Author(s): FELSTEAD, A., AGGETT, J., ORDISH, C., BATES, M., HEFFERNAN, S.M., WALDRON, M., Institution: SWANSEA UNIVERSITY , Country: UNITED KINGDOM, Abstract-ID: 2269

INTRODUCTION:
Near-Infrared Spectroscopy (NIRS) has been extensively used for the measurement of post-occlusive reactive hyperaemia (PORH), which reflects the time-course and magnitude of skeletal muscle microvasculature reperfusion following a brief period of limb ischemia. It has been proposed that the oxyhaemoglobin concentration response ([O2Hb]) to PORH was best described by a Gompertz model [1]; however, the reliability of all model parameters and the goodness of fit among different Gompertz models has not been investigated. This study evaluated the test-retest reliability of NIRS-derived PORH tests of the lower-limb, using Gompertz models to characterise the [O2Hb] response.
METHODS:
Five healthy participants (1 female, 4 males; 23 ± 1 years) undertook two PORH measurements (45-min apart). After 10-min of supine rest, a 10 Hz NIRS device was positioned on the right medial calf and raised above the level of the heart. An occlusion cuff was fitted immediately proximal to the knee and inflated to occlude the lower-leg for 5-min at 300 mmHg. Following release of the cuff, changes in [O2Hb] were measured and the reperfusion kinetics were analysed using a custom-written script in R-Studio. The period between the start of the reperfusion and the peak hyperaemic response was used across all participants. Both three and four-parameter Gompertz models were used to characterise the response. Two reperfusion parameters were selected: Amplitude (A) as the peak vascular response relative to baseline, and rate constant (k). Goodness of Fit (GoF) statistics were calculated. Systematic differences between test 1 and test 2 were identified using a paired samples t-test and agreement between tests was calculated using Coefficient of Variation (CV) and 95% confidence intervals.
RESULTS:
A four-parameter Gompertz model would not converge for the current data, thus a three-parameter model was used. There were no differences in the means between the two trials (A: p=0.591, k: p=0.694), demonstrating no mean bias for both variables. The CV% of the A was 2.9%; 95% CI 0.7 to 5.0% and k was 19.0%; 95% CI 12.2 to 25.7%. The GoF, denoted by R2 values, across all three-parameter models was 0.983±0.015.
CONCLUSION:
The NIRS-derived A parameter from a three-parameter Gompertz model was more reliable than the k parameter during a PORH test. Based on reports that trained males have a ~30% and ~45% increase in the NIRS-derived amplitude and recovery rate during a PORH test, respectively, compared to untrained males [2] demonstrates that the measurement errors of the current test would be acceptable and permit identification of larger differences between phenotypes. However, potential users of this test should consider whether it could be used to identify analytical goals of smaller magnitude, particularly if k values are to be used, as the CV% was appreciably larger for this parameter.

References
1. Bopp et al. 2011, Eur. J. Appl. Physiol.
2. Rasica et al. 2022, Med Sci Sport Exerc