Heart rate variability in diabetic children: sensitivity of the time- and frequency-domain methods

Abstract

Heart rate variability (HRV) is a noninvasive index of the neural activity of the heart. Although also influenced by the sympathetic activity of the heart, HRV is essentially determined by the vagal stimulation of the heart. Several HRV abnormalities have been described in adults with diabetes mellitus. However, there are few data on HRV in children with diabetes mellitus. In the present study, HRV was assessed in seven healthy children, 10 diabetic children with good glycemic control and 11 diabetic children with poor glycemic control. All had normal standard cardiac autonomic function tests, obtained from 24-h Holter tapes. HRV was measured by calculating six time-domain (mean R-R interval (RR), standard deviation of the R-R interval [SDRR], standard deviation of the mean of 288 R-R intervals [SDANN], the mean of the 288 standard deviations computed for each 5-rain period [SD], percentage of differences of adjacent R-R intervals of >50 msec for the entire 24 h [pNN50], and the root mean square of successive differences [rMSSD]) and four frequency-domain (low frequency [LF], high frequency [HF], total heart rate power spectra, and LF/HF ratio) indexes. SD, pNN50, rMSSD, LF, HF and total heart rate power spectra were markedly and significantly reduced in diabetic children with poor metabolic control. The 24-h variation of low- and highfrequency components of heart rate power spectra of the latter children had a different shape. Thus, diabetic children with poor metabolic control (elevated HbAlc and B2M levels) have a low HRV compared to those diabetic children with good control and healthy children. These results can be interpreted as evidence of cardiac autonomic neuropathy in diabetic children with asymptomatic diabetic autonomic neuropathy.