Abstract

The LeanScreenTM app uses photographs and touchscreen technology of an iPad or iPhone to estimate body composition using the Department of Defense (DoD) prediction equations that use cirumference measurements of the neck, abdomen, waist, and hips. The purpose of this study was to determine the validity and reliability of the LeanScreenTM app in 148 weight-stable adults (82 men, 66 women) who were normal weight, overweight, or obese as defined by body mass index. The percent body fat (%BF) of each subject was estimated during one visit using dual-energy x-ray absorptiometery (DXA) as the criterion measure, and three field methods: the LeanScreenTM app, manually measured circumferences, and an OMRON bioelectical impedance (BIA) device. The %BF of each subject was determined once using DXA. Each of two administrators assessed the %BF of each subject twice using the LeanScreenTM app, manually measured circumferences, and the OMRON BIA device. When using the LeanScreenTM app, administrators assessed body composition using photographs they had taken and the photographs taken by the other administrator. Validity was established by comparing estimates of %BF from the LeanScreenTM app, manually measured circumferences, and the OMRON BIA device to %BF values obtained from DXA. Inter- and intrarater reliability was determined using mutliple measurements taken by each of two administrators. The three field methods were compared to DXA using mixed model ANOVA and Bland-Altman analyses. Analysis of the data revealed that the LeanScreenTM app, manually measured circumferences, and the OMRON BIA device significantly underestimated (p < 0.05) the %BF determined by DXA by an average of -3.26 ± 3.57 %BF, -4.82 ± 3.45 %BF, and -8.45 ± 3.48 %BF, respectively. Limits of agreement (LOA) for the LeanScreenTM app (6.99 %BF), manually measured circumferences (6.76 %BF), and the OMRON BIA device (6.82 %BF) were large. Slopes of the line-of-best-fit through the data in the Bland-Altman plots indicate that bias of %BF estimates using the LeanScreenTM app (slope = 0.06; p = 0.008) and the OMRON BIA device (slope = 0.15; p < 0.0001) increased as %BF increased. For each method of assessment, minimal variance could be attributed to different administrators peforming the assessment and each administrator performing multiple assessments. All inter- and intrarater reliability coefficients of the LeanScreenTM app, manually measured circumferences, and OMRON BIA estimates of %BF exceeded 0.99. The results of this study indicate that all three field methods of body composition assessments were highly reliable, however, these field measures are not recommended for use in the assessments of %BF due to a significant bias and large limits of agreements.

Degree

College and Department

Life Sciences; Exercise Sciences

Rights

http://lib.byu.edu/about/copyright/