Dear Editor,
Let me express our gratitude for Dr Körblein's interest and deep analysis of our article. The authors have tried to answer every comment, that were very helpful to us.
A few words about the methods of mathematical statistics that we use for analysis. Initially, many of our analyzes used linear regression, Poisson regression, quadratic and other types of de novo congenital malformations frequency regression in newborns depending on the levels of radioactive and chemical contamination in different areas of the Bryansk region. However, these initial tests showed that these types of regressions did not approximate the available data satisfactorily: the standard deviation is large and subsequent predictions are unreliable. Therefore, we applied inversely proportional regression, which is more flexible and more closely matches the original data. Yearly data varies greatly from year to year, and this does not provide a good regression (with little deviation from the data). Therefore, we decided to take 3-year periods that smooth out these fluctuations.
The calculation of p-values (t-tests) in the table presented by Dr Körblein fully confirms our conclusions about significance of p-values. Although in our case it is not suitable for testing the hypothesis about differences in the incidence of congenital malformations, since the Student's test (t-tests) and also F-test can only be applied in the case of a normal distribution, but not in our article.
Since we obtained a non-normal distribution of the incidence of congenital malformations, we used the nonparametric Wilcoxon T-test to test the hypothesis about the differences in the incidence of congenital malformations.
Of course, Dr Körblein is absolutely right that we received a large discrepancy between our data obtained in the control region and the data of EUROCAT, and we will certainly take this into account and discuss it in our future studies. The proposed approach seems interesting to us, and we will try to work in this direction in the future. We express our gratitude for the link to the article by Calzolari et al. (2014), which provides data on the prevalence of multiple congenital anomalies according to the data of 19 European registries. Of course the EUROCAT data is extremely important. However, we want to emphasize that in this study, the goal was to identify differences in the population frequency of the prevalence of de novo (radiation-induced) congenital malformations between the areas of the Bryansk region, which significantly differ in the level of radiation, chemical and combined environmental contamination. The prevalence of congenital malformations was collected on the basis of reliable data - statistical materials from the medical and genetic consultation of the Bryansk Clinical Diagnostic Center. Information on radiation contamination and chemical pollution of the environment was also collected from official scientific sources and the authors have no reason not to trust these data. The data for the districts are characterized by relative uniformity; their volume is sufficient for the purposes of this article. Based on the collected material, the attempt was made to solve the goal – to assess the dynamics of the frequency de novo of congenital malformations in newborns of ecologically different territories of the Bryansk region. The results obtained are quite objective and will become the basis for furthermore detailed and in-depth study of the frequency and structure of congenital malformations in the Bryansk region with varying degrees of radioactive and chemical pollution, taking into account socio-economic and other factors. Therefore, we included in the discussion the main significant limitations of our research.
Concerning the assessment for the entire Bryansk region, Dr Körblein's remark is true and we have thought about it many times. However, it is practically impossible to do this, because it is difficult to classify part of regions as territories of certain type (the so-called intermediate areas in terms of the level of radioactive contamination and, especially, chemical pollution). If one consider entire Bryansk region, then the meaning of ecological zoning will be lost, since contamination boundaries will be "blurred".
As for the additive or synergistic effects of radioactive and combined contamination. In our article this effect was revealed only for multitype congebital malformations. For polydactyly, limb reduction defects and total congenital malformations no such effects were found.
We fully agree with Dr. Körblein that we should continue these research on more type of congenital malformations (for example, for all congenital malformations of strict accounting), and to compare our results with results on Ukraine and Belarus. We hope to do this in future.
Declarations
Author contribution statement
Anton V. Korsakov, Emilia V. Geger, Dmitry G. Lagerev, Leonid I. Pugach, Timothy A. Mousseau: Wrote the paper.
Competing interest statement
The authors declare no conflict of interest.