Polfuntikova Anastasia Viktorovna

Introduction. The degree of comparability of body composition assessment results using bioimpedance analysis and anthropometry, their combined use, and their interchangeability remain controversial. The aim of this study was to evaluate the relationships, level of bias, and consistency of time courses of body composition indices obtained repeatedly using bioimpedance analysis and anthropometry in elite athletes.

Materials and Methods. Study participants included 18 athletes (12 males, 6 females) aged 19–23 years, specializing in fire and rescue sports. The following were used: an ABC-01 Medass body water sector analyzer (Russia); an anthropometer, medical scales, a measuring tape, and a Lange caliper (Cambridge, Maryland). Muscle and fat mass were calculated using the ABC-03612 basic body composition assessment program and the Matiegka scheme (1921). Five repeated measurements were taken at intervals of 35–40 days; absolute and relative muscle and fat mass were estimated. The paired Wilcoxon test, Spearman's rank coefficients, intraclass correlation coefficient (ICC2), Bland–Altman analysis for levels and interassay changes, and linear mixed-effects models (LMM) were used.

Results and discussion. In the athlete group, BIA muscle mass values were systematically lower and fat mass values were systematically higher compared to anthropometric results. Differences were significant at all points (p = 0.002 in male, p = 0.028 in female). A strong correlation (0.63–0.93) was found between absolute values; ICC2, which takes into account systematic bias and intermethod variance, was significantly lower (0.05–0.65). Proportional and significant systematic bias was found for most parameters. The methods consistently reflect group time trends; their interchangeability is limited in individual longitudinal monitoring.

Conclusion. The results highlight the need to assess intermethod agreement by taking into account intermethod bias, its robustness, the possible proportional component, and differences between group and individual levels of analysis.

Funding. The study was carried out within the framework of the state assignment of the Federal Scientific Center of Physical Culture No. 777-00021-26 (topic No. 001-25/3).