Physical exercise is beneficial, but if intensive, it can cause potentially deleterious cardiac remodeling.1 The characteristics that define “athlete's heart”2,3 include dilatation of the chambers of the heart.4 Dilatation of the left atrium (LA) in competitive athletes can be considered an adaptive mechanism, which is not secondary to an increase in left ventricular filling pressures or mitral valve disease.5 However, there is a clear association between LA dilatation and an increased rate of cardiovascular events, even in individuals with no supraventricular tachyarrhythmias or valve disease.6,7 Recent studies have analyzed the risks associated with LA dilatation in athletes, principally due to the potential predisposition to atrial fibrillation.8

Echocardiography is the most frequently used method for assessing LA volume,9 and the LA volume indexed to the body surface area (LAVi) has been shown to have prognostic value.10 Previous studies have indicated a poor correlation between electrocardiographic P wave criteria and LA size in the general population.11 The association between LA dilatation and atrial physiology characteristics has not been analyzed in competitive athletes, a group in which the upper limit for normal LA size is unclear.4,5 Our objective was to analyze LA dilatation, P-wave duration, and interatrial block (IAB) in competitive athletes.

ALMUDAINA (In Spanish, Análisis y Lectura de Mediciones Uniformes de Dilatación Auricular Izquierda Notificadas en Atletas; in English, Analysis and Reading of Uniform Measurements of Left Atrial Dilatation Reported in Athletes) is a national retrospective observational study with 9 participating Spanish hospitals. The study period was from January to September 2020. Individuals who met the study participation criteria were those aged ≥ 16 years, who had a digitalized electrocardiogram (ECG) and a transthoracic echocardiogram done on the same day. The exclusion criteria were the following: a) ischemic heart disease, moderate or severe valve disease, left ventricular ejection fraction <50% or known primary cardiomyopathy (hypertrophic, dilated, arrhythmogenic, or noncompaction); b) significant arrhythmias; c) implanted cardiac device; d) no measurable P waves on surface ECG; e) poor transthoracic window; f) chest-wall abnormality or deformity; and g) chronic obstructive pulmonary disease.

Patients were selected if they met the consensus definition of competitive athlete (participation in an individual or team sport that involves regular competition against other athletes, places great importance on excellence and success, and demands regular and generally intensive training)12,13 in a skilled, strength, resistance, or mixed activity, at a national or international level.

The selected controls were individuals not participating in competitive sports (recreational sport was allowed) and not taking part in regular training programs.

A resting 12-lead surface ECG was obtained for each participant in line with established standards (25mm/s and 10mm/50mV), with amplitude filtering of 0.05-150Hz and a 50-Hz filter. The following P-wave parameters were analyzed: a) duration, b) voltage in lead I, and c) MVP (morphology-voltage-P-wave duration) score.14 We analyzed the presence of IAB, whether partial (defined as a P wave duration ≥ 120ms, with positive polarity in the inferior leads) or advanced (P wave duration ≥ 120ms with a biphasic morphology [±] in the inferior leads).15 The MVP score is a risk scale that can help predict de novo atrial fibrillation and is calculated by assigning points according to P-wave morphology in the inferior leads (0-2), its voltage in lead I (0-2), and its duration (0-2).14 The P-wave duration is measured manually in a central laboratory using GeoGebra 4.2 software after amplifying the original size of the ECG by a factor of 20. All measurements were performed by an observer who was blinded to the patients’ clinical and electrocardiographic data and, if there was any uncertainty, a second observer was consulted.

Two-dimensional transthoracic Doppler echocardiograms were analyzed by experienced echocardiogram specialists and acquired in line with the guidelines of the American Society of Echocardiography/European Association of Cardiovascular Imaging for cardiac chamber quantification.9 LA size was quantified at end-systole with the following measurements averaged after 3 consecutive cycles: a) anteroposterior LA diameter (parasternal long-axis view), b) LA area (apical 4-chamber view), and c) LA volume (on apical 2-chamber and 4-chamber views, using the modified biplane method of disc summation). The LA volume was indexed to body surface area to obtain the LAVi. An LAVi ≥ 34mL/m2 was used to define LA dilatation.9,16 If the LAVi could not be calculated, LA area (≥ 20cm2) was used as the criterion to define LA dilatation.9 Right atrial dilatation was defined using area (> 18cm2) or indexed volume (25mL/m2).

Baseline clinical characteristics were recorded for all participants: age weight, height, cardiovascular risk factors, comorbidities, family history of sudden cardiac death, and pharmacological treatment. Blood pressure was measured with automated devices after a 5-minute rest period.

The study complied with the Declaration of Helsinki and was approved by the local ethics committee of the coordinating hospital. Being a strictly observational, descriptive, retrospective study, with no participants subject to an intervention, the ethics committee exempted the requirement for an informed consent form.

The final population included 356 Caucasian participants: 308 cases and 48 controls. Table 1 summarizes the participants’ baseline characteristics. Compared with controls, the athlete group was younger (36.4±11.6 vs 49.3±16.1 years) and had a lower prevalence of cardiovascular risk factors, lower systolic and diastolic blood pressure, and lower heart rate. In the competitive athlete group, the most common type of activity was resistance sports (n=221, 71.8%), followed by strength (n=35 [11.4%]) and a combination of both (n=52 [16.9%]).

Table 2 summarizes the echocardiographic characteristics of the two groups. Compared with controls, the athletes had a higher LAVi (29.8±8.6 vs 25.6±8.0mL/m2) and higher prevalence of LA dilatation (36.7% vs 10.4%). Interventricular septum thickness, left ventricular volumes, and tricuspid annular plane systolic excursion were also higher in athletes than controls.

Table 3 shows data on ECG evaluation. There were no significant differences between athletes and controls in the electrocardiographic indices: P-wave duration (106.3±12.5 vs 108.2±7.7ms), IAB prevalence (40 [13.0%] vs 4 [8.3%]) and MVP score (1.8±0.84 vs 1.5±0.8). Only 1 case of advanced IAB was detected in an athlete, who was 37 years old, had no comorbidities, trained for 16hours per week, and did not have LA dilatation.

Table 4 shows the comparison of participants with and without LA dilatation. Those with LA dilatation had higher mean age, P-wave duration, IAB frequency, and MVP score (1.9±1.0 vs 1.7±0.7; P=.01). As only 5 controls had LA dilatation, the differences in the electrocardiographic parameters of atrial activation were mainly due to the athletes (table 5).

There was no significant correlation between P-wave duration and LAVi, either in the whole sample (linear correlation coefficient r=0.01; P=.901) or in the athlete group (r=0.004; P=.966).

Table 6 shows the multivariate analysis for predictors of LA dilatation. Competitive training was independently associated with LA dilatation (odds ratio [OR], 14.7; 95% confidence interval [95%CI], 4.7-44.0). We also forced inclusion in the model of cardiovascular risk factors, ventricular diameters, PR interval duration, and P-wave duration, which produced no significant changes in the association between participation in competitive sports and LA dilatation.

However, competitive training was not associated with P-wave duration (OR,1.02; 95%CI, 0.99-1.04; P=.19), MVP score (OR,1.4; 95%CI, 0.9-2.2; P=.14), or the presence of IAB (OR,1.4; 95%CI, 0.7-3.1; P=.34). The only variables that were independently associated with IAB were age in years (OR,1.05; 95%CI, 1.01-1.09; P=.015) and left ventricular mass in grams (OR,1.013; 95%CI, 1.004-1.023; P=.06).

Table 7 shows the main differences in the clinical, echocardiographic, and electrocardiographic characteristics of the athletes according to the presence of IAB and LA dilatation. Compared with the other 3 groups, athletes with IAB and LA dilatation were older and more frequently female, and had a greater left ventricular mass.

Table 8 shows the association of LA dilatation with competitive sport using propensity score adjustment.

Our study shows that, despite being younger and healthier than controls, athletes had a far higher prevalence of LA dilatation, although this finding correlated poorly with electrocardiographic changes.

Left atrial dilatation and remodeling in athletes are practically defining characteristics of athlete's heart.3,4 However, the threshold for defining normal LA size in athletes is a matter of debate.5,17 Two meta-analyses have demonstrated that LA volume is significantly larger in athletes than in controls. However, the studies analyzed yielded heterogeneous results on the prevalence of LA dilatation, presumably due to differences in the sporting discipline (strength or resistance training), level of training, and age.18,19 Rates of LA dilatation of up to 42% to 8% have been described in elite rowers20 and professional basketball players,21,22 although severe dilatation is an uncommon finding. Our data confirm both the association between competitive sport and LA dilatation and that severe LA dilatation is rare (only 1.6% of our athletes).

The main contribution of the ALMUDAINA study is that it demonstrates that competitive training in athletes is not associated with longer P-wave duration or presence of IAB. Despite a high prevalence of LA dilatation, we found no significant changes in the atrial electrocardiographic parameters. A recent study in young athletes showed that less than 1% had a P-wave duration ≥ 120ms and a negative component ≥ 1mm deep in lead V1.23 Even in the general population, the electrocardiographic criteria for LA dilatation have shown low sensitivity.11,24,25 Furthermore, changes in P-wave duration and morphology may be due to an IAB in normal-sized atria. This has been demonstrated experimentally26 and in situations such as the presence of an atrial tumor, which can cause IAB without LA dilatation.27 Our study also supports the argument that LA dilatation and IAB are 2 different entities, as most of our athletes with LA dilatation did not have IAB.

However, the consensus documents of the American Heart Association regarding athletes recommend that any change in the P-wave, whether due to LA dilatation or IAB, should be considered a “P-wave abnormality”28,29; we think that this may actually increase confusion regarding LA assessment. Furthermore, the ECG index with the highest diagnostic sensitivity for LA dilatation (Morris index) has a low reproducibility,30 and P wave negativity increases when lead V1 is positioned in the second intercostal space.31 In addition, this pattern in V1 could simply represent an interatrial conduction defect.25

Nonetheless, compared with athletes without LA dilatation, those with LA dilatation had a longer P-wave duration (110 vs 104ms) and a higher prevalence of IAB (20% vs 9%). Therefore, we could hypothesize that it is likely that there is also a certain degree of electrical remodeling in athletes with LA dilatation. However, the fact that none of the athletes with LA dilatation had advanced IAB also indicates a less pathological LA dilatation in association with sports training.

The 2017 international recommendations for ECG interpretation in athletes emphasized that LA dilatation on ECG remains a borderline variant not requiring further cardiological assessment, unless accompanied by other abnormalities.29 This recommendation is supported by the findings of our study (competitive training was not associated with P-wave ECG indices).

Regarding the association of other variables with LA dilatation, our data are in agreement with previous studies reporting an association between age32,33 and female sex and LA dilatation.32,34

Left atrial dilatation is common in athletes, but is not accompanied by significant changes in electrocardiographic parameters. Our data support the hypothesis that LA dilatation and IAB are 2 different entities. These findings may not be extrapolable to younger athletes, in whom prospective targeted studies of LA dilatation and electrocardiographic parameters should be performed.

None.

Definition of study objectives and design (M. Martínez-Sellés, C. Herrera), data collection (all authors), data analysis, interpretation of results, and writing first draft of article (M. Martínez-Sellés, C. Herrera). All authors have reviewed and approved the final version of the manuscript.

The authors have no conflicts of interest. The authors would like to thank José María Bellón Cano, research statistician at the Instituto de Investigación Sanitaria Gregorio Marañón for performing the propensity score adjustment.

.