Trial Outcomes & Findings for Evaluation of the Accuracy of a Computer Vision-based Tool for Assessment of Total Body Fat Percentage (NCT NCT04854421)
NCT ID: NCT04854421
Last Updated: 2024-08-22
Results Overview
Comparison of Body Fat Percentage estimates to DXA references classified by Body Mass Index
Recruitment status
COMPLETED
Target enrollment
138 participants
Primary outcome timeframe
within a single, 3-hour evaluation
Results posted on
2024-08-22
Participant Flow
Participants were recruited from 8/12/2020 through 10/21/2020 to undergo body composition assessments at one of two locations, Pennington Biomedical Research Center (PBRC) in Baton Rouge, Louisiana or the Weight Center of Massachusetts General Hospital (MGH) in Boston Massachusetts. Participants assessed at PBRC were recruited from the surrounding community using advertisements promoted by the facility's Recruitment Department. Those assessed at MGH were ongoing patients seen at the center.
Participant milestones
| Measure |
Male Participants
Community dwelling and have no life-threatening conditions or diseases that would alter body composition from what is typical for their age, sex, ethnicity, and BMI.
|
Female Participants
Community dwelling and have no life-threatening conditions or diseases that would alter body composition from what is typical for their age, sex, ethnicity, and BMI.
|
|---|---|---|
|
Overall Study
STARTED
|
54
|
84
|
|
Overall Study
COMPLETED
|
52
|
82
|
|
Overall Study
NOT COMPLETED
|
2
|
2
|
Reasons for withdrawal
| Measure |
Male Participants
Community dwelling and have no life-threatening conditions or diseases that would alter body composition from what is typical for their age, sex, ethnicity, and BMI.
|
Female Participants
Community dwelling and have no life-threatening conditions or diseases that would alter body composition from what is typical for their age, sex, ethnicity, and BMI.
|
|---|---|---|
|
Overall Study
Removed from the final analysis due to poor image quality.
|
2
|
2
|
Baseline Characteristics
Evaluation of the Accuracy of a Computer Vision-based Tool for Assessment of Total Body Fat Percentage
Baseline characteristics by cohort
| Measure |
Male Participants
n=52 Participants
Community dwelling and have no life-threatening conditions or diseases that would alter body composition from what is typical for their age, sex, ethnicity, and BMI.
|
Female Participants
n=82 Participants
Community dwelling and have no life-threatening conditions or diseases that would alter body composition from what is typical for their age, sex, ethnicity, and BMI.
|
Total
n=134 Participants
Total of all reporting groups
|
|---|---|---|---|
|
Age, Continuous
|
43.1 Years
STANDARD_DEVIATION 14.4 • n=39 Participants
|
43 Years
STANDARD_DEVIATION 14.8 • n=41 Participants
|
43 Years
STANDARD_DEVIATION 14.7 • n=35 Participants
|
|
Sex: Female, Male
Female
|
0 Participants
n=39 Participants
|
82 Participants
n=41 Participants
|
82 Participants
n=35 Participants
|
|
Sex: Female, Male
Male
|
52 Participants
n=39 Participants
|
0 Participants
n=41 Participants
|
52 Participants
n=35 Participants
|
|
Ethnicity (NIH/OMB)
Hispanic or Latino
|
2 Participants
n=39 Participants
|
2 Participants
n=41 Participants
|
4 Participants
n=35 Participants
|
|
Ethnicity (NIH/OMB)
Not Hispanic or Latino
|
47 Participants
n=39 Participants
|
76 Participants
n=41 Participants
|
123 Participants
n=35 Participants
|
|
Ethnicity (NIH/OMB)
Unknown or Not Reported
|
3 Participants
n=39 Participants
|
4 Participants
n=41 Participants
|
7 Participants
n=35 Participants
|
|
Race (NIH/OMB)
American Indian or Alaska Native
|
0 Participants
n=39 Participants
|
1 Participants
n=41 Participants
|
1 Participants
n=35 Participants
|
|
Race (NIH/OMB)
Asian
|
5 Participants
n=39 Participants
|
4 Participants
n=41 Participants
|
9 Participants
n=35 Participants
|
|
Race (NIH/OMB)
Native Hawaiian or Other Pacific Islander
|
0 Participants
n=39 Participants
|
0 Participants
n=41 Participants
|
0 Participants
n=35 Participants
|
|
Race (NIH/OMB)
Black or African American
|
11 Participants
n=39 Participants
|
21 Participants
n=41 Participants
|
32 Participants
n=35 Participants
|
|
Race (NIH/OMB)
White
|
31 Participants
n=39 Participants
|
50 Participants
n=41 Participants
|
81 Participants
n=35 Participants
|
|
Race (NIH/OMB)
More than one race
|
2 Participants
n=39 Participants
|
2 Participants
n=41 Participants
|
4 Participants
n=35 Participants
|
|
Race (NIH/OMB)
Unknown or Not Reported
|
3 Participants
n=39 Participants
|
4 Participants
n=41 Participants
|
7 Participants
n=35 Participants
|
|
Height
|
175.6 Centimeters
STANDARD_DEVIATION 8.7 • n=39 Participants
|
163.0 Centimeters
STANDARD_DEVIATION 7.6 • n=41 Participants
|
167.8 Centimeters
STANDARD_DEVIATION 10.2 • n=35 Participants
|
|
Weight
|
92.8 Kilograms
STANDARD_DEVIATION 19.7 • n=39 Participants
|
78.3 Kilograms
STANDARD_DEVIATION 19.4 • n=41 Participants
|
84.0 Kilograms
STANDARD_DEVIATION 20.8 • n=35 Participants
|
|
Body Mass Index (BMI)
|
30.0 Kilogram per square meter
STANDARD_DEVIATION 5.6 • n=39 Participants
|
29.5 Kilogram per square meter
STANDARD_DEVIATION 7.0 • n=41 Participants
|
29.7 Kilogram per square meter
STANDARD_DEVIATION 6.5 • n=35 Participants
|
|
Waist Circumference
|
102.2 Centimeters
STANDARD_DEVIATION 14.3 • n=39 Participants
|
92.2 Centimeters
STANDARD_DEVIATION 16.3 • n=41 Participants
|
98.5 Centimeters
STANDARD_DEVIATION 15.7 • n=35 Participants
|
|
Hip Circumference
|
108.2 Centimeters
STANDARD_DEVIATION 11.5 • n=39 Participants
|
110.1 Centimeters
STANDARD_DEVIATION 14.3 • n=41 Participants
|
109.4 Centimeters
STANDARD_DEVIATION 13.2 • n=35 Participants
|
|
Waist-to-hip Ratio
|
0.94 Ratio
STANDARD_DEVIATION 0.08 • n=39 Participants
|
0.87 Ratio
STANDARD_DEVIATION 0.07 • n=41 Participants
|
0.90 Ratio
STANDARD_DEVIATION 0.08 • n=35 Participants
|
|
DXA %BF
|
28.6 Percentage
STANDARD_DEVIATION 6.4 • n=39 Participants
|
39.4 Percentage
STANDARD_DEVIATION 7.2 • n=41 Participants
|
35.2 Percentage
STANDARD_DEVIATION 8.7 • n=35 Participants
|
PRIMARY outcome
Timeframe: within a single, 3-hour evaluationPopulation: A total of 406 adults were screened for this study. Of those,199 met all inclusion and exclusion criteria and were considered eligible. 138 participants were enrolled into the clinical study and 134 participants (64 from MGH and 70 from PBRC) were included in the final analysis; four participants (2.9%) were removed from the final analysis due to poor image quality. Participants at PBRC underwent ADP (N = 70). Note that DXA images were never used, only their derived Body Fat percentage.
Comparison of Body Fat Percentage estimates to DXA references classified by Body Mass Index
Outcome measures
| Measure |
Smartphone Visual Body Composition (VBC)
n=134 Participants
VBC: Body Composition was measured using the VBC system that takes under 60 seconds per scan. Image captured for VBC included taking a series of four photos using a smartphone camera.
|
FitBit Aria 2 (cBIA 1)
n=123 Participants
Three consumer weight scales capable of BIA-based body composition analysis were included in the protocol: FitBit Aria 2 (Fitbit, San Francisco, CA); Tanita BF-684W (Tanita, Tokyo, Japan); and Renpho ES-24M-W/B (Joicom Corporation, Anaheim, CA). These scales are designated as cBIA 1, cBIA 2, and cBIA 3.(Consumer Bio-Impedance Analysis)
|
Tanita BF-684W (cBIA 2)
n=131 Participants
Three consumer weight scales capable of BIA-based body composition analysis were included in the protocol: FitBit Aria 2 (Fitbit, San Francisco, CA); Tanita BF-684W (Tanita, Tokyo, Japan); and Renpho ES-24M-W/B (Joicom Corporation, Anaheim, CA). These scales are designated as cBIA 1, cBIA 2, and cBIA 3.(Consumer Bio-Impedance Analysis)
|
Renpho ES-24M-W/B (cBIA 3)
n=131 Participants
Three consumer weight scales capable of BIA-based body composition analysis were included in the protocol: FitBit Aria 2 (Fitbit, San Francisco, CA); Tanita BF-684W (Tanita, Tokyo, Japan); and Renpho ES-24M-W/B (Joicom Corporation, Anaheim, CA). These scales are designated as cBIA 1, cBIA 2, and cBIA 3.(Consumer Bio-Impedance Analysis)
|
InBody S10 (pBIA1)
n=70 Participants
All participants also underwent professional BIA (pBIA) at PBRC with an InBody S10 (InBody Co., Seoul, Korea) and at MGH with a RJL system (Quantum IV, RJL Systems, Clinton Township, MI, USA.). The RJL system uses a single frequency (50 Khz) and four gel adhesive electrodes. Instead, InBody is multi-frequency and uses contact electrodes. InBody and RJL are designated as pBIA1 and pBIA 2, respectively, in the sections that follow and were analyzed separately. (Professional Bio-Impedance Analysis).
|
Quantum IV (pBIA 2)
n=64 Participants
All participants also underwent professional BIA (pBIA) at PBRC with an InBody S10 (InBody Co., Seoul, Korea) and at MGH with a RJL system (Quantum IV, RJL Systems, Clinton Township, MI, USA.). The RJL system uses a single frequency (50 Khz) and four gel adhesive electrodes. Instead, InBody is multi-frequency and uses contact electrodes. InBody and RJL are designated as pBIA1 and pBIA 2, respectively, in the sections that follow and were analyzed separately. (Professional Bio-Impedance Analysis).
|
BOD POD ADP
n=70 Participants
Participants who were evaluated at PBRC (N = 70) also had %BF assessed with the BOD POD Air displacement plethysmography (ADP) device (BodPod Gold Standard Body Composition Tracking System, COSMED, Rome, Italy).
|
|---|---|---|---|---|---|---|---|
|
Comparison of Body Fat Percentage Estimates to DXA References Classified by Body Mass Index
BMI < 25
|
2.50 Total Body Fat Percentage
Standard Deviation 1.8
|
6.36 Total Body Fat Percentage
Standard Deviation 3.5
|
5.75 Total Body Fat Percentage
Standard Deviation 3.3
|
7.66 Total Body Fat Percentage
Standard Deviation 4.6
|
4.29 Total Body Fat Percentage
Standard Deviation 2.2
|
4.41 Total Body Fat Percentage
Standard Deviation 2.8
|
3.31 Total Body Fat Percentage
Standard Deviation 2.2
|
|
Comparison of Body Fat Percentage Estimates to DXA References Classified by Body Mass Index
BMI 25-29.9
|
1.90 Total Body Fat Percentage
Standard Deviation 1.4
|
3.02 Total Body Fat Percentage
Standard Deviation 2.3
|
5.04 Total Body Fat Percentage
Standard Deviation 14.3
|
6.00 Total Body Fat Percentage
Standard Deviation 4.9
|
2.49 Total Body Fat Percentage
Standard Deviation 1.4
|
3.82 Total Body Fat Percentage
Standard Deviation 3.1
|
2.30 Total Body Fat Percentage
Standard Deviation 1.9
|
|
Comparison of Body Fat Percentage Estimates to DXA References Classified by Body Mass Index
BMI > 30
|
2.16 Total Body Fat Percentage
Standard Deviation 1.4
|
4.44 Total Body Fat Percentage
Standard Deviation 1.8
|
4.31 Total Body Fat Percentage
Standard Deviation 3.6
|
4.60 Total Body Fat Percentage
Standard Deviation 4.6
|
2.63 Total Body Fat Percentage
Standard Deviation 2.0
|
5.58 Total Body Fat Percentage
Standard Deviation 2.6
|
3.67 Total Body Fat Percentage
Standard Deviation 2.3
|
PRIMARY outcome
Timeframe: within a single, 3-hour evaluationPopulation: total of 406 adults were screened for this study. Of those,199 met all inclusion and exclusion criteria and were considered eligible. 138 participants were enrolled into the clinical study and 134 participants (64 from MGH and 70 from PBRC) were included in the final analysis; four participants (2.9%) were removed from the final analysis due to poor image quality. Participants at PBRC underwent ADP (N = 70). Note that DXA images were never used, only their derived Body Fat percentage.
Comparison of Body Fat Percentage estimates to DXA references classified by Sex
Outcome measures
| Measure |
Smartphone Visual Body Composition (VBC)
n=134 Participants
VBC: Body Composition was measured using the VBC system that takes under 60 seconds per scan. Image captured for VBC included taking a series of four photos using a smartphone camera.
|
FitBit Aria 2 (cBIA 1)
n=123 Participants
Three consumer weight scales capable of BIA-based body composition analysis were included in the protocol: FitBit Aria 2 (Fitbit, San Francisco, CA); Tanita BF-684W (Tanita, Tokyo, Japan); and Renpho ES-24M-W/B (Joicom Corporation, Anaheim, CA). These scales are designated as cBIA 1, cBIA 2, and cBIA 3.(Consumer Bio-Impedance Analysis)
|
Tanita BF-684W (cBIA 2)
n=131 Participants
Three consumer weight scales capable of BIA-based body composition analysis were included in the protocol: FitBit Aria 2 (Fitbit, San Francisco, CA); Tanita BF-684W (Tanita, Tokyo, Japan); and Renpho ES-24M-W/B (Joicom Corporation, Anaheim, CA). These scales are designated as cBIA 1, cBIA 2, and cBIA 3.(Consumer Bio-Impedance Analysis)
|
Renpho ES-24M-W/B (cBIA 3)
n=131 Participants
Three consumer weight scales capable of BIA-based body composition analysis were included in the protocol: FitBit Aria 2 (Fitbit, San Francisco, CA); Tanita BF-684W (Tanita, Tokyo, Japan); and Renpho ES-24M-W/B (Joicom Corporation, Anaheim, CA). These scales are designated as cBIA 1, cBIA 2, and cBIA 3.(Consumer Bio-Impedance Analysis)
|
InBody S10 (pBIA1)
n=70 Participants
All participants also underwent professional BIA (pBIA) at PBRC with an InBody S10 (InBody Co., Seoul, Korea) and at MGH with a RJL system (Quantum IV, RJL Systems, Clinton Township, MI, USA.). The RJL system uses a single frequency (50 Khz) and four gel adhesive electrodes. Instead, InBody is multi-frequency and uses contact electrodes. InBody and RJL are designated as pBIA1 and pBIA 2, respectively, in the sections that follow and were analyzed separately. (Professional Bio-Impedance Analysis).
|
Quantum IV (pBIA 2)
n=64 Participants
All participants also underwent professional BIA (pBIA) at PBRC with an InBody S10 (InBody Co., Seoul, Korea) and at MGH with a RJL system (Quantum IV, RJL Systems, Clinton Township, MI, USA.). The RJL system uses a single frequency (50 Khz) and four gel adhesive electrodes. Instead, InBody is multi-frequency and uses contact electrodes. InBody and RJL are designated as pBIA1 and pBIA 2, respectively, in the sections that follow and were analyzed separately. (Professional Bio-Impedance Analysis).
|
BOD POD ADP
n=70 Participants
Participants who were evaluated at PBRC (N = 70) also had %BF assessed with the BOD POD Air displacement plethysmography (ADP) device (BodPod Gold Standard Body Composition Tracking System, COSMED, Rome, Italy).
|
|---|---|---|---|---|---|---|---|
|
Comparison of Body Fat Percentage Estimates to DXA References Classified by Sex
Female Participants
|
2.34 Total Body Fat Percentage
Standard Deviation 1.6
|
4.45 Total Body Fat Percentage
Standard Deviation 3.3
|
4.10 Total Body Fat Percentage
Standard Deviation 3.1
|
6.89 Total Body Fat Percentage
Standard Deviation 5.3
|
2.98 Total Body Fat Percentage
Standard Deviation 1.8
|
4.54 Total Body Fat Percentage
Standard Deviation 2.9
|
2.90 Total Body Fat Percentage
Standard Deviation 2.1
|
|
Comparison of Body Fat Percentage Estimates to DXA References Classified by Sex
Male Participants
|
1.88 Total Body Fat Percentage
Standard Deviation 1.3
|
4.53 Total Body Fat Percentage
Standard Deviation 5.0
|
6.23 Total Body Fat Percentage
Standard Deviation 13.4
|
4.11 Total Body Fat Percentage
Standard Deviation 3.3
|
3.37 Total Body Fat Percentage
Standard Deviation 2.4
|
5.01 Total Body Fat Percentage
Standard Deviation 3.0
|
3.50 Total Body Fat Percentage
Standard Deviation 2.4
|
Adverse Events
Male Participants
Serious events: 0 serious events
Other events: 0 other events
Deaths: 0 deaths
Female Participants
Serious events: 0 serious events
Other events: 0 other events
Deaths: 0 deaths
Serious adverse events
Adverse event data not reported
Other adverse events
Adverse event data not reported
Additional Information
Dr. Steven Heymsfield
Pennington Biomedical Research Center
Phone: (225) 763-2541
Email: [email protected]
Results disclosure agreements
- Principal investigator is a sponsor employee
- Publication restrictions are in place