Abstract
Introduction & Purpose
Interpersonal postural synchronization is well-documented (Oullier et al., 2008; Richardson et al., 2007) and often attributed to motor resonance, an unconscious neural mirroring of observed actions (Bisio et al., 2014; Bolzoni et al., 2023). However, the specific alignment of shared movement strategies between partners during complex postural tasks remains unclear. Therefore, this study examined postural behavior during an interactive, partner-based balance task under three conditions: no interaction, visual interaction, and tactile interaction. Grounded in dynamic systems theory, which suggests that seemingly independent systems can synchronize through a self-organized, interaction-driven process (in the visual condition through perception, and in the tactile condition additionally through mechanical coupling), we hypothesized that visual and tactile interactions would elicit synchronized (i.e. correlating) postural balancing movements between partners compared to the non-interactive baseline.
Methods
Thirty-eight participants (19 pairs, 25 females, mean age = 24.4 ± 2.8 years) maintained a bipedal, hip-width stance on balance boards (MFT Challenge Disc 2.0, Vienna, Austria) across three 15-second conditions: (1) back-to-back (baseline); (2) facing with touching palms (tactile); and (3) facing while alternately displaying numbers with both hands and reading them aloud (visual). Each pair completed one 15 s trial per condition, with condition order randomized across pairs. Markerless whole-body kinematics were recorded at 100 Hz using a motion capture system (Theia Markerless Inc., Kingston ON, Canada) and exported at 200 Hz after Theia3D's internal interpolation. The resulting 3D trajectories were globally aligned by mirroring one partner's horizontal coordinates, allowing concatenation across trials and subjects. After centering on each joint's mean and scaling by Euclidean distance, a single PCA was performed on the pooled data, so that the resulting principal movement (PM) variables represent identical aspects of the movement across participants and can be analyzed accordingly (Federolf, 2016). The analysis was limited to the first four PMs, as they together explained approximately 80.1% of the total variance. The PMs were animated as stick figures in the original 3D space to interpret the underlying movement strategies (Federolf, 2016). Interpersonal PM correlations were assessed via Pearson correlations of the scores, Fisher z-transformed, and compared across conditions using repeated-measures ANOVAs (α = 0.05), with Mauchly's sphericity test. P-values were Bonferroni-corrected both across the four PMs and for the pairwise post-hoc comparisons within each. Coupling was temporally stable across the trial, as confirmed by a windowed cross-correlation analysis.
Results
Figure 1 illustrates the interpersonal correlations for the first four PMs (PM1–PM4), which varied significantly across conditions (all p <.004). Post-hoc pairwise comparisons revealed significant differences between all three conditions for PM1 and PM2 (all p <.001). For PM3 and PM4, both the baseline and visual conditions differed significantly from the tactile condition (all p <.001), while no significant difference was found between the baseline and visual condition.
Discussion
Our findings support the hypothesis that visual and tactile interactions between partners lead to an increased correlation in movement strategies compared to the non-interactive baseline. However, this effect appears limited to specific movement strategies. The two most dominant movement strategies (PM1 and PM2), characterized by anterior-posterior whole-body shifts combined with lateral upper-body shifts, accounted for 70.1% of the total variance. These primary strategies showed a strong inverse correlation not only during physical contact but, remarkably, also under the visual condition. Since the participants' body orientation was face-to-face and their horizontal coordinates were inverted, the inverse correlations indicate a functional convergence of their movements (e.g., an anterior shift by one partner corresponded to a posterior shift by the other). This suggests that visual information alone can contribute to interpersonal coupling of balance movements, a process potentially driven by unconscious motor resonance. In contrast, PM3 and PM4, representing arm and shoulder movements, showed a strong correlation only during the tactile condition, likely due to the direct mechanical linkage between partners' hands. Since PM3 involves simultaneous vertical (upward-downward) arm and shoulder movements primarily along the z-axis, a positive correlation was observed, as the z-coordinates were not inverted. Conversely, as PM4 involves opposing vertical arm and shoulder movements, the functional convergence here is reflected by an inverse correlation. While previous research has established the existence of general postural coordination (Oullier et al., 2008; Richardson et al., 2007), the present PCA-based approach allowed us to identify the specific movement strategies where this coupling emerged, and the specific movement components where it did not emerge.
Conclusion
Interactive balance does not emerge as a global synchronization of postural movements, but rather as the dynamic alignment of specific, context-dependent movement strategies. By showing that visual and tactile interaction affect strategy-specific interpersonal coupling, these findings may be relevant for partner-based rehabilitation exercises or for dyadic sports.
References
Bisio, A., Sciutti, A., Nori, F., Metta, G., Fadiga, L., Sandini, G., & Pozzo, T. (2014). Motor Contagion during Human-Human and Human-Robot Interaction. PLOS ONE, 9(8), e106172. https://doi.org/10.1371/journal.pone.0106172
Bolzoni, F., Montanari, B., Farinelli, V., Adamo, P., Temporiti, F., Marino, G., Fesce, R., & Gatti, R. (2023). Motor resonance to non-visible postural adaptation: A novel aspect of the mirror mechanism. European Journal of Neuroscience, 58(2), 2515–2522. https://doi.org/10.1111/ejn.16055
Federolf, P. A. (2016). A novel approach to study human posture control: “Principal movements” obtained from a principal component analysis of kinematic marker data. Journal of Biomechanics, 49(3), 364–370. https://doi.org/10.1016/j.jbiomech.2015.12.030
Oullier, O., de Guzman ,Gonzalo C., Jantzen ,Kelly J., Lagarde ,Julien, & and Scott Kelso, J. A. (2008). Social coordination dynamics: Measuring human bonding. Social Neuroscience, 3(2), 178–192. https://doi.org/10.1080/17470910701563392
Richardson, M. J., Marsh, K. L., Isenhower, R. W., Goodman, J. R. L., & Schmidt, R. C. (2007). Rocking together: Dynamics of intentional and unintentional interpersonal coordination. Human Movement Science, 26(6), 867–891. https://doi.org/10.1016/j.humov.2007.07.002