Table 4.

Regression analysis for the association between training hours/week and ECG parameters in the two pre-specified age groups and interaction testing

ECG parameter	Age subgroup (years old)	Odds ratio	95% CI	P	P for interaction
LVH	8–11	0.95	0.87–1.03	0.23	0.21
	12–18	1.01	0.96–1.07	0.64
RVH	8–11	0.95	0.87–1.05	0.33	0.12
	12–18	0.86	0.78–0.94	<0.001
iRBBB	8–11	1.03	0.93–1.08	0.96	0.89
	12–18	1.01	0.97–1.05	0.68
ER	8–11	1.11	1.04–1.20	0.003	0.61
	12–18	1.14	1.09–1.18	<0.001
TWI V1–V3	8–11	1.05	0.97–1.15	0.24	0.77
	12–18	1.07	0.99–1.17	0.10
HR <60	8–11	0.89	0.76–1.05	0.17	0.01
	12–18	1.11	1.05–1.16	<0.001
HR <55	8–11	0.96	0.79–1.17	0.71	0.03
	12–18	1.12	1.06–1.19	<0.001
Ectopic AR	8–11	0.81	0.59–1.12	0.20	0.67
	12–18	0.88	0.72–1.09	0.25
1st AVB	8–11	0.84	0.34–2.09	0.70	0.60
	12–18	1.07	0.95–1.20	0.27
RAD	8–11	1.06	0.80–1.39	0.68	0.57
	12–18	1.15	1.05–1.26	0.001

1st AVB, first-degree atrioventricular block; Ectopic AR, ectopic atrial rhythm; ER, early repolarization; HR, heart rate; iRBBB, incomplete right bundle branch block; LVH, voltage criteria for left ventricular hypertrophy; RAD, right-axis deviation; RVH, voltage criteria for right ventricular hypertrophy; TWI V1–V3, T-wave inversion in leads V1–V3 in children ≤16 years old.

Classification of electrocardiographic abnormalities based on prevalence

A classification of ECG abnormalities based on their prevalence and relation to training in our paediatric athletes sample is shown in the Table 5. According to this classification, 987 (40%) had a completely normal ECG, 1147 (46.7%) had only common ECG abnormalities, 208 (8.5%) had one or more uncommon but no rare ECG abnormalities and 116 (4.7%) had one or more rare abnormalities.

Table 5.

Classification of electrocardiographic findings according to prevalence and relation to training

	<12 years old		≥12 years old
	Prevalence	Relation to training	Prevalence	Relation to training
Rhythm and heart rate
Heart rate 55–60 b.p.m.	C	Yes	C	Yes
Heart rate 50–55 b.p.m.	R	No	U	Yes
Heart rate <50 b.p.m.	R		R
Ectopic atrial rhythm	U	No	U	No
Junctional escape rhythm	R		R
Atrial enlargement
Left atrial enlargement	R		R
Right atrial enlargement	R		R
Atrioventricular conduction
1st degree AV block	R	No	U	No
Advanced (>400 ms) 1st degree AV block	R		R
2nd and 3rd degree AV block	R		R
Ventricular pre-excitation	R		R
Depolarization abnormalities
LBBB	R		R
Incomplete RBBB	C	No	C	No
Complete RBBB	R		R
Non-specific IVCD	R		R
Left-axis deviation	R		R
Right-axis deviation	R	Yes	U	Yes
QRS fragmentation	R		R
Pathological Q-wave	R		R
QRS voltages
Low QRS voltage in limb leads	R		R
LVH voltage criteria	C	No	C	No
RVH voltage criteria	C	Yes	C	Yes
Ventricular repolarization
Early repolarization	C	Yes	C	Yes
ST-segment depression	R		R
T-wave inversion in V1–V3	C	No	U	No
T-wave inversion in inferior and/or lateral leads	R		R
Long-QT	R		R
Brugada type 1 pattern	R		R
Arrhythmias
PVB ≥1	R		R
Atrial arrhythmias	R		R

1st degree AV block, first-degree atrioventricular block; 2nd degree AV block, second-degree atrioventricular block; 3rd degree AV block, third-degree atrioventricular block; C, common (>5%); IVCD, intraventricular conduction delay; LBBB, left bundle branch block; LVH, left ventricular hypertrophy; PVB, premature ventricular beats; R, rare (<1%); RBBB, right bundle branch block; RVH, right ventricular hypertrophy; U, uncommon (1–5%).

Further investigations, diagnoses, and clinical management

Further investigations were prescribed for abnormalities at PPS in 153 individuals (6.2%) (51/1384, 3.6% in those <12 years old and 102/1074, 9.5% in those ≥12 years old) and consisted in echocardiography (N = 138), 24-h ambulatory ECG monitoring (N = 84), maximal exercise testing (N = 75), cardiac magnetic resonance (N = 5), coronary computed tomography (N = 2), and electrophysiological study (N = 3). In addition, all 225 Hungarian athletes underwent echocardiography as part of PPS. Reasons for further investigations are detailed in the Table 6.

Table 6.

Reasons for prescription of additional investigations and diagnostic findings

Abnormalities	Number with further testing	Number with diagnosis
Isolated abnormalities
Family history	N = 4
Personal history	N = 17
Physical examination	N = 7	N = 2
ECG	N = 83	N = 3
Screening exercise testing (Italian athletes)	N = 28	N = 1
Screening echo (Hungarian athletes)	N = 1
Multiple abnormalities
Family history + ECG	N = 2	N = 1
Personal history + Exercise testing	N = 3
Physical examination + Exercise testing	N = 2
Physical examination + Echo	N = 1
ECG + Exercise testing	N = 5	N = 3

At the end of the diagnostic work-up, a potentially life-threatening cardiovascular disease was identified in 10 (0.4%) individuals (9 from Italy and 1 from Hungary), of those 8 belonged to the ≥12-year-old group. Specifically, one isolated non-ischaemic left ventricular scar (with RAD on the ECG); one arrhythmogenic right ventricular cardiomyopathy (with TWI in V1–V3 at the age of 14 years on the ECG); one hypertrophic cardiomyopathy (with ST-segment depression and lateral TWI on the ECG); one long-QT syndrome (with prolonged QT on the ECG); one Wolff–Parkinson–White syndrome with low-refractory period of the accessory pathway (with ventricular pre-excitation on the ECG); two bicuspid aortic valve with aortic root dilation (both with normal basal ECG); one anomalous partial venous return (with incomplete RBBB and LAD on the ECG); two arrhythmic mitral valve prolapse (one with inferior TWI and the other with normal ECG). Therefore, the basal ECG had rare or uncommon abnormalities in 7/10 cases; however, three of seven ECGs would be classified as normal according to the 2017 IC (isolated RAD, isolated LAD, TWI in V1–V3 before the age of 16).

The athlete with ventricular pre-excitation and the one with anomalous partial venous return underwent catheter ablation and surgical intervention, respectively, and later resumed competitive sport. One athlete with bicuspid aortic valve plus aortic root dilation and one with arrhythmic mitral valve prolapse were prescribed beta-blockers therapy but were allowed to continue practicing competitive sports because of the low cardiovascular demand of their disciplines under close (every 6 months) follow-up. The other six athletes with at-risk conditions and playing sports at moderate to high cardiovascular demands were not considered eligible to competitions but were offered a tailored non-competitive exercise prescription program.

Discussion

Main findings

The study aimed to analyse the ECG findings of a large cohort of Caucasian children and adolescents engaged in competitive sports who underwent PPS. The population was divided into two subgroups using 12 years as the age cut-off, corresponding to the age at which the 2017 IC for Interpretation of the Athlete’s ECG started to apply.⁴ The main findings of the study were: (i) several electrocardiographic alterations, listed in the normal category by the 2017 IC, were found to be uncommon (prevalence 1–5%) or rare (<1%) in the paediatric population practicing sports; (ii) specifically, classic markers of electrical remodelling in athletes’ hearts, such as SB and first-degree AVB, were uncommon or rare, particularly among children under 12 years of age; (iii) for some abnormalities with intermediate prevalence (1–5%), we identified modulators such as age, sex, and training intensity that can aid in their clinical interpretation; (iv) the PPS led to the diagnosis of conditions associated with a risk of sudden death in 10 athletes, 7 of whom had uncommon or rare ECG findings. However, three out of seven would have been classified as normal under the 2017 IC.

Rationale and limitations of the 2017 International Criteria

In the context of PPS for asymptomatic athletes, two apparently conflicting needs must be addressed: on the one hand, diagnosing cardiac conditions with the highest possible sensitivity; on the other hand, reducing the number of unnecessary and costly second-level investigations. Some of the ECG changes induced by training can be interpreted as indicative of potential structural heart disease and lead to an unacceptable number of false positives PPS results if ECG criteria for the general population are used.^3,4 For this reason, specific criteria for interpreting athletes’ ECGs have been developed and the most recent update, resulting from a consensus conference in Seattle in 2015, was published in 2017.^4,16,17 These criteria categorize ECG abnormalities into three groups. The first group includes so-called ‘normal’ changes, which are common and/or training-related, carry a low probability of disease, and typically do not require further investigations in asymptomatic individuals. The second group comprises ‘borderline’ abnormalities, which are uncommon but carry a low probability of disease when present in isolation. For this category, it has been proposed that additional investigations are needed only when at least two abnormalities are present at the same time. The third group includes rare abnormalities that are not training-related and necessitate further examinations to exclude underlying pathology if any of those are present. Different studies have demonstrated that the use of these criteria is associated with excellent sensitivity and an acceptable false-positive rate.^18–20 However, both the development and validation of these criteria have been predominantly studied in populations outside the paediatric age range. Moreover, in the decade following the formulation of these criteria, new scientific evidence on athletes’ cardiac conditions and their ECG characteristics has accumulated.^{14,15,21–23}

Peculiarities of the paediatric athlete’s ECG

A child or adolescent engaging in competitive sports exhibits several unique characteristics. Firstly, this is a developmental phase where training-induced changes and those due to pubertal maturation interact.²⁴ Secondly, most participate in team sports, while endurance sports are those that induce the most pronounced heart remodelling and subsequent ECG changes.¹³ Several studies published in recent years analysing paediatric athletes’ ECGs have confirmed the limited applicability of the 2017 IC in this population.^{6–9,14,25–29} In particular, classic electrical anomalies associated with training, such as marked SB, first-degree AVB, junctional rhythm, or Mobitz type I second-degree AVB, were found to be rare, especially among younger children.^7,8

Recent large-scale studies confirmed this perspective. Doumparatzi et al.⁶ observed age-related reductions in heart rate in male football players aged 5–18 years, with a modest contribution from training. Similarly, in a cohort of over 6000 male soccer players aged 5–16 years, Diaz-Gonzalez et al.⁹ reported a prevalence of SB (HR <60 b.p.m.) of 5.9%, but no cases of HR <30 b.p.m. and only 0.1% with first-degree AVB. Our results align with these observations but also expand them: we included both sexes and a variety of sports, and found that male sex and weekly training volume were independently associated with bradycardia, although the effect of training was modest. Most importantly, our interaction analysis showed for the first time that the association between training and HR reduction is limited to adolescents (12–18 years), with no effect in younger children. This suggests that training-induced autonomic adaptation becomes physiologically relevant only after a certain threshold of biological maturation.

In our cohort, the PR interval increased significantly with age (from 130 ms in children <12 years to 140 ms in adolescents ≥12 years). First-degree AVB was rare overall (1.0%) and markedly more common in adolescents (2.0%) than in younger children (0.2%). Age was the only independent predictor of first-degree AVB, while weekly training volume lost significance in multivariable analysis. These findings suggest that the progressive lengthening of PR interval in paediatric athletes is primarily driven by physiological maturation, and that training-related conduction changes, when present, remain within normal limits. These results are in line with previous findings by Doumparatzi et al.⁶ who observed that the PR interval was more strongly associated with training age than with chronological age, and by Cavarretta et al.⁸ who reported correlations between PR duration and somatometric measures in adolescent athletes. However, our study adds new evidence by quantifying, for the first time, the age-dependent increase in the prevalence of first-degree AV block in a large, mixed-sex cohort of paediatric athletes, and by demonstrating that this conduction adaptation is primarily associated with chronological age, rather than training exposure.

Finally, since the publication of the 2017 IC, new evidence has emerged regarding ECG changes in recently defined conditions, such as non-ischaemic left ventricular scar (also termed non-dilated left ventricular cardiomyopathy), an emerging substrate for sudden death in athletes.^{14,15,21–23} Some of these anomalies were considered normal (e.g. the presence of a single PVB), borderline (e.g. LAD and RAD) or simply not considered (e.g. low QRS voltages in limb leads or fragmented QRS) by the 2017 IC.⁴ Finally, RBBB, classified as borderline by the IC, can be associated with congenital pathologies, particularly those involving right-sided lesions with volume loading (such as Ebstein’s anomaly and atrial septal defects) that are particularly relevant in the paediatric setting.³⁰

These findings highlight the limitations of the IC for interpreting athletes’ ECGs in paediatric populations and suggest the need for tailored criteria.

Prevalence and determinants of the young athlete’s electrocardiographic abnormalities

Based on the results of this study, we could identify three groups of paediatric athletes’ ECG findings (Graphical abstract).

Common (>5%)

This category includes electrocardiographic alterations frequently observed in both children and adolescents. Examples include mild SB (HR 55–60 b.p.m.), incomplete RBBB, voltage criteria for LVH and RVH and ER. All these abnormalities are also considered normal by the 2017 IC. In our study, 87% of young athletes had either a completely normal ECG or common ECG findings.

Uncommon (1–5%)

This category includes abnormalities with intermediate prevalence that were intermediate in the overall sample (1–5%), often varying depending on modulatory factors such as age, sex, and training intensity. Examples include a heart rate between 50 and 55 b.p.m., first-degree AVB, and RAD, which are more prevalent in older athletes, particularly males and those undergoing more intensive training. This category also includes EAR, which has a well-known benign nature in children.³¹

With a prevalence of 5.4%, negative T waves in V1–V3 are at the border between common and uncommon. However, this anomaly is much more common in paediatric populations (under the age of 12) but markedly decreases in prevalence with increasing age.^24,26,27 Its interpretation in the context of PPS of young athletes is extremely challenging. The original 2010 European Society of Cardiology criteria listed all TWI as abnormal,¹⁶ but further studies showed that this would lead to an unacceptably high number of false positives, particularly in Black athletes, females, and pre-pubertal children.^17,24,32,33 For this reason, the IC listed as normal the pattern of TWI in V1–V4 preceded by ST-segment elevation in Black athletes, because this was considered an ethnic ER variant. In the other cases, the IC set an age threshold of 16 years for differentiating normal vs. abnormal TWI in V1–V3. However, most Caucasian children exhibit a normal repolarization pattern before that age. In particular, a previous longitudinal study showed that in young Italian competitive athletes, positivization of TWI in V1–V3 occurred at a mean age of 13.0 ± 2.0 years.²⁷ On the other hand, arrhythmogenic cardiomyopathy, which typically begins to develop during puberty, may account for anterior repolarization abnormalities: dismissing all TWI in V1–V3 as normal before the age of 16 years may preclude to make a diagnosis of one of the most dangerous diseases in terms of sudden death risk.³⁴ For this reason, TWI in V1–V3 in young Caucasian athletes 12–16 years old but who have already completed their pubertal development may require further investigations, particularly in males and with a previous ECG showing a normal repolarization pattern, a finding that is not compatible with a ‘persistence of the juvenile pattern’.

Rare (<1%)

This group encompasses all other ECG abnormalities, including those originally classified as abnormal by the International and abnormalities that gained attention recently (e.g. PVB ≥1, low QRS voltage in limb leads, fragmented QRS) based on emerging knowledge.^14,35,36 Notably, despite the expanded list of ECG abnormalities falling in the rare category, fewer than 5% of our study population exhibited one or more of these abnormalities. Among the 10 individuals identified with a disease through screening, three would have been classified as normal according to the 2017 Criteria but showed rare findings.

Study limitations

This study has several limitations that must be acknowledged. Firstly, it focused solely on Caucasian athletes, while ethnic background is known to be a significant determinant of athletes’ ECGs. Similarly, although a high percentage of participants practiced at least 6 h of sports per week, only a minority could be considered ‘elite junior athletes’ (>10 h per week of training) and this limitation may have hindered the evaluation of the effect of intense cardiac remodelling on the ECG. The lack of data on pubertal development prevented us from confirming findings from our previous study on the independent role of sexual maturation in normalizing anterior repolarization. Interpretation of ECG in athletes <12 years was not standardized due to the lack of specific guidelines, and this may have introduced variations on the prescription of further investigations depending on the screening physician. Finally, the absence of second-level assessments, such as echocardiography, for all participants aligns with the observational nature of this study, reflecting current clinical practice, but does not allow us to establish the diagnostic accuracy of the proposed new criteria.

Conclusions

The results of our study, along with previous research, support the need for specific criteria for the interpretation of the paediatric athletes’ ECGs, particularly in children under 12 years old. Indeed, several abnormalities that are relatively common in adult athletes were found to be rare and potentially pathological in children. For certain ECG findings, such as negative T waves in V1–V3, evaluation of variables such as age, sex, training status, and degree of pubertal development may be helpful for selecting young athletes requiring second-level evaluations.

Supplementary material

Supplementary material is available at Europace online.

 .

Funding

TKP2021-NKTA-46 has been implemented with the support provided by the Ministry of Innovation and Technology of Hungary from the National Research, Development and Innovation Fund, financed under the TKP2021-NKTA funding scheme.

Data availability

The data underlying this article will be shared on reasonable request to the corresponding author.

Generative AI statement

ChatGPT ver 4.0 was used to improve the quality of the English language of the introduction and discussion sections only after a complete draft of the manuscript including all the intellectual content was written by the authors.