New Models for Predicting Body Fat from BMI
Author Information
Author(s): David G Levitt, Steven B Heymsfield, Richard N Pierson Jr., Sue A Shapses, John G Kral
Primary Institution: University of Minnesota
Hypothesis
Can we develop more accurate regression relations for predicting body fat fraction from BMI and body density, especially for morbidly obese individuals?
Conclusion
New regression relations for predicting body fat fraction from BMI and body density provide more accurate insights into body composition changes in obesity.
Supporting Evidence
- The study included a diverse sample of 1356 subjects to enhance the validity of the regression models.
- New regression equations were developed that are more accurate than existing models for predicting body fat.
- Significant differences in fat fraction were observed among different ethnic groups for the same BMI.
Takeaway
This study created better ways to figure out how much fat someone has based on their height and weight, especially for people who are very overweight.
Methodology
The study analyzed body fat using 3 compartment measurements of body density and total body water from a large database of subjects.
Potential Biases
Potential bias in the applicability of BMI-based predictions for different ethnic groups and age ranges.
Limitations
The study assumes a constant value of body density for morbidly obese subjects, which may not be accurate.
Participant Demographics
Participants included 1356 adults aged 18 and over, with a BMI range of 17 to 65, across various ethnic groups.
Statistical Information
P-Value
p<0.01
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website