Increased longitudinal contractility and diastolic function at rest in well-trained amateur Marathon runners: a speckle tracking echocardiography study

The organizers of the (2006 and 2007) Berlin Marathon invited local registered amateur runners by e-mail to participate in our study. Runners from Berlin-Brandenburg area who had completed at least one marathon previously were included into this study. All participants underwent clinical examination and echocardiography at least 10 days before the marathon at rest. Exclusion criteria were known cardiovascular disorders (atrial fibrillation, pacemaker, bypass surgery, prosthetic valves or congenital heart disease), signs and symptoms of coronary artery disease, recent pathologic stress test, hypertension and diabetes. All participants had a normal standard echocardiography with no signs of cardiac pathology. In addition, an ECG and a blood sample were taken from every participant. The study protocol was approved by the ethics committee of the Charité University Hospital and conducted according to the Helsinki Declaration. Written informed consent was obtained from each participant. Twenty two healthy non-athlete adults served as controls.

Standard transthoracic echocardiography was performed according to the guidelines of the American Society of Echocardiography (ASE) and the European Association of Cardiovascular Imaging (EACVI) on Vivid 7 Dimension (GE Vingmed, Horton, Norway, M3S 1.5-4.0 MHz transducer) [18]. Segmental Color Doppler recordings were performed to exclude valvular dysfunction. Real-time 2D ultrasound data from the left ventricle at the three apical views with a frame rate greater than 50 frames per second (fps) were stored digitally for offline 2D strain and strain rate analysis (EchoPac PC, GE Vingmed, Horton Norway) (Figure 1). Aortic valve opening and closure was set manually by the opening and closing artefact using a transaortic continuous-wave Doppler signal in order to define end systole for 2D strain analysis. For segmental longitudinal strain and strain rate analysis an 18 segment- model of the LV were used by dividing each LV wall into 3 segments. The region of interest (ROI) was corrected manually in order to analyze the myocardium and to exclude the pericardium. For each segment the quality of speckle tracking was analyzed automatically. Segments with poor tracking were excluded for further measurements. To achieve average 2D peak systolic strain the mean of every wall and then the average of all walls was calculated. Strain rate was measured at the most negative point of SR during the ejection phase. The early (E’) and late (A’) diastolic strain rate was measured as a marker of the diastolic function in this cohort.

Results are expressed as mean value ± standard deviation (SD). Outliers and/or skew distributions differences in interesting groups of individuals were analyzed using nonparametric statistical tests (Mann–Whitney U-test). Spearman’s correlation coefficient was calculated in order to test for variable dependencies of 2D strain parameters. Statistical analyses were performed using SPSS for Windows (Release 19.0, Copyright ® SPSS Inc.). The comparison of our results and the published values in the literature was performed by the t-test. Significance was assessed at the p < 0.05 level.

The intra- and interobserver variability was measured on different days by three different cardiologists blinded to the results of the other cardiologists. 10 healthy individuals were examined. We calculated the mean longitudinal 2D strain in each segment. We have then calculated the delta of each individual measurement and the mean. Finally the average of the deltas was calculated to measure the intra and inter observer variability.

Eighty-four marathon runners, 43 women and 41 men participated in this study (mean age 50 ± 11 years). All participants had a normal ECG and blood test including NTproBNP in the normal age and gender-adjusted range according to Hess [19]. The normal NTproBNP levels underline the echocardiography results of cardiac healthy population in our study. The systolic and diastolic blood pressure as well as heart rate and weekly average training differ not significantly between female and male amateur runners. The male runners had a higher running experience explained in a higher amount of previous marathons. The female runners were slightly younger and there was a significant difference in body mass index between the groups. The baseline characteristics, training level and running experience are shown in Table 1.

1450 LV wall segments (96%) of a total of 1512 segments had sufficient speckle tracking quality and could be analyzed. The best tracking was detectable in the septum and in the basal and medial segments of all LV walls. The worst quality was seen in the anterior wall and in the apex. The number of analyzed segments is summarized in Table 2.

Table 3 shows the mean values of peak systolic 2D strain of all LV segments. The average peak systolic strain in the basal segments was significantly lower than in the mid-ventricular segment in all LV walls except posterior (P = 0.751) and inferior (P = 0.282). The average peak systolic strain in the mid-ventricular segments was significantly lower than the values of the apical segment in all LV walls. The average global longitudinal peak systolic strain of the left ventricle was -23 ± 2% with peak systolic strain rate -1.39 ± 0.21/s, early diastolic strain rate 2.0 ± 0.40/s and late diastolic strain rate 1.21 ± 0.31/s (Table 4). The global longitudinal peak systolic strain of the left ventricle differed significantly between female and male amateur runners with lower systolic deformation indices in male runner group. This phenomenon is due to significantly different average strain values in the apical long axis (posterior and anteroseptal wall) and two chamber (inferior and anterior wall) view, while in the four chamber view no significant difference were seen between the groups (Table 5).There is no correlation between age and average global longitudinal 2D Strain the whole study group (Figure 2).

The interobserver variability for longitudinal 2D strain measurements was 5,0% and the intraobserver variability was 8,4%.

Table 4 shows the results of our cohort compared to previous published normal 2D strain values in healthy individuals. For comparison we used the data published by Marwick et al. and from the large HUNT Study [12, 13]. The global left ventricular longitudinal peak systolic strain and systolic strain rate of our cohort are significantly higher compared to the previous published values of normal individuals. The same phenomenon was observable for the early and late diastolic strain rate values, E’ and A’ (Table 4). The average longitudinal 2D strain at rest of the healthy non-athlete adult controls (n = 22) was -21.1 ± 2.5% (p < 0.001, compared with athletes).

There is no age and gender matched control group in this study. The comparisons were drawn to historic controls in the literature. Secondly, we have only analyzed longitudinal 2D Strain and do not have invasive data. Our study did not examine age and gender differences.