Dual-energy X-ray absorptiometry (DXA)

Dual-energy X-ray absorptiometry (DXA), also called a DEXA scan, divides the body into bone, lean soft-tissue mass and fat mass.^{1Schoenfeld, Science and Development of Muscle Hypertrophy, 2nd ed., p. 71.} A total-body scan uses two very low-energy X-ray beams and typically gives less radiation than a cross-country flight.^{2Ozemek et al., ACSM’s Guidelines for Exercise Testing and Prescription, 12th ed., p. 193.} It can also report fat and muscle by region, such as the trunk and limbs.^{2Ozemek et al., ACSM’s Guidelines for Exercise Testing and Prescription, 12th ed., p. 193.} Often called a gold standard for body composition, DXA stays mostly in clinical and research settings because of its cost, specialized equipment and staffing needs.^{3Helms et al., The Muscle & Strength Pyramid: Nutrition, 2nd ed., p. 177.},^{4Ozemek et al., ACSM’s Guidelines for Exercise Testing and Prescription, 12th ed., p. 191.}

Reported prediction error runs from 1.2% to 4.8%.^{2Ozemek et al., ACSM’s Guidelines for Exercise Testing and Prescription, 12th ed., p. 193.} Body positioning, hydration, food intake and manufacturers’ formulas all add error.^{2Ozemek et al., ACSM’s Guidelines for Exercise Testing and Prescription, 12th ed., p. 193.} Even under good conditions, repeat body fat percentage readings in one person rarely get below ±2% error, so a 10% reading represents roughly 8% to 12%.^{3Helms et al., The Muscle & Strength Pyramid: Nutrition, 2nd ed., p. 177.}

In practice

• Scan fasted: eating can raise DXA lean-mass estimates by up to 1.7% for the whole body and 3% for regions, and extra glycogen and its water can add several kilograms.^{5Schoenfeld, Science and Development of Muscle Hypertrophy, 2nd ed., p. 72.}
• Use the same scanner and conditions each time; accuracy varies between machine types and populations, with biases of up to about 2 kg of fat-free mass.^{5Schoenfeld, Science and Development of Muscle Hypertrophy, 2nd ed., p. 72.}
• Be wary of frequent scans for small changes, because DXA tracks change over a training program less well than it measures a single time point.^{5Schoenfeld, Science and Development of Muscle Hypertrophy, 2nd ed., p. 72.} Inconsistent testing conditions can produce variation of ±10%.^{3Helms et al., The Muscle & Strength Pyramid: Nutrition, 2nd ed., p. 177.}