Movement Control and Long-Latency Reflexes Are Reproducible Measures of Shoulder Neuromuscular Control

Background: Human physiology-based biomarkers, such as transcortical long-latency reflexes (LLRs) and movement control performance, are measurements used to evaluate human performance. We developed a method to assess human performance variables using a custom-designed visuomotor control device with the capability to examine performance accuracy and neurophysiological responses to unexpected perturbations. We assessed the internal consistency and reproducibility of this device during a shoulder tracking task including the performance accuracy and the transcortical long-latency reflexes during unexpected perturbations. Methods: 86 healthy young adults (49 females, right-handed, mean age 25.8 ± 9.5 years) were assessed for internal consistency across varying resistance and velocity conditions. We next determined test-retest reliability among 31 participants (17 females, right-handed, mean age 24.13 ± 3.8 years). We then determined the predictability of the LLRs and performance error during perturbations using the receiver operating characteristic analysis (ROC). Results: Our results supported excellent internal consistency, fair-to-good test-retest reliability for task performance accuracy, and fair-to-good transcortical LLR responses to perturbations (McDonald's omega > 0.9; intraclass correlation coefficients (ICCs, 0.63-0.82)). Tracking accuracy, changes in movement velocity, and infraspinatus LLRs were effective predictors of perturbation conditions (receiver operating characteristics: AUC 0.72-0.90). Conclusions: These findings support that performance-based biomarkers have moderate-to-good reliability and neurophysiology-based biomarkers have fair-to-good reliability when assessing human shoulder performance among healthy adults. Studies are currently underway to determine if these measures are reproducible across other joint movements and among people with musculoskeletal and central nervous system injury.