Safe and independent mobility at home and in the community requires control of walking while accomplishing other functional tasks. A hallmark of healthy control of walking is automaticity, defined as the ability of the nervous system to successfully coordinate movement with minimal use of attention-demanding executive resources [1]. Recent evidence indicates that walking disorders are often characterized by a shift in the locomotor control strategy from healthy automaticity to compensatory executive control. This shift is potentially detrimental to walking performance as an executive control strategy is not optimized for locomotor control and it places excessive demands on a limited pool of cognitive reserve.

Here, the investigators hypothesize that walking automaticity, as measured by the prefrontal cortex activity while walking, will be improved by donepezil (a cholinesterase inhibitor).

A challenge to studying automaticity is that central nervous system circuitry for locomotion cannot be directly assessed in humans so shifts between automaticity and executive control have been inferred with a dual-task paradigm. If gait slows or deteriorates during a concurrent cognitive task, gait is deemed less automatic and using more cortical control. Recently, a more direct measure of automaticity has been studied with functional near infrared spectroscopy (fNIRS). fNIRS is used to measure changes in cortical oxygenated hemoglobin (HbO2) levels using light-tissue interaction properties of light within the near infrared range, similar to the way the HbO2 is detected with magnetic resonance imaging. Increased oxygenated hemoglobin are related to increased blood flow, which, in turn, reflects increased cortical activity.

A recent study shows higher than normal activation of the prefrontal cortex during usual walking in patients with PD, an indicator of increased cortical control and reduced automaticity. An enhanced understanding of the neurochemistry underlying gait automaticity in PD will have implications for pharmacologic management of gait impairment in PD. If reduced gait automaticity is due to depressed central dopamine, then more aggressive dopaminergic strategies may be in order. On the other hand, if reduced gait automaticity is due to depressed central cholinergic tone (as the investigators hypothesize), then cholinergic strategies will be more effective for optimizing gait in PD. Few studies have explored the positive effects of cholinergic augmentation on gait and falls but none have examined its direct effects on gait automaticity.

Aim I. Examine the effect of donepezil or levodopa on walking automaticity, as measured by prefrontal cortex activity while walking. This study will employ a double-blind, two-way crossover design in which 20 participants with idiopathic PD will be tested either after 14 days of placebo and their regular dose of levodopa or after 14 days of donepezil (5 mg/day oral) and their regular dose of levodopa. Levodopa doses are not determined or controlled by the investigators, as the recruited participants are already on levodopa specific to their own needs. The investigators will measure automaticity by measuring the activity of the prefrontal cortex while walking in single- and dual-task conditions with a wireless fNIRS system. The investigators predict that donepezil plus levodopa will decrease the prefrontal cortex activity, indicating more automaticity, during walking compared to placebo and levodopa.

Aim II. Relate walking automaticity with spatio-temporal measures of walking. In addition to walking automaticity measured with fNIRS while walking, subjects will be equipped with 3 inertial sensors to measure spatio-temporal features of gait, such as stride time, stride length and their variability. The investigators hypothesize that variability of gait will be related to level of prefrontal cortex activity. The investigators predict that variability of gait will not change with levodopa, reflecting prefrontal control of walking. In contrast, the investigators predict that donepezil will improve variability of gait, both with and without a concurrent, cognitive task.

Inclusion Criteria:

• Subjects must be able to stand unassisted for a minute and to walk continuously for 2 minutes without assistance or assistive devices.
• Diagnosis of idiopathic Parkinson's disease with sensitivity to levodopa and off-medication Hoehn & Yahr scores of III-IV.
• Subjects must be currently taking levodopa, and not already taking donepezil
• The subjects must be able to appreciate the purpose of the research, give informed consent to participate, be able to cooperate with the testing and be compliant with taking the experimental medications.

Exclusion Criteria:

• Other factors affecting gait (hip replacement, musculoskeletal disorder, uncorrected vision or vestibular problem), or an inability to stand or walk for 2 minutes at a time. Major depression, hallucinations or other psychiatric disturbances will be exclusions.
• Medical problems that might be worsened by donepezil are exclusion criteria and include tachycardia, bradycardia, arrhythmias, and peptic ulcer disease.
• Use of anticholinergics for parkinsonism, cholinesterase inhibitors for cognitive problems, bladder antispasmodics for urinary urgency or tricyclic antidepressants for depression are contraindications.