| The measurement of heart rate variability (HRV), the oscillations in the interval between consecutive heart beats, is a noninvasive method to assess cardiac autonomic activity quantitatively. Low, or impaired, HRV is associated with morbidity and mortality. While there has been recent interest in using recordings from the standard twelve-lead electrocardiogram to capture HRV, there is little information on the validity and repeatability of these measures, or of measures derived from longer two- or six-minute records. Furthermore, most previous studies have only measured HRV at one point in time, and thus have had difficulty addressing the temporal sequence between HRV and various impairments of metabolism and homeostasis. The aim of this dissertation was to assess the repeatability of HRV measures, to describe the change in HRV over the course of nine years, and to investigate the influence of hypertension, blood pressure, diabetes, and pre-diabetic metabolic impairment on the nine-year change in HRV.; In general, the repeatability of HRV measures derived from two- and six-minutes records was quite high, with intraclass correlation coefficients ≥0.70 for six-minute measures and time domain two-minute measures, and ≥0.50 for frequency domain two-minute measures. Although HRV measures from ten-second records had lower repeatability, taking the mean from multiple records increased the repeatability considerably.; Over nine years, HRV decreased and R-R interval increased. Analyses revealed cross-sectional association between blood pressure and HRV, and HRV predicted incident hypertension. The HRV of normotensives and hypertensives converged over nine years, suggesting that while HRV appears to be predictive of hypertension, hypertensives may experience a normalization of HRV with time.; There were cross-sectional associations between lower HRV and diabetes and elevated fasting insulin. After adjustment for baseline HRV, a greater decrease in HRV was observed among diabetics. However, neither baseline fasting glucose or fasting insulin levels led to a greater decline in HRV. Most, but not all, of these findings are consistent with decreases in autonomic function at early stages of pre-diabetic metabolic impairment, followed by a progressive worsening of autonomic function. Additional experimental, clinical, and epidemiologic studies are needed to further elucidate the interrelationships between diabetes, hypertension, and HRV. |
