The effects of blood glucose changes on frequency-domain measures of HRV signal in type 1 diabetes

Abstract

The analysis of time duration between consecutive R waves of electrocardiogram (ECG) is a standard method to evaluate the variations in heart rate. The physiological literature reveals that blood glucose levels modulate the autonomic nervous system (ANS) activity and heart rate variability (HRV) is representative of the cardiovascular autonomic function. In the research described here, a pilot investigation was carried out to investigate the relationship between HRV signal measures derived from ECG and arterial blood glucose changes in a female subject with type 1 diabetes mellitus (T1DM) subject during normoglycemic and mildly hyperglycemic conditions. A CleveLabs BioCapture wireless device was used to acquire ECG signals from a 160 Kg, 59.6 year old female volunteer with type 1 diabetes. The PhysioToolkit Software was used to extract the HRV signal and the Kubios software package was deployed to perform comprehensive HRV signal analysis. This software has an easy-to-use graphical user interface that displays the HRV signal and provides three options to calculate: Time-domain, Frequency-domain and Nonlinear Dynamics parameters from raw HRV signals. In its Frequency-domain analysis section, it provides frequency bands such as VLF (Hz), LF (Hz), and HF (Hz), with LF/HF as an index that reflects the sympathovagal balance of the ANS. ECG data were acquired for 30 minutes during normoglycemic condition and for another 30 minutes during mildly hyperglycemic conditions, while blood glucose levels were measured manually by the subject using a glucometer every 5 minutes. ECG signal segments of 5 minute durations were then processed to extract HRV signals and these in turn were analyzed to provide frequency-domain measures. The results indicated that blood glucose changes were inversely related to LF/HF. For this dataset, it was observed that mean ± std of the LF/HF decreased from 6.0 ± 1.04 to 0.91 ± 0.17 when blood glucose levels increased from 156 ± 22 mg/dl to 202 ± 29 mg/dl. Further investigation is underway to recruit more diabetic subjects to acquire a large dataset and explore the relationships between different HRV signal parameters and blood glucose changes under different gylcemic conditions in a comprehensive way.

Effects of hyperglycemia on variability of RR, QT and corrected QT intervals in Type 1 diabetic patients

“…Mean RR and LF/HF ratio were also reduced but did not achieve statistical significance. This result concurs with other reports that described lower HRV measures in diabetic patients than in normal subjects [3,12,17]. Table II demonstrated that BGLs were inversely and significantly correlated with SDNN, RMSSD, pNN50 and LF and HF (p<0.05).…”

“…The power of LF, HF and VLF components is usually expressed in absolute units (ms 2 ). The ratio between LF and HF components (LF/HF ratio) is used to reflect the balance between sympathetic and parasympathetic parts of the autonomic system and hence shows the level of sympathetic dominance (stress) in a subject [17]. These time and frequency domains parameters are considered to be most physiologically linked to the autonomic nervous system.…”

Hyperglycemia Identification Using ECG in Deep Learning Era

“…Other works have shown that blood heart rate variability (HRV) can be modulated by blood glucose levels [ 27 ]; therefore, several studies investigating HRV and glucose levels have been developed. Amanipour et al [ 27 ] analyzed the HRV frequency domain components of a 59-year old diabetic female subject under normoglycemic and hyperglycemic conditions. They noticed a 6-fold decrease in the low frequency/high frequency ratio.…”

A non-invasive approach for the diagnosis of Type 2 diabetes using HRV parameters

“…The ratio of low frequency component to that of the high frequency component decreased when the blood glucose levels are changed from normoglycaemic to hypoglycaemic condition. This pilot study brought out only the brief relationship between the HRV spectral parameters and sugar levels during normoglycaemic and mildly hyperglycaemic conditions (Amanipour et al, 2012). In another study, the HRV in healthy individuals and diabetic patients is compared to find that a negative correlation exists between HbA1c values and the HRV parameters (Mirza and Laxmi, 2012).…”