Body fat percentage mathematical model based on a Croatian tennis players morphology

Sastav tijela važan je faktor koji utječe na izvedbu profesionalnih sportaša. Raznolikost sastava tkiva i različitost tipova tijela izravno mjerenje tjelesnog sastava čini izazovnim zadatkom. Skup podataka morfoloških obilježja hrvatskih tenisača upotrebljen je kako bi se modelirao odnos između morfoloških obilježja tenisača i tjelesne masnoće. Analizom skupa podataka otkrivena je značajna korelaciju između debljine kožnih nabora, dobi, spola i postotka tjelesne masti. Dodatno, PCA analiza otkrila je kritičnu dob u kojoj se muška i ženska populacija počinju diferencirati tijekom puberteta, pomažući time modeliranje trenažnog procesa sportaša juniora. PCA analiza omogućava i usporedbu pojedinca i populacije kako bi se rano uočila potencijalna odstupanja u razvoju. Modeli linearne regresije razvijeni su kako bi procjenjivali ukupan postotak potkožnog masnog tkiva, kao i parcijalne postotke za određene dijelove tijela. Razvijeni model za procjenu ukupnog postotka potkožnog masnog tkiva uspoređen je s do sada razvijenim metodama za procjenu istoga te su razlike statistički značajne. Rad naglašava značaj debljine kožnih nabora, dobi i spola u procjeni postotka masnoće u tijelu tenisača. Razvijeni modeli pružaju trenerima isplativu metodu za procjenu i praćenje sastava tijela korištenjem mjerenja kožnih nabora umjesto skupe laboratorijske opreme.Body composition is an important factor affecting the performance of professional athletes. The difference among tissue compositions and diversity of body types makes direct measurement of body composition a challenging task. Croatian tennis player morphology dataset was used to model the relationship between morphological characteristics of tennis players and body fat percentage. Dataset analysis revealed a significant correlation between the skin-fold thickness, age, gender and body fat percentage. Additionally, PCA analysis revealed a critical age at which male and female players begin to differentiate during puberty, thus helping to model the training process of junior athletes. PCA analysis also enables the comparison of individuals and populations in order to detect potential deviations in development. Linear regression models were developed to estimate total percentage of subcutaneous fat as well as partial percentages for specific body parts. The model developed for estimating the total percentage of subcutaneous fat tissue was compared against the already established methods for the same task, and the differences proved to be statistically significant. This work emphasizes the importance of the skin-fold thickness, age and gender in estimating the percentage of body fat in tennis players. The developed models provide trainers with a cost-effective method to assess and monitor body composition using skin-fold measurements instead of expensive laboratory equipment