Objective: To search for features discriminating different gait patterns during intercritical FOG-free phases in patients with anamnestic FOG, using spatiotemporal and kinematic variables and machine learning (ML).

Background: Freezing of gait (FOG) is defined as a brief and anecdotical absence of or marked reduction in walking in Parkinson’s disease (PD) patients, whose prevalence is often underestimated ^[1], because its clinical assessment may be quite challenging due to its episodic and unpredictable nature ^{[2, 3]}, especially in the first stage of disease and during intercritical phases.

Method: The study population included 52 PD patients classified as freezing-experienced (PD+FOG) or non-freezing-experienced patients (PD-FOG) based on self-report through an anamnestic question with a possible dichotomous response format (presence-absence), modeled on analogous instruments and previously published studies in the literature ^{[4, 5]}. All patients were assessed with complete clinical examination, dedicated scales and cognitive screening tests. The experiment was carried out with an optoelectronic system on patients performing a single walking task and a cognitive dual-task walking. Univariate statistical analysis and machine learning algorithms were employed to identify the features most contributing to the classification.

Results: Comparing PD+FOG vs PD-FOG patients (p<0.0001), the first ones had a higher trunk obliquity (p=0.035); a reduced hip and knee flexion angle and a more intra-rotated knee (p=0.009); consequently, PD+FOG had a reduced range of motion. Regarding the ML results, Decision Tree was the best algorithm in classifying the groups with an accuracy of 83.7% and the variables most involved in the dichotomic classification were related to pelvic tilt and knee internal-external rotation.

Conclusion: PD patients with anamnestic FOG exhibit a peculiar gait pattern on kinematic measures already in intercritical FOG-free walking: these findings highlight the potential of kinematic analysis as new biomarkers in identifying gait abnormalities in PD patients with anamnestic FOG, even beyond freezing episodes and potentially underlying different phenotypes.

References: 1. M. Amboni, F. Stocchi, G. Abbruzzese, L. Morgante, M. Onofrj, S. Ruggieri, M. Tinazzi, M. Zappia, M. Attar, D. Colombo, L. Simoni, A. Ori, P. Barone, A. Antonini, Prevalence and associated features of self-reported freezing of gait in Parkinson disease: The DEEP FOG study, Parkinsonism Relat. Disord. 21 (2015) 644–649. https://doi.org/10.1016/j.parkreldis.2015.03.028.
2. Factor SA et al., A possible pathway to freezing of gait in Parkinson’s disease. J Parkinsons Dis. 2025. Mar;15(2):282-290.
3. M. Mancini, K. Smulders, R.G. Cohen, F.B. Horak, N. Giladi, J.G. Nutt, The clinical significance of freezing while turning in Parkinson’s disease, Neuroscience 343 (2017) 222–228. https://doi.org/10.1016/j.neuroscience.2016.11.045.Factor SA et al.
4. Nanhoe-Mahabier W, Snijders AH, Delval A, Weerdesteyn V, Duysens J, Overeem S, Bloem BR. Walking patterns in Parkinson’s disease with and without freezing of gait. Neuroscience. 2011 May 19;182:217-24. doi: 10.1016/j.neuroscience.2011.02.061. Epub 2011 Mar 5. PMID: 21382449.
5. Castelli Gattinara Di Zubiena F, Zampogna A, Patera M, Cusolito G, Apa L, Mileti I, Cannuli A, Suppa A, Paoloni M, Del Prete Z, Palermo E. Uncovering the Kinematic Signature of Freezing of Gait in Parkinson’s Disease Through Wearable Inertial Sensors. Sensors (Basel). 2025 Aug 14;25(16):5054. doi: 10.3390/s25165054. PMID: 40871914; PMCID: PMC12390408.

« Back to 2026 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/beyond-freezing-of-gait-episodes-in-parkinsons-disease-intercritical-kinematic-gait-pattern-matters/