Gait disorders and falls are part of the clinical spectrum of parkinsonian syndromes. In progressive supranuclear palsy - Richardson’s syndrome (PSP-RS), falls are considered as a cardinal symptom (1). In other atypical parkinsonian (AP) syndromes, falls also occur but later during the course of the disease. In the Caribbean islands of Guadeloupe and Martinique, atypical parkinsonism is abnormally frequent, with levodopa-resistant motor symptoms, gait disorders and falls, dementia, and supranuclear-gaze-palsy (2). Unlike the classical forms of PSP, these patients have hallucinations, rapid eye movement sleep behaviour disorders (RBD), dysautonomia and cortical myoclonus. In PSP-RS patients, the basal ganglia, brainstem, cerebellum, and at the cortical level mainly the frontal lobes showed tau pathology and neurodegeneration, correlated with reduced gait velocity or stride length and postural imbalance (4). Degeneration of cholinergic neurons within the pedunculopontine nucleus (PPN) with decreased metabolic activity of the thalamic centromedian-parafascicular nucleus (CM-Pf) have also been linked to falls and postural instability (3). In Caribbean AP patients, neuropathological studies show a wider cortical distribution of tau pathology, with relative mesencephalic sparing (4). Finally, even though these two groups of patients suffer gait disorders and falls, the specific assessment of gait and postural control in Caribbean AP patients has never been performed and the link between gait and balance disorder severity and brain lesions has not been explored.
Here, we characterised gait and balance disorders in 16 Caribbean AP patients (mean age = 64.3 ± 6.8 years, 9F/7M), which we compared to 15 PSP-RS patients (mean age = 66.2 ± 5.9 years, 6F/9M), and 17 controls (mean age = 64.6 ± 5.1 years, 12F/5M) (Table S1), using clinical scales and gait initiation recordings. We also recorded oculomotor parameters. We measured brain atrophy using MRI imagery to calculate the Magnetic Resonance Parkinsonism Index (MRPI 2.0)(5), and performed Voxel Based Morphometry (VBM) analysis of T1 images to examine the changes in the grey and white matter volumes (GMV and WMV). To analyse the link between brain atrophy and gait and balance disorders, we performed principal and multiple factor analysis, and multi-table Partial Least square correlations. Detailed methods are provided in Supplementary File 1.
We found no significant differences between patient groups for age, sex, disease duration and the median delay for first fall (eTable 1). Caribbean AP had a significantly lower educational level and more frequent Rapid-eye movement Behaviour Disorders (RBD), and PSP-RS patients with more frequent pyramidal syndrome (eTable 1). No difference in clinical severity of gait and balance disorders was found between patients (eTables 1–2). Cognitive status was significantly altered in both patient groups compared to controls, with lower scores for MMS, backward span and total recall tests in the Caribbean relative to PSP-RS patients (eTable 1). Caribbean AP and PSP-RS patients showed altered gait initiation parameters compared to controls, with higher anticipatory postural adjustments (APAs) and double-stance phase durations and lower APAs, posterior centre of foot pressure (CoP) displacement, length, velocity of the first step, and braking index (see eMethods and eFig.1). PSP-RS patients also had higher double-stance duration, mediolateral CoP displacement and step width relative to Caribbean patients (eFig.1). Both patients groups had altered oculomotor parameters relative to controls (eTable 3).
For brain MRI, we found that the MRPI was higher in PSP-RS relative to both Caribbean AP patients and controls (eTable 4). We found significant decreases in GMV and WMV for both patients’ groups relative to controls in the striatum, insula, temporal, parietal and frontal cortices (Fig. 1, eTables 5 and 6), with additional decreases in WMV in the occipital lobe and the vermis of the cerebellum in Caribbean AP patients, and in the thalamus, mesencephalon, and cerebellar hemispheres bilaterally in PSP-RS (Fig. 1, eTables 5 and 6). We also found significantly lower GMV and WMV mainly in the right temporal pole and parietal superior lobule in Caribbean AP relative to PSP-RS patients; and in the thalamus, mesencephalon, striatum, cerebellar hemispheres, frontal and parietal areas, and the right temporal lobe in PSP-RS relative to Caribbean AP patients (Fig. 1, eTables 5 and 6).
Using multifactor analysis, we found two significant components (Fig. 1). The first component separated healthy controls from patients, with stronger loading on almost all clinical variables and gait and oculomotor parameters. Significant brain VBM correlations were found in the cerebellum, brainstem, parahippocampal gyrus, mid, inferior, superior and inferior orbital frontal gyri, posterior cingulate and supramarginalis gyri (Fig. 1). The second component separated Caribbean AP from PSP-RS patients with stronger loading on VBM data, including the cerebellum, brainstem, subcalloseal area, posterior and mid cingulate gyri, parahippocampal gyrus, anterior insula, mid, inferior and superior frontal and posterior orbital gyri, the frontal pole, medial premotor and primary motor cortices, left angular and supramarginalis gyri and the cuneus (Fig. 1). Clinical variables that distinguished Caribbean AP and PSP-RS patients were some cognitive scores (Fig. 1); postural control parameters (APAs, double-stance and step durations, APAs mediolateral displacement, step width and length, and vertical velocities); and oculomotor upward and downward gains (Fig. 1). In both patient groups, the severity of gait and balance disorders correlated with lateral cerebellum, caudate nucleus, inferior frontal gyrus and operculum, posterior cingulum, angular, postcentral and supramarginalis gyri GMV (Fig. 1). In Caribbean AP patients, the severity of gait and balance disorders also correlated with frontal areas (frontal pole, midfrontal, medial precentral gyrus), temporal tractus, anterior insula and cuneus GMV. In PSP-RS patients, the severity of gait and balance disorders also correlated with cerebral, cerebellar, and midbrain WMV, the MRPI, and the frontal (superior frontal, posterior orbital) and parahippocampal gyri GMV (Fig. 1).
In our study, both Caribbean AP and PSP-RS patients had severe gait and balance disorders, with precocious falls and altered gait and postural parameters (2), suggesting that these signs could not formally distinguish these two diseases. In both patient groups, gait and postural deficits correlated with GMV in the caudate nucleus, inferior frontal and pars orbicularis gyri, frontal operculum, posterior cingulate, postcentral and supramarginalis gyri (4, 6). These brain areas are implicated in processing and integration of vestibular, visual and somatosensory information, and changes in postcentral and supramarginal gyrus activity have been found during real or imagined walking in healthy subjects. Reduced GMV and dysfunction in these deep and cortical brain areas have also previously been associated with lower gait speed and/or longer double support time in PSP patients and cohorts of community-dwelling adults (7). The only significant difference in gait between patients was higher mediolateral APAs (8), step width and double-stance duration in PSP-RS relative to Caribbean AP patients, suggesting that PSP-RS patients had an enlarged support base with higher postural instability and ataxia. This could be due to the cerebellar atrophy found in PSP-RS patients (6), but also to that of the basal ganglia and mesencephalon with higher MRPI index (4) and cholinergic neuronal degeneration previously related to postural instability and falls (3). We also found that premotor and sensorimotor area volumes distinguished PSP-RS from Caribbean AP patients, suggesting that lesions of the SMA-PPN network, in addition to the cerebellar damage, could be the main contributors to the additional balance deficit of PSP-RS patients (6). In Caribbean AP patients, gait and balance disorders also correlated to atrophy of the medial precentral gyrus, anterior insula, angular gyrus and cuneus (4). The medial precentral gyrus is part of the primary motor cortex controlling leg muscle activity and is activated during mental imagery of walking. The anterior insula cortex has been involved in awareness and interoception, and the angular gyrus and cuneus are two interconnected structures involved in sensory processing and spatio-temporal self-location ability (9). This could indicate that gait and balance disorders in Caribbean AP patients are mainly related to cognitive deficit and changes in sensory processing and self-perception. Our Caribbean AP patients, also had a lower educational level, and more severe cognitive deficit with lower GMV and WMV in temporal lobe brain areas involved in numerous cognitive processes, i.e. the temporal pole, hippocampus, fusiform and inferior temporal gyri (10), in addition to brain areas involved in executive function such as the caudate nucleus, frontal pole and inferior frontal gyrus.
This study has some limitations. First, the sample size for each group was relatively low. However, these two disorders are rare, and we used strict inclusion criteria. Another limitation is the fact that diagnoses of Caribbean AP and PSP-RS were based on clinical features and not pathologically confirmed. However, the two groups could be accurately distinguished using PCA analysis suggesting that the baseline patient classification is accurate.
In conclusion, our results indicate that gait and balance disorders observed in Caribbean AP patients are mainly related to dysfunction of cortical brain areas involved in visuo-sensorimotor processing and self-awareness, whereas these signs mainly result from premotor-brainstem-cerebellar network dysfunction in PSP-RS patients, brain areas involved in initiation and maintenance of locomotor patterns and postural adaptation. Further longitudinal cohort studies are needed to examine the link between brain imagery changes and the occurrence of gait and balance disorders.