Table 3.

Results relating to variability movement trajectories.

	Main Study Outcomes	Outcomes Related to Movement Variability and the Freezing-to-Releasing Process	Interpretation of Results Relating to Freezing and/or Releasing DoF
Caty et al. (2009) [30]	Good reliability was established for the kinematic gait variables among stroke patients.	NR
Lofrumento et al. (2021) [31]	UCM hypothesis rejected: VORT consistently greater than UCM variability.	No treatment/time effects, except a slight increase in VUCM at the end of the swing phase.	The increase in VUCM at late swing is consistent with releasing DoF.
Papi et al. (2015) [32]	UCM method successfully applied: with rehabilitation, the UCM ratio becomes ‘like normal’ when using an ankle–foot orthosis. More research is required to confirm the findings from this N = 1 study.	Stroke patients consistently show higher VUCM than VORT with progression towards a ratio waveform similar to healthy controls at 6 months.	Both VUCM and VORT strongly decrease between the baseline and 3-month follow-up, consistent with a freezing DoF. In measures without an ankle–foot orthosis, a peak in VUCM disappears between 3 and 6 months, which is inconsistent with releasing DoF.
Shin et al. (2020) [33]	A relationship was established between the amount of movement around the lower limb joints and gait speed at follow-up.	NR
Roby-Brami et al. (2003) [34]	Stroke patients established different levels of coordination in reaching compared to healthy controls.	Patients with greater impairment recruited additional DoF (trunk bending) to compensate for the limited range of motion in distal joints.	Recruitment of more trunk activity is inconsistent with freezing DoF.

NR: None reported; V_UCM: variability in the uncontrolled manifold, which does not impact task performance; V_ORT: variability orthogonal to the uncontrolled manifold, which does affect the task performance; marked in bold: highlights the main result of this study.