Homeopathy and Heart Rate Variability: Clarification About Concerns and Issues

Dear Editor,

The interest generated post publication of the article on heart rate variability and blood flow variability is an encouraging signal to explore and adopt modern technological devices for homeopathic research. ¹ The authors would like to clarify the concerns and issues raised by the readers. ²

This was an exploratory study and therefore it was planned as single blind. Generally results of such studies are useful to provide a basis for planning studies in the future. The team explored various ideas on experimental methodology and considered that the same person would be a better control for the study of physiologic variability, which is more or less stationary within the person in a given time duration (unless modified by intervention or external stimulus). Taking into account that autonomic functions are largely not under voluntary control, the change, if significant as predefined, would be valid.

Identification and quantification of a peak in a given band reflects the existence of specific oscillations in the given system. Higher peak reflects higher specificity in the spectral studies, since smaller amplitude can also be noise. There are examples from electroencephalography studies, where peak frequency has been identified to determine oscillators. ^3,4 In the present study, our emphasis has been on addressing the question of specificity in the autonomic oscillator. Considering that both of the recordings were made in a single setting, the question of extraneous factor with respect to diurnal variation does not arise (i.e., the signal can be taken to be stationary).

We have observed changes in peak amplitude after giving the medicine. This method of analysis is based on scientific principles used in analysis of radio transmission and reception. There are two factors: sensitivity and specificity. Sensitivity refers to broadband, whereas specificity refers to narrow band; that is, for detecting a specific signal in the presence of noise, peak amplitude for a given frequency band is better. Such deviations were by design and not because of limitations of our processing and analyzing tools. Here we are proposing a new method based on scientific thought.

As per task force recommendations, ⁵ the very low frequency (VLF) band has been defined as being from 0.003 Hz to 0.04 Hz; the lowest frequency is 0.003 Hz with a time period of 333.3 seconds. The Nyquist criterion for sampling frequency is double that of the highest frequency contents in the signal. Extending the same argument for the lowest frequency, the minimum signal length needs to be 666.7 seconds. Further application of windowing function (Hanning, etc.) necessitates that the signal length be further increased 1.5 times. Therefore, to cover the entire band of VLF, the minimum recording time needs to be 1000 seconds (16.7 minutes). This is probably the reason why short-term variability (5 minutes) should not be used for analysis of the VLF band. We have used the part of the VLF band ranging from 0.02 to 0.04 Hz. Applying the above considerations, the minimum recording length needs to be 150 seconds, which is well below the recording length of 256 seconds used by us. We have followed standard practice for processing the data, which includes manual editing. Artifacts were replaced by interpolated values.