Test-retest reliability of a laboratory-based head acceleration testing protocol

To evaluate the test-retest reliability of a novel laboratory-based protocol for inducing and measuring head acceleration in multiple directions under anticipated and unanticipated conditions.

Laboratory-based test-retest reliability study.

Thirty physically active adults (50% female) completed standardized head perturbations using a custom-built apparatus. Perturbations were applied in flexion, extension, lateral flexion, and rotation under both anticipated (cervical muscle pre-activation) and unanticipated (no cervical pre-activation) conditions across two sessions conducted within one week. Peak linear head acceleration (g) and rotational head acceleration (rad/s²) were recorded using a motion capture system. Reliability was assessed using intraclass correlation coefficients (ICC_3,1), standard error of measurement (SEM%), and Bland-Altman plots.

For anticipated perturbations, both linear head acceleration and rotational head acceleration demonstrated good to excellent test-retest reliability (intraclass correlation coefficient 0.75-0.90), with SEM% ranging from 3.52% to 8.54%. For unanticipated perturbations, reliability was within the moderate to good range (intraclass correlation coefficient 0.72-0.85), with SEM% ranging from 3.99% to 11.72%. Bland-Altman plots indicated no systematic differences between sessions and no evidence of heteroscedasticity.

Linear and rotational head acceleration showed moderate to excellent reliability across multiple directions and anticipatory conditions, supporting the utility of this laboratory protocol for head impact biomechanics research. Unanticipated perturbations and rotational accelerations were more variable, likely reflecting the complexity of reflexive and asymmetric neuromuscular responses. Together, these findings provide a methodological foundation for future studies on head impact biomechanics and cervical neuromuscular function.