3DBODY.TECH 2021 - Extended abstract 21.36

M. C. Wong et al., "3D Optical Body Composition Accuracy across Subgroups of BMI and Race/Ethnicity", Proc. of 3DBODY.TECH 2021 - 12th Int. Conf. and Exh. on 3D Body Scanning and Processing Technologies, Lugano, Switzerland, 19-20 Oct. 2021, #36, https://doi.org/10.15221/21.36.


3D Optical Body Composition Accuracy across Subgroups of BMI and Race/Ethnicity


Michael C. WONG 1,2, Yong En LIU 2, Samantha F. KENNEDY 3, Nisa N. KELLY 2, Gertraud MASKARINEC 2, Andrea K GARBER 4, Julia MW. WONG 5, Cara B. EBBELING 5, David S. LUDWIG 5, Ethan WEISS 4, Sergi PUJADES 6, Steven B. HEYMSFIELD 3, John A. SHEPHERD 2

1 Graduate Program in Human Nutrition, University of Hawai'i Manoa, Honolulu, Hawaii, USA;
2 University of Hawai'i Cancer Center, Honolulu, Hawaii, USA;
3 Pennington Biomedical Research Center, Louisiana State University, Baton Rouge, Louisiana, USA;
4 University of California, San Francisco, California, USA;
5 Boston Children's Hospital, Boston, Massachusetts, USA;
6 Inria, Univ. Grenoble Alpes, CNRS, Grenoble INP, LJK, Grenoble, France


Three-dimensional optical (3DO) body scanning is emerging as an alternative for health assessments. Body composition models predicted from 3DO body shape have been shown to correlate strongly with criterion methods like dual-energy X-ray absorptiometry (DXA) in highly diverse sample sets. Further examination of 3DO in subgroups is needed to increase generalizability and establish wider clinical applicability. Therefore, the objective of this study was to evaluate the accuracy of these 3DO body composition models by subgroups of sex, body mass index (BMI), and Race/ethnicity.

Participants were recruited from the Shape Up! Adults, FB4, and Louisiana State University Athlete's Studies. BMI categories included underweight, normal, overweight, and obese. Race/ethnic groups included White, Black, Hispanic, Asian, and Native Hawaiian and other Pacific Islanders (NHOPI). Each participant received whole-body 3DO and DXA scans on a Fit3D Proscanner (Fit3D Inc., San Mateo, CA, USA) and Hologic Horizon/A or Discovery/A system (Hologic Inc., Marlborough, MA, USA), respectively. 3DO scans were templated with a 110,000-vertex mesh for standardization and reposed through Meshcapade (Meshcapade GmbH, Tubingen, Germany). Principal component (PC) analysis was performed on the 3DO scans to reduce the dimensionality of the data to explain the shape variance in the sample with minimal PCs. These PCs were used to create fat and lean models for whole-body, arms, legs, and torso as well as visceral adipose tissue (VAT) using stepwise forward linear regression with 5-fold cross-validation. 3DO body composition estimates were subtracted from DXA measures to obtain the difference. Students t-test of the differences in each subgroup were considered significant if the P-value was <0.05. Percent (%) mean differences were categorized low (< 5%), moderate (5–10%), and large (>10%). For this analysis, we present whole-body results and VAT.

In total, 723 participants aged 18-89 years were included in this study (381 females). Female and male total fat mass achieved a coefficient of determination (R2) of 0.95 and 0.94 and a root mean square error (RMSE) of 2.74 kg and 3.01 kg, respectively. Total fat mass ranged from 6.3 kg to 72.7 kg for females and 5 kg to 67.3 kg for males. Significant total fat mass differences were found in underweight females, Asian females, Black females, NHOPI females, and NHOPI males (% mean differences = 6.8%, 2.9%, -2.6%, -4.8%, and -9.3%, respectively; all P-values < 0.03). A significant VAT difference was also found in underweight males (%mean difference = -1.2%, P-value = 0.01).

3DO assessment appears to be comparable to DXA measures not only on the population level but also in most ethnic and weight subgroups except for underweight females and NHOPI males, possibly due to their low representation in our study sample. Using this overall model will provide accurate results in respects to DXA, but these two subgroups may need specific calibration.


Extended abstract: 2136wong-eabs.pdf
Proceedings: 3DBODY.TECH 2021, 19-20 Oct. 2021, Lugano, Switzerland
Paper id#: 36
DOI: 10.15221/21.36
Presentation video: 3DBodyTech2021_36_Wong.mp4

Copyright notice

© Hometrica Consulting - Dr. Nicola D'Apuzzo, Switzerland, hometrica.ch.
Reproduction of the proceedings or any parts thereof (excluding short quotations for the use in the preparation of reviews and technical and scientific papers) may be made only after obtaining the specific approval of the publisher. The papers appearing in the proceedings reflect the author's opinions. Their inclusion in these publications does not necessary constitute endorsement by the editor or by the publisher. Authors retain all rights to individual papers.

Note: click the + on the top left of the page to open/close the menu.