Influence of geomagnetic activity and earth weather changes on heart rate and blood pressure in young and healthy population

Abstract

There are many references in the literature related to connection between the space weather and the state of human organism. The search of external factors influence on humans is a multi-factor problem and it is well known that humans have a meteo-sensitivity. A direct problem of finding the earth weather conditions, under which the space weather manifests itself most strongly, is discussed in the present work for the first time in the helio-biology. From a formal point of view, this problem requires identification of subset (magnetobiotropic region) in three-dimensional earth’s weather parameters such as pressure, temperature, and humidity, corresponding to the days when the human body is the most sensitive to changes in the geomagnetic field variations and when it reacts by statistically significant increase (or decrease) of a particular physiological parameter. This formulation defines the optimization of the problem, and the solution of the latter is not possible without the involvement of powerful metaheuristic methods of searching. Using the algorithm of differential evolution, we prove the existence of magnetobiotropic regions in the earth’s weather parameters, which exhibit magneto-sensitivity of systolic, diastolic blood pressure, and heart rate of healthy young subjects for three weather areas (combinations of atmospheric temperature, pressure, and humidity). The maximum value of the correlation confidence for the measurements attributable to the days of the weather conditions that fall into each of three magnetobiotropic areas is an order of 0.006, that is almost 10 times less than the confidence, equal to 0.05, accepted in many helio-biological researches.

Appendix

We have the measurements of physiological parameters of 197 people (70 women and 127 men). Let us show that if there is a dependency of the physiological parameters from Kp-index in the general population, it must be at least in one of the sub-populations. The proof will be done by contradiction.

Let us assume nonstationary mixture of samples, consisting of men and women, while in each of the subsamples relationship between the studied parameters is missing. It is easy to show that the mixing of these subsamples will not cause a statistically significant relationship to the resultant mixture. It follows that at least one of subsamples dependency is present.

Let us denote Kp-index as ξ, and physiological parameter as η. The two-dimensional distribution of this pair of options for women will be denoted as the p_ξη (x, y), and for women as q_ξη (x, y). Under the assumption of the lack of correlation between ξ and η, we can write the system of equations:

$$ \left\{\begin{array}{c}\hfill {\displaystyle \iint }{p}_{\xi \eta}\left(x,y\right)\left[x-{\overline{x}}_p\right]\left[y-{\overline{y}}_p\right]=0\hfill \\ {}\hfill {\displaystyle \iint }{q}_{\xi \eta}\left(x,y\right)\left[x-{\overline{x}}_q\right]\left[y-{\overline{y}}_q\right]=0\hfill \end{array}\right. $$

Here, ${\overline{x}}_p$, ${\overline{y}}_p$, ${\overline{x}}_q$, and ${\overline{y}}_q$ indicate the average values of the parameters ξ and η on the distributions p_ξη and q_ξη, respectively. Due to the similarity of the distribution of Kp-index for the subsamples of men and women: ${\overline{x}}_p={\overline{x}}_q=\overline{x}$

Let the proportion of women in the resulting sample is α, and men is β, α + β = 1. As follows from the formula, the total probability distribution of the resulting mixture will be αp _ξη (x, y) + βq _ξη (x, y), while the mean values of ξ and η on the resulting distribution will be correspondingly

$$ {\displaystyle \iint}\left[\alpha {p}_{\xi \eta}\left(x,y\right)+\beta {q}_{\xi \eta}\left(x,y\right)\right]\mathrm{xdxdy}=\alpha {\overline{x}}_p+\beta {\overline{x}}_q=\left(\alpha +\beta \right)\overline{x}=\overline{x} $$

and

$$ {\displaystyle \iint}\left[\alpha {p}_{\xi \eta}\left(x,y\right)+\beta {q}_{\xi \eta}\left(x,y\right)\right]\mathrm{ydxdy}=\alpha {\overline{y}}_p+\beta {\overline{y}}_q $$

Covariance parameters ξ and η on the new distribution is:

$$ {\displaystyle \iint}\left[\alpha {p}_{\xi \eta}\left(x,y\right)+\beta {q}_{\xi \eta}\left(x,y\right)\right]\left[x-\overline{x}\right]\left[y-\alpha {\overline{y}}_p-\beta {\overline{y}}_q\right]\mathrm{dxdy} $$

This integral splits into a sum:

$$ \begin{array}{c}\hfill \alpha {\displaystyle \iint }{p}_{\xi \eta}\left(x,y\right)\left[x-\overline{x}\right]\left[y-\alpha {\overline{y}}_p-\beta {\overline{y}}_q\right]\mathrm{dxdy}+\hfill \\ {}\hfill +\beta {\displaystyle \iint }{q}_{\xi \eta}\left(x,y\right)\left[x-\overline{x}\right]\left[y-\alpha {\overline{y}}_p-\beta {\overline{y}}_q\right]\mathrm{dxdy}\hfill \end{array} $$

Let us prove the vanishing of the first sum of the components (the vanishing of the second one can be proved similarly):

$$ \begin{array}{l}{\displaystyle \iint {p}_{\xi \eta }}\left(x,\overline{y}\right)\left[x-\overline{x}\right]\left[y-a\overline{y_p+}\beta {\overline{y}}_q\right]\mathrm{dxdy}={\displaystyle \iint {p}_{\xi \eta }}\left(x,y\right)\left[x-\overline{x}\right]\left[y-{\overline{y}}_p+{\overline{y}}_p-a\overline{y_p-\beta \overline{y}}\right]\mathrm{dxdy}=\\ {}\kern1.5em ={\displaystyle \iint {p}_{\xi \eta }}\left(x,y\right)\left[x-\overline{x}\right]\left[y-\overline{y_p}\right]\mathrm{dxdy}+{\displaystyle \iint {p}_{\xi \eta }}\left(x,y\right)\left[x-\overline{x}\right]\left[{\overline{y}}_p-a{\overline{y}}_p-\beta {\overline{y}}_q\right]\mathrm{dxdy}=\\ {}\kern1em =0+\left[{\overline{y}}_p-a{\overline{y}}_p-\beta {\overline{y}}_q\right]{\displaystyle \iint {p}_{\xi \eta }}\left(x,y\right)\left[x-\overline{x}\right]\mathrm{dxdy}=0+\left[\overline{y_p}-a{\overline{y}}_q-\beta {\overline{y}}_q\right]\left(\overline{x}-\overline{x}\right)=0\end{array} $$

Thus, we have proved that the absence of correlation with respect to distributions in both subsamples leads to absence of correlation in the resultant mixture, therefore we can come to the fact that in our case at least in one of the sub-samples.