Freezing of gait is a complex, heterogeneous, and highly variable phenomenon whose pathophysiology and neural signature remains enigmatic. Evidence suggests that freezing is associated with impairments across cognitive, motor and affective domains; however, most research to date has focused on investigating one axis of freezing of gait in isolation. This has led to inconsistent findings and a range of different pathophysiological models of freezing of gait, due in large part to the tendency for studies to investigate freezing of gait as a homogeneous entity. To investigate the neural mechanisms of this heterogeneity, we used an established virtual reality paradigm to elicit freezing behaviour in 41 Parkinson’s disease patients with freezing of gait and examined individual differences in the component processes (i.e. cognitive, motor and affective function) that underlie freezing of gait in conjunction with task-based functional MRI. First, we combined three unique components of the freezing phenotype: impaired set-shifting ability, step time variability, and self-reported anxiety and depression in a principal components analysis to estimate the severity of freezing behaviour with a multivariate approach. By combining these measures, we were then able to interrogate the pattern of task-based functional connectivity associated with freezing (compared to normal foot tapping) in a sub-cohort of 20 participants who experienced sufficient amounts of freezing during task functional MRI. Specifically, we used the first principal component from our behavioural analysis to classify patterns of functional connectivity into those that were associated with: (i) increased severity; (ii) increased compensation; or (iii) those that were independent of freezing severity. Coupling between the cognitive and limbic networks was associated with ‘worse freezing severity’, whereas anti-coupling between the putamen and the cognitive and limbic networks was related to ‘increased compensation’. Additionally, anti-coupling between cognitive cortical regions and the caudate nucleus were ‘independent of freezing severity’ and thus may represent common neural underpinnings of freezing that are unaffected by heterogenous factors. Finally, we related these connectivity patterns to each of the individual components (cognitive, motor, affective) in turn, thus exposing latent heterogeneity in the freezing phenotype, while also identifying critical functional network signatures that may represent potential targets for novel therapeutic intervention. In conclusion, our findings provide confirmatory evidence for systems-level impairments in the pathophysiology of freezing of gait and further advance our understanding of the whole-brain deficits that mediate symptom expression in Parkinson’s disease.

Individuals with Parkinson's disease (PD) commonly experience freezing of gait under time constraints, in narrow spaces, and in the dark. One commonality between these different situations is that they may all provoke anxiety, yet anxiety has never been directly examined as a cause of FOG. In this study, virtual reality was used to induce anxiety and evaluate whether it directly causes FOG. Fourteen patients with PD and freezing of gait (Freezers) and 17 PD without freezing of gait (Non-Freezers) were instructed to walk in two virtual environments: (i) across a plank that was located on the ground (LOW), (ii) across a plank above a deep pit (HIGH). Multiple synchronized motion capture cameras updated participants' movement through the virtual environment in real-time, while their gait was recorded. Anxiety levels were evaluated after each trial using self-assessment manikins. Freezers performed the experiment on two separate occasions (in their ON and OFF state). Freezers reported higher levels of anxiety compared to Non-Freezers (p<0.001) and all patients reported greater levels of anxiety when walking across the HIGH plank compared to the LOW (p<0.001). Freezers experienced significantly more freezing of gait episodes (p = 0.013) and spent a significantly greater percentage of each trial frozen (p = 0.005) when crossing the HIGH plank. This finding was even more pronounced when comparing Freezers in their OFF state. Freezers also had greater step length variability in the HIGH compared to the LOW condition, while the step length variability in Non-Freezers did not change. In conclusion, this was the first study to directly compare freezing of gait in anxious and non-anxious situations. These results present strong evidence that anxiety is an important mechanism underlying freezing of gait and supports the notion that the limbic system may have a profound contribution to freezing in PD.

The main objective of this study was to determine how manipulating the amount of sensory information available about the body and surrounding environment influenced freezing of gait (FOG), while walking through a doorway. It was hypothesized that the more limited the sensory information, the greater the occurrence of freezing of gait. Nineteen patients with Parkinsoǹs disease who experience freezing of gait (PD-FOG) walked through a doorway or into open space in complete darkness. The three doorway conditions included: (i) FRAME (DARK) – walking through the remembered door frame; (ii) FRAME - walking through the door with the door frame illuminated; (iii) FRAME+BODY - walking through the door (both the door and the limbs illuminated). Additionally, two conditions of walking away from the doorway included: (iv) NO FRAME (DARK) - walking into open space; (v) NO FRAME+BODY - walking into open space with the limbs illuminated, to evaluate whether perception (or fear) of the doorway might account for FOG behaviour. Key outcome measures included: the number of freezing of gait episodes recorded, total duration of freezing of gait, and the percentage of time spent frozen. Significantly more freezing of gait episodes occurred when participants walked toward the doorway in complete darkness compared to walking into open space (p<0.05). Similar to previous studies, velocity (p<0.001) and step length (p<0.0001) significantly decreased when walking through the door in complete darkness, compared to all other conditions. Significant increases in step width variability were also identified but only when walking into open space (p<0.005). These results support the notion that sensory deficits may have a profound impact on freezing of gait that need to be carefully considered.