Skip to main content
NCBI home page
Annals of Noninvasive Electrocardiology logoLink to Annals of Noninvasive Electrocardiology
. 2018 Nov 7;24(3):e12614. doi: 10.1111/anec.12614

Vectorcardiogram in athletes: The Sun Valley Ski Study

Jason A Thomas 1,2,[Link], Erick A Perez‐Alday 1,[Link], Allison Junell 3, Kelley Newton 3, Christopher Hamilton 1, Yin Li‐Pershing 1, David German 1, Aron Bender 1, Larisa G Tereshchenko 1,4,✉
PMCID: PMC6476648  NIHMSID: NIHMS993451  PMID: 30403442

Abstract

Background

Global electrical heterogeneity (GEH) is associated with sudden cardiac death (SCD) in adults of 45 years and above. However, GEH has not been previously measured in young athletes. The goal of this study was to establish a reference for vectorcardiograpic (VCG) metrics in male and female athletes.

Methods

Skiers (n = 140; mean age 19.2 ± 3.5 years; 66% male, 94% white; 53% professional athletes) were enrolled in a prospective cohort. Resting 12‐lead ECGs were interpreted per the International ECG criteria. Associations of age, sex, and athletic performance with GEH were studied.

Results

In age and training level‐adjusted analyses, male sex was associated with a larger T vector [T peak magnitude +186 (95% CI 106–266) µV] and a wider spatial QRS‐T angle [+28.2 (17.3–39.2)°] as compared to women. Spatial QRS‐T angle in the ECG left ventricular hypertrophy (LVH) voltage group (n = 21; 15%) and normal ECG group did not differ (67.7 ± 25.0 vs. 66.8 ± 28.2; p = 0.914), suggesting that ECG LVH voltage in athletes reflects physiological remodeling. In contrast, skiers with right ventricular hypertrophy (RVH) voltage (n = 26, 18.6%) had wider QRS‐T angle (92.7 ± 29.6 vs. 66.8 ± 28.2°; p = 0.001), larger SAI QRST (194.9 ± 30.2 vs. 157.8 ± 42.6 mV × ms; p < 0.0001), but similar peak SVG vector magnitude (1976 ± 548 vs. 1939 ± 395 µV; p = 0.775) as compared to the normal ECG group. Better athletic performance was associated with the narrower QRS‐T angle. Each 10% worsening in an athlete's Federation Internationale de’ Ski downhill ranking percentile was associated with an increase in spatial QRS‐T angle by 2.1 (95% CI 0.3–3.9) degrees (p = 0.013).

Conclusion

Vectorcardiograpic adds nuances to ECG phenomena in athletes.

Keywords: athletes, electrocardiogram, global electrical heterogeneity, vectorcardiogram

1. INTRODUCTION

Sudden cardiac death (SCD) in young athletes is a major cause of death during sport and exercise (Harmon et al., 2015; Maron, Doerer, Haas, Tierney, & Mueller, 2009), and a tragic event that devastates family, community, and society at large. Significant progress has been made in recent years toward understanding causes and mechanisms of SCD as well as in establishing a new paradigm—one that recognizes the value and limitations of the electrocardiogram (ECG)—for cardiovascular screening in athletes (Drezner et al., 2016). International criteria for ECG interpretation in athletes (Sharma et al., 2018) were endorsed by 16 professional medical and sport societies. International consensus has established standards of ECG interpretation in athletes and provided guidance on further testing and clinical evaluation of athletes (Sharma et al., 2018). Recent application of these international ECG criteria to differing athlete populations has yielded findings of abnormal ECG as low as 3% (Dhutia et al., 2017) and up to 15.6% (Waase et al., 2018) of those tested, a discrepancy that highlights knowledge gaps in understanding of the continuum of sex‐, race‐, and age‐specific characteristics of electrophysiological (EP) substrate in athletes (Heidbuchel, 2017). The vectorcardiogram (VCG) provides additional information to routine 12‐lead ECG evaluation (Waks et al., 2016). Tereshchenko and colleagues recently demonstrated that global electrical heterogeneity (GEH), measured by five VCG metrics [spatial QRS‐T angle, spatial ventricular gradient (SVG) magnitude, azimuth, elevation, and sum absolute QRST integral (SAI QRST)] was associated with SCD in a community cohort study of more than 20,000 adults of age 45 and above (Waks et al., 2016). However, detailed VCG analysis has not been performed in athletes. To address this knowledge gap, we initiated a prospective cohort study of athletes, the Sun Valley Ski Study. We aimed to describe VCG characteristics in athletes and to evaluate associations of VCG parameters with age, sex, athletic training level and performance, and traditional ECG findings (Sharma et al., 2018).

2. METHODS

2.1. Study population

A prospective cohort study of athletes was initiated at Oregon Health & Science University (OHSU). The study is approved by the OHSU Institutional Review Board. All athletes and their parents/guardians signed informed consent and/or assent, as appropriate, before entering the study. The study was conducted at the Sun Valley (Idaho) official Olympic & Paralympic Training Site for Alpine, Freestyle, Nordic, FreeSkiing, and Snowboarding.

The Sun Valley Ski study was designed as a prospective cohort study. Athletes as defined by (Drezner et al., 2017) 14–30 years of age were invited to participate. Each athlete's medical and family history as well as a detailed athletic performance profile was collected at enrollment via questionnaire. Past medical and family history was obtained via the first ten questions of the American Heart Association 14‐element screening (Maron et al., 2014). Athletic history captured participation in any of the 59 sports classified previously for cardiovascular demand (Mitchell et al., 2005) and further captured the following training levels: recreational, high school, college division I‐III, and professional. A single supine, resting 12‐lead ECG was recorded at baseline at the Sun Valley Site.

To characterize VCG presentation of ventricular remodeling in athletes, we utilized the International criteria for ECG interpretation in athletes (Drezner et al., 2017) to quantitatively compare VCG in three study groups: athletes with ideally normal traditional ECG (free from any traditional ECG findings), athletes with right ventricular hypertrophy (RVH) voltage criteria as the only ECG finding, and athletes with left ventricular hypertrophy (LVH) voltage criteria as the only ECG finding. Athletes with abnormal or borderline ECG findings, athletes with several normal ECG findings (those who met both LVH and RVH voltage criteria, or those who met LVH or RVH voltage criteria and early repolarization (ER) ECG pattern), and athletes with family history of SCD comprised a special, 4th study group, which due to its heterogeneity underwent only qualitative descriptive analysis.

Some athletes in this study were competitors in the US National Alpine Championships, which represents the country's elite national annual competition for alpine skiing. Federation Internationale de’ Ski (FIS, the skiing and snowboarding international governing body) top percentile rankings were calculated for each of the four disciplines (slalom, giant slalom, super giant slalom, and downhill) analyzed from individual athlete rankings at the end of the 2015–2016 season ("13th FIS points list 2015/2016," 2016) for all officially ranked athletes enrolled in the study. Top percentile was calculated by taking each athlete's international ranking (position) by discipline and dividing it by the maximum ranking for that discipline by sex. Therefore, a top percentile of 1.0 signifies an athlete performed better than 99% of all officially ranked sex and discipline‐matched competitors over the course of the 2015–2016 season.

A comparison was performed with age‐, sex‐, and race‐matched nonathlete controls from two separate studies. Healthy participants of the OHSU ECG patch study (Kabir, Perez‐Alday, Thomas, Sedaghat, & Tereshchenko, 2017) who were free from cardiovascular disease and its risk factors were included in this analysis as one source of controls. For the second source of controls, we included digital ECGs of the Intercity Digital Electrocardiogram Alliance (IDEAL) study, which was previously described (Couderc, Xiaojuan, Zareba, & Moss, 2005; Sur, Han, & Tereshchenko, 2013).

2.2. ECG recording and analysis

Athletes underwent a 10‐s resting, supine, 12‐lead ECG recorded using a MAC 5,500 HD ECG machine (General Electric Healthcare, Milwaukee, WI, USA). Athlete ECG recordings were performed during the winter ski season at an elevation of 1,800 m. In healthy nonathlete controls, modified (5th intercostal space) Frank XYZ ECGs were recorded at rest, as previously described (Kabir et al., 2017; Sur et al., 2013). Digital ECG files (sampling rate 500 Hz; amplitude resolution 1 µV) were exported for VCG analysis. The routine clinical 12SL algorithm (General Electric Marquette, Milwaukee, WI) was used for measurement of traditional ECG parameters.

Athlete ECGs were interpreted in accordance with the International ECG criteria (Sharma et al., 2018) by investigators (AJ, KN, DG, AB, LGT), blinded to automated VCG analysis. At least two investigators adjudicated each 12‐lead ECG. In case of disagreement, the 3rd investigator (LGT) provided final adjudication. The κ‐statistic was used to evaluate interobserver agreement between 12‐lead ECG adjudicators. ECG LVH voltage criteria (Sharma et al., 2018) were used as recommended by the consensus document, based on QRS complex amplitudes (SV1 + RV5 or RV6 > 3.5 mV). ECG RVH voltage criteria were defined (Sharma et al., 2018) as RV1 + SV5 or SV6 > 1.1 mV.

Automated VCG analysis of all ECGs was performed by investigators (EAPA, ALP) blinded to the clinical interpretation of ECGs. Custom MATLAB (MathWorks, Inc, Natick, MA) software used for VCG analysis was developed in the Tereshchenko Laboratory. A detailed description of the measurements is provided in Supplemental Materials. The software code for VCG analysis is provided at https://physionet.org/physiotools/global electrical heterogeneity/. The 12‐lead ECG was transformed into orthogonal XYZ leads using Kors transformation (Kors, van Sittig, & van Bemmel, 1990). The median sinus beat was constructed. Sinus beats before and after premature ventricular and atrial complexes, as well as artifact‐distorted beats, were excluded from analysis. The electrical origin point of VCG was defined as the time interval when the electrical heart vector does not move in three‐dimensional space (i.e., is electrically silent). Vector magnitude was calculated, and fiducial points were automatically detected. Accuracy of the origin point and fiducial point detection was evaluated using visual aid (EAPA, ALP, CH, LGT); correction was performed, if needed. Spatial peak and area QRS, T, and SVG vectors were defined (Figure 1). Direction (azimuth and elevation) and magnitude of each vector were measured. Wilson's ventricular gradient was calculated (Waks et al., 2016). Scalar value of SVG was measured by two approaches. First, as previously described (Sur et al., 2013; Tereshchenko et al., 2011), SAI QRST was calculated as the arithmetic sum of areas under the QRST curve on XYZ leads, with baseline defined as the voltage at the end of the T wave. In addition, scalar value of SVG was calculated as a QT integral on vector magnitude signal (iVMQT), as an area under the vector magnitude signal curve from the QRS onset to T offset.

Figure 1.

Figure 1

Open in a new tab

Methods of vectorcardiographic analysis: (a) Measurement of vector magnitude QT integral. (b) Measurement of Sum absolute QRST integral. (c) Vectorcardiographic loops and QRS, T, and spatial ventricular gradient vectors. (d) Measurement of azimuth and elevation of the vectors

2.3. Statistical analyses

After confirmation of normality, normally distributed continuous variables were presented as means ± standard deviation (SD) and were compared using t test. Fisher's exact test was used to compare categorical variables. VCG parameters were compared by age category (14 to <16 years, 16 to <18 years, 18 to <22 years, and 22–30 years) for consistency with known effects of age on ECG morphology (Rijnbeek, Witsenburg, Schrama, Hess, & Kors, 2001). ANOVA was used for unadjusted comparison of participants across age categories. ANOVA analyses were stratified by sex to study the effect of sex on association of age with VCG characteristics. Athletes without variant ECG findings were compared with athletes with normal findings of ECG RVH and ECG LVH by voltage criteria.

Effect of age, sex, and athletic training level on VCG parameters was studied in a subgroup of athletes with normal ECG. Multiple linear regression models for each continuous VCG variable outcome were constructed. Age, sex, and athletic training category (recreational‐high school‐collegiate‐professional) were included as predictors. Association of VCG parameters with athletic performance in all four alpine ski disciplines was tested in age‐ and sex‐adjusted linear regression analyses.

Circular statistics were used to compare circular variables (azimuth and elevation of QRS, T, and SVG vectors) across study groups. To describe direction of the vector, mean circular direction and 95% confidence interval (CI) are reported. Nonuniformity of the circular variable distribution was confirmed by the Rayleigh test and the Kuiper test for all studied circular variables. Circular variables in male versus female athletes were compared using the Watson U‐square statistic and the Kuiper k* statistics. Comparison of circular variables across age subgroups was performed using the Wheeler‐Watson‐Mardia statistic.

The agreement between SAI QRST and iVMQT was assessed via Bland–Altman analysis (Bland & Altman, 1986). The degree of agreement was expressed as the bias (the mean difference) with 95% limits of agreement (mean ± 2 SD). Pearson's correlation coefficient r, and Lin's concordance correlation coefficient ρc (rho_c) were calculated. Due to expected differences in absolute values of SAI QRST and iVMQT, we used detrended log‐transformed SAI QRST and iVMQT.

Statistical analysis was performed using STATA MP 15.1 (StataCorp LP, College Station, TX). To account for multiple comparisons, Bonferroni‐adjusted p‐value <0.002 was considered statistically significant. Circular statistics was performed using Oriana–Circular Statistics version 4 (Kovach Computing Services, Pentraeth, Wales, UK).

3. RESULTS

3.1. Study population

The study flowchart is shown in Figure 2. The study population (Table 1) included 140 skiers (mean age 19.2 ± 3.5 years; 66% male, 94% white) and 43 healthy nonathlete controls (mean age 21.7 ± 7.8 years; 61% male, 88% white). According to the administered questionnaire, 25% of athletes were also involved in a variety of sports other than skiing (7% American football; 5% soccer, 2% baseball, basketball, volleyball, and snowboarding; ~1% luge, boxing, kayaking, cycling, motorcycling, lacrosse, badminton, ice hockey, running, water skiing, race walking, weight lifting, and golf). More than a half of the athletes were professional or collegiate division skiers (n = 74; 53%), while 53 skiers (38%) were high school athletes. A majority (n = 82; 37% female) were officially ranked in at least one alpine skiing discipline for the 2015–2016 season, 44 of which competed in the US National Alpine Championship. Median top percentile rankings and interquartile ranges by sex for downhill, slalom, giant slalom, and super giant slalom, respectively, for officially ranked athletes were as follows: [female: 39% (23–68), 26% (15–45), 26% (15–43), 27% (14–55)] and [male: 48% (26–72), 18% (9–46), 19% (11–36), 27% (14–44)]. Seven skiers had a family history of SCD due to a heart disease in relative(s) of age <50 years.

Figure 2.

Figure 2

Open in a new tab

The study flowchart

Table 1.

Characteristics of the study participants

Characteristics All athletes (n = 140) Male [n = 93, (66%)] Female [n = 47, (34%)] p value
Age (SD), year 19.2 (3.5) 19.4 (3.5) 18.8 (3.70) 0.327
Age 13–16y, n (%) 25 (17.9) 13 (14.0) 12 (25.5) 0.420
Age 16–18y, n (%) 39 (27.9) 27 (29.0) 12 (25.5)
Age 18–22y, n (%) 44 (31.4) 30 (32.3) 14 (29.8)
Age 22–30y, n (%) 32 (22.9) 23 (24.7) 9 (19.2)
White, n (%) 132 (94.3) 88 (94.6) 44 (93.6) 1.00
Exertional chest pain/discomfort 10 (7.4) 6 (6.7) 4 (8.7) 0.733
Exertional syncope/near‐syncope 3 (2.3) 2 (2.3) 1 (2.3) 1.00
Excessive exertional and unexplained fatigue/fatigue with exercise 6 (4.4) 4 (4.4) 2 (4.4) 1.00
Prior recognition of a heart murmur 10 (7.4) 6 (6.7) 4 (8.7) 0.733
Elevated systemic blood pressure 1 (0.7) 1 (1.1) 0 1.00
Prior restriction from participation in sports 5 (3.7) 1 (1.1) 4 (8.7) 0.046
Prior testing for the heart ordered by a physician 7 (5.2) 7 (7.8) 0 0.095
Palpitations, n (%) 7 (5.2) 5 (5.6) 2 (4.4) 1.00
SCD before age 50 years due to heart disease, in one or more relatives, n (%) 7 (5.2) 5 (5.6) 2 (4.4) 1.00
Disability from heart disease in a close relative <50 years, n (%) 4 (3.0) 4 (4.4) 0 0.301
Specific knowledge of certain cardiac conditions in family members, n (%) 3 (2.2) 3 (3.3) 0 0.551
Recreational level of athletic activity, n (%) 13 (9.7) 10 (11.2) 3 (6.7) 0.830
High school athlete, n (%) 50 (37.3) 34 (38.2) 16 (35.6)
Collegiate division I‐III, n (%) 24 (17.9) 15 (16.9) 9 (20.0)
Professional, n (%) 47 (35.1) 30 (33.7) 17 (37.8)
Open in a new tab

SD, standard deviation.

3.2. ECG findings in athletes using international consensus standards for ECG interpretation

There was high (kappa statistic 0.96) agreement between investigators in adjudication of 12‐lead ECGs. No athlete had abnormal ECG findings, as defined by the international consensus standards for ECG interpretation in athletes (Drezner et al., 2017). One athlete had a borderline ECG finding: right axis deviation (>120°) with QRS axis of +154°. We observed several normal ECG findings in athletes (Drezner et al., 2017): isolated voltage criteria for ECG LVH (n = 21; 15%) and ECG RVH (n = 26; 19%), ER, and incomplete right bundle branch block or rSr’ pattern in V1.

3.3. Comparison of athletes and nonathlete controls

Nonathlete controls had slightly narrower QRS and slightly shorter PR interval, as compared to athletes with normal ECG. However, no differences in direction of vectors or QRS‐T angle were observed (Table 2).

Table 2.

Comparison of ECG parameters in athletes with normal ECG versus athletes with ECG RVH and ECG LVH criteria

ECG and VCG parameters Control (C; n = 43) p (C‐N) Normal (N) ECG (n = 80) ECG RVH (n = 21) p (N‐RVH) ECG LVH (n = 16) p (N‐LVH)
Age (SD), year 21.7 (7.8) 0.051 19.2 (3.4) 19.4 (3.8) 0.818 19.4 (3.4) 0.821
Heart rate (SD), bpm 68.3 (12.9) 0.820 67.6 (12.8) 68.8 (14.5) 0.741 67.1 (13.7) 0.886
QTc interval (SD), ms 409.4 (17.2) 0.261 413.8 (17.0) 411.2 (19.7) 0.590 410.0 (25.2) 0.571
QRS duration (SD), ms 80.8 (16.5) 0.0003 91.3 (9.7) 97.1 (7.1) 0.004 97.4 (7.5) 0.001
PR interval (SD), ms 109.3 (49.3) <0.0001 145.6 (20.6) 149.0 (20.1) 0.504 147.3 (20.7) 0.773
Spatial peak QRS‐T angle (SD),° 51.0 (27.8) 0.134 43.1 (27.0) 71.4 (40.7) 0.006 52.4 (18.5) 0.103
Spatial area QRS‐T angle (SD) 69.5 (27.9) 0.608 66.8 (28.2) 92.7 (29.6) 0.001 67.7 (25.0) 0.914
Spatial peak QRS vector magn (SD), µV 1983 (838) 0.004 1588 (294) 1726 (356) 0.113 2,341 (301) <0.0001
Spatial QRS area (SD), mV × ms 42.2 (20.7) 0.001 30.7 (8.6) 32.5 (11.1) 0.486 55.3 (12.5) <0.0001
Spatial peak T‐vector magnitude (SD), µV 453 (236) <0.0001 487 (195) 568 (157) 0.053 601 (203) 0.051
Spatial T area (SD), mV × ms 39.5 (21.0) 0.626 41.3 (19.4) 52.8 (14.7) 0.005 51.9 (18.6) 0.053
Spatial peak SVG vector magn (SD), µV 2,260 (936) 0.037 1939 (395) 1976 (548) 0.775 2,744 (345) <0.0001
Wilson's ventricular gradient (SD), mV × ms 65.0 (28.6) 0.221 59.0 (19.1) 60.0 (20.0) 0.835 88.4 (24.9) 0.0003
SAI QRST (SD), mV × ms 179.7 (82.4) 0.110 157.8 (42.6) 194.9 (30.2) <0.0001 216.3 (42.3) 0.0001
Vector magnitude QT integral (SD), mV × ms 115.4 (50.9) 0.178 104.0 (27.6) 129.3 (18.8) <0.0001 141.1 (24.6) <0.0001
Spatial peak QRS vector azimuth (95% CI),° 26.7 (18.1–35.3) 0.500 25.5 (19.0–31.9) 45.5 (20.3–70.7) 0.0001* 25.9 (17.0–34.8) 0.200
Spatial area QRS vector azimuth (SD),° 39.9 (31.5–48.3) <0.05 47.7 (41.3–54.1) 71.0 (56.8–85.2) 0.0001 * 33.3 (23.1–43.5) 0.200
Spatial peak T‐vector azimuth (SD),° −28.2 (−34.2 to −22.3) 0.200 −26.1 (−30.2 to−22.0) −31.8 (−40.7 to −23.0) 0.500 −38.7 (−46.0 to −31.4) <0.05
Spatial area T‐vector azimuth (SD),° −32.3 (−38.7 to −25.9) 0.200 −32.9 (−36.8 to −28.9) −36.8 (−44.7 to −29.0) 0.500 −44.8 (−51.9 to −37.8) 0.05
Spatial peak SVG vector azimuth (SD),° 16.7 (9.1–24.4) 0.200 12.9 (7.4–18.5) 29.4 (7.9–50.9) 0.0001 * 12.2 (3.7–20.7) 0.500
Spatial area SVG vector azimuth (SD),° 0.2 (−6.1 to 6.6) 0.500 −4.9 (−9.1 to −0.7) −6.0 (−16.8 to 4.8) 0.500 −8.9 (−16.1 to −1.8) 0.500
Spatial peak QRS vector elevation (SD),° 50.1 (45.2–55.0) 0.500 43.2 (40.1–46.4) 52.9 (42.3–63.5) 0.500 43.0 (39.5–46.4) 0.500
Spatial area QRS vector elevation (SD),° 56.6 (51.1–62.1) 0.500 50.5 (46.8–54.1) 55.9 (45.7–66.1) 0.500 46.2 (41.5–50.9) 0.500
Spatial peak T‐vector elevation (SD),° 61.6 (57.5–65.8) 0.500 56.7 (54.3–59.1) 63.5 (60.4–66.6) 0.100 61.1 (56.7–65.6) 0.500
Spatial area T‐vector elevation (SD),° 67.0 (62.8–71.1) 0.200 62.8 (60.4–65.1) 68.0 (64.5–71.5) 0.100 68.3 (62.8–73.9) 0.200
Spatial peak SVG vector elevation (SD),° 49.0 (44.1–53.8) 0.200 43.2 (40.3–46.1) 51.0 (41.1–60.9) 0.200 42.7 (39.3–46.2) 0.200
Spatial area SVG vector elevation (SD),° 53.4 (49.4–57.4) 0.500 48.8 (46.2–51.4) 50.4 (44.9–55.9) 0.500 48.3 (44.6–52.0) 0.200
Open in a new tab
a

Both Watson U‐square statistic and Kuiper k* statistic <0.002.

3.4. Right and Left ventricular remodeling

Nearly one‐fifth of the athletes (26 out of 140 athletes, 18.6%) reached the RVH voltage threshold (Sharma et al., 2018). RVH by voltage criteria was more frequent in male versus female athletes [24 (25.8%) vs. 2 (4.3%); p = 0.002]. VCG activation pattern was characterized by the QRS vector at the second half of QRS loop pointing to the right, backward, and upward, reflecting activation of lateral right ventricle (RV) and RV outflow tract (Figure 3). In athletes with RVH ECG voltage, spatial QRS and SVG vectors pointed farther backward as compared to the athletes free from any ECG findings, which explained wider QRS‐T angle (Table 2). Athletes with ECG RVH voltage had significantly larger SAI QRST and iVMQT, whereas QRS and SVG vector magnitudes did not differ.

Figure 3.

Figure 3

Open in a new tab

Representative normal electrocardiogram and vectorcardiogram of an athlete with right ventricular hypertrophy voltage. (a) 12‐lead electrocardiogram. (b) vectorcardiogram. (c) Frontal plane. (d) Transverse plane. (e) Sagittal plane. Color progression from red to purple reflects progression of vectorcardiographic loop from QRS onset

Twenty‐one (15%) athletes met 12‐lead ECG LVH criteria. Male athletes were more likely to demonstrate ECG LVH voltage than female athletes [19 (20.4%) vs. 2 (4.3%); p = 0.011]. Athletes with ECG LVH voltage were characterized by significantly larger magnitudes of all vectors: QRS, T, and SVG, including Wilson's SVG, and SAI QRST and iVMQT (Table 2). However, direction of vectors and QRS‐T angle did not differ from that in athletes free from any ECG findings.

3.5. Sex and age differences

In our study, there were 80 athletes free from any 12‐lead ECG findings and without family history of SCD. Approximately half were male (n = 43; 54%). Male athletes had wider QRS‐T angle due to significant differences in the azimuth of QRS and T vectors. In male athletes, QRS vector pointed farther backward (Figure 4), whereas T vector pointed farther forward, as compared to female athletes, which resulted in wider QRS‐T angle. In female athletes, QRS and T vectors were closer to each other, forming a narrower QRS‐T angle. Male athletes had larger T‐vector magnitude, SAI QRST, and iVMQT (Supporting information Table S1). However, there were no differences in QRS and SVG vector magnitudes between male and female athletes.

Figure 4.

Figure 4

Open in a new tab

Comparison of normal vectorcardiogram in female (a) versus male (b) athlete. Peak QRS (red), T (green), and spatial ventricular gradient (blue) vectors are shown

In male (but not female) athletes, QRS duration increased, whereas QRS‐T angle decreased with age (Figure 5). Narrowing of QRS‐T angle with age in male athletes was due to forward rotation of QRS vector. Consistently, SVG vector also rotated forward in male (but not female) athletes. T‐vector elevation decreased with age in female (but not male) athletes. PR interval slightly increased with age (Supporting information Table S2).

Figure 5.

Figure 5

Open in a new tab

Boxplot of QRS duration (a), mean QRS‐T angle (b), peak QRS‐T angle (c), peak SVG azimuth (d), peak QRS azimuth (e), peak T elevation (f) in male and female athletes of four age categories (14 to <16 years, 16 to <18 years, 18 to <22 years, and 22–30 years). Median (dark horizontal line crossing the box) and interquartile range [IQR] (box) are plotted. Whiskers specify the adjacent values, defined as the most extreme values within 1.5 IQR of the nearer quartile

In multiple linear regression analyses, sex remained the main statistically significant predictor of ECG and VCG parameters, whereas age and training level of athletes did not associate with ECG or VCG characteristics in this study (Table 3), with one exception. PR interval significantly increased with age [+2.7 (95% CI 1.1–4.3) ms per year of age; p = 0.001]. In age‐ and training level‐adjusted analyses, male sex was associated with lager T‐vector magnitudes, SAI QRST, and iVMQT, and broader QRS‐T angle (Table 3).

Table 3.

Age‐ and training level‐adjusted associations of sex with ECG/VCG parameters

Difference in the following ECG/VCG parameters Male versus Female (95% CI) p‐value
QTc interval, ms −7.6 (−15.2 to 0.05) 0.052
QRS duration, ms +8.7 (4.7–12.7) <0.0001
Spatial peak QRS‐T angle,° +20.8 (9.6–31.9) <0.0001
Spatial mean QRS‐T angle,° +28.2 (17.3–39.2) <0.0001
Spatial peak T‐vector magnitude, µV +186 (106–266) <0.0001
Spatial mean T‐vector magnitude, µV +91 (61–120) <0.0001
SAI QRST, mV × ms +51.6 (35.6–67.6) <0.0001
iVMQT, mV × ms +33.8 (23.5–44.1) <0.0001
Wilson's Ventricular Gradient, mV × ms +11.5 (3.0–20.0) 0.008
Open in a new tab

3.6. Effect of athletic performance

In a multiple linear regression analysis, after adjustment for sex and age, top percentile rank of the athletes’ FIS ranking in the downhill discipline—and none of the other three disciplines—was nominally (p < 0.05) associated with spatial peak QRS‐T angle (Figure 6). Better athletic performance was associated with narrower QRS‐T angle. Each 10% worsening in an athlete's FIS downhill ranking percentile was associated with an increase in spatial QRS‐T angle by 2.1 (95% CI 0.3–3.9) degrees (p = 0.013). The effect of athletic performance was independent from the effect of age: An increase in age by one age category was associated with narrowing of QRS‐T angle by −9.1 (95% CI −16.9 to −1.2)°.

Figure 6.

Figure 6

Open in a new tab

Association of the athletic performance (downhill race top percentile) with spatial peak QRS‐T angle. Sex‐ and age‐adjusted predictive margins of spatial peak QRS‐T angle with 95% confidence interval (Y‐axis) are plotted against the downhill race top percentile (X‐axis)

3.7. Scalar value of spatial ventricular gradient vector magnitude

SAI QRST perfectly correlated with iVMQT (r = 0.99; p < 0.0001), and their detrended log‐transformed values were in perfect agreement (Supporting information Figure S1).

3.8. VCG manifestation of ECG phenomena

The clinical value of VCG review is illustrated by two cases. In a first case, a 17 y.o. white female, professional alpine skier, without history of cardiovascular symptoms, or family history of SCD, presented with a classical ER 12‐lead ECG pattern (Figure 7). Her VCG was characterized by an unusual activation pattern, with concealed septal activation: in the beginning of QRS ventricular activation, the vector pointed to the back, leftward, and downward. Consistently, there was complete absence of the q wave and an abrupt upstroke of a narrow QRS complex on 12‐lead ECG. VCG demonstrated that a “J wave” in V4‐V6 corresponded with a normal activation pattern of inferoposterior left ventricular (LV) basal region and RV outflow tract at the second half of QRS.

Figure 7.

Figure 7

Open in a new tab

A case of an early repolarization in an athlete. (a) 12‐lead electrocardiogram. (b) vectorcardiogram. (c) Frontal plane. (d) Transverse plane. (e) Sagittal plane. “J wave” location on QRS loop is shown. Color progression from red to purple reflects progression of vectorcardiographic loop from QRS onset

In a second case, a 22 y.o. white female, professional alpine skier, presented with a bifid T wave in V2‐V4, and otherwise normal 12‐lead ECG with QTc of 431 ms (Figure 8). She was free from symptoms and had no personal history of cardiovascular disease, but had a family history of SCD. Her VCG was characterized by a special curved appearance of a folding narrow T‐loop (“lizard tongue”), which was responsible for an appearance of bifid T wave on 12‐lead ECG.

Figure 8.

Figure 8

Open in a new tab

A case of a bifid T wave in an athlete. (a) 12‐lead electrocardiogram. (b) Vectorcardiogram. (c) Transverse plane. (d) Sagittal plane. (e) Frontal plane. Color progression from red to purple reflects progression of vectorcardiographic loop from QRS onset

4. DISCUSSION

This prospective Sun Valley Ski study of predominantly white skiers showed that sex is the major determinant of ECG and VCG characteristics in athletes. Male athletes were characterized by larger T‐vector magnitude, and wider QRS‐T angle than female athletes, but similar direction and magnitudes of QRS and SVG vectors. QRS‐T angle gradually decreased with age in male athletes, due to forward rotation of QRS vector. Athletes with ECG LVH voltage criteria were characterized by larger amplitudes of all VCG vectors without widening of QRS‐T angle, suggesting likely physiological remodeling of the athlete's LV. In athletes with ECG RVH voltage criteria, QRS and SVG vectors rotated farther backwards, leading to wider QRS‐T angle, which characterized RV remodeling. Better athletic performance was associated with narrower QRS‐T angle. Half of athletes with ECG RVH voltage criteria had VCG parameters consistent with previously reported abnormal GEH (Waks et al., 2016). Further studies of VCG in athletes are needed, to determine personalized threshold of GEH‐manifested RV remodeling associated with the risk of RV dysfunction development (Heidbuchel, 2017; Heidbuchel, Prior, & La Gerche, 2012).

4.1. VCG manifestation of the athlete's heart

The athlete's heart is characterized by distinctive electrical, structural, and functional adaptations (Sharma, Merghani, & Mont, 2015). All four chambers of the heart undergo remodeling. In our study, RVH voltage criteria were observed more frequently than previously reported (19% vs. 13%) (Drezner et al., 2017), whereas LVH voltage criteria were observed less frequently (15% vs. 50%; Sharma et al., 1999). Observed differences may be due to the differences in the type of sport activity and training. As alpine skiing is a class IIIB sport with a high static and moderate dynamic demand (Mitchell et al., 2005), we observed ECG RVH in nearly one in five athletes within our study. Importantly, in our study, there were no differences between nonathletes and athletes with normal ECG, which supports validity of the study findings. Evidence of RV injury associated with bolus of intense exercise has been previously reported (Claessen & La Gerche, 2016; Heidbuchel & La Gerche, 2012). However, it remains unknown whether transient myocardial injury can lead to a chronic myocardial damage and clinical events, and it is unknown how to identify athletes at risk (Claessen & La Gerche, 2016). The results of our study help to generate hypotheses to be tested in future studies. It is known that signaling pathways and molecular mechanisms responsible for mediating pathological versus physiological cardiac hypertrophy induced by long‐term exercise training are different (Weeks, Bernardo, Ooi, Patterson, & McMullen, 2017). The insulin‐like growth factor 1 (IGF1) signaling pathway regulates exercise‐induced physiological cardiac hypertrophy and cardiac protection (Weeks et al., 2017). Tereshchenko et al recently identified a GEH‐associated genetic polymorphism in the intron of the IGF1‐receptor (IGF1R) gene (Tereshchenko et al., 2018). As demonstrated by an IGF1R locus‐associated GEH phenotype (Figure 4e in (Tereshchenko et al., 2018)), physiological cardiac remodeling in the athlete's heart is characterized by sizable increase in SAI QRST and SVG magnitude with only mild increase in SVG azimuth and QRS‐T angle and simultaneous decrease in SVG elevation. Thus, GEH presentation can potentially distinguish athletes with physiological RV remodeling from those who are overtrained and are at risk to develop RV dysfunction.

ECGs that fulfilled LVH voltage criteria were characterized by increased size of VCG loops but preserved normal loop morphology; normal direction of QRS, T, and SVG vectors; and normal QRS‐T angle. Observed VCG pattern was consistent with physiological LV remodeling.

4.2. Sex and age differences in normal VCG

Allowing for differences in cardiac repolarization between men and women, we observed significantly larger T vector, SAI QRST, and iVMQT in male, as compared to female athletes. Male (but not female) athletes demonstrated a gradual decrease in QRS‐T angle with age, frequently with forward rotation of QRS vector. T‐vector elevation slightly decreased with age in female (but not male) athletes. Observed differences were likely explained by sex hormones and different growth patterns in young male versus female athletes, which underscores the importance of sex‐specific criteria for ECG/VCG interpretation.

4.3. Effect of an athletic training and athletic performance

In this study, we observed an association of athletic performance, but not athletic training, with spatial QRS‐T angle. Higher athletic performance in the downhill discipline was associated with narrower QRS‐T angle. Downhill is the highest speed and longest alpine skiing discipline with participants reaching speeds up to 130 km/h for 2–3 min and thus requires greater cardiac output, in contrast to the shorter (Super Giant Slalom) and more technical (giant slalom and slalom) disciplines (Turnbull, Kilding, & Keogh, 2009). Validation of this finding in larger studies of athletes is needed.

4.4. VCG manifestation of ECG patterns

Vectorcardiograpic presentation of classical 12‐lead ECG patterns of ER and bifid T wave helped to interpret the 12‐lead ECG appropriately. Discrepant results of the studies that investigated association of an ER 12‐lead ECG pattern (Macfarlane et al., 2015) with clinical outcomes (Haissaguerre et al., 2008; Uberoi et al., 2011) suggest heterogeneity of associated conditions. In our study, we observed a young athlete with a typical ER 12‐lead ECG pattern. Review of VCG morphology clearly demonstrated that a “J wave” in V4‐V6 and QRS slurring in II, III, and aVF corresponds to a QRS loop and not the repolarization phase. Direction of the QRS vector at the time of a “J wave” is consistent with a normal pattern of ventricular activation of inferoposterior LV and RV outflow tract, as expected at the second half of QRS. This subject's unusual pattern of ventricular activation was observed in the beginning and not the end of QRS. In this individual, activation spreads very fast in all directions in the LV, likely simultaneously toward the endo‐ and epicardium, which formed the initial QRS vector pointing backward, leftward, and downward, forming a narrow, abrupt QRS with concealed septal activation. We speculate that increased trabeculation of the subendocardium may allow the Purkinje network to more rapidly conduct through theoretically more porous LV walls, thus offering an explanation for the observed phenomenon. John P. Boineau previously described ER with narrow QRS complexes and discussed possible mechanisms (Boineau, 2007). Future studies of the “ER”‐subtype due to concealed septal ventricular activation are needed in order to determine its clinical significance.

In another case, we described a bifid T wave in V2‐V4 appearance due to curved T‐loop. Bifid T wave is a characteristic feature of long QT syndrome type 2 (Takenaka et al., 2003). At the same time, bifid T wave was reported as a benign age‐related ECG phenomenon (Calabro et al., 2009). In an Italian population, 50% of children of 5 years of age had a bifid T wave. The rate of bifid T wave decreased as children got older; it was observed in less than 10% of children above 12 years of age (Calabro et al., 2009). The high prevalence of bifid T wave in V2‐V3 highlights the need to develop additional features that might help distinguish between “malignant” and benign bifid T waves. Experiments have shown that interventricular dispersion of repolarization could be a mechanism behind the bifid T wave appearance (Meijborg et al., 2015). However, clinical investigators frequently experience challenges in discrimination of T and U waves on a 12‐lead ECG (Kurokawa et al., 2010). We observed that a bifid T wave presentation on a 12‐lead ECG corresponds to a narrow, curved, “lizard tongue”‐like appearance of three‐dimensional T‐loop. VCG presentation allows for clear differentiation of U‐ and T‐loops as two separate and easily distinguishable phenomena. The “notch” on a T wave in V2‐V4 in our study athlete is a result of the V2‐V4 leads’ axes crossing both base and peak areas of the T‐loop. Projection of a curved T‐loop on the V2‐V4 leads’ axes (Figure 8) is presented as a T vector moving toward, then away, and then again toward the V2‐V4 leads’ electrodes, forming the “T‐notch” or bifid T wave. Further studies of narrow, curved, T‐loop are needed to determine its’ diagnostic and prognostic value. While our study athlete with narrow, curved, T‐loop and bifid T wave had a family history of SCD, which raised suspicion of long QT syndrome type 2, we had no data to diagnose or rule out long QT syndrome type 2 or any other condition in this case.

4.5. Quantitative characterization of VCG

In this study, we observed a perfect correlation (r = 0.99) and agreement of detrended values between SAI QRST and iVMQT, which means that both measures reflect the same underlying phenomenon and either can be used to measure scalar value of SVG. However, as SAI QRST is not numerically equal to iVMQT, thresholds for abnormal values for each of these metrics should be determined separately.

4.6. Limitations

Limitations of this study must be considered. Most of the study participants were white, which limited study generalizability. Further studies of different races and ethnicities are needed. Statistical power of the study was small, which limited subgroups analyses, and did not allow a systematic study of observed VCG subgroups of typical ECG phenomena of ER and bifid T wave. However, the observed hypothesis‐generating findings are important and will inform future studies.

5. CONCLUSION

Review of VCG morphology in addition to traditional 12‐lead ECG presentation provides an opportunity to characterize additional features of ventricular activation and repolarization and delivers nuanced interpretation of EP substrate. GEH presentation can potentially distinguish athletes with physiological RV remodeling from those who are overtrained and develop RV dysfunction. Our study provided sex‐ and age‐specific reference data for VCG measures in male and female athletes and will serve as a hypothesis‐generating foundation for future studies.

CONFLICT OF INTERESTS

None.

Supporting information

 

Click here for additional data file. (171.7KB, docx)

ACKNOWLEDGMENT

The authors thank Dr. Albert Starr, MD, and the Sun Valley Ski Education Foundation for the study support.

Thomas JA, Perez‐Alday EA, Junell A, et al. Vectorcardiogram in athletes: The Sun Valley Ski Study. Ann Noninvasive Electrocardiol. 2019;24:e12614 10.1111/anec.12614

Funding information

This work was partially supported by the National Institutes of Health R01HL118277 (LGT) and the National Library of Medicine Training Grant T15LM007442 (JAT).

REFERENCES

  1. Bland, J. M. , & Altman, D. G. (1986). Statistical methods for assessing agreement between two methods of clinical measurement. Lancet, 1(8476), 307–310. [PubMed] [Google Scholar]
  2. Boineau, J. P. (2007). The early repolarization variant–normal or a marker of heart disease in certain subjects. Journal of Electrocardiology, 40(1), 3–6. 10.1016/j.jelectrocard.2006.04.002. [DOI] [PubMed] [Google Scholar]
  3. Calabro, M. P. , Barberi, I. , La Mazza, A. , Todaro, M. C. , De Luca, F. L. , Oreto, L. , … Oreto, G. (2009). Bifid T waves in leads V2 and V3 in children: A normal variant. Italian Journal of Pediatrics, 35(1), 17 10.1186/1824-7288-35-17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Claessen, G. , & La Gerche, A. (2016). Exercise‐induced cardiac fatigue: The need for speed. The Journal of Physiology, 594(11), 2781–2782. 10.1113/JP272168. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Couderc, J. P. , Xiaojuan, X. , Zareba, W. , & Moss, A. J. (2005). Assessment of the stability of the individual‐based correction of QT interval for heart rate. Annals of Noninvasive Electrocardiology, 10(1), 25–34. 10.1111/j.1542-474X.2005.00593.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Dhutia, H. , Malhotra, A. , Finocchiaro, G. , Merghani, A. , Papadakis, M. , Naci, H. , … Sharma, S. (2017). Impact of the International Recommendations for Electrocardiographic Interpretation on Cardiovascular Screening in Young Athletes. Journal of the American College of Cardiology, 70(6), 805–807. 10.1016/j.jacc.2017.06.018. [DOI] [PubMed] [Google Scholar]
  7. Drezner, J. A. , O'Connor, F. G. , Harmon, K. G. , Fields, K. B. , Asplund, C. A. , Asif, I. M. , … Roberts, W. O. (2016). AMSSM position statement on cardiovascular preparticipation screening in athletes: Current evidence, knowledge gaps, recommendations, and future directions. Clinical Journal of Sport Medicine, 26(5), 347–361. 10.1097/JSM.0000000000000382. [DOI] [PubMed] [Google Scholar]
  8. Drezner, J. A. , Sharma, S. , Baggish, A. , Papadakis, M. , Wilson, M. G. , Prutkin, J. M. , … Corrado, D. (2017). International criteria for electrocardiographic interpretation in athletes: Consensus statement. British Journal of Sports Medicine, 51(9), 704–731. 10.1136/bjsports-2016-097331. [DOI] [PubMed] [Google Scholar]
  9. Haissaguerre, M. , Derval, N. , Sacher, F. , Jesel, L. , Deisenhofer, I. , de Roy, L. , … Clementy, J. (2008). Sudden cardiac arrest associated with early repolarization. New England Journal of Medicine, 358(19), 2016–2023. 10.1056/NEJMoa071968 [DOI] [PubMed] [Google Scholar]
  10. Harmon, K. G. , Asif, I. M. , Maleszewski, J. J. , Owens, D. S. , Prutkin, J. M. , Salerno, J. C. , … Drezner, J. A. (2015). Incidence, cause, and comparative frequency of sudden cardiac death in National Collegiate Athletic Association athletes: A decade in review. Circulation, 132(1), 10–19. 10.1161/CIRCULATIONAHA.115.015431. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Heidbuchel, H. (2017). The athlete's heart is a proarrhythmic heart, and what that means for clinical decision making. Europace, 20(9), 1401–1411. 10.1093/europace/eux294. [DOI] [PubMed] [Google Scholar]
  12. Heidbuchel, H. , & La Gerche, A. (2012). The right heart in athletes. Evidence for exercise‐induced arrhythmogenic right ventricular cardiomyopathy. Herzschrittmacherther Elektrophysiol, 23(2), 82–86. 10.1007/s00399-012-0180-3. [DOI] [PubMed] [Google Scholar]
  13. Heidbuchel, H. , Prior, D. L. , & La Gerche, A. (2012). Ventricular arrhythmias associated with long‐term endurance sports: What is the evidence? British Journal of Sports Medicine, 46(Suppl 1), i44–i50. 10.1136/bjsports-2012-091162. [DOI] [PubMed] [Google Scholar]
  14. Kabir, M. M. , Perez‐Alday, E. A. , Thomas, J. , Sedaghat, G. , & Tereshchenko, L. G. (2017). Optimal configuration of adhesive ECG patches suitable for long‐term monitoring of a vectorcardiogram. Journal of Electrocardiology, 50(3), 342–348. 10.1016/j.jelectrocard.2016.12.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Kors, J. A. , & van, H. G., Sittig, A. C., & van Bemmel, J. H., (1990). Reconstruction of the Frank vectorcardiogram from standard electrocardiographic leads: Diagnostic comparison of different methods. European Heart Journal, 11(12), 1083–1092. 10.1093/oxfordjournals.eurheartj.a059647 [DOI] [PubMed] [Google Scholar]
  16. Kurokawa, S. , Niwano, S. , Kiryu, M. , Murakami, M. , Ishikawa, S. , Yumoto, Y. , … Izumi, T. (2010). Importance of morphological changes in T‐U waves during bepridil therapy as a predictor of ventricular arrhythmic event. Circulation Journal, 74(5), 876–884. 10.1253/circj.CJ-09-0937 [DOI] [PubMed] [Google Scholar]
  17. Macfarlane, P. W. , Antzelevitch, C. , Haissaguerre, M. , Huikuri, H. V. , Potse, M. , Rosso, R. , … Yan, G. X. (2015). The early repolarization pattern: A consensus paper. Journal of the American College of Cardiology, 66(4), 470–477. 10.1016/j.jacc.2015.05.033. [DOI] [PubMed] [Google Scholar]
  18. Maron, B. J. , Doerer, J. J. , Haas, T. S. , Tierney, D. M. , & Mueller, F. O. (2009). Sudden deaths in young competitive athletes: Analysis of 1866 deaths in the United States, 1980–2006. Circulation, 119(8), 1085–1092. 10.1161/CIRCULATIONAHA.108.804617. [DOI] [PubMed] [Google Scholar]
  19. Maron, B. J. , Friedman, R. A. , Kligfield, P. , Levine, B. D. , Viskin, S. , Chaitman, B. R. , … Thompson, P. D. (2014). Assessment of the 12‐lead ECG as a screening test for detection of cardiovascular disease in healthy general populations of young people (12–25 Years of Age): A scientific statement from the American Heart Association and the American College of Cardiology. Circulation, 130(15), 1303–1334. 10.1161/cir.0000000000000025. [DOI] [PubMed] [Google Scholar]
  20. Meijborg, V. M. F. , Chauveau, S. , Janse, M. J. , Anyukhovsky, E. P. , Danilo, P. R. , Rosen, M. R. , … Coronel, R. (2015). Interventricular dispersion in repolarization causes bifid T waves in dogs with dofetilide‐induced long QT syndrome. Heart Rhythm: The Official Journal of the Heart Rhythm Society, 12(6), 1343–1351. 10.1016/j.hrthm.2015.02.026. [DOI] [PubMed] [Google Scholar]
  21. Mitchell, J. H. , Haskell, W. , Snell, P. , & Van Camp, S. P. (2005). Task Force 8: Classification of sports. Journal of the American College of Cardiology, 45(8), 1364–1367. 10.1016/j.jacc.2005.02.015. [DOI] [PubMed] [Google Scholar]
  22. Rijnbeek, P. R. , Witsenburg, M. , Schrama, E. , Hess, J. , & Kors, J. A. (2001). New normal limits for the paediatric electrocardiogram. European Heart Journal, 22(8), 702–711. 10.1053/euhj.2000.2399. [DOI] [PubMed] [Google Scholar]
  23. Sharma, S. , Drezner, J. A. , Baggish, A. , Papadakis, M. , Wilson, M. G. , Prutkin, J. M. , … Corrado, D. (2018). International recommendations for electrocardiographic interpretation in athletes. European Heart Journal, 39(16), 1466–1480. 10.1093/eurheartj/ehw631. [DOI] [PubMed] [Google Scholar]
  24. Sharma, S. , Merghani, A. , & Mont, L. (2015). Exercise and the heart: The good, the bad, and the ugly. European Heart Journal, 36(23), 1445–1453. 10.1093/eurheartj/ehv090. [DOI] [PubMed] [Google Scholar]
  25. Sharma, S. , Whyte, G. , Elliott, P. , Padula, M. , Kaushal, R. , Mahon, N. , & McKenna, W. J. (1999). Electrocardiographic changes in 1000 highly trained junior elite athletes. British Journal of Sports Medicine, 33(5), 319–324. 10.1136/bjsm.33.5.319. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Sur, S. , Han, L. , & Tereshchenko, L. G. (2013). Comparison of sum absolute QRST integral, and temporal variability in depolarization and repolarization, measured by dynamic vectorcardiography approach, in healthy men and women. PLoS ONE, 8(2), e57175 10.1371/journal.pone.0057175. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Takenaka, K. , Ai, T. , Shimizu, W. , Kobori, A. , Ninomiya, T. , Otani, H. , … Horie, M. (2003). Exercise stress test amplifies genotype‐phenotype correlation in the LQT1 and LQT2 forms of the long‐QT syndrome. Circulation, 107(6), 838–844. 10.1161/01.Cir.0000048142.85076.A2. [DOI] [PubMed] [Google Scholar]
  28. Tereshchenko, L. G. , Cheng, A. , Fetics, B. J. , Butcher, B. , Marine, J. E. , Spragg, D. D. , … Berger, R. D. (2011). A new electrocardiogram marker to identify patients at low risk for ventricular tachyarrhythmias: Sum magnitude of the absolute QRST integral. Journal of Electrocardiology, 44(2), 208–216. 10.1016/j.jelectrocard.2010.08.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Tereshchenko, L. G. , Sotoodehnia, N. , Sitlani, C. M. , Ashar, F. N. , Kabir, M. , Biggs, M. L. , … Arking, D. E. (2018). Genome‐wide associations of global electrical heterogeneity ECG phenotype: The ARIC (Atherosclerosis Risk in Communities) study and CHS (Cardiovascular Health Study). Journal of the American Heart Association, 7(8), e008160 10.1161/JAHA.117.008160. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. 13th FIS points list 2015/2016 (2016). https://data.fis-ski.com/alpine-skiing/fis-points-lists.html.
  31. Turnbull, J. R. , Kilding, A. E. , & Keogh, J. W. (2009). Physiology of alpine skiing. Scandinavian Journal of Medicine and Science in Sports, 19(2), 146–155. 10.1111/j.1600-0838.2009.00901.x. [DOI] [PubMed] [Google Scholar]
  32. Uberoi, A. , Jain, N. A. , Perez, M. , Weinkopff, A. , Ashley, E. , Hadley, D. , … Froelicher, V. (2011). Early repolarization in an ambulatory clinical population. Circulation, 124(20), 2208–2214. 10.1161/CIRCULATIONAHA.111.047191. [DOI] [PubMed] [Google Scholar]
  33. Waase, M. P. , Mutharasan, R. K. , Whang, W. , DiTullio, M. R. , DiFiori, J. P. , Callahan, L. , … Engel, D. J. (2018). Electrocardiographic findings in National Basketball Association athletes. JAMA Cardiology, 3(1), 69–74. 10.1001/jamacardio.2017.4572. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Waks, J. W. , Sitlani, C. M. , Soliman, E. Z. , Kabir, M. , Ghafoori, E. , Biggs, M. L. , … Tereshchenko, L. G. (2016). Global electric heterogeneity risk score for prediction of sudden cardiac death in the general population: The atherosclerosis risk in communities (ARIC) and cardiovascular health (CHS) studies. Circulation, 133(23), 2222–2234. 10.1161/CIRCULATIONAHA.116.021306. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Weeks, K. L. , Bernardo, B. C. , Ooi, J. Y. Y. , Patterson, N. L. , & McMullen, J. R. (2017). The IGF1‐PI3K‐Akt signaling pathway in mediating exercise‐induced cardiac hypertrophy and protection. Advances in Experimental Medicine and Biology, 1000, 187–210. 10.1007/978-981-10-4304-8_12. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

 

Click here for additional data file. (171.7KB, docx)

Articles from Annals of Noninvasive Electrocardiology : The Official Journal of the International Society for Holter and Noninvasive Electrocardiology, Inc are provided here courtesy of International Society for Holter and Noninvasive Electrocardiology, Inc. and Wiley Periodicals, Inc.

RESOURCES