Are Cholinesterase Inhibitors Effective in Improving Balance in Parkin- son’s Disease?

Most falls in Parkinson’s disease (PD) appear to be due to a primary balance dysfunction rather than an environmental factor [1]. Accumulating evidence suggests that balance dysfunction cannot only be attributed to changes in the dopaminergic system that seems to be responsible for most of the other motor symptoms in PD [1]. The interventions often helpful in improving other PD symptoms, dopaminergic medications and deep brain stimulation, do not consistently improve balance and falls [2,3]. Degeneration in other brain areas, such as cholinergic cortex, the adrenergic locus coeruleus or the cholinergic/glutaminergic pedunculopontine nucleus may be the origin of the balance problems associated with PD [4].

network [12]. The frontoparietal attention network has been implicated as a mediator of space-based attention that may be critical for switching attention among sensory systems for spatial orientation for postural control [13]. The cholinergic system, in this case, may play an indirect role on balance via executive function. In fact, patients with executive function deficits show more balance and gait deficits and falls [14]. It is possible that cholinesterase inhibitors could help reduce falls through direct effects on balance and/or via indirect effects on cognition.
The current work is a pilot study with a primary aim to determine the effects of donepezil (a centrally acting cholinesterase inhibitor) on cognition and balance in non-demented persons with PD compared to a placebo. weeks of treatment with either placebo or donepezil with assessments at the beginning and end of the 6-week period, followed by a month washout, then 6 weeks of the other treatment again with assessments at the beginning and end.
Inclusion criteria were 1) idiopathic PD (as determined by UK Brain Bank Criteria) with a Hoehn and Yahr score of 2 to 4, 2) treated with levodopa for at least a year and on a stable antiparkinsonian regiment for at least one month, 3) abnormal dynamic posturography on screening with a composite score of 70 or less. Exclusion criteria were 1) dementia, defined as an MMSE<27, 2) other medical conditions affecting gait 3) inability to stand unassisted, 4) already on cholinesterase inhibitors or anticholinergic medications, or 5) other medical or psychiatric comorbidities that could interfere with compliance or safety.

Data collection
Balance was measured using the sensory organization test (SOT) on the NeuroCom Balance Master Clinical Research System platform (Neurocom International, Inc), which tests sway in 6 sensory conditions: 1) eyes open, 2) eyes closed, 3) sway referenced visual surround with a stable platform, 4) sway referenced platform with eyes open and 5) sway referenced visual surround and platform. Center of Pressure (CoP) was calculated from the force plate recordings. Forces and moments were recorded at 100Hz sampling frequency, and data were exported and analyzed with a custom algorithm in Matlab (MathWorks, Inc). Unlike commercial posturography, we measured medio-lateral postural sway dispersion because it is most highly related to falls [15,16].
In addition, Parkinson's severity scales included United Parkinson's Disease Rating Scale Motor Section (UPDRS-III) and cognitive testing included the MMSE at baseline, Stroop, and Trail Making Tests. A general physical and neurological exam, heart rate, and blood pressure measures where completed at each visit. During the active treatment phase subjects were given donepezil 5mg for three weeks and if tolerated, increased to 10mg for weeks four through six.
A total of twenty one subjects were enrolled in the study; ten withdrew because of adverse events and one subject did not meet enrollment criteria (MMSE was too low). Five withdrawals were related to the study drug with 4 due to nausea, vomiting, or gastrointestinal discomfort and one who felt freezing and balance was worse after starting the study drug. One dropped out after a possible transient ischemic attack, one after a serious fall, two due to scheduling concerns, and one due to worsened radicular pain.

Data analysis
The Root Mean Square distance (RMS) was calculated for condition 1-6 of the SOT, in both the antero-posterior (AP) and medio-lateral (ML) direction from the COP measurements, as measure of sway dispersion. Due to the small sample size, the difference in RMS, UPDRS motor, and cognitive function between the end and the beginning of each treatment phase (placebo or active drug) were calculated and we used a Student's t-test to assess if the difference significantly differed from 0 (where 0 represents no change). A Pearson product-moment correlation was used to assess the relationship between balance and cognitive performances. Data analysis and statistics were performed by a blinded researcher (MM).

Results
Balance performance improved with donepezil. Specifically, the change in ML RMS in condition 4 of the SOT, during the donepezil phase (beginning-end), was significantly different from zero (p=0.03), reflecting a significant decrease in the ML RMS. In contrast ML RMS did not change during the placebo phase (beginning-end), as depicted in Figure 1, left panel. Individual subject results are showed in Figure 1, right panel. Specifically, out of 10 participants, 4 improved more than 40%, 4 between 15% and 30%, and 2 didn't show changes in postural sway during the active phase of the study. There were no significant differences seen in the AP RMS, nor in conditions 1-3, 5 or 6 of the SOT, in either the donepezil or placebo phase.
A significant correlation between improvement in the Trail B-A time and improvement in the ML RMS was found in the active phase, r=0.8, p=0.001 (Figure 2). In addition, the change in the Stroop conflict time and trails B-A were -13.3 ± 24s and 5.4 ± 44s after the placebo phase versus -14 ± 23s and -13.8 ± 20s after the donepezil phase. None of those changes reached statistical significance.

Conclusion
In this small, pilot study, we did find that the cholinesterase inhibitor, donepezil, improved a measure of balance in persons with PD without dementia, and that there was a relationship between changes in executive function and changes in balance. These findings are in keeping with previous intervention studies that have showed cholinesterase inhibitors improve cognitive function in persons with PD [17] and Alzheimer's disease [18,19]. The improvement in the Trailmaking executive function might suggest that donepezil improved the ability to rapidly switch attention between letters and numbers.
The relationship between balance improvement and improvement in executive function may be causal or parallel effects. It is possible that attentional set-switching cognitive mechanisms are directly involved in switching attention among sensory systems for postural control, such as between somatosensory and vision in Condition 4 of the SOT. It is also possible that cholinergic innervation affects both executive function and balance in parallel and both factors influence fall risk. Perhaps the fall reduction seen in the study of Chung et al. could be attributable to improvements in both balance and executive function since an increase in mediolateral postural sway has been associated with falls [20,21] and cognitive impairments have been associated with falls [14]. Increased postural sway has been associated with thalamic cholinergic denervation in subjects with PD [22]. The effects specifically on SOT condition 4 support a role for the PPN-thalamic cholinergic projections in sensory integration [22]. The decreased ML RMS in the active phase of the study during one of the most problematic balance conditions for PD is when the sensory integration is between somatosensory altered information and visual-vestibular information. Control of standing balance depends on the ability to switch attentional focus among visual, proprioceptive, and vestibular systems, as needed when the environmental conditions change [23]. Previous studies indicated that integration of somatosensory with visual information is affected in PD [24], and that patients with PD show excessive postural sway in conditions with limited or inappropriate sensory feedback [25,26]. Therefore, we hypothesized that cholinergic treatment in PD may ameliorate postural instability, particularly in condition 4, in which somatosensory information is altered. In addition, the conditions including standing on a moving support surface have been associated with increased falls [20]. This is promising for future studies and suggests that cholinergic treatment in PD subjects may ameliorate postural instability and reduce falls through improved postural sensory integration function.
There are multiple limitations of this study. Our dropout rate was very high, which is concerning but seems atypical relative to other studies with donepezil in PD. Dubois completed a large study with 195 randomized to donepezil 5mg and another 182 to donepezil 10mg. Dropout rates in the active arms were 24.1% and 23.6% versus 17.9% in the placebo group with adverse events reported in 13.8%, 17%, and 11% respectively [17].
Also we only saw changes in one of the SOT conditions. This may represent a bias due to our small population or it could support previous results finding that poorer postural control has been reported in PD in situations where somatosensory information was impaired, perhaps reflecting difficulties to use or integrate visual or vestibular feedback, as reflected in condition 4 of the SOT [24,27]. This small pilot study is meant to inform future study direction and design. The improvements seen in certain measures of balance are promising and suggest further investigation is certainly warranted and should include in-depth measures of balance, specifically SOT4, and cognitive function as well as perhaps some dual task testing. With a lack of pharmacological or surgical treatments to improve balance and falls in PD these results are certainly exciting and deserve further investigation.