Upper extremity soft and rigid tissue mass prediction using segment anthropometric measures and DXA

Area of Science:

• Biomedical Engineering
• Anthropometry
• Human Physiology

Background:

• Existing regression equations for predicting body composition are limited to lower extremities.
• Accurate in-vivo tissue mass estimation for upper extremities is needed for biomechanical modeling.

Purpose of the Study:

• To develop and validate regression equations for predicting bone mineral content (BMC), fat mass (FM), lean mass (LM), and wobbling mass (WM) of the arm and forearm using anthropometric measures.
• To assess the accuracy of these equations in a young adult population.

Main Methods:

• Multiple linear stepwise regression was used to derive prediction equations from anthropometric data (segment lengths, circumferences, breadths, skin folds).
• Dual-energy X-ray Absorptiometry (DXA) scans provided actual tissue masses for equation development and validation.
• Validation was performed on an independent sample of 24 healthy university-aged participants.

Main Results:

• Prediction equations demonstrated high adjusted R(2) values (0.854–0.968), with greater variance explained for LM and WM compared to BMC and FM.
• Scatter plots showed strong relationships between actual and predicted masses (R(2) 0.681–0.951).
• Validation group errors ranged from -2.2% to 15.5%, with RMS errors from 7.92g to 180.26g.

Conclusions:

• Simple anthropometric measurements can accurately predict in-vivo upper extremity tissue masses in young adults.
• These validated equations facilitate more precise biomechanical models for predicting dynamic responses to impact.
• The study provides essential data for upper limb body composition analysis.