PURPOSE: Female athletes are known to have lower rates of sudden cardiac arrest/death (SCA/D) than male athletes, although overall incidence is low. Athlete electrocardiogram (ECG) criteria have been developed and refined over the past 15 years to more accurately distinguish between physiology and pathology, including as part of pre-participation screening. The International Criteria for athlete ECG interpretation (2017) has reduced the rate of false positives, particularly in males. This study describes sex differences in elite athlete cardiac screening ECGs using study databases collated from various sports in Australia and New Zealand. It involves collaboration between sports medicine, epidemiology and sports cardiology.

METHODS: Records of elite athletes who underwent cardiac screening (including resting 12 lead ECG) were collated from sports in Australia and New Zealand as part of retrospective cohort studies (n=2905 ECGs, 49% female, mean age 21 years). Sports included: cricket, football (soccer), netball and Olympic sports (summer and winter). All ECGs were reviewed according to the International Criteria and comparisons made between sexes.

RESULTS: Several athlete’s heart features (considered normal in athletes) were significantly less common in female cricketers. These include: left ventricular hypertrophy (by voltage criteria) (1.6% 25.0% vs; p<0.0001); sinus bradycardia (27.0% vs 37.4%; p=0.002) and left axis deviation (0.3% vs 2.7%; p=0.01). However, T-wave inversion (TWI) (excluding leads aVR, III and V1) was more common in females than males (8.0% vs 1.2%; p<0.0001). Anterior TWI beyond V2 was also more common in females compared to males (1.3% vs 0.2%; p = 0.049). Lateral TWI was uncommon in both groups (0.0% in females vs 0.4% in males; p = 0.27). 

In the New Zealand dataset, 3.5% of ECGs were abnormal, with ECGs of female athletes more commonly abnormal (4.4% vs 2.5%, p=0.02). ECGs of female athletes also had a higher proportion of ECGs with abnormal TWI (3.1% vs 0.9%, p<0.001) compared to males. Of the abnormal TWI in females (all aged ≥16 years), 47% was limited to V1-V3 with no other abnormalities.

CONCLUSION: Analysis of Australian and New Zealand cardiac screening datasets show some sex differences in ECG features in line with previous studies. While the proportion of abnormal ECGs was low, it was higher in females. The major cause was anterior TWI in V1-V3 which was not associated with diagnoses of conditions associated with SCA/D. This issue has been flagged as a consideration for future updates to athlete ECG criteria. The most recent criteria reduced the rate of false positive ECGs in male athletes, but further improvements may be possible for female athletes. Future work should ensure females are sufficiently represented in datasets to better understand features of the female athlete’s heart.

DISCLOSURES: JJO is funded by a National Health and Medical Research Council Investigator Grant No 2016730