Ultrasound-Derived Skinfolds in Anthropometric Predictive Equations Overestimate Fat Mass: A Validation Study Using a Four-Component Model

Background: The evaluation of body composition is considered a key factor for assessing nutritional status. In several settings, ultrasound (US) has been used as a useful tool in nutritional practice by estimating body composition parameters, such as the whole-body fat mass (FM). The estimation of FM can be carried out by using predictive equations that generally require measurements of skinfold thickness, which can be measured directly via US imaging. The main aim of this study was to evaluate the validity of US-derived skinfolds within anthropometric equations for estimating whole-body FM. Methods: Skinfold thickness was measured in 37 active individuals (19 males, age 24.2 ± 4.3 years, and 18 females, age 25.3 ± 4.2 years) using both anthropometry and US. The skinfolds obtained from anthropometry and US were entered into Evans’ equation to estimate the FM and were validated against a four-component model (4C) as a reference. Results: The use of US-derived skinfolds within anthropometric equations resulted in an overestimation of FM (4.8%, p < 0.01). An agreement analysis between the FMs estimated with US-derived skinfolds and the 4C model revealed a concordance correlation coefficient of 0.33, 95% limits of agreements ranging from −3.4% to 0.6%, and a positive trend (r = 0.8; p < 0.01). Conclusions: The practice of doubling the US thickness to approximate skinfold thickness leads to an overestimation of FM by ~5%, and it should be avoided. This results in a lack of agreement with the 4C model at both the group and individual levels. New equations based on US measurements should be developed to enhance the accuracy of body composition evaluation and help optimize nutritional strategies.