RELIABILITY OF BODY COMPOSITION MEASURED USING A SMARTPHONE APPLICATION AND DIFFERENT CAMERA RESOLUTIONS
Mary E. Lovelady, Casey J. Metoyer, Katherine Sullivan, Jacob Broeckel, Michael R. Esco, FACSM, Michael V. Fedewa. University Of Alabama, Tuscaloosa, AL.
BACKGROUND: Many traditional laboratory methods of measuring body fat percentage (%Fat) are inaccessible to healthcare professionals due to the cost, complexity, time, and portability. Recently, a new smartphone application was developed that allows for an accurate estimate of %Fat by analyzing a single 2-dimensional digital image. Although the validity of the application has been previously examined, the reliability across different camera resolutions has not been thoroughly tested. PURPOSE: The aim of this study was to evaluate the reliability of %Fat estimates from digital images captured using devices with different megapixel cameras. METHODS: A convenience sample of adult participants was recruited for the study (n=12, 83.33% female, 83.33% Caucasian, 31.25±10.49 yrs., 69.44±11.77 kg/m2). Age, gender, and race/ethnicity were assessed via self-report. Height was measured to the nearest 0.1 cm using a stadiometer (SECA 213, Seca Ltd., Hamburg, Germany). Weight was measured to the nearest 0.1 kg using a calibrated digital scale (Tanita BWB-800, Tanita Corporation, Tokyo, Japan). A full-body digital image was taken from the posterior view, with participants standing in front of a white background, using a 12-megapixel iPhone 12 (Apple Inc., Cupertino, CA) (%Fat12mp) and an 8-megapixel iPad Air 2 (Apple Inc., Cupertino, CA) (%Fat8mp). %Fat was derived using an automated smartphone application and a proprietary algorithm (made Health and Fitness, LLC, Birmingham, AL). A paired samples t-test was used to examine potential mean differences between %Fat12mp and %Fat8mp. The reliability was also measured using Pearson’s r, and described as weak, moderate, strong, or near-perfect (r=0.2, 0.5, 0.8, or 0.9 respectively). Data are presented as mean±standard deviation, with an alpha level set to p<0.05. RESULTS: No significant mean differences in %Fat were observed between %Fat12mp and %Fat8mp (26.92±4.96 %Fat and 27.16±3.08 %Fat, respectively; p=0.37). Near-perfect correlations were observed between %Fat12mp and %Fat8mp (r=0.99, p<0.001). DISCUSSION: Based on the results of this study, the smartphone application provides a reliable estimate of %Fat across devices with different megapixel cameras. Future studies should explore other conditions, including different lighting, different color backgrounds, and other devices, as well as within a larger more diverse sample.
Lovelady, ME; Metoyer, CJ; Sullivan, K; Broeckel, J; Esco, FACSM, MR; and Fedewa, MV
"RELIABILITY OF BODY COMPOSITION MEASURED USING A SMARTPHONE APPLICATION AND DIFFERENT CAMERA RESOLUTIONS,"
International Journal of Exercise Science: Conference Proceedings: Vol. 16:
2, Article 94.
Available at: https://digitalcommons.wku.edu/ijesab/vol16/iss2/94