P4419The association between ECG voltage and left-ventricular mass, sex, body size and the distance between the heart and chest wall in college athletes

Abstract

Abstract Background The ECG is widely used in pre-participation evaluation (PPE) of athletes (ATH). While it is assumed that greater than normal QRS voltages reflect physiologically increased left ventricular mass (LVM), this has not been adequately demonstrated in ATH. Purpose To examine the relation between QRS voltage on surface ECG and LVM and explore if the distance from the chest wall to mid-LV (CWLVdis) affects QRS voltage in ATH. Methods We examined digitized ECG data and echocardiograms in college ATH, obtained as part of routine PPE in years 2010–16. ECG parameters included R and S-wave voltage components of the Sokolow-Lyon (S-L) and Cornell criteria for LV hypertrophy (i.e. SV1 + RV5-V6 and RaVL + SV3, respectively). Transthoracic 2D echocardiography was used to determine LVM (area-length method) and the CWLVdis (detailed in Fig1A). S-L positive (SV1 + RV5-V6 >35 mV or RaVL >11 mV) ATH were compared to S-L negative by t-test, and univariate correlation and multivariable regression analysis was used to explore independent effects of body characteristics, sex, LVM and CWLVdis on QRS voltage. Results Included were 227 ATH (age 18.6±0.7 yr; 85% male; 60%/33% Caucasian/Afro-american). Of these, 66% played American football, 18% volleyball and 16% basketball. Overall, mean LVM was 174±37 g (range 96–284 g), and BSA-indexed LVM was 78±12 g/m2 (range 49–108 g/m2). Mean CWLVdis was 8.5±1.1 cm (range 5.6–11.3 cm) and was greater in males (p<0.001, Fig1B). Forty-six ATH (24%, all male) were S-L positive and no ATH were positive according to Cornell criteria. S-L positive ATH had lower BMI (25.3±3.5 vs 26.9±4.9, p=0.012), greater absolute LVM (189.1±31.3 vs. 170.1±37.4 g, p=0.002) and greater BSA-indexed LVM (85.3±10.3 vs. 76.6±11.7 g/m2, p<0.001) than S-L negative ATH. The CWLVdis was similar between S-L positive and negative ATH (8.4±1.2 vs. 8.6±1.1, respectively, p=0.213). CWLVdis was more strongly correlated to body mass (r=0.73, p<0.001, Fig. 1C) than to height (r=0.34, p<0.001). LVM correlated weakly to ECG voltage as combined in the S-L or Cornell criteria (Fig. 1C). CWLVdis was weakly correlated with R in aVL, V5 and V6 (r=0.21, 0.16 and 0.16, all p<0.02). In multivariate analysis, male sex (β=0.31), LVM (β=0.45) and body mass index (β=-0.37) were independently associated with the S-L summed voltage (R2 0.26, p<0.001). For Cornell summed voltage, only sex was an independent predictor (β=0.48, R2 0.22, p<001). Figure 1 Conclusion The R and S wave ECG amplitudes used in the two most common ECG criteria for LV hypertrophy were weakly related in the highest to lowest order to sex, LVM, body size and the distance from the LV to the chest wall in our college ATH.