Reliability of arterial spin labelling measurements of perfusion within the quadriceps during steady-state exercise

Arterial spin labelling (ASL) provides a potential method to non-invasively determine muscle blood flow and examine the impact of interventions such as supplementation and training. However, it's a method with intrinsically low signal, leading to limitations in accuracy and temporal resolution. To examine these limitations, the current study measured perfusion via ASL on three occasions in the rectus femoris of 10 healthy adults, during light and moderate exercise, over three different exercise durations. For data sampled over 9 min, light intensity exercise gave an average perfusion of 35.0 ± 5.1 ml/min.100g(-1) with a coefficient of variation (COV) of 16% and single intraclass correlation coefficient (ICC) of 0.67. For the moderate bout, perfusion was 51.3 ± 5.6 ml/min.100g(-1) (COV 10%, ICC 0.82). When the same data were analyzed over 5 min 24 s, perfusion was 37.8 ± 11.13 (COV 30%, ICC 0.13) during light and 49.5 ± 8.8 ml/min.100g(-1) (COV 18%, ICC 0.52) during moderate exercise. When sampling was reduced to 1 min 48 s, perfusion was 41.2 ± 13.7 (COV 33%, ICC 0.26) during light and 49.5 ± 13.6 ml/min.100g(-1) (COV 28%, ICC 0.04) during moderate exercise. For 9 min a significant perfusion difference was found between the exercise intensities; however, this was not the case for sampling over 5 min 24 s or 1 min 48 s. Such findings illustrate the potential of ASL to non-invasively monitor muscle perfusion under steady-state conditions, but highlight that extended exercise protocols are necessary in order to generate date of sufficient reliability to be able to discriminate intervention dependent perfusion differences.