Effect of ramadan fasting on the sympathovagal balance through a study of heart rate variability

II.3.1. In the time domain

Time domain analysis is a simpler analysis than spectral analysis. Measurements in the time domain are produced from arithmetic calculations. There are two classes: on the one hand, the measurements derived directly from the normal-to-normal intervals between two beats and, on the other hand, the measurements derived from the differences of the normal-to-normal intervals themselves, among the parameters which can be measured by the analysis in the field of time:

· NN 50: number of successive RR intervals greater than 50 ms.

· PNN50: NN50 divided by the total number of intervals that expresses the high frequency variability mainly of modulated parasympathetic origin.

· RMSSD: square root of the squared differences of the successive RR intervals (the squared root of the mean of the sum of the squares of differences between adjacent NN intervals) which also expresses the high frequency variability mainly of parasympathetic origin, modulated by the breathing. This measurement is preferable to pNN50 and NN50.

· SDNN: (standard deviation of the RR interval over the entire recording period, standard deviation of all NN intervals) which gives information on the overall variability.

25

These indices are therefore a non-invasive method for studying the cardiac response to stimulation of the autonomic nervous system. They constitute a global approach to the influence of the autonomic nervous system. However, some methodological precautions should be emphasized. Many of these clues depend on the length of the recording. It is therefore necessary to standardize this length in order to be able to compare these different parameters. Consequently, it is imperative to only compare these parameters for an identical recording length (Jourdan.G, 2008).

II.3.2. In the frequency domain

In recent years, the spectral analysis of cardiac variability, based on the analysis of variations of RR intervals, has become the reference tool for the study of the dynamic interactions between parasympathetic and sympathetic controls (Malliani et al. 1991). Spectral analysis then breaks down a complex signal like heart rate into its constituents of frequency and quantifies the relative power of these components (Jourdan.G, 2008).After mathematical processing, a periodic signal of any shape (such as the heart rate, for example) appears in fact as the superposition of a sum of sinusoids or elementary oscillations. The fast Fourier transform allows the mathematical decomposition of a complex record into its constituent or elementary elements without loss of information. Each elementary sinusoid is mathematically defined by its amplitude and frequency. The set of sinusoids then constitutes the spectrum. The resulting graph shows on the abscissa, a frequency scale (in hertz, Hz) and on the y-axis, an amplitude scale. It allows the study of different oscillations of specific frequencies. In humans, the spectrum of the heart rate ranges from 0 to 0.4 Hz and can be divided into 3 areas of interest (on a recording of short duration, 2 to 5 minutes) or in 4 areas of interest (on a long-term recording, 24 hours) (Anonymous, 1996).

The parameters that can be calculated from the spectral analysis:

· Total power (ms²): Normal-to-normal interval variance of the entire record.

· Ultrafast frequencies (ULF): from 0.0001 to 0.003 Hz (only if 24 hours recording).

· Very low frequencies (VLF): from 0.003 to 0.04 Hz.

· Low Frequencies (LF): 0.04 to 0.15 Hz Oscillation in this frequency band is known as Traube-Hering waves.

· High frequencies (HF): 0.15 to 0.4 Hz. The oscillation in this frequency band is known as the Mayer wave.

·

26

VLF (ms²): Power in very low frequencies.

· LF (ms²): Power in the low frequencies.

· HF (ms²): Power in high frequencies.

LF and HF can also be in so-called normalized values, which corresponds to the power of the frequency band considered divided by the total power of the spectrum less VLF:

· HF (normalized): HF nu = 100 X HF / (HF + LF).

· LF (normalized): LF nu = 100 X LF / (HF + LF). The values thus standardized and the LF / HF ratio then make it possible to quantify, albeit in a simplified way, the sympathetic and vagal contribution to the variability of the heart rate (Neto et al., 2005).