Integrative Approach to Assessing the Complex of Correlations between Indicators of Physiological Functions
- PDF / 386,236 Bytes
- 4 Pages / 594 x 792 pts Page_size
- 32 Downloads / 156 Views
rative Approach to Assessing the Complex of Correlations between Indicators of Physiological Functions Y. N. Smolyakov1,2 and B. I. Kuznik1,2
Translated from Byulleten’ Eksperimental’noi Biologii i Meditsiny, Vol. 170, No. 8, pp. 256-259, August, 2020 Original article submitted April 11, 2020 We present the results of application of the combined correlation matrix method in some clinical studies. A practical algorithm is proposed for constructing a correlation matrix that compactly reflects a large number of interconnections for the situations when several diagnostic methods are used in an experimental clinical or preclinical trial. Several approaches to assessing and displaying the relationships are demonstrated for comparison. Key Words: correlation; matrix; preclinical trial; clinical trial
One of the most important approaches in the description of physiological state, under both normal and pathological conditions is assessment of the relationships between the individual parameters obtained in laboratory tests and clinical examinations. In mathematical terms, these problems are most often solved by correlation analysis methods [7,9-11]. In preclinical and clinical trials, the identification and thorough analysis of the relationships allows us to understand the subtle adaptive physiological mechanisms of the body functioning. However, the results of correlation analysis can be very extensive, which greatly complicates the understanding and interpretation of the data and, as a result, reduces credibility of the published results. The aim of this research was to develop a compact and informative technique for representing a large number of paired relationships and facilitating perception of the complex structure of dynamic physiological processes. This study combines the results of the application of the combined correlation matrix method in 1 Chita State Medical Academy, Ministry of Health of the Russian Federation, Chita, Russia; 2Innovative Clinic “Academy of Health”, Chita, Russia. Address for correspondence: [email protected]. Y. N. Smolyakov
some clinical studies [1-6], where the correlations between the content of plasma proteins (GDF11, GDF15, JAMA, and CCL11) and parameters of systolic (SBP), diastolic (DBP), and mean (MBP) BP were assessed. Statistical analysis and graphical constructions were performed using the specialized language R (http://cran.r-project.org) version 3.6.3. In the presented examples of correlation analysis, the Spearman’s rank correlation method was applied [13]. The significance level for mapping the relationships was 0.05. Single relationship between 2 quantitative variables is usually expressed by a linear correlation coefficient (r). Other forms of assessing the strength of correlations (nonlinear methods, contingency tables) can also be integrated into the proposed methodology. It is known that the correlation coefficient ranges from -1 (negative relationship) to 1 (positive relationship). In addition to the strength of the relationship (correlation coefficient), it is import
Data Loading...
