Heart rate variability shows significant correlations with the echocardiographic markers of the subclinical myocardial involvement in type 1 diabetes mellitus

Background Even in type 1 diabetes mellitus (T1DM) patients without manifest heart disease, subclinical myocardial involvement of the disease may be present. Besides, reduced heart rate variability (HRV) was reported as early marker of diabetes-related cardiac autonomic neuropathy. Nevertheless, no data are available whether impairment of the HRV reflects the severity of myocardial involvement in this disease. Purpose Our aim was to investigate the potential correlations between the echocardiographic markers of the subclinical myocardial involvement and the HRV parameters in asymptomatic T1DM patients. Methods Seventy T1DM patients (38.2 ± 11.7 years, 46 female, HbA1c: 7.6±1.3%, BMI: 24.8±4.2 kg/m²) were enrolled. Hypertension did not, but other known cardiovascular diseases served as exclusion criteria. In addition to the conventional and tissue Doppler measurements, myocardial mechanics was investigated with 2D speckle tracking technique. For HRV analysis, time-domain (SDNN [standard deviation of the normal-normal intervals] and rMSSD [root mean square differences of successive normal-normal intervals]), frequency-domain (VLF [very low-frequency component], LF [low-frequency component], HF [high-frequency component], Total Power of the spectrum and LF/HF ratio), and nonlinear (SD1 and SD2 reflecting the dispersion along minor and major axis of the fitted ellipse) indices were calculated. HRV parameters were obtained from the less noisy 5-minute segment of the 30-minutes ECG recording. Since HRV parameters did not show normal distribution, logarithmic transformation (ln) was performed before executing linear regression analysis. Results Mitral A (e.g. with ln SDNN: r=-0.470, p<0.001), average mitral annular e’ (e.g. with ln LF: r=0.500, p<0.001), mitral E/e’ (e. g. with ln SDNN: r=-0.373, p=0.002) as well as left atrial reservoir (e.g. with ln SDNN: r=0.322, p=0.007) and conduit strain (e.g. with ln SDNN: r=0.473; p<0.001) showed significant correlations with all HRV parameters, excluding ln LF/HF. Regarding the right heart, various significant correlations were found between tricuspid E (e. g. with ln SD1: r=0.376, p=0.001), tricuspid A (e.g. with ln rMSSD: r=-0.438, p<0.001), tricuspid annular e’ (e. g. with ln LF: r=0.253, p=0.037) as well as right atrial reservoir (e.g. with ln SDNN: r=0.327, p=0.007) and conduit strain (e.g. with ln SDNN: r=0.465, p<0.001), and the HRV parameters. Conclusions HRV parameters show significant correlations with the echocardiographic markers of the early, subclinical myocardial involvement in T1DM population. Thus, measuring HRV may be a useful screening tool for identifying patients with a high suspicion of myocardial involvement in asymptomatic T1DM population.