CLUSTER ANALYSIS AS AN INTEGRATOR OF DIFFERENT TECHNIQUES FOR ASSESSING THE PHYSICAL DEVELOPMENT OF HEALTHY ADOLESCENTS

The influence of external factors can significantly deviate the distribution of somatotypes from the indicators of the general population, which makes researchers select the most suitable scheme of so-matotyping from the already known ones or construct their own. Applicants of military universities are a particular case of such samples, formed as a result of a complex of factors, including the work of military medical commissions. The aim was to use cluster analysis to develop an integrative classifica-tion for assessing the physical development of practically healthy adolescents. Materials and meth-ods. The object of the study was 280 applicants (234 boys and 46 girls), who were examined by spe-cialists in the medical commissions of military commissariats and were considered to be practically healthy. Anthropometric data were analyzed: body length, body weight, length of the torso, limbs, girth dimensions of the torso and limbs, as well as body component composition. Hierarchical cluster analysis by the Ward method and Kendall's correlation coefficient were used in the study. Results. By means of cluster analysis of the physical development indicators of the virtually healthy individuals, we identified a significant contribution of body mass index, weight, body length in the formation of three major groups, which we characterized as: the male macrosomatic phenotype, the male mesoso-matic phenotype, and the female phenotype. Based on the proportion of muscle and fat tissue, visceral fat, distal diameter of the forearm as an indicator of bone development, body mass index and thick-ness of the skin-fat fold of the anterior abdominal wall, the following subgroups were identified: male macrosomatic phenotype, with hyperstenic features, male macrosomatic phenotype, with normosthen-ic features, male mesosomatic phenotype with digestive features, male mesosomatic phenotype with mesotonic features, male mesosomatic muscle phenotype, female megalosomic athletic and subathletic phenotype, female stenoplastic phenotype. Conclusion. The used statistical approach allows to change the classical paradigm of searching for a universal methodology of somatotyping, to the creation of algorithms of compilation of well-known schemes of physical development evaluation in the selected population groups.