Exercise might bias skeletal-muscle fat fraction calculation from Dixon images.

We examined the influence of a single exercise session on quantitative muscle fat fraction MRI measurements. Ten healthy volunteers were scanned on a 3T body scanner before and after a session of bilateral squats until muscular fatigue. Axial in- and opposed phase images were acquired at a fixed distance from the knee joint and fat fractions were calculated using a 2-point Dixon technique as well as muscle cross sectional area at the same position. After the squat session, calculated fat fraction in the quadriceps bilaterally appeared to be significantly decreased, while all but one non-exercised muscles showed no change. In conclusion exercise might modify the measured apparent fat fraction. Trials using quantitative MRI should consider the timing of scanning sessions and physical examinations to avoid bias caused by the influence of exercise on measurements.