Short- and long-term reproducibility of heart rate variability in patients with long-standing type I diabetes mellitus.

Heart rate variability (HRV) has been used to assess cardiac autonomic function noninvasively, understand the pathophysiologic mechanisms of heart disease, evaluate therapy, and assess long-term prognosis. We examined both the short- and long-term reproducibility of the time and frequency domain HRV parameters in 23 type I diabetics over a 12-month interval. Entry criteria included juvenile onset diabetes before age 35 years, >24-year duration of diabetes, diabetes difficult to control, and albuminuria. Standardized noninvasive autonomic testing and 24-hour ambulatory electrocardiographic recordings were obtained. Fifteen men and 8 women (mean age 36.7 years) were enrolled. Fifty-three percent of the men and 75% of the women were smokers, and women had higher cholesterol than men. All HRV parameters were markedly decreased when compared with normal persons. Using Pearson correlation, the time domain indicators of parasympathetic activity demonstrated very strong correlations at 3 and 6 months compared with baseline, with good correlations at 1 year. The average SD of all 5-minute RR intervals maintained a very strong correlation for the entire year (r >0.94). In the frequency domain, the measures of parasympathetic and sympathetic activity maintained a solid correlation for the entire study period. Reproducibility of HRV was also examined using repeated-measures analysis of variance. The time and frequency domain parameters demonstrated very little variation over the study period of 12 months. Thus, our investigation demonstrated that HRV in long-term diabetics using 24-hour ambulatory recordings is abnormal and reproducible over a 12-month interval; very little variation in all HRV parameters, especially in parameters of parasympathetic activity, occurred during the study period.