Single-sensor inertial assessment reveals divergent postoperative motor adaptation after major surgery: increased postural sway but guarded trunk-stiffened gait

Background: Early postoperative recovery is usually evaluated using complications, length of stay, pain, and clinical judgement, whereas objective information on postural control and gait is rarely available at the bedside. Wearable inertial sensors may provide simple digital biomarkers of functional reserve and postoperative motor adaptation. Methods: This paired observational study analysed adults undergoing major surgery who had complete preoperative and postoperative inertial measurement unit (IMU) recordings. A single WitMotion IMU was placed over C7 during four standardised stance tasks and three gait tasks. Root mean square (RMS) accelerations and angular velocities were extracted as primary IMU parameters. Preoperative and postoperative values were compared using paired Wilcoxon signed-rank tests with rank-biserial effect sizes and Hodges-Lehmann 95% confidence intervals. Benjamini-Hochberg false-discovery-rate correction was applied across 61 IMU comparisons. Spearman rank correlations with 95% confidence intervals explored associations between preoperative IMU metrics, age, surgical duration, haemoglobin drop, and postoperative function. Results: Thirty-one patients had complete paired preoperative and postoperative recordings. After Benjamini–Hochberg false-discovery-rate correction across 61 IMU comparisons, 13 comparisons remained significant. The strongest FDR-surviving stance effects were observed during eyes-open narrow stance, with increased RMS angular velocity in pitch (rank-biserial r = 0.67), roll (r = 0.54), and yaw (r = 0.65), all p_FDR ≤ 0.04. During gait, RMS trunk accelerations and angular velocities decreased across all three axes, most prominently in gyro yaw (r = 0.61, p_FDR = 0.031) and gyro roll (r = 0.67, p_FDR = 0.020) during eyes-open walking, consistent with a guarded, trunk-stiffened strategy. Preoperative metrics were strongly correlated with postoperative metrics (Spearman ρ = 0.68–0.80). Older age was associated with greater postoperative balance impairment. Conclusion: A single C7-mounted IMU identified a clinically interpretable postoperative motor phenotype: impaired standing balance combined with cautious trunk-stiffened gait. Preoperative trunk motion metrics may provide objective markers of functional reserve and could help target postoperative mobilisation and prehabilitation pathways. Larger studies with clinical endpoints are required before implementation.