Prediction of segmental percent fat using anthropometric variables

Area of Science:

• Human physiology
• Body composition analysis
• Biostatistics

Background:

• Accurate assessment of body fat distribution is crucial for health monitoring.
• Traditional methods for measuring segmental body fat can be complex or inaccessible.
• Developing practical prediction equations from anthropometric data is needed.

Purpose of the Study:

• To develop and validate prediction equations for segmental percent fat (arms, legs, trunk).
• To utilize readily available anthropometric measurements for body fat estimation.
• To assess the accuracy and practicality of these new prediction models.

Main Methods:

• 107 adults (77 males, 30 females) aged 21-82 years participated.
• Anthropometric measurements included height, weight, waist/hip circumference, BMI, waist-hip ratio, and subcutaneous fat thickness (SFT) at 14 sites.
• Segmental percent fat was measured using dual-energy absorptiometry (DXA) as the reference standard.
• Stepwise multiple regression analysis was employed to develop prediction equations using sex, age, and anthropometric variables.
• Bland-Altman analysis was used to evaluate systematic error and limits of agreement.

Main Results:

• Prediction equations demonstrated high accuracy for %SF(arms) (R=0.919, SEE=3.333%, LA=6.5%) and %SF(legs) (R=0.915, SEE=3.468%, LA=6.5%).
• %SF(trunk) prediction showed good accuracy (R=0.858, SEE=4.944%, LA=9.7%), outperforming previous methods for trunk fat estimation.
• The developed equations utilized 5-7 predictors and met established standards for body fat prediction.

Conclusions:

• The developed prediction equations are useful and practical for estimating segmental percent fat.
• These equations provide a valuable tool for assessing body fat distribution.
• The findings support the use of anthropometric data for non-invasive body composition analysis.